MORPHOLOGY FOR IMPROVED SELECTIVITY

TECHNICAL FIELD

[001] The technical field generally relates to catalytic methods for carbon monoxide (CO) reduction, and more particularly to electrocatalysts composed of metallic material such as Cu and associated methods of manufacture and use in electrochemical CO reduction for the production of C3+ alcohols. The technical field also relates to REDOX active catalysts and related methods where the catalysts have a confinement cavity morphology.

BACKGROUND

[002] The electrosynthesis of higher-order alcohols from carbon dioxide (CO2) and carbon monoxide (CO) addresses the need for long-term storage of renewable electricity. Unfortunately, present-day performance remains below that needed for practical applications. There is a need for improved techniques and catalyst materials for efficient electrochemical reduction of gases, such as CO, and related methods and systems of producing chemical compounds, such as C3+ alcohols.

SUMMARY

[003] In some implementations, there is provided an electrocatalyst for electroreduction of carbon monoxide (CO) to produce C3 alcohols, the electrocatalyst comprising metal nanoparticles comprising copper (Cu) and having a morphology including open nanocavities sized and configured to provide confinement of reactive intermediates within the nanocavities to promote selective production of the C3 alcohols.

[004] In some implementations, the metal nanoparticles have Cu as exclusive catalytic metal therein. In some implementations, the open nanocavities have a total open angle between about 35° and about 90°. In some implementations, the open nanocavities have a total open angle between about 45° and about 90°, between about 50° and about 80°, or between about 55° and about 65°. In some implementations, the metal nanoparticles have an average size between about 50 nm and about 200 nm, or between about 70 nm and about 150 nm. In some implementations, the metal nanoparticles have an average size between about 90 nm and nm 130 nm. In some implementations, the metal nanoparticles have an average aspect ratio between about 1.5 and about 0.5, or between about 1.2 and about 0.8. In some implementations, the metal nanoparticles are generally spheroid in shape. In some implementations, the metal nanoparticles have a shell structure walls defining the respective nanocavities and the walls have a wall thickness between about 1 nm and about 10 nm, or between about 1.5 nm and about 5 nm. In some implementations, most of the metal nanoparticles each have multiple openings to provide communication with the nanocavities. In some implementations, the openings are generally circular in shape. In some implementations, the openings each have a rough edge. In some implementations, the nanocavities have a volume between about 5,000 nm ² and about 10,000 nm ² , or between about 7,000 nm ² and about 9,000 nm ² . In some implementations, the metal nanoparticles are each entirely composed of Cu, which means that only trace or impurity amounts of other compounds are present. In some implementations, the metal nanoparticles have exposed nanocavity surfaces that consist of Cu. In some implementations, the metal nanoparticles further comprise exposed outer surfaces that comprise Cu and catalyse electroreduction reactions. In some implementations, the exposed outer surfaces consist of Cu. In some implementations, the metal nanoparticles are enclosed over about 70% to about 90% of an external surface area thereof.

[005] In some implementations, the open nanocavities are sized and configured to provide a propanol Faradaic efficiency of at least 15 at a conversion rate of 7.8 ± 0.5 mA cm ^-2 at -0.56 V versus a reversible hydrogen electrode. In some implementations, the open nanocavities are sized and configured to provide a propanol Faradaic efficiency of at least 20 at a conversion rate of 7.8 ± 0.5 mA cm ^-2 at -0.56 V versus a reversible hydrogen electrode.

[006] In some implementations, the electrocatalysts are formed as a deposited catalyst layer on a first side of gas diffusion membrane, wherein the deposited catalyst layer is configured to be in direct contact with an electrolyte and wherein a second opposed side of the gas diffusion membrane is configured to be in direct contact with a CO-containing gas. In some implementations, the C3 alcohols comprises propanol.

[007] In some implementations, there is provided an electrocatalyst for electroreduction of a carbon-containing gas to produce C3+ multi-carbon compounds, the electrocatalyst comprising metal nanoparticles comprising a catalytic metal and having a morphology including open nanocavities sized and configured to provide confinement of reactive intermediates within the nanocavities to promote selective production of the C3+ multi carbon compounds.

[008] In some implementations, carbon-containing gas comprises carbon monoxide (CO) or carbon dioxide (CO2). In some implementations, the C3+ multi-carbon compounds comprise C3 alcohols. In some implementations, the catalytic metal comprises copper (Cu). In some implementations, the electrocatalyst has one or more features as defined in any one of the sections above or herein.

[009] In some implementations, there is provided a method of fabricating a copper (Cu) electrocatalyst for electroreduction of carbon monoxide (CO) into C3 alcohols, comprising: forming Cu ₂ 0 nanoparticles having a closed and hollow structure; treating the Cu ₂ 0 nanoparticles to form open nanocavities therein; and subjecting the Cu ₂ 0 nanoparticles to electroreduction conditions in the presence of an electrolyte and of CO to reduce the Cu ₂ 0 to Cu and thereby form the Cu electrocatalyst.

[0010] In some implementations, the Cu ₂ 0 nanoparticles are formed via nucleation and growth of Cu ₂ 0 nanocrystals. In some implementations, the forming of the Cu ₂ 0 nanoparticles comprises dissolving a sulfate in distilled water to form a solution; and adding a copper salt to the solution to induce the nucleation. In some implementations, the sulfate comprises sodium dodecyl sulfate. In some implementations, the copper salt comprises CuCI ₂ . In some implementations, the treating of the Cu ₂ 0 nanoparticles comprises etching the Cu ₂ 0 nanoparticles to form the open nanocavities. In some implementations, the etching comprises acidic etching. In some implementations, the etching is performed under gentle etching conditions. In some implementations, the etching is performed within the solution. In some implementations, the etching comprises adding a strong acid to the solution. In some implementations, the strong acid is HCI. In some implementations, the etching further comprises aging the solution for a residence time sufficient to provide the nanocavities having a total open angle between about 35° and about 90°. In some implementations, the method includes adding a hydroxide base to the solution. In some implementations, the hydroxide base comprises NaOH. In some implementations, the solution is aged between about 4 hours and about 6 hours after acid addition. In some implementations, the subjecting of the Cu ₂ 0 nanoparticles to electroreduction conditions comprises depositing the Cu ₂ 0 nanoparticles onto a first side of a gas diffusion membrane to form an electrode having a catalyst layer; exposing the catalyst layer to the electrolyte; exposing an opposed second side of the gas diffusion membrane to a CO-containing gas which passes therethrough; and applying a voltage to provide a current density to cause the CO gas contacting the catalyst layer to be electrochemically converted into a multi-carbon compound. In some implementations, the gas diffusion membrane is composed of a carbon substrate. In some implementations, the electrolyte is a solution of KOH. In some implementations, the depositing of the Cu ₂ 0 nanoparticles onto the gas diffusion membrane is performed by spray-coating. In some implementations, the Cu ₂ 0 nanoparticles are deposited in an amount between about 0.001 and about 0.0015 grams per cm ² . In some implementations, the Cu electrocatalyst is formed within 100 seconds of applying the electroreduction conditions. In some implementations, the Cu electrocatalyst further has one or more features as defined herein.

[0011] In some implementations, there is provided a method of fabricating a metal electrocatalyst for electroreduction of a carbon-containing gas into C3+ multi-carbon compounds, comprising forming metal oxide nanoparticles having a closed and hollow structure; treating the metal oxide nanoparticles to form open nanocavities therein; and subjecting the metal oxide nanoparticles to electroreduction conditions in the presence of an electrolyte and of a carbon-containing gas to reduce the metal oxides to the corresponding metal and thereby form the metal electrocatalyst.

[0012] In some implementations, there is provided a use of the electrocatalyst as defined above or herein for electrocatalytic reduction of CO into at least one C3 alcohol compound. In some implementations, the at least one C3 alcohol compound comprises propanol.

[0013] In some implementations, there is provided a process for electrochemical production of a C3 multi-carbon compound from CO, comprising: contacting CO gas and an electrolyte with an electrode comprising the electrocatalyst as defined herein, such that the CO contacts the electrocatalyst; applying a voltage to provide a current density to cause the CO gas contacting the electrocatalyst to be electrochemically converted into the C3 hydrocarbon; and recovering the C3 multi-carbon compound. [0014] In some implementations, the electrolyte comprises an alkaline compound. In some implementations, the electrolyte comprises KOH and/or other alkaline solutions. In some implementations, the process is conducted in a three-electrode flow-cell.

[0015] In some implementations, there is provided a system for CO or CO2 electroreduction to produce a multi-carbon compound, comprising an electrolytic cell configured to receive a liquid electrolyte and CO or CO2 gas; an anode; a cathode comprising an electrocatalyst as defined above or herein; and a voltage source to provide a current density to cause the CO or CO2 gas contacting the electrocatalyst to be electrochemically converted into the multi-carbon compound.

[0016] In some implementations, there is provided a catalyst precursor comprising metal oxide nanoparticles comprising cuprous oxide (CU2O) and having a morphology including open nanocavities having an open angle between about 35° and about 90°, and being electro-reduceable to form Cu surfaces at least within the open nanocavities to facilitate electroreduction of CO to form C3 alcohols.

[0017] In some implementations, the open nanocavities have a total open angle between about 45° and about 90°, between about 50° and about 80°, or between about 55° and about 65°. In some implementations, the metal oxide nanoparticles have an average size between about 50 nm and about 200 nm, or between about 70 nm and about 150 nm. In some implementations, the metal oxide nanoparticles have an average size between about 90 nm and nm 130 nm. In some implementations, the metal oxide nanoparticles have an average aspect ratio between about 1.5 and about 0.5, or between about 1.2 and about 0.8. In some implementations, the metal oxide nanoparticles are generally spheroid in shape. In some implementations, the metal oxide nanoparticles have a shell structure walls defining the respective nanocavities and the walls have a wall thickness between about 1 nm and about 10 nm, or between about 1.5 nm and about 5 nm. In some implementations, most of the metal oxide nanoparticles each have multiple openings to provide communication with the nanocavities. In some implementations, the openings are generally circular in shape. In some implementations, the openings each have a rough edge. In some implementations, the nanocavities have a volume between about 5,000 nm ² and about 10,000 nm ² , or between about 7,000 nm ² and about 9,000 nm ² . In some implementations, the metal oxide nanoparticles are enclosed over about 70% to about 90% of an external surface area thereof. In some implementations, the precursor is formed as a deposited layer on a first side of gas diffusion membrane.

[0018] In some implementations, there is provided an electrocatalyst for electrocatalyzing a REDOX reaction of a carbon-containing reactant to produce a multi-carbon product, the electrocatalyst comprising metal nanoparticles comprising a catalytic metal and having a morphology including open nanocavities sized and configured to provide confinement of reactive intermediates within the nanocavities to promote selective production of the multi carbon product.

[0019] In some implementations, the carbon-containing reactant comprises carbon monoxide (CO) or carbon dioxide (CO2) and the REDOX reaction is electroreduction to form C3+ compounds, and the catalytic metal comprises Cu. In some implementations, the electrocatalyst has one or more features described above.

[0020] In some implementations, electrocatalyst the carbon-containing reactant comprises ethylene and the REDOX reaction is electro-oxidation to form ethylene glycol, and the catalytic metal comprises Au, Pt or Pd. electrocatalyst the open nanocavities have a total open angle between about 35° and about 90°, between about 45° and about 90°, or between about 50° and about 80°, or between about 55° and about 65°. In some implementations, the metal nanoparticles have an average size between about 50 nm and about 200 nm, between about 70 nm and about 150 nm. In some implementations, the metal nanoparticles have an average size between about 90 nm and nm 130 nm. In some implementations, the metal nanoparticles have an average aspect ratio between about 1.5 and about 0.5. In some implementations, the metal nanoparticles have an average aspect ratio between about 1.2 and about 0.8. In some implementations, the metal nanoparticles are generally spheroid in shape. In some implementations, the metal nanoparticles have a shell structure walls defining the respective nanocavities and the walls have a wall thickness between about 1 nm and about 10 nm. In some implementations, the wall thickness is between about 1.5 nm and about 5 nm. In some implementations, most of the metal nanoparticles each have multiple openings to provide communication with the nanocavities. In some implementations, the openings are generally circular in shape. In some implementations, the openings each have a rough edge. In some implementations, the nanocavities have a volume between about 5,000 nm ² and about 10,000 nm ² or between about 7,000 nm ² and about 9,000 nm ² . In some implementations, the metal nanoparticles are enclosed over about 70% to about 90% of an external surface area thereof. In some implementations, the electrocatalyst is formed as a deposited catalyst layer on a first side of gas diffusion membrane, wherein the deposited catalyst layer is configured to be in direct contact with an electrolyte and wherein a second opposed side of the gas diffusion membrane is configured to be in direct contact with the carbon- containing reactant.

[0021] In some implementations, there is provided a method of fabricating a metal electrocatalyst for REDOX conversion of a carbon-containing reactant into a multi-carbon product, comprising forming precursor nanoparticles having a closed and hollow structure and being composed of a metal oxide; treating the precursor nanoparticles to form open nanocavities therein; and subjecting the precursor nanoparticles to conditions to convert the metal oxide to the corresponding metal.

[0022] In some implementations, there is provided a process for electrochemical production of a multi-carbon compound from a carbon-containing reactant, comprising: contacting the carbon-containing reactant and an electrolyte with an electrode comprising the electrocatalyst as defined above or herein, such that the carbon-containing reactant contacts the electrocatalyst; applying a voltage to provide a current density to cause the carbon-containing reactant contacting the electrocatalyst to be electrochemically converted into the multi-carbon compound by corresponding REDOX reactions; and recovering the multi-carbon compound.

[0023] In some implementations, the REDOX reactions comprise electrooxidation and the electrocatalyst has a catalytic metal comprising Au, Pt or Pd; the reaction can be the conversion of ethylene into ethylene glycol. In some implementations, the REDOX reactions comprise electroreduction and the electrocatalyst has a catalytic metal comprising Cu; the reaction can be the conversion of CO or CO2 into a multi-carbon compound, such as an alcohol like propanol.

[0024] In some implementations, there is provided a system for electro-conversion of a carbon-containing reactant to produce a multi-carbon compound, comprising an electrolytic cell configured to receive a liquid electrolyte and the carbon-containing reactant; an anode and a cathode, one of which comprises an electrocatalyst as defined herein; and a voltage source to provide a current density to cause the carbon-containing reactant contacting the electrocatalyst to be electrochemically converted into the multi carbon compound via REDOX reactions.

DESCRIPTION OF DRAWINGS

[0025] Fig. 1 : Computed concentration and flux distribution of species a-c, CO, C2, and C3 concentrations (color map) and flux distributions (arrows) on the cavity confinement structure d, Ratio of C3/C2 productivity, measured by the total out-flux of C2 and C3 products, as a function of cavity open angle e, A schematic showing how the cavity confinement effect promotes C2 species binding and further conversion to C3. f, Energy profile of C3 formation intermediate. Geometries of intermediate states and transition states are shown as insets (Only the CO species in the reaction are illustrated). Red, grey and orange balls stand for oxygen, carbon, and copper atoms, respectively.

[0026] Fig. 2: Structural characterization of the 3D cavity confinement in nanocatalysts a, Scanning electron microscopy (SEM) image of initial Cu ₂ 0 nanomaterials, showing the cavity structures in particles b, Scanning transmission electron microscopy annular dark field (STEM-ADF) image. Inset shows the magnified image: the red and yellow balls represent Cu and O atoms, respectively c, Corresponding Fast Fourier Transform (FFT) image d, The fabrication of cavity Cu nanocatalyst via in-situ electroreduction from initial CU2O. e-j, SEM, STEM-ADF, FFT, TEM, and Energy dispersive X-ray spectroscopy (EDS) mapping images of derived Cu nanocatalysts, showing the retaining cavity structure.

[0027] Fig. 3: Characterization of the as-prepared Cu ₂ 0 (yellow curves) and post-CO reduction reaction Cu (red curves) nanocatalysts synthesized via in-situ electroreduction a-b, X-ray photoelectron spectroscopy (XPS) spectra of Cu 2p (a) and Cu LMM (b), showing the change from Cu ⁺ to Cu°. c-e, Powder X-ray diffraction (PXRD) spectra and grazing incidence wide angle X-ray scattering (GIWAXS) 2D maps of the nanocavity electrodes, showing the material phase change from Cu ₂ 0 to Cu. f-g, In-situ X-ray absorption spectra measurement of Cu ₂ 0 nanocavity electrode with time evolution (0-90 seconds) in a flow cell system, further verifying the pure Cu in 3D nanocavity electrode after electroreduction f, In-situ Cu K-edge X-ray absorption near edge spectra (XANES). g, Extended X-ray absorption fine structure (EXAFS). Ex-situ Cu K-edge XAS of Cu ₂ 0 and metallic Cu were performed as standards. [0028] Fig. 4: CO electrochemical reduction performance in a flow cell system a, Plot of current density versus applied potential on Cu-based catalysts with different morphology structures, including solid, cavity-1, cavity-11, and fragment b, Propanol Faradaic efficiencies on the catalysts at different applied potentials, showing the best selectivity of the Cavity-ll catalyst compared to other samples c, Faradaic efficiencies of C2 products (acetate, ethanol, and ethylene) and C3 propanol on the Cavity-ll Cu catalyst under a range of applied potentials d, Representative SEM images of these catalysts. Scale bars are 100 nm. e, Faradaic efficiencies of C2 and C3 products on these catalysts at the applied potential of -0.56 V vs RHE. f, The experimental and FEM simulation results of C3/C2 product selectivity on different catalysts, showing the good agreement. Error bars correspond to the standard deviation of three or more measurements.

[0029] Fig. 5. Finite-element method (FEM) simulations a, FEM 2D axisymmetric modeling b, CO diffuses to the catalyst surface and gets adsorbed c, Two CO* can produce C2*. d, CO* and C2* to further get C3*. e, Production of C2 and C3, desorbed from the catalyst surface.

[0030] Fig. 6. Computed concentrations of CO, C2, and C3 species with different open levels for cavity a, Solid model (fully closed, 0°). b-f, cavity models with different open angles (10°, 30°, 60°, 90°, 180°). g, Fragment model (fully open, 270°).

[0031] Fig. 7. Simulated C2 and C3 species in different parts of the particle a, Schematic showing the interior and exterior of a cavity b, Interior and exterior surface areas of a cavity (c) Total and (d) area specific C2 and C3 species productivity in the interior and exterior of the particle.

[0032] Fig. 8. Geometries of CO dimerization. Top views of a, initial state, b, transition state, c, final state, and side views of d, initial state, e, transition state, f, final state. Red, grey and orange balls stand for oxygen, carbon and copper atoms, respectively. Water molecules are shown as lines. These notations are used throughout this document.

[0033] Fig. 9. Geometries of C1 and C2 coupling. Top views of a, initial state, b, transition state, c, final state, and side views of d, initial state, e, transition state, f, final state.

[0034] Fig. 10. Structural characterization of the 3D cavity confinement in nanocatalysts a, Energy dispersive X-ray spectroscopy (EDS) mapping images of the initial Cu ₂ 0. b-c, Scanning electron microscopy (SEM) and Scanning transmission electron microscopy (STEM) images of the nanocavity Cu.

[0035] Fig. 1 1. The open angle distribution of the derived Cu nanocavity.

[0036] Fig. 12. In-situ Cu K-edge X-ray absorption near edge spectra (XANES) and extended X-ray absorption fine structure (EXAFS) measurement of Cu ₂ 0 nanocavity electrode at -0.56 V vs RHE in 1 M KOH under CO reduction reaction in a flow cell system.

[0037] Fig. 13. Morphology and phase characterizations: SEM and grazing incidence wide angle X-ray scattering (GIWAX) of the initial Cu ₂ 0 electrodes a-b, Solid; c-d, Cavity I; e- f, Fragment. The structure characterizations of cavity II sample are shown in Fig. 2-3.

[0038] Fig. 14. a, Schematic illustration of the cathode flow cell system using a gas- diffusion layer coated nanocavity Cu electrode for CO electroreduction to propanol b, The photo of used flow cell.

[0039] Fig. 15. Electrochemical surface area measurement. Determination of double layer capacitance for the Cu catalysts in 1 M KOH with different morphologies: a-b, Solid; c-d, Cavity I; e-f, Cavity II; g-h, Fragment a, c, e, g, CVs taken over a range of scan rates b, d, f, h, Current due to double-layer charging plotted against CV scan rate i, Capacitance values and surface roughness factors j-k, The geometric current densities and the normalized current densities using different samples at -0.56 V vs RHE in 1 M KOH.

[0040] Fig. 16. Local electric potential around the nanocavity. The inner and outer surfaces of the nanocavity have a uniform electric potential. This is expected because of the high conductivity of the copper metal. The uniform electric potential on the surface of the catalyst ensures that the inner surface will react with the same driving force as the outer surface, and therefore the inner surface will be just as active in the catalytic reaction as the outer surface as long as it receives sufficient CO concentration by diffusion. Scale bar: 50 nm.

[0041] Fig. 17. Nuclear magnetic resonance spectra a, Representative ¹ H-NMR spectrum obtained from the catholyte of CO reduction on Cavity II Cu nanocatalyst at - 0.56 V vs RHE in 1 M KOH. DMSO (Dimethyl sulphoxide) is used as an internal standard for quantification of liquid products b, Faradic efficiency over reaction time for propanol at -0.56 V vs RHE in 1 M KOH. c, Faradaic efficiency distribution of C2 products (acetate, ethanol, and ethylene) and C3 propanol with different reaction time at -0.56 V vs RHE in 1 M KOH.

[0042] Fig. 18. Structural characterization of the cavity Cu catalyst after CO reduction reaction a, Post-10-min run. b, Post-30-min run. c, Post-1.5-hour run. d-e, Post-2.4-hour run. SEM and TEM images show the cavity morphology gradually changes to aggregation particles with time evolution. XRD spectrum in f demonstrates the catalyst phase is still metallic copper after reaction.

[0043] Fig. 19. Parameter sweep of Kf_C2 and Keq_C3. We obtain qualitative agreements between simulations and experiments by applying some simple fitting using the equilibrium and rate constants. There were five reaction-related parameters that we considered: three equilibrium coefficients, Keq_CO, Keq_C2 and Keq_C3, of adsorption- desorption of the CO, C2 and C3 species, as well as two rate constants, Kf_C2 and Kf_C3, of the C-C coupling and CO-C2 coupling steps. The results are shown in Supplementary Figs. 15-18. The parameters Keq_C3 and Kf_C2 imposed no change on the simulation results. This is expected because these parameters are independent of the CO-C2 coupling, which is the most critical step of the simulation (Supplementary Fig. 15). Varying the parameters, Keq_CO (Supplementary Fig. 16), Keq_C2 (Supplementary Fig. 17) and Kf_C3 (Supplementary Fig. 18), each in a range as wide as two orders of magnitude, does not significantly change the finding that the cavity geometry with an opening angle of 45 - 90° maximizes the C3/C2 ratio.

[0044] Fig. 20. Graphs showing parameter sweep of Keq_CO.

[0045] Fig. 21. Graphs showing parameter sweep of Keq_C2.

[0046] Fig. 22. Graphs showing parameter sweep of Kf_C3.

DETAILED DESCRIPTION

[0047] The present description relates to metal nanocatalyst materials having nanocavities shaped and configured to provide confinement of reactive species to promote the formation of C3+ compounds from reactants, such as carbon monoxide (CO) gas, in electroreduction conditions as well as related processes for producing the C3+ compounds and for manufacturing the nanocatalyst material. The present description particularly relates to electroreduction copper (Cu) catalysts having nanocavities that confine intermediates for efficient electrosynthesis of C3 alcohol fuels or fuel precursors from CO or CO2 as well as related methods and uses. Various aspects of the nanocatalyst, the processes and system that use such nanocatalysts, and the methods for manufacturing such nanocatalysts will be described in detail below.

[0048] Furthermore, the present description describes a catalyst design strategy that promotes C3 formation via nanoconfinement of C2 intermediates, thereby promoting C2:C1 coupling inside a reactive nanocavity. The present work employed finite-element method simulations to assess the potential for retention and binding of C2 intermediates as a function of cavity structure. The work then developed a method to synthesize open Cu nanocavity structures with tunable geometry via electroreduction of Cu ₂ 0 cavities formed via acidic etching. The example nanocavities showed a morphology-driven shift in selectivity from C2 to C3 products during CO electroreduction, reaching a propanol Faradaic efficiency of 21 ± 1 % at a conversion rate of 7.8 ± 0.5 mA cm ^-2 at -0.56 V versus a reversible hydrogen electrode.

[0049] For additional context, electrocatalytic reduction of carbon dioxide (CO2) to valuable carbon-based chemical feedstocks offers a route to long-term storage of renewable electricity that closes the carbon cycle. N-propanol, a higher-order alcohol, is desired for its high volumetric energy density (27 M J/L): cost-effective renewable propanol would offer a sustainable liquid fuel for existing internal combustion engines. It is of interest to explore means to increase selectivity in favour of electrochemical propanol production. With CC>2-to-CO conversion now well-established, subsequent electrocatalytic CO-to- propanol conversion shows promise; however, even here, prior methods have shown selectivities that remain in the vicinity of 10%, due primarily to preferred selectivity for C2 products.

[0050] Given that C3 formation can proceed via C-C coupling between C2 and C1 intermediates, it was hypothesized that modifying catalysts to target C2 intermediate binding could potentially be used to promote C3 production. The present work used finite- element method simulations (FEM) to study how a nanocavity confinement structure affects the binding and retention of C2 intermediates. Thus, it was found, suppresses net C2 loss that otherwise curtails C3 production. The present work then implemented 3D nanocavity Cu catalysts employing an in-situ electroreduction strategy, starting from Cu ₂ 0 nanoparticles (precatalyst) exhibiting an open structure. [0051] Prior studies of porous catalysts have exploited confined intermediates to some degree. One study used the confinement effect to explain a selectivity shift from C1 to C2 (see Yang K.D. et al). The present work further applied the confinement effect to boost C3 production by extending the retention of C2 species, and a comprehensive model that tracks the key species has been developed. The clearly mapped confinement model guides the design of catalysts, and enables shifting selectivity away from C2 products, leading to higher C3 production. In the present work, the optimal nanocavity catalyst reduces CO to propanol with 21 ± 1 % Faradaic efficiency at a conversion rate of 7.8 ± 0.5 mA cm ^-2 at -0.56 V versus a reversible hydrogen electrode.

Electrocatalyst and catalyst precursor

[0052] The electrocatalyst can be for electroreduction of CO to produce C3 alcohols, such as propanol, with relatively high selectivity compared to previous methods. The electrocatalyst can include metal nanoparticles comprising Cu and having a morphology including open nanocavities sized and configured to provide confinement of reactive intermediates within the nanocavities to promote selective production of the C3 alcohols.

[0053] As described herein, the metal nanoparticles are preferably Cu-based, and can be composed of Cu as the exclusive catalytic metal. However, in other alternative applications, the metal nanoparticles can be composed of other catalytic metals and/or a combination of multiple catalytic metals. Thus, the catalyst can include metal alloys. In addition, the catalytic metal can be doped. Depending on the particular reactant gas and the target C3+ multi-carbon product, the catalytic metal can be chosen accordingly. For example, the catalytic material could include transition metals (Mo, Fe, Co, and Ni)-based catalysts, post-transition metal (Al and Sn)-based catalysts, noble metals (gold, silver, platinum, palladium and rhodium)-based catalysts, metal-based heterogeneous electrocatalysts, metal-based homogeneous electrocatalysts, and doped catalysts. The catalytic metal could be alloyed with Pd, Zn, Ag, for example. In addition, depending on the particular reaction to be catalysed and the conditions to be used (e.g., electroreduction, electro-oxidation), the catalytic metal could be chosen accordingly. As an example, electro-oxidation catalysts could include Pd, Au, Pt or other metals.

[0054] As shown in the figures, the open nanocavities can be configured to have a certain open angle that promotes the production of certain products due to confinement of the intermediates. For example, in some implementations, the open nanocavities have a total open angle between about 35° and about 90°, between about 45° and about 90°, between about 50° and about 80°, between about 55° and about 65°. As shown in Fig 5, the open angle can be determined for spherical nanoparticles with respect to the center of the sphere. In addition, for nanoparticles with a single opening, the open angle can be determined by obtaining an image of the nanoparticles (e.g., see Fig 10) and then estimating the size of the particle and the opening and then determining the open angle with the vertex being at the approximate centre of the particle. For nanoparticles with multiple openings, the total open angle can be determined by adding together the open angles of the openings. This method of determining and providing openness of the nanoparticles is particularly suitable for spheroid and ovoid nanoparticles and those that can be generally approximated by a spheroid or ovoid shape. However, for other nanoparticle geometries, the openness of the nanocavities can be measured or determined in other ways.

[0055] In addition, the openness of the nanocavities for substantially spherical nanoparticles can also be measured and defined in other ways, such as the ratio between the diameter of the sphere and the diameter of the opening or the percentage of the surface area (outer or inner) that is open (e.g., 15% to 30% open for example). These openness characteristics can also be determined by visual inspection and estimation from images of the nanoparticles (e.g. see Fig 10). Other structures of the nanomaterial could include mesoporous structure that includes pores that act as nanocavities having certain openness to surrounding volumes.

[0056] In some implementations, the metal nanoparticles have an average size between about 50 nm and about 200 nm, between about 70 nm and about 150 nm, or between about 90 nm and nm 130 nm, which for spherical nanoparticles can be determined as an average diameter of the particles. For particles with other shapes, the size can be determined or defined in other ways, such as having an average length and width between 50 nm and about 200 nm.

[0057] The electrocatalyst nanoparticles can have different shapes, such as spheroid or ovoid. In some implementations, the metal nanoparticles have an average aspect ratio between about 1.5 and about 0.5, between about 1.2 and about 0.8, or close to 1 for spheres. In addition, the metal nanoparticles are preferably shaped to have a lower portion that has a closed bottom and an open upper part that defines the largest width or diameter of the particle, and a top portion that extends from the upper part inwardly and defines the opening. By extending inwardly from the upper part, the top portion provides an overhang in order to facilitate confinement of species within the nanocavities to promote formation of C3 multi-carbon compounds.

[0058] The metal nanoparticles can also be enclosed over about 70% to about 90% of an external surface area thereof.

[0059] In some implementations, the metal nanoparticles have a shell structure with walls defining the respective nanocavities. The walls have a wall thickness between about 1 nm and about 10 nm or between about 1.5 nm and about 5 nm. The walls of a given nanoparticle can have a relatively constant thickness or different thicknesses at different locations around the nanoparticle. The nanoparticles can each have one nanocavity, but it should be noted that it is also possible for single nanoparticles to include multiple nanocavities, where each nanocavity has one or more corresponding openings.

[0060] In some implementations, some of the metal nanoparticles each have respective single openings to provide communication with the nanocavities and the rest have multiple openings. In some implementations, most of the metal nanoparticles each have multiple openings to provide communication with the nanocavities. In some implementations, about 5% to about 50% or about 10% to about 30% of the particles have single openings. It is noted that the nanoparticles can have multiple openings, e.g., two, three or more openings. It was found that both single and multiple openings provided enhanced selectivity for C3 production from CO using the Cu nanoparticles in the present work. Single opening particles were convenient for simulation work, and the experimentation showed that multiple opening particles with a similar total open angle provided similar results to the simulations.

[0061] As illustrated in certain figures, in some implementations, the openings in the nanoparticles are generally circular in shape. The openings can be defined by smooth or rough edges. Of course, depending on the manufacturing method and the shape of the nanoparticles, the openings can have different sizes and shapes.

[0062] In some implementations, the nanocavities have a volume between about 5,000 nm ² and about 10,000 nm ² or between about 7,000 nm ² and about 9,000 nm ² . [0063] As mentioned above, in one example implementation, the metal nanoparticles are each entirely composed of Cu (except for impurities and trace amounts of other materials). This means that the entire nanoparticle, including the inner and outer surfaces of the particle and the walls are composed essentially of Cu. Preferably, the metal nanoparticles have exposed nanocavity surfaces that consist of Cu. The outer surfaces can be composed of Cu as well, but could also be composed of another substance, depending on the manufacturing method for example. Nevertheless, in some implementations, the metal nanoparticles include exposed outer surfaces that also comprise Cu and catalyse electroreduction reactions, the exposed outer surfaces optionally consisting of Cu. The ratio of the inner and outer surface areas of the nanoparticles can be close to one (e.g., the case of a sphere with thin walls), although it is also noted that higher surface area in the nanocavities compared to the outer surface area could further promote selective production of C3 multi-carbon compounds.

[0064] In some implementations, the open nanocavities are sized and configured to provide a propanol Faradaic efficiency of at least 15 or at least 20 or at least 21 at a conversion rate of 7.8 ± 0.5 mA cm ^-2 at -0.56 V versus a reversible hydrogen electrode.

[0065] As will be described in greater detail below, the electrocatalyst can be fabricated in various ways. In one implementation, the electrocatalyst is formed as a deposited catalyst layer on a first side of gas diffusion membrane, wherein the deposited catalyst layer is configured to be in direct contact with an electrolyte and wherein a second opposed side of the gas diffusion membrane is configured to be in direct contact with a CO-containing gas. In addition, a catalyst precursor optionally composed of Cu ₂ 0 and including etched hollow Cu ₂ 0 nanoparticles having openings can be deposited, and then the Cu ₂ 0 catalyst precursor can be subjected to in situ electroreduction conditions to form the Cu-based electrocatalyst.

[0066] While some implementations of the electrocatalyst can be composed of Cu and provided to convert CO into C3 alcohols by electroreduction, other implementations can use other metals and can be provided to convert other carbon-containing gases (e.g., C02) into various C3+ multi-carbon compounds. Broadly, the electrocatalyst includes metal nanoparticles comprising a catalytic metal and having a morphology including open nanocavities sized and configured to provide confinement of reactive intermediates within the nanocavities to promote selective production of the C3+ multi-carbon compounds. [0067] The carbon-containing gas can include carbon monoxide (CO) or carbon dioxide (CO2). The reactant gas can also be a mixture of CO and CO2, in various proportions, and can optionally include inert gas such as N2. Other reactants can also be used depending on the particular catalytic metal and reaction to be catalyzed. For example, for an electro oxidation reaction, the reactant can be ethylene and the product can be ethylene glycol, with the catalytic metal comprising Pd, Au or Pt, for example.

[0068] The C3+ multi-carbon compounds can include one or more C3 alcohols and optionally C4 alcohols as well. If different products are to be targeted, then the operating conditions, the reactants and the catalytic metal could be adapted accordingly.

[0069] As discussed above, the catalytic metal can include copper (Cu), particularly for producing multi-carbon products in CO2/CORR conditions. For other reactions, such as the electro-oxidation of ethylene to ethylene glycol, the catalytic metal can include Pd, Pt or Au, for example, which can be alloyed with other metals or doped, if desired. For other reactions, the catalytic metal can be chosen accordingly to favour the production of a target compound that cab benefit from confinement of reaction intermediates within nanocavities of the catalytic nanoparticles.

[0070] As mentioned above, one useful way to provide the metal electrocatalyst is to manufacture a catalyst precursor that can be implemented in an electroreduction process and in so doing the electrocatalyst with metal nanoparticles is formed in situ. The catalyst precursor can be a metal oxide (e.g., CU2O) including nanoparticles that have many of the properties that were mentioned above for the metal nanoparticles in terms of shape, size and configuration. The metal oxide nanoparticles can be formed by deposition to form hollow nanoparticles that are then subjected to a treatment to provide openings, to thereby form the nanocavities. After the catalyst precursor with nanocavities is formed, it can be subjected to in situ electroreduction to convert the oxide into the corresponding metal.

Methods of fabrication

[0071] There are various methods that can be used to fabricate the electrocatalyst. In one implementation, the method includes of fabricating a Cu electrocatalyst for electroreduction of CO into C3 alcohols includes forming CU2O nanoparticles having a closed and hollow structure; treating the CU2O nanoparticles to form open nanocavities therein; and subjecting the CU2O nanoparticles to electroreduction conditions in the presence of an electrolyte and of CO to reduce the Cu ₂ 0 to Cu and thereby form the Cu electrocatalyst.

[0072] The Cu ₂ 0 nanoparticles can be formed via nucleation and growth of Cu ₂ 0 nanocrystals. The forming of the Cu ₂ 0 nanoparticles can include dissolving a sulfate in distilled water to form a solution; adding a copper-based salt to the solution; adding an acid to the solution; aging the solution for an aging time sufficient to provide the nanocavities; adding a hydroxide based to the solution for neutralization. The precursor nanoparticles can then be deposited onto a substrate and then be subjected to electroreduction conditions. The method can be adapted for different metals other than Cu to enable the formation of precursor particles and etching to provide the openings and nanocavities.

[0073] In terms of the chemicals employed in the fabrication, various compounds and quantities can be used. For example, the sulfate can include sodium dodecyl sulfate, the copper salt can include CuCI ₂ , the acid can include HCI or any other etching acid, and the hydroxide base can include NaOH. Of course, other compounds can be used.

[0074] The aging time can be tailored to provide openings of a desired size. The aging can depend on the chemicals that are used, the concentration of the chemicals, the timing of addition of the different chemicals, and also on the wall thickness of the precursor nanoparticles, for example. The aging time can be, for example, between about 3 hours and about 6 hours, or between 4 and 5 hours. The aging time can be selected to generate an open angle of the nanocavities between about 45° and about 90°, or other angles disclosed herein, particularly when the particles are generally spherical. The treating of the Cu ₂ 0 nanoparticles to provide the openings can include etching the Cu ₂ 0 nanoparticles to form the open nanocavities. The etching can include acidic etching, which can be performed under gentle etching conditions. In one implementation, the etching uses an etching acid, such as HCI, added into the solution of the sulfate, water and Cu salt. The etching acid can be added simultaneously with the other ingredients or can be added afterword. The etching can be done using relatively mild or gentle conditions, which means using relatively low concentration of the etching acid and higher residence times, rather than high concentrations which can tend to produce nanocavities that are too large for the enhanced electivity effect. In one example, about 5 ml_ of 2 M of HCI was added to about 500 ml_ of solution, and it is noted that the mild etching can include the addition of a strong acid, such as HCI, in this concentration range ± 10 to 20%. The etching or aging time at such concentration ranges can be between about 4 hours to 6 hours, for example. Mild etching can enhance the consistency of the opening sizes, rather than having relatively irregular openings across the nanoparticles. Still, it is noted that various etching conditions can be used such that the selection of acid type, concentration, etching time, and other relevant operating conditions of the etching process are provided to given a desired range of open angles or openness of the particles.

[0075] In some implementations, subjecting the Cu ₂ 0 nanoparticles to electroreduction conditions includes depositing the Cu ₂ 0 nanoparticles onto a first side of a gas diffusion membrane to form an electrode having a catalyst layer; exposing the catalyst layer to the electrolyte; exposing an opposed second side of the gas diffusion membrane to a CO- containing gas which passes therethrough; and applying a voltage to provide a current density to cause the CO gas contacting the catalyst layer to be electrochemically converted into a multi-carbon compound. The in situ electroreduction preferably is performed at substantially the same or similar conditions as the eventual process for converting CO into C3 alcohols, although different operating conditions could also be used for the in situ electroreduction step of the manufacturing process.

[0076] During this step of electroreduction, the gas diffusion membrane is composed of a carbon substrate or another type of material. The electrolyte can be a solution of KOH or another appropriate base. The depositing of the Cu ₂ 0 nanoparticles onto the gas diffusion membrane can be performed, for example, by spray-coating. The Cu ₂ 0 nanoparticles can deposited in a certain amount, which can be defined in terms of thickness or weight per surface area. In some implementations, the Cu ₂ 0 nanoparticles can deposited between 0.001-0.0015 grams per cm ² .

[0077] In addition, the Cu electrocatalyst can be formed within 100 seconds of applying the electroreduction conditions. The electroreduction conditions can be provided accordingly.

[0078] Also provided is a method of fabricating a metal electrocatalyst for electroreduction of a carbon-containing gas into C3+ multi-carbon compounds. This method includes forming metal oxide nanoparticles having a closed and hollow structure; treating the metal oxide nanoparticles to form open nanocavities therein; and subjecting the metal oxide nanoparticles to electroreduction conditions in the presence of an electrolyte and of a carbon-containing gas to reduce the metal oxides to the corresponding metal and thereby form the metal electrocatalyst.

Process for producing C3 multi-carbon compounds such as alcohols

[0079] The electrocatalyst can be used to convert carbon-containing gases into multi carbon compounds. The carbon-containing gas can be or include CO, CO2, a mixture thereof that may or may not have other gases that may be inert. The carbon-containing gas could include other compounds depending on the reaction system and conditions that are being implemented. In one implementation, there is a process for electrochemical production of a C3 multi-carbon compound from CO, the process including contacting CO gas and an electrolyte with an electrode comprising the Cu-based electrocatalyst as defined herein, such that the CO contacts the electrocatalyst; applying a voltage to provide a current density to cause the CO gas contacting the electrocatalyst to be electrochemically converted into the C3 multi-carbon compound; and recovering the C3 multi-carbon compound.

[0080] Preferably, the C3 multi-carbon compound is an alcohol, such as propanol. The electrolyte can include an alkaline compound, such as KOH and/or other alkaline solutions.

[0081] The process can also be performed in a system CO electroreduction to produce a multi-carbon hydrocarbon or oxygenate. The system includes an electrolytic cell configured to receive a liquid electrolyte and CO gas; an anode; a cathode comprising an electrocatalyst as herein; and a voltage source to provide a current density to cause the CO gas contacting the electrocatalyst to be electrochemically converted into the multi carbon compound.

[0082] Various operating conditions can be used for the process and system in order to selectively produce a target C3 compound. Some values and ranges of such operating conditions are reported in the present description, but it should be understood that other operating conditions can be used depending on various factors.

[0083] It is also noted that the process and system mentioned above can also be applied with electrocatalysts that are composed of metals other than Cu, reactant gases other than CO, to produce a C3+ compound other than propanol. The catalytic metal can be adapted based on the desired reactions to be favoured.

EXPERIMENTATION & EXAMPLES

[0084] A number of experiments were conducted in the present work to assess and develop catalyst materials for the selective conversion of CO to C3 alcohol fuels. Details regarding experiments, simulations and results are provided below.

Results

[0085] Finite-element method simulations. It was hypothesized that a nanocavity structure could potentially concentrate C2 species via steric confinement, thereby limiting the desorption of C2 intermediates and promoting further conversion into a C3 product. The present work used FEM simulations to explore the prospects of cavity-enhanced C3 selectivity. Hollow spherical shells ^{21 23} (with an outer diameter of 100 nm and an inner diameter of 60 nm) with circular openings of various central angles were used to represent nanocavities immersed in an aqueous electrolyte (see Fig. 5a). In these simulations, CO molecules diffuse to the surface (Fig. 1a and Fig. 5b), are adsorbed, and are converted into C2 species at both interior and exterior surfaces of the nanoparticle (Fig. 1 b and Fig. 5c). The C2 species may then either desorb from the reactive surface as a C2 product, or be coupled with a CO molecule to form a C3 product (Fig. 1c and Fig. 5d-e). This work found that the cavity restricts the outflow of locally-produced C2 species (Fig. 1 b, arrows), leading to higher local C2 intermediate concentrations inside the cavity (Fig. 1 b, color map). The desorption of C2 intermediates is then reduced, leading to increased surface coverage and residence of the intermediates necessary for C3 production, and ultimately generating a heightened C3 production rate inside the cavity (Fig. 1 c).

[0086] In contrast, solid nanoparticles are not predicted to restrict materially the transport of C2 reactants away from their surfaces, and this leads to lower C2:C1 coupling rates (see Fig. 6a). To study the bounds of the cavity-enhancement effect, this work quantitatively assessed the angular-dependence of the nanocavity on C2 versus C3 selectivity from fully-closed to fully-open structures (see Fig. 6). Specifically, this work monitored the out-flux of C2 and C3 products from both the interior and exterior surfaces of the structures (see Fig. 7a). The productivity of both species in the interior cavity of the particles shows a strong dependence on the opening angle, in strong contrast to that at the exterior surfaces (see Fig. 7b-d). At small opening angles (< 30°), C2 and C3 productivities are low as CO reagent transport into the cavity is unduly restricted, resulting in CO limitation. At large opening angles (> 180°), the cavity does not succeed in containing the generated C2, and exhibits lower C3/C2 selectivity than all other cases. At intermediate angles (45 - 90°), the C3/C2 selectivity and overall C3 productivity appreciably exceed those found in the comparison cases (Fig. 1 d and Supplementary Table 1), with a maximum at ~ 60°. The interior and exterior in these cases produce a similar amount of total C2+ species (sum of C2 and C3) - it is the ratio of C3 to C2 that is significantly higher for the cavity interior. In these cases, the enrichment of C2 is not restricted by the reduced CO influx, resulting in an approximate 2.8-fold enhancement of C3/C2 production for this optimized cavity opening. Altogether, these simulations point to nanocavity intermediate confinement as a promising means to enhance the production of higher carbon products (Fig. 1 e).

[0087] This work also calculated the formation of C3 on Cu(11 1) using density functional theory (DFT) assuming that C3 is formed as a result of CO:C2 coupling. C2 formation is first calculated based on the dimerization of CO (Fig. 1f, Fig. 8 and Supplementary Table 2). In CO reduction, CO is abundant on the surface due to the strong adsorption energy of CO compared to C0 ₂ ; thus, the C2 and CO coupling was assumed as one of the most likely pathways for C3 formation (Fig. 1 f, Fig. 9, and Supplementary Table 2). This work’s calculations suggest that the coupling of C2 with a CO is favorable both thermodynamically and kinetically. The findings also suggest that the low stability of the C2 species will lead to low C2 surface coverage, a fact that reduces the likelihood that CO and C2 will meet to form C3; this accounts for the need to concentrate the C2 species via the nanocavity strategy.

[0088] Catalyst preparation and characterization. This work then sought to fabricate a tunable nanocavity copper catalyst. First, Cu ₂ 0 nanoparticles (with average sizes of 110 ± 20 nm, comparable to the scale of simulations) were synthesized via the nucleation and growth of nanocrystals. By applying a gentle acidic etching technique, the work then produced open structures, controlling the size of the hole via exposure time (details in Methods). The work characterized the resulting Cu ₂ 0 particles using scanning electron microscopy (SEM), and witnessed an open morphology in the majority of nanoparticles (see Fig. 2a). This work further characterized the particles via scanning transmission electron microscopy annular dark field (STEM-ADF), corresponding Fast Fourier Transform (FFT), and Energy Dispersive X-ray spectroscopy (EDS) mapping (see Fig. 2b- c and Fig. 10a).

[0089] Following synthesis, the Cu ₂ 0 particles were deposited onto a carbon substrate, and then the final nanocavity Cu catalyst was produced (Fig. 2d) via in-situ CO electrochemical reduction. SEM and TEM images confirm that the catalyst that exists following CO reduction reaction (CO-RR) retains the particle size and the open morphology of the original Cu ₂ 0; whereas the material structure has been electrochemically reduced to pure Cu as demonstrated by STEM-ADF, FFT, and EDS mapping (Fig. 2e-j, Figs. 10b-c and 11).

[0090] To characterize the in-situ electroreduced electrodes, and gain insight into the chemical state of the active catalyst, a series of spectral measurements were made on as- prepared and electrochemically derived samples. The spectra of Cu 2p and Auger Cu LMM conducted by X-ray photoelectron spectroscopy (XPS), taken together, show the valence state transition from Cu ⁺ to Cu° (Fig. 3a-b). The work used powder X-ray diffraction (PXRD) and grazing incidence wide angle X-ray scattering (Gl WAXS) to confirm the phase change from cuprite Cu ₂ 0 to cubic Cu (Fig. 3c-e).

[0091] The Cu K-edge X-ray absorption near edge spectra (XANES) and extended X-ray absorption fine structure (EXAFS) of the catalyst and corresponding reference standard materials demonstrate that the Cu ₂ 0 has been reduced to Cu in less than 100 seconds after the reaction, reconfirming the final nanocavity catalyst is metallic Cu (Fig. 3f-g).

[0092] Taken together, these studies (Fig. 2-3 and Fig. 12) indicate that the feature of Cu nanocavity catalyst provides the increased C3 selectivity regardless of other established factors known to influence eletroreduction performance, such as oxidation state, Cu defects, oxygen doping, or grain boundary density as documented in prior studies.

[0093] Electrochemical CO reduction performance. The present work then explored the CO-RR activity of the Cu catalysts derived from Solid, Cavity-I, Cavity-ll, and Fragment morphologies (see Fig. 13). The catalyst was deposited onto a carbon gas-diffusion electrode via spray-coating of a material ink (details in Methods) and the samples were tested in an engineered flow cell configuration (see Fig. 14). Compared to conventional H-cells, flow cells with gas-diffusion electrodes increase gas reactant availability at the electrode surface. [0094] The current density recorded on the Cavity-11 sample was larger than those of other opening angles at the same applied potential (Fig. 4a). Once normalized by the electrochemical surface area (ECSA, Supplementary Fig. 11 a-i), the current densities presented similar values (Fig. 15j-k) across samples and across the potential range applied. It was concluded that the overall activity towards CO-RR is substantially independent of the cavity morphology; and the attention of the work thus turned to selectivity.

[0095] The work evaluated electrocatalytic CO reduction performance in the potential range of -0.36 V to -1.76 V vs RHE in 1 M KOH solution. Propanol was detected using ¹ H nuclear magnetic resonance (NMR). The Cavity-ll nanocatalyst showed the highest Faradaic efficiency (FE) of propanol compared to other samples over the entire potential range, demonstrating the structural effect of cavity confinement on the propanol selectivity. Since Cu is an electrical conductor, the cavity structures is not expected to introduce a difference in the electric potential over the entire surface (including the interior and the exterior) of the cavity (see modeling result in Fig. 16). At -0.56 V vs RHE, the FE/propanol of Cavity-ll sample reaches 21 ± 1 % (Fig. 4b) with a partial current density of 7.8 ± 0.5 mA cm ^-2 (Supplementary Table 3). In Supplementary Table 4 this performance is compared with that of catalysts previously reported in the published literature.

[0096] The carbon-based product distribution, obtained using NMR (Fig. 17a) and gas chromatography, including C2 (acetate, ethanol, and ethylene) and propanol from the Cavity-ll nanocatalyst, is shown in Fig. 4c. The figure compares Cavity-ll with the Solid, Cavity-I, and Fragment controls (see Table 5).

[0097] The nanoconfinement of carbon-based intermediates within cavity structures of Cu catalysts (Fig. 4d) enabled a shift of product selectivity to C3 propanol. The combined product distributions at the optimized applied potential -0.56 V vs RHE (Fig. 4e), show that the increased propanol production in the Cavity-ll structure corresponds to decreased ethylene formation, with the mix acetic acid and ethanol remaining similar to the other catalyst structures. These results indicate that the C2 intermediate for propanol production via C-C coupling with CO inside the cavity is related to the ethylene, in agreement with previously reported mechanisms for propanol production, and as modeled in the FEM simulations performed in this work. The performance and structure change of the catalyst with time evolution were examined (Figs. 17b-c and 18). After significant reaction time, the catalyst has reconstructed to form aggregated Cu particles (Fig. 18d-f) that decreased the FE of propanol from 21 % to 14%, shifting the selectivity from C3 to C2 products (Fig. 17c). This result further confirms the nanocavity scenario as the source of the high C3 content.

[0098] The work also compared the experimental and simulated ratio of C3/C2 products. The FEM model qualitatively agrees with the experimental data in terms of the best geometry for C3 production, regardless of the adsorption/desorption equilibrium constants and reaction rate constants used in the model (Figs. 19-22). The FEM results can be further fitted to the experimental data to obtain qualitative agreements between the two (Fig. 4f), and the resulting parameters follow the trend predicted by the DFT calculations. This relatively simple model captures the key transport and geometric aspects of the nanocavities and provides a physical picture of the diffusive trapping of C2 intermediates and the resulting enhancement in C3 production.

Comments

[0099] This work provides a catalyst structuring approach wherein desired intermediates are concentrated to direct selectivity along a desired reaction pathway. FEM simulations, material structure characterization, and electrochemical measurements attest to the role of the naonocavity catalyst in improving catalytic performance and directing electrons to higher carbon products. These findings provide a physical route to tune chemical selectivity and enable the electroproduction of renewable liquid fuels and feedstocks.

Methods

[00100] Finite-element method simulations. Finite-element method (FEM) simulations was performed using the COMSOL Multiphysics software package Chtps.://vywW _; com sol.com/). Three modules were used to establish a comprehensive chemistry-mass transport model of the nanocavity structure. Detailed information regarding the FEM simulations can be found in Supplementary Methods further below.

[00101] Density functional theory calculations. In this work, all the density functional theory (DFT) calculations were carried out with a periodic slab model using the Vienna ab initio simulation program (VASP) (htjps;//www.yasp.at/). Detailed theoretical methods can be found in Supplementary Methods. [00102] Initial Cu ₂ 0 nanoparticles synthesis. This work synthesized the Cu ₂ 0 nanoparticles with the morphology, including Solid, Cavity-I, Cavity-ll, and Fragment according to a previously reported method (see Wu, H.-L. et al.). In a typical procedure, 4.5 g sodium dodecyl sulfate was dissolved in 450 ml_ distilled water in a glass reactor, and then added 65 mg CuCI ₂ . After that, 22.5 ml_ hydroxylamine hydrochloride (310 mg) solution and 4.5 ml_ 2 M HCI solution were injected into the reactor. Next, 12.5 ml_ NaOH (500 mg) solution was quickly added into the solution. The solution was aged at room temperature for 3 hours to prepare the Cavity-I, for 5 hours to prepare the Cavity-ll, and for 9 hours with 6 mL HCI to prepare the Fragment. The Solid Cu ₂ 0 was synthesized from the above procedure without adding HCI by aging 3h.

[00103] Derived Cu nanocatalyst synthesis. We prepared the derived Cu nanocatalyst electrode with different morphologies via in-situ CO electroreduction from the corresponding initial Cu ₂ 0 electrode, which can be obtained after initial running (2-5 min).

[00104] Electrochemical measurements. Electrocatalytic measurements were carried out in a three-electrode system using an electrochemical station (AUT50517). All potentials were measured against an Ag/AgCI reference electrode (3 M KCI, BASi) and converted to the reversible hydrogen electrode (RHE) reference scale using:

(1) E ( versus RHE ) = E ( versus Ag/AgCV) + 0.197 V + 0.0591 x pH

[00105] CO reduction product analysis. Gas-phase and liquid-phase products were quantified by gas chromatography (GC) and nuclear magnetic resonance (NMR) spectroscopy, respectively.

[00106] The GC, running Argon (Linde, 99.999%) as a carrier gas, contained a Molecular Sieve 5A and Carboxen-1000 columns. A thermal conductivity detector (TCD) was used to quantify hydrogen (H ₂ ) and a flame ionization detector (FID) was used to quantify ethylene (C ₂ H ₄ ).

[00107] The liquid products were quantified using nuclear magnetic resonance spectroscopy (NMR). ¹ H NMR spectra of freshly acquired samples were collected on Agilent DD2 500 spectrometer in 10% D ₂ 0 using water suppression mode, with Dimethyl sulfoxide (DMSO) as an internal standard. Sixteen second relaxation time between the pulses was used to allow for complete proton relaxation. The faradic efficiency (FE) of the liquid products was calculated from the total amount of charge Q (in units of coulombs) passed through the sample and the total amount of the liquid products produced n (in moles). Q = I x t, where I (in amperes) is the reduction current at a specific applied potential and t is the time (in seconds) for the constant reduction current.

[00108] The FE of the liquid products can be calculated as follows as examples:

where F is the Faraday constant.

[00109] Working electrode preparation and CO reduction measurements. To prepare a catalyst coated in a flow cell system, the work deposited 10 mg of catalyst mixed with 20 ul of 5 wt% Nafion in 1 ml_ methanol on a carbon gas-diffusion layer with the amount of ~1 mg cm ^-2 using the air-brush. The work combined the diffusion layer coated catalyst, anion exchange membrane, and nickel anode together using PTFE spacers such that a liquid electrolyte could be introduced into the chambers between the anode and membrane as well as the membrane and the cathode. Gaseous CO was passed through the gas chamber at the back side of the gas-diffusion layer coated catalysts. The electrolytes (KOH solution of various concentrations, 20 ml_) were circulated through both anode and cathode chamber in. The electrolyte flow was kept at 10 mL min ¹ . The CO (Linde, 99.99%) flow was kept constant at 50 mL min ^-1 using a mass flow controller.

[00110] Electrochemical active surface area (ECSA) measurement. Surface roughness factors for the four catalytic electrodes relative to the polycrystalline Cu foil were estimated from double layer capacitances. The capacitance (C _di ) was determined by measuring the geometric current at a potential window where no Faradaic process was occurring as a function of the scan rate of cyclic voltammetry stripping (CV). For this, CV was performed in 1 M KOH electrolyte with an anion exchange membrane. The potential window was between 0.19 V and 0.26 V vs RHE. The scan rates were from 40 mV/s to 160 mV/s with an interval of 20 mV/s. C _di was estimated by plotting the Aj = (j _a - j _c )l2 at 0.225 V vs RHE (where j _a and j _c are the anodic and cathodic current densities, respectively) against the scan rate. The slope gives the C _di . Supplementary Methods

[00111] Materials. Gas-diffusion layer (Commercial from FuelCellStore, Freudenberg H14C9), anion exchange membrane (Fumasep FAB-PK-130), nickel foam (1.6 mm thickness, MTI Corporation) anode, copper (II) chloride (CuC , 203149 sigma- aldrich) sodium dodecyl sulfate (SDS, 74255 sigma-aldrich), hydroxylamine hydrochloride (NH2OH HCI, 379921 sigma-aldrich), hydrogen chloride (HCI, 37%), and sodium hydroxide (NaOH, 795429 sigma-aldrich).

[00112] Finite-element method simulations. First, the Chemistry’ module was used to define the CO-RR intermediate steps. Three chemical species, CO, C2, and C3, each in a bulk and a surface adsorbed form, were defined to represent the CO feedstock, the C2 intermediate, and the C3 product. Five reactions were defined: three surface adsorption-desorption equilibrium reactions for the three chemical species, as well as two irreversible reactions for the CO-CO dimerization into C2, and the CO-C2 coupling into C3. Second, the Transport of Diluted Species’ module was used to solve the mass transport of the three species. The work determined the equilibrium coefficients and rate constants of chemical reactions by sweeping these parameters over a large range (2 orders of magnitude) and fitting to the experimental data. There were five reaction-related parameters that were swept: three equilibrium coefficients, Keq_CO, Keq_C2 and Keq_C3, of adsorption-desorption of the CO, C2 and C3 species, as well as two rate constants, Kf_C2 and Kf_C3, of the C-C coupling and CO-C2 coupling steps. After fitting to the experimental data, the equilibrium coefficient for the surface adsorption-desorption was set to 0.03, the rate constants for the CO-CO dimerization and the CO-C2 coupling was set to 100 and 6. The diffusion constants of CO, C2 (C2H4 assumed) and C3 (n- PrOH) were obtained from the literatures. Finally, the‘Surface Reactions’ defines the catalyst surface where the surface reactions take place.

[00113] In the electric potential simulation, the‘Electric Current’ module in Comsol was used. A cathode at -0.56 V (the potential with the highest experimental C3/C2 selectivity) was connected to the outer surface of the nanocavity, and an anode at 1.23 V was placed 2000 nm away from the nanocavity. The conductivity of the nanocavity (Cu) and the electrolyte (1 M KOH) was taken from literature.

[00114] Density functional theory calculations. In this work, all the DFT calculations were carried out with a periodic slab model using the Vienna ab initio simulation program (VASP). The generalized gradient approximation (GGA) was used with the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional. The projector- augmented wave (PAW) method was utilized to describe the electron-ion interactions, and the cut-off energy for the plane-wave basis set was 450 eV. In order to illustrate the long- range dispersion interactions between the adsorbates and catalysts, this work employed the D3 correction method by Grimme et al. Brillouin zone integration was accomplished using a 3*3x 1 Monkhorst-Pack k-point mesh. All the adsorption geometries were optimized using a force-based conjugate gradient algorithm, while transition states (TSs) were located with a constrained minimisation technique. At all intermediate and transition states, one charged layer of water molecules was added to the surface to take the combined field and solvation effects into account. In the CO dimerization, there is no proton or electron transfer, thus the computational hydrogen electrode was not used in this work.

[00115] In-situ X-ray absorption spectroscopy. In-situ XAS measurements were conducted at 9BM of the Advanced Photon Source (Argonne National Laborator, IL), in partnership with the Canadian Light Source (Saskatoon, SK).

[00116] Grazing incidence wide angle X-ray scattering. GIWAXS measurements were conducted at the Hard X-ray MicroAnalysis (HXMA) beamline of the Canadian Light Source (CLS). An energy of 17.998 keV (l = 0.6888 A) was selected using a Si(1 11) monochromator. Patterns were collected on a SX165 CCD camera (Rayonix) placed at a distance of 155 mm from the sample. A lead beamstop was used to block the direct beam. Images were calibrated using LaB6 and processed via the Nika software package and the GIXSGUI MATLAB plugin.

[00117] Characterization. Powder X-ray diffraction patterns (PXRD) were measured on a Philips X’Pert Pro Super X-ray diffractometer equipped with graphite- monochromatized Cu Ka radiation. X-ray photoelectron spectroscopy (XPS) was carried out with the Thermo Scientific K-Alpha XPS system. An Al Ka source with a 400pm spot size was used for measurements to detect photo-electrons at specific energy ranges to determine the presence of specific elements. Scanning electron microscopy (SEM) was performed with a field emission scanning electron microanalyzer (Zeiss Supra 40). Transmission electron microscopy (TEM) observations were performed on Hitachi H-7650 and JEOL-2010F with an acceleration voltage of 200 kV. Scanning transmission electron microscopy (STEM) was carried out using a JEM-ARM 200F Atomic Resolution Analytical Microscope operating at an accelerating voltage of 200 kV. Elemental mapping was collected by a Gatan GIF Quantum 965 instrument. Electron tomography tilt series were acquired on an aberration corrected cubed FEI Titan operating at 300 kV.

[00118] The following tables are presented and are referred to above:

Table 1. Details for ratio of C3/C2 productivity as a function of cavity open angle.

Table 2. Activation energies (Ea) and enthalpy changes (DH) of CO dimerization and C1&C2 coupling. All the energies are in eV.

Table 3. Faradaic efficiencies of propanol at different applied potentials (V vs RHE) using different samples.

Table 4. Comparison of electroreduction of propanol from C1 feedstocks (CO2 and CO) using copper-based catalysts. Catalysts 1-6 are data in the literatures.

Table 5. Faradaic efficiencies of product distribution in 1M KOH electrolyte at different applied potentials using different nanocatalysts in a flow cell system.

Potential Hydrogen Ethylene Acetate Ethanol Propanol Total

(VvsRHE) (FE,%) (FE, %) (FE, %) (FE, %) (FE, %) (FE, %)

-036 12.2+0.6 149 + 1 20.6+2 19.7+1 67 + 05 -741

-056 252 + 2 300 + 3 98 + 05 128— 08 136 = 05 -914

Solid

-0.76 31.2 + 3 236 + 2 7.1+05 11.5 + 0.6 13.5 = 0.5 -86.9

-096 33.9 + 3 283 + 2 8.5 + 05 10.1 = 0.8 117 + 05 -925

-036 9.2 + 0.5 137 + 1 220 + 2 20.7+1 72 + 0.5 -728

-056 242 + 2 275 + 2 88 + 05 126-08 162 + 1 -893

Cavity I

-0.76 30.2 + 3 292 + 3 7.6 + 05 15.6+1 14.5 = 0.8 -97.1

-096 36.1+3 312 + 2 81+05 14.1=0.8 121 0.6 101.6

-036 78 + 05 120 + 1 240 + 2 217+1 110+08 -765

-046 152 + 1 165 + 1 153 + 1 153-08 185 + 1 -808

-0.56 23.2 + 2 210 + 1 7.8 + 05 12.5 = 0.6 21.0 + 1 -85.5 0,66 25.1+2 241+2 10.2 + 0.5 9, 0 + 0, 5 17.7 + 1 -861

Cavity II _,086 294+3 256 + 2 128+08 100=05 158=08 -936

-0.96 341 + 3 263 + 2 9.0 + 05 151-08 124-06 -96.9

-116 38.3 + 3 250 + 2 71+05 12.0 = 0.5 63 + 05 -887

-1,36 40.5 + 3 272 + 2 9.5 + 06 22.0+1 59 + 05 -105.1

-176 410 + 3 281+2 108 + 05 175+1 57 + 05 -1031

-0.36 14.8 + 0.5 15.9 + 0.6 12.5 + 0.8 11.9 + 0.6 86 + 08 -63.7

-056 24.6 + 1 320 + 3 102 + 0.5 18.2+1 10.6 = 0.6 -956

Fragment

-076 35.2 + 3 375 + 3 6.8 + 05 135 = 0,6 78 + 06 -100.8

-096 391+3 353+2 58 + 05 121=08 43 + 05 -966

[00119] As seen in the results, the“selectivity” of C3 over C2 products does not have to correspond to the production of C3 above 50% of the total products. In the context of the present disclosure, the selectivity of C3 production using the nanoparticle catalysts can be viewed as corresponding to an increase in C3 production compared to the catalyst without the confinement-morphology modified catalyst material. The test results provide some examples of the increase in selectivity that the nanoparticles can facilitate. [00120] The following is a list of references the entire contents of which are hereby incorporated herein by reference. It is also noted that the entire contents of all documents mentioned herein are incorporated herein by reference.

• Lin, S. et al. Covalent organic frameworks comprising cobalt porphyrins for catalytic CO2 reduction in water. Science 349, 1208-1213 (2015).

• Asadi, M. et al. Nanostructured transition metal dichalcogenide electrocatalysts for CO2 reduction in ionic liquid. Science 353, 467-470 (2016).

• Jiang, K. et al. Metal ion cycling of Cu foil for selective C-C coupling in electrochemical CO2 reduction. Nat. Catal. 1 , 111-119 (2018).

• Zhuang, T.-T. et al. Steering post-C-C coupling selectivity enables high efficiency electroreduction of carbon dioxide to multi-carbon alcohols. Nat. Catal. 1 , 421-428 (2018).

• Gao, S. et al. Partially oxidized atomic cobalt layers for carbon dioxide electroreduction to liquid fuel. Nature 529, 68-71 (2016).

• Hahn, C. et al. Engineering Cu surfaces for the electrocatalytic conversion of CO2:

Controlling selectivity toward oxygenates and hydrocarbons. Proc. Natl. Acad. Sci. 114, 5918-5923 (2017).

• Xiao, H., Goddard, W.A., Cheng, T. & Liu, Y. Cu metal embedded in oxidized matrix catalyst to promote CO2 activation and CO dimerization for electrochemical reduction of CO2. Proc. Natl. Acad. Sci. 114, 6685-6688 (2017).

• De Luna, P. et al. Catalyst electro-redeposition controls morphology and oxidation state for selective carbon dioxide reduction. Nat. Catal. 1 , 103-110 (2018).

• Schouten, K.J.P., Calle-Vallejo, F. & Koper, M. A step closer to the electrochemical production of liquid fuels. Angew. Chem. Int. Ed. 53, 10858-10860 (2014).

• Loiudice, A. et al. Tailoring Copper Nanocrystals towards C2 Products in Electrochemical CO2 Reduction. Angew. Chem. Int. Ed. 55, 5789-5792 (2016).

• Ren, D., Wong, N.T., Handoko, A.D., Huang, Y. & Yeo, B.S. Mechanistic Insights into the Enhanced Activity and Stability of Agglomerated Cu Nanocrystals for the Electrochemical Reduction of Carbon Dioxide to n-Propanol. J. Phy. Chem. Lett. 7, 20-24 (2016).

• Kim, D., Kley, C.S., Li, Y. & Yang, P. Copper nanoparticle ensembles for selective electroreduction of CO2 to C2-C3 products. Proc. Natl. Acad. Sci. 114, 10560- 10565 (2017).

• Verma, S., Lu, X., Ma, S., Masel, R.l. & Kenis, P.J. The effect of electrolyte composition on the electroreduction of CO2 to CO on Ag based gas diffusion electrodes. Phys. Chem. Chem. Phys. 18, 7075-7084 (2016). • Jhong, H.R.M. et al. A Nitrogen-Doped Carbon Catalyst for Electrochemical CO2 Conversion to CO with High Selectivity and Current Density. ChemSusChem 10, 1094-1099 (2017).

• Kuhl, K.P., Cave, E.R., Abram, D.N. & Jaramillo, T.F. New insights into the electrochemical reduction of carbon dioxide on metallic copper surfaces. Energy Environ. Sci. 5, 7050-7059 (2012).

• Kortlever, R., Shen, J., Schouten, K.J.P., Calle-Vallejo, F. & Koper, M.T.M.

Catalysts and Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide. J. Phy. Chem. Lett. 6, 4073-4082 (2015).

• Seifitokaldani, A. et al. Hydronium-lnduced Switching between CO2 Electroreduction Pathways. J. Am. Chem. Soc. 140, 3833-3837 (2018).

• Li, C.W., Ciston, J. & Kanan, M.W. Electroreduction of carbon monoxide to liquid fuel on oxide-derived nanocrystalline copper. Nature 508, 504 (2014).

• Verdaguer-Casadevall, A. et al. Probing the active surface sites for CO reduction on oxide-derived copper electrocatalysts. J. Am. Chem. Soc. 137, 9808-9811 (2015).

• Xiao, H., Cheng, T. & Goddard III, W.A. Atomistic mechanisms underlying selectivities in C1 and C2 products from electrochemical reduction of CO on Cu (111). J. Am. Chem. Soc. 139, 130-136 (2016).

• Yang, K.D. et al. Morphology-Directed Selective Production of Ethylene or Ethane from CO2 on a Cu Mesopore Electrode. Angew. Chem. Int. Ed. 56, 796-800 (2017).

• Zheng, G. et al. Interconnected hollow carbon nanospheres for stable lithium metal anodes. Nat. Nanotechnol. 9, 618 (2014).

• Li, Z. et al. A sulfur host based on titanium monoxide@ carbon hollow spheres for advanced lithium-sulfur batteries. Nat. commun. 7, 13065 (2016).

• Wu, H.-L. et al. Formation of pseudomorphic nanocages from CU2O nanocrystals through anion exchange reactions. Science 351 , 1306-1310 (2016).

• Kuo, C.-H. & Huang, M.H. Fabrication of truncated rhombic dodecahedral CU2O nanocages and nanoframes by particle aggregation and acidic etching. J. Am. Chem. Soc. 130, 12815-12820 (2008).

• Mistry, H. et al. Highly selective plasma-activated copper catalysts for carbon dioxide reduction to ethylene. Nat. Commun. 7, 12123 (2016).

• Sen, S., Liu, D. & Palmore, G.T.R. Electrochemical reduction of CO2 at copper nanofoams. ACS Catal. 4, 3091-3095 (2014).

• Favaro, M. et al. Subsurface oxide plays a critical role in CO2 activation by Cu (111) surfaces to form chemisorbed CO2, the first step in reduction of CO2. Proc. Natl. Acad. Sci. 114, 6706-6711 (2017). • Rasul, S. et al. A Highly Selective Copper-Indium Bimetallic Electrocatalyst for the Electrochemical Reduction of Aqueous CO2 to CO. Angew. Chem. Int. Ed. 7, 2146- 2150 (2015).

• Thang Dinh. C et. al CO2 electroreduction to ethylene via hydroxide-mediated copper catalysis at an abrupt interface. Science 360, 783-787 (2018).

• Hori, Y., Takahashi, R., Yoshinami, Y. & Murata, A. Electrochemical reduction of CO at a copper electrode. J. Phy. Chem. S 101 , 7075-7081 (1997).

• Unver, A. & Himmelblau, D. Diffusion Coefficients of CO ₂ , C ₂ H ₄ , C ₃ H ₆ and C ₄ H ₈ in Water from 6 to 65 C. Journal of Chemical & Engineering Data 9, 428-431 (1964).

• Hao, L. & Leaist, D.G. Binary mutual diffusion coefficients of aqueous alcohols.

Methanol to 1-heptanol. Journal of Chemical & Engineering Data 41 , 210-213 (1996).

• Cussler, E.L. Diffusion: mass transfer in fluid systems. (Cambridge university press, 2009).

• Waszczuk, P., Zelenay, P. & Sobkowski, J. Surface interaction of benzoic-acid with a copper electrode. Electrochim. Acta 40, 1717-1721 (1995).

• Unver, A. & Himmelblau, D. Diffusion Coefficients of CO2, C2H4, C3H6 and C4H8 in Water from 6 to 65 C. Journal of Chemical & Engineering Data 9, 428-431 (1964).

• Hao, L. & Leaist, D.G. Binary mutual diffusion coefficients of aqueous alcohols.

Methanol to 1-heptanol. Journal of Chemical & Engineering Data 41 , 210-213 (1996).

• Matula, R.A. Electrical resistivity of copper, gold, palladium, and silver. J. Phys.

Chem. Ref. Data 8, 1147-1298 (1979).

• Gilliam, R., Graydon, J., Kirk, D. & Thorpe, S. A review of specific conductivities of potassium hydroxide solutions for various concentrations and temperatures. Int. J. Hydrogen Energy 32, 359-364 (2007).

• Kresse, G. & Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169-11186 (1996).

• Kresse, G. & Furthmuller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 6, 15-50 (1996).

• Kresse, G. & Hafner, J. Ab-Initio Molecular-Dynamics Simulation of the Liquid- Metal Amorphous-Semiconductor Transition in Germanium. Phys. Rev. B 49, 14251-14269 (1994).

• Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev.

B 47, 558-561 (1993).

• Perdew, J.P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77, 3865-3868 (1996). • Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 1758-1775 (1999).

• Blochl, P.E. Projector augmented-wave method. Phys. Rev. B 50, 17953-17979 (1994).

• Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, 154104 (2010).

• Michaelides, A. et al. Identification of general linear relationships between activation energies and enthalpy changes for dissociation reactions at surfaces. J. Am. Chem. Soc. 125, 3704-3705 (2003).

• Liu, Z.P. & Hu, P. General rules for predicting where a catalytic reaction should occur on metal surfaces: A density functional theory study of C-H and C-0 bond breaking/making on flat, stepped, and kinked metal surfaces. J. Am. Chem. Soc. 125, 1958-1967 (2003).

• Alavi, A., Hu, P.J., Deutsch, T., Silvestrelli, P.L. & Hutter, J. CO oxidation on Pt( 111): An ab initio density functional theory study. Phys. Rev. Lett. 80, 3650- 3653 (1998).

• Montoya, J.H., Shi, C., Chan, K. & Norskov, J.K. Theoretical insights into a CO dimerization mechanism in CO2 electroreduction. J. Phys. Chem. Lett. 6, 2032- 2037 (2015).

• Handoko, A.D. et al. Mechanistic insights into the selective electroreduction of carbon dioxide to ethylene on Cu ₂ 0-derived copper catalysts. J. Phy. Chem. C 120, 20058-20067 (2016).

• Li, C.W. & Kanan, M.W. CO2 reduction at low overpotential on Cu electrodes resulting from the reduction of thick CU2O films. J. Am. Chem. Soc. 134, 7231- 7234 (2012).

• Li, C.W., Ciston, J. & Kanan, M.W. Electroreduction of carbon monoxide to liquid fuel on oxide-derived nanocrystalline copper. Nature 508, 504 (2014).

• Ren, D., Wong, N.T., Handoko, A.D., Huang, Y. & Yeo, B.S. Mechanistic insights into the enhanced activity and stability of agglomerated Cu nanocrystals for the electrochemical reduction of carbon dioxide to n-propanol. J. Phys. Chem. Lett. 7, 20-24 (2015).

• Rahaman, M., Dutta, A., Zanetti, A. & Broekmann, P. Electrochemical Reduction of CO2 into Multicarbon Alcohols on Activated Cu Mesh Catalysts: An Identical Location (IL) Study. ACS Catal. 7, 7946-7956 (2017).

• Jiang, K. et al. Metal ion cycling of Cu foil for selective C-C coupling in electrochemical CO2 reduction. Nat. Catal. 1 , 111-119 (2018).