Molten sulfide electrolysis

Abstract

Methods comprising molten sulfide electrolysis are generally described. In some cases, the method comprises performing electrolysis on a molten sulfide electrolyte in an electrochemical cell. In some instances, the molten sulfide electrolyte comprises greater than 0.5 wt% oxygen. In some cases, the molten sulfide electrolyte comprises a metal M, wherein the method comprises removing a gas comprising sulfur formed at an anode of the electrochemical cell and producing metallic M at a cathode of the electrochemical cell. Described herein is an electrolysis cell, with its anode, cathode, refractory and electrolyte components, to support the removal of sulfur via the anodic reaction and the production of metal at the cathode.

Description

[0001] MOLTEN SULFIDE ELECTROLYSIS

[0002] RELATED APPLICATIONS

[0003] This application claims priority under 35 U. S. C. § 119(e) to U.S. Provisional Application No. 63 / 779,516, filed March 28, 2025, and entitled “Molten Sulfide Electrolysis,” which is hereby incorporated herein by reference in its entirety for all purposes.

[0004] TECHNICAL FIELD

[0005] Methods comprising molten sulfide electrolysis are generally described.

[0006] SUMMARY

[0007] Methods comprising molten sulfide electrolysis are generally described. In some embodiments, the method comprises performing electrolysis on a molten sulfide electrolyte in an electrochemical cell. In certain embodiments, the molten sulfide electrolyte comprises oxygen, such as greater than 0.5 wt%, greater than or equal to 1 wt%, less than or equal to 12 wt%, and / or less than or equal to 8 wt%. According to some embodiments, the method comprises removing a gas comprising sulfur, such as S2gas, formed at an anode of the electrochemical cell. In some cases, the anode comprises carbon, such as graphite. In certain instances, the molten sulfide electrolyte comprises metal M, and the method comprises producing metallic M at a cathode of the electrochemical cell. In some instances, the metallic M produced is a liquid. According to some embodiments, the method comprises performing electrolysis continuously for a period of time by replenishing the molten sulfide electrolyte. According to certain embodiments, the method comprises forming the molten sulfide electrolyte by sulfidation of two or more compounds. The subject matter of the present disclosure involves, in some cases, interrelated products, alternative solutions to a particular problem, and / or a plurality of different uses of one or more systems and / or articles. This Summary introduces a selection of concepts in simplified form that are described further below in the Detailed Description. This Summary neither identifies key or essential features, nor limits the scope, of the claimed subject matter.

[0008] Certain embodiments relate to methods. In some embodiments, the method comprises a. performing electrolysis on a molten sulfide electrolyte in an electrochemical cell, wherein the molten sulfide electrolyte comprises greater than 0.5 wt% oxygen and the molten sulfide electrolyte comprises M, wherein M is a metal; b. removing a gas comprising

[0009] 1

[0010] #15087612vlsulfur formed at an anode of the electrochemical cell; and c. producing metallic M at a cathode of the electrochemical cell.

[0011] In some embodiments, the method comprises a. forming a molten sulfide electrolyte, wherein forming the molten sulfide electrolyte comprises sulfidation of MaXband at least one other compound, wherein the at least one other compound comprises NcYd, wherein M and N are each independently a metal, X and Y are each independently an electronegative species, and a, b, c, and d are each independently a number, and wherein the molten sulfide electrolyte comprises greater than 0.5 wt% oxygen; b. performing electrolysis on the molten sulfide electrolyte in an electrochemical cell; c. removing a gas comprising sulfur formed at an anode of the electrochemical cell; and d. producing metallic M at a cathode of the electrochemical cell.

[0012] One aspect of the disclosure herein is a method comprising:

[0013] a. forming a molten sulfide electrolyte by sulfidation of BaSO4, La2O3, La2(SO4)3, and FeS;

[0014] b. removing sulfur (S) via electrolysis in an anodic reaction; and

[0015] c. producing Fe metal at the cathode.

[0016] One aspect of the disclosure herein is an electrolysis cell comprising an anode, cathode, refractory and electrolyte components, and a molten sulfide electrolyte comprising BaSO4, La2O3, La2(SO4)3, and FeS.

[0017] In one embodiment, the disclosed electrolysis cells supports the removal of sulfur via an anodic reaction and the production of metal at the cathode.

[0018] Other advantages and novel features of the present disclosure will become apparent from the following detailed description of various non-limiting embodiments of the disclosure when considered in conjunction with the accompanying figures. In cases where the present specification and a document incorporated by reference include conflicting and / or inconsistent disclosure, the present specification shall control. The following Detailed Description references the accompanying drawings which form a part this application, and which show, by way of illustration, specific example implementations. Other implementations may be made without departing from the scope of the disclosure.

[0019] BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Non-limiting embodiments of the present disclosure will be described by way of example with reference to the accompanying figures, which are schematic and are not

[0021] 2

[0022] #15087612vlintended to be drawn to scale unless otherwise indicated. In the figures, each identical or nearly identical component illustrated is typically represented by a single numeral. For purposes of clarity, not every component is labeled in every figure, nor is every component of each embodiment of the disclosure shown where illustration is not necessary to allow those of ordinary skill in the art to understand the disclosure. In the figures:

[0023] FIG. 1 is an E-pO2−diagram for the Ba-La-Fe-S system at 1573 K (1300 C).

[0024] FIG. 2 is a reference range of stability of O-S-Fe-Ba-La-S.

[0025] FIG. 3 are results of synthesis of O-S-Fe-Ba-La-S.

[0026] FIG. 4 shows that between 0.01 ppm and 8% of oxygen could be tolerated (weight %) as a result of using Ba-La-Fe-S.

[0027] FIG. 5 shows results of successful stability of the anode.

[0028] FIG. 6 shows a cross-section of the support (graphite), electrolyte (porous grey) showing Fe production at the graphite / electrolyte interface.

[0029] FIG. 7 shows very efficient sulfur removal on the carbon anode despite the presence of oxygen.

[0030] FIGs. 8A-8B show a route to liquid cast iron from iron oxide utilizing sulfidation of iron ore to FeS (FIG. 8A) and electrolysis of FeS contained in a molten sulfide electrolyte (FIG. 8B), in accordance with some embodiments.

[0031] FIG. 9 shows a conceptual diagram of the sulfidation process for BaSO4in the presence of FeS.

[0032] FIGs. 10A-10C show chemical potential diagrams for Ba-O-S (FIG. 10A), Ba-Fe-O-S (FIG. 10B), and La-Fe-O-S systems at 1473 K (FIG. 10C).

[0033] FIG. 11 shows a schematic diagram for a carbosulfidation method, in accordance with some embodiments.

[0034] FIGs. 12A-12C show example photographs of samples in a graphite tray before (FIG. 12A) and after (FIG. 12B) sulfidation, with FIG. 12C showing a side view of the samples.

[0035] FIG. 13 shows X-ray diffraction (XRD) spectra of the BaSO4–FeS samples before and after sulfidation.

[0036] FIG. 14A shows a cross-sectional optical micrograph (OM) (with polarized light) after sulfidation. FIG. 14B, FIG. 14C, FIG. 14D, and FIG. 14E show Scanning Electron Microscopy (SEM) images of the BaSO4–FeS sample after sulfidation (Sample no. 04_1).

[0037] 3

[0038] #15087612vlFIGs. 15A-15B provide Wavelength Dispersive Spectroscopy (WDS) analysis results for the sulfidized BaSO4–FeS samples plotted on the Ba–Fe–O–S compositional tetrahedral diagram (FIG. 15A) and the Ba–Fe–(S+O) compositional triangular diagram (FIG. 15B).

[0039] FIG. 16 shows a cross-sectional SEM image of the BaSO4–FeS sample after the shortest sulfidation duration (Sample no. 07_1).

[0040] FIG. 17A and FIG. 17B show XRD spectra of the BaSO4-La2O3-FeS samples before and after sulfidation.

[0041] FIG. 18A, FIG. 18B, FIG. 18C, and FIG. 18D show cross-sectional SEM images of the sulfidized BaSO4-La2O3-FeS samples. FIGs. 18A-18B show images for Sample no. 06_2; FIGs. 18C-18D show images for Sample no. 06_3.

[0042] FIGs. 19A-19B show EDS analysis results for the sulfidized BaSO4-La2O3-FeS samples. FIG. 19A shows 4.5 wt% FeS addition. FIG. 19B shows 8.6 wt% FeS addition.

[0043] FIG. 20A shows a cross-sectional SEM image of the sulfidized BaSO4-La2O3-FeS sample. FIG. 20B, FIG. 20C, and FIG. 20D show EDS elemental mappings of the sulfidized BaSO4-La2O3-FeS samples (Sample, no. 08_l). FIG. 20E shows an enlarged SEM image.

[0044] FIG. 20F shows an La elemental mapping. FIG. 20G shows an Optical Microscopy (OM) image (with polarized light) of the same sample.

[0045] FIG. 21A and FIG. 21B show XRD spectra of the BaSO4-La2(SO4)3-FeS samples before and after sulfidation.

[0046] FIGs. 22A-22D show cross-sectional SEM images of the sulfidized BaSO4-La2(SO4)3-FeS samples. FIG. 22A and FIG. 22B show images for Sample no. 08_2. FIG.

[0047] 22C and FIG. 22D show images for Sample no. 08_3.

[0048] FIG. 23 shows EDS analysis results for the sulfidized BaSO4-La2(SO4)3-FeS samples.

[0049] FIG. 24 shows a process flow diagram of an ironmaking route utilizing molten sulfide electrolysis, integrated with an electrolyte production pathway, in accordance with some embodiments.

[0050] FIG. 25 shows a process flow chart depicting the mass balance to produce 1 metric ton molten iron, in accordance with some embodiments.

[0051] FIG. 26 shows a plot depicting the variation of enthalpy and Gibbs energy change with the temperature for sulfidising iron oxide to FeS to electrolytic ally produce one ton of Fe, in accordance with some embodiments.

[0052] 4

[0053] #15087612vlFIG. 27 shows a plot depicting the variation of enthalpy change and Gibb’s free energy change with the temperature for sulfidising iron oxide to FeS in kWh / ton of Fe produced, in accordance with some embodiments.

[0054] FIG. 28 shows a plot depicting the variation of enthalpy change and Gibbs energy change with the temperature for the electrolysis reaction, in kWh / ton of molten iron produced, in accordance with some embodiments.

[0055] FIG. 29 shows a bar chart depicting the energy requirements for steel making via Blast Furnace-Basic Oxygen Furnace (BF-BOF), molten oxide electrolysis (MOE) and molten sulfide electrolysis (MSE) processes, in accordance with some embodiments.

[0056] FIG. 30 depicts a schematic of the vertical tube furnace used for sulfidation experiments.

[0057] FIGs.31A-31B show photos of graphite electrodes used for molten sulfide electrolysis (MSE) experiments: (FIG. 31A) Cathode (FIG. 31B) Anode.

[0058] FIG. 32 is a schematic of the 2-electrode experimental set up used in the thermal imaging furnace (TIF) used for MSE experiments.

[0059] FIG. 33 shows an in-situ demonstration of the molten sulfide electrolysis experiment.

[0060] FIG. 34 shows an x-ray diffraction (XRD) scan of the product upon sulfidation of iron (III) oxide.

[0061] FIG. 35A and FIG. 35B are optical micrographs of the cross-section of the electrolyte sample before electrolysis (pre-melt) depicting different colors for the 2 main phases Ba rich and La rich identified.

[0062] FIG. 36 shows scanning electron microscope (SEM) micrographs of the cross-section of the electrolyte sample after pre-melt.

[0063] FIG. 37A shows an SEM micrograph and FIG. 37B shows an element intensity overlay of the part of the pre-melt cross-section depicting the 2 phases: Ba rich and La rich.

[0064] FIG. 38 shows evidence of gas bubbles observed during electrolysis, plausibly indicating S2 evolution.

[0065] FIG. 39 shows galvanostatic measurements carried out measuring variation of Voltage v / s V ref with time at different current values.

[0066] FIG. 40 shows variation of Voltage (V) v / s V ref with time - Chronopotentiometry curve for 0.15 A.

[0067] FIG. 41 shows variation of Voltage (V) v / s V ref with time - Chronopotentiometry curves for 0.3 A current.

[0068] 5

[0069] #15087612vlFIGs. 42A-42B show variation of impedance with (FIG. 42A) distance between the electrodes (FIG. 42B) and applied potential (in volts).

[0070] FIG. 43 shows a plot comparing the experimentally obtained resistance values and those calculated using COMSOL.

[0071] FIG. 44 shows a photo of the top part of the quartz tube showing a yellowish tinge, serving as evidence of S2 evolution (also confirmed by typical sulfur gas odor), presumably some of which condensed on the tube surface.

[0072] FIG. 45 shows a chart depicting effect of time and current on the faradaic efficiency calculated for the electrolysis experiments.

[0073] FIG. 46 shows a chart depicting effect of current density on the faradaic efficiency.

[0074] FIGs. 47A-47C show optical micrographs depicting cross-section of the electrolyte on a graphite cathode showing Fe deposition after electrolysis at (FIG. 47A) 2.5X (FIG. 47B) 10X (FIG. 47C) 25X.

[0075] FIG. 48 shows a WDS-SEM (Wavelength Dispersive Spectroscopy in a Scanning Electron Microscope) micrograph II of the electrolyte sample after electrolysis distinctly showing the light and dark phases and the metallic Fe deposits.

[0076] FIG. 49 shows line scans performed during EDS (Energy Dispersive X-ray Spectroscopy) analysis of the electrolyte sample through a Fe deposit.

[0077] FIG. 50 shows solidification of the electrolyte as observed after 120 secs of passing 0.5 A.

[0078] FIGs. 51A-51C show an (FIG. 51A) optical micrograph; (FIG. 51B) SEM micrograph; and (FIG. 51C) element intensity map of the electrolyte sample after electrolysis distinctly showing the metallic Fe deposits.

[0079] FIG. 52A and FIG. 52B show optical micrographs of Fe deposits, after etching the sample with 4% Nital solution metallic deposits at 50 X.

[0080] FIGs. 53A-53D show SEM micrographs of Fe deposits, after etching the sample with 4% Nital solution, showing carbide inclusions (FIG. 53A) deposit -1 at 3000 X (FIG. 53B) deposit -2 at 6000 X (FIG. 53C) deposit-3 at 2200 X (FIG. 53D) deposit-4 at 2000 X.

[0081] FIG. 54 shows a flowsheet of a proposed process for producing steel using MSE integrated with present steel making infrastructure, in accordance with some embodiments.

[0082] FIG. 55 shows an electrolyte cross-section (SEM) after a thermal decomposition test, showing a phase segregation.

[0083] 6

[0084] #15087612vlFIG. 56 shows an optical micrograph of electrolyte cross-section (5 X) after a thermal decomposition test, showing a phase segregation.

[0085] FIG. 57 shows results of the thermal decomposition test carried out recording mass loss versus time.

[0086] FIGs. 58A-58D show a cross section of the electrolyte after electrolysis, showing a phase segregation as well as Fe deposition (FIG. 58A) SEM image (FIG. 58B) Fe (FIG. 58C) Ba (FIG. 58D) La.

[0087] FIG. 59 shows a schematic of an assembly for electrolysis, in accordance with some embodiments, with the molten electrolyte droplet supported by the graphite cathode stand, with the graphite anode tip in contact from above, and with the electrical connection provided by sheathed molybdenum rods threaded to the anode and cathode.

[0088] FIG. 60 shows a plot comparing the obtained resistance values and those calculated using the finite element model of the primary distribution (COMSOL).

[0089] FIG. 61A shows an optical micrograph of an electrolyte cross-section at the graphite cathode interphase depicting iron deposits after electrolysis, with a rectangle showing the frame for SEM-BEC and SEM-EDS imaging. FIG. 61B shows a SEM-BEC micrograph of iron deposition in electrolyte. FIG. 61C shows a SEM-EDS element intensity map of iron deposition in electrolyte.

[0090] FIG. 62 shows SEM micrographs of Fe deposits, after etching the sample with 4% Nital solution, showing carbide inclusions.

[0091] FIG. 63 shows a process flow chart depicting the stoichiometric mass balance to produce one tonne molten iron (Fe-4.3wt%C), in accordance with some embodiments. The heat produced (sulfidation), and electricity needed (MSE) and produced (sulfuric acid production) per tonne cast iron are shown on the vertical axis.

[0092] FIG. 64 shows a reported range (high and low) for energy consumption of various iron and steel making technologies: BF-BOF energy consumption; MOE energy consumption; H2-DRI and electric arc furnace (EAF) energy consumption; Alkaline electrowinning (EW) and electric arc furnace (EAF) energy consumption. MSE is shown for the case of molten iron production at 1300°C.

[0093] FIG. 65A, FIG. 65B, and FIG. 65C show images of electrolytes formed in accordance with some embodiments. FIG. 65D and FIG 65E show additional analysis thereof.

[0094] 7

[0095] #15087612vlFIG. 66A and FIG. 66B show the experimental setup used in FIGs. 65A-65E with a set temperature of 1410°C and the internal temperature reached 1350 °C.

[0096] FIG. 67 shows a Ternary Diagram with La2S3, BaS, and FeS.

[0097] FIG. 68A, FIG. 68B, and FIG. 68C show WDS results, analyzing a liquid sulfide electrolyte and the amount of oxygen therein. Samples were fully molten starting at 12 Wt% FeS. The samples were prepared at 1300 °C with 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3. FIG. 68C shows the mass percentage of various elements in various phases. The samples had high oxygen content as oxygen entered the system during testing and was present in the powdered samples.

[0098] FIG. 69A, FIG. 69B, and FIG. 69C show SEM / EDS results, analyzing a liquid sulfide electrolyte and the amount of oxygen therein. Samples were fully molten starting at 12 Wt% FeS. The samples were prepared at 1300 °C with 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3. FIG. 69C shows the mass percentage of various elements in various phases. The samples had low oxygen content, as the furnace system was completely leak tested and the chemicals were newly purchased.

[0099] FIG. 70 shows a molten sulfide electrolysis schematic, in accordance with some embodiments.

[0100] FIG. 71 shows a schematic of sulfidation reaction in a rotary kiln, in accordance with some embodiments.

[0101] FIG. 72 shows a schematic of selective sulfidation of iron oxide ores, in accordance with some embodiments.

[0102] FIG. 73 shows a schematic of an SM furnace, in accordance with some embodiments.

[0103] FIG. 74 is an image showing gas evolution as current was passed when using molten sulfide electrolysis for iron production.

[0104] FIG. 75 shows characterization studies after etch results.

[0105] FIG. 76 shows a Sankey diagram, showing raw materials and final products, in accordance with some embodiments.

[0106] DETAILED DESCRIPTION

[0107] Methods comprising molten sulfide electrolysis are generally described. In some embodiments, the method comprises performing electrolysis on a molten sulfide electrolyte in an electrochemical cell. In certain embodiments, the molten sulfide electrolyte comprises oxygen, such as greater than 0.5 wt%, greater than or equal to 1 wt%, less than or equal to 12

[0108] 8

[0109] #15087612vlwt%, and / or less than or equal to 8 wt%. According to some embodiments, the method comprises removing a gas comprising sulfur, such as S2gas, formed at an anode of the electrochemical cell. In some cases, the anode comprises carbon, such as graphite. In certain instances, the molten sulfide electrolyte comprises metal M, and the method comprises producing metallic M at a cathode of the electrochemical cell. In some instances, the metallic M produced is a liquid. According to some embodiments, the method comprises performing electrolysis continuously for a period of time by replenishing the molten sulfide electrolyte. According to certain embodiments, the method comprises forming the molten sulfide electrolyte by sulfidation of two or more compounds.

[0110] Certain embodiments relate to methods. In some embodiments, the method comprises performing electrolysis on a molten sulfide electrolyte in an electrochemical cell.

[0111] In some cases, the method and / or the electrolysis is tolerant for oxygen. For example, in certain instances, the method and / or the electrolysis is not significantly affected (e.g., changed by less than 10%, less than 5%, less than 3%, less than 1%, less than 0.1%, or is not changed, in one or more, or all, ways) by oxygen (e.g., in amounts disclosed herein) in the molten sulfide electrolyte, the feedstock, and / or the gas-handling system. For example, in some instances, the degradation of electrochemical cell components (e.g., one or more anode components) and / or the energetic efficiency of the electrolysis is not significantly affected (e.g., changed by less than 10%, less than 5%, less than 3%, less than 1%, less than 0.1%, or is not changed) by oxygen (e.g., in amounts disclosed herein) in the molten sulfide electrolyte, the feedstock, and / or the gas-handling system. Without wishing to be bound by theory, it is believed that the method and / or the electrolysis being tolerant for oxygen provides numerous advantages, in certain embodiments, such as compatibility with more container materials and / or electrode materials, allowing for the use of more abundant, less expensive, and / or a larger variety of container materials and / or electrode materials, in some cases, and / or making it easier to perform the method and / or electrolysis in ambient environments.

[0112] In certain embodiments, the molten sulfide electrolyte comprises oxygen. For example, in some cases, the molten sulfide electrolyte comprises greater than 0.5 wt%, greater than or equal to 0.55 wt%, greater than or equal to 0.6 wt%, greater than or equal to 0.65 wt%, greater than or equal to 0.7 wt%, greater than or equal to 0.75 wt%, greater than or equal to 0.8 wt%, greater than or equal to 0.85 wt%, greater than or equal to 0.9 wt%, greater than or equal to 0.95 wt%, greater than or equal to 1 wt%, greater than or equal to 1.1 wt%,

[0113] 9

[0114] #15087612vlgreater than or equal to 1.2 wt%, greater than or equal to 1.3 wt%, greater than or equal to 1.4 wt%, greater than or equal to 1.5 wt%, greater than or equal to 1.75 wt%, greater than or equal to 2 wt%, greater than or equal to 2.25 wt%, greater than or equal to 2.5 wt%, greater than or equal to 2.75 wt%, greater than or equal to 3 wt%, greater than or equal to 3.25 wt%, greater than or equal to 3.5 wt%, greater than or equal to 3.75 wt%, greater than or equal to 4 wt%, greater than or equal to 4.5 wt%, greater than or equal to 5 wt%, greater than or equal to 5.5 wt%, greater than or equal to 6 wt%, greater than or equal to 6.5 wt%, greater than or equal to 7 wt%, or greater than or equal to 7.5 wt% oxygen. In accordance with some embodiments, the molten sulfide electrolyte comprises less than or equal to 12 wt%, less than or equal to 11.75 wt%, less than or equal to 11.5 wt%, less than or equal to 11.25 wt%, less than or equal to 11 wt%, less than or equal to 10.5 wt%, less than or equal to 10 wt%, less than or equal to 9.5 wt%, less than or equal to 9 wt%, less than or equal to 8.5 wt%, less than or equal to 8 wt%, less than or equal to 7.5 wt%, less than or equal to 7 wt%, less than or equal to 6.5 wt%, less than or equal to 6 wt%, less than or equal to 5.5 wt%, less than or equal to 5 wt%, less than or equal to 4.5 wt%, less than or equal to 4 wt%, less than or equal to 3.5 wt%, or less than or equal to 3 wt% oxygen. Combinations of these ranges are also possible (e.g., greater than 0.5 wt% and less than or equal to 12 wt%, greater than 0.5 wt% and less than or equal to 8 wt%, greater than or equal to 1 wt% and less than or equal to 12 wt%, or greater than or equal to 1 wt% and less than or equal to 8 wt%).

[0115] In accordance with some embodiments, the molten sulfide electrolyte comprises M. In some cases, M is a metal. In certain embodiments, M comprises aluminum, silicon, iron, titanium, chromium, manganese, vanadium, niobium, beryllium, zirconium, hafnium, cobalt, nickel, copper, zinc, gallium, molybdenum, silver, cadmium, indium, tin, tantalum, tungsten, rhenium, lead, and / or a rare earth element, such as lanthanum, cerium, neodymium, praesodymium, terbium, and / or dysprosium. For example, in some instances, M comprises iron, cobalt, nickel, copper, zinc, molybdenum, silver, tantalum, tungsten, and / or a rare earth element (e.g., neodymium, praesodymium, terbium and / or dysprosium).

[0116] In accordance with certain embodiments, the method comprises forming the molten sulfide electrolyte. In some instances, forming the molten sulfide electrolyte comprises sulfidation of MaXband at least one other compound, wherein the at least one other compound comprises NcYd.

[0117] According to some embodiments, M is a metal. For example, in certain cases, M comprises aluminum, silicon, iron, titanium, chromium, manganese, vanadium, niobium,

[0118] 10

[0119] #15087612vlberyllium, zirconium, hafnium, cobalt, nickel, copper, zinc, gallium, molybdenum, silver, cadmium, indium, tin, tantalum, tungsten, rhenium, lead, and / or a rare earth element, such as lanthanum, cerium, neodymium, praesodymium, terbium, and / or dysprosium. For example, in some instances, M comprises iron, cobalt, nickel, copper, zinc, molybdenum, silver, tantalum, tungsten, and / or a rare earth element (e.g., neodymium, praesodymium, terbium and / or dysprosium).

[0120] In some embodiments, N is a metal. In certain embodiments, N comprises an alkali metal, such as sodium and / or potassium; an alkaline earth metal, such as magnesium, calcium, strontium, and / or barium; and / or a rare earth metal, such as lanthanum and / or cerium. For example, in some cases, N comprises barium and / or lanthanum.

[0121] According to some embodiments, M and N are each independently a metal. In some cases, M and N comprise the same metal. In certain instances, M and N comprise a different metal.

[0122] In some instances, X is an electronegative species. For example, in certain embodiments, X comprises a sulfide anion (S2-), a selenide anion (Se2-), and / or an arsenide anion (e.g., As4-or As5-). For example, in some cases, X comprises a sulfide anion (S2-).

[0123] In some embodiments, Y is an electronegative species. According to certain embodiments, Y comprises a sulfate anion (SO42-), a phosphate anion (PO43-), and / or an oxide anion (O2-).

[0124] In accordance with some embodiments, X and Y are each independently an electronegative species. In some cases, X and Y comprise the same electronegative species. In certain instances, X and Y comprise a different electronegative species.

[0125] In accordance with certain embodiments, a, b, c, and d are each independently a number. For example, in some embodiments, a, b, c, and / or d are each independently a whole number. In certain cases, a, b, c, and d are each independently a number greater than or equal to 1 and less than or equal to 9. In some cases, a, b, c, and d are each independently selected from 1, 2, 3, 4, 5, 6, 7, 8, or 9. For example, in accordance with some embodiments, a is 1, 2, 3, or 9. According to certain embodiments, b is 1, 2, 3, 4, or 8. In certain embodiments, c is 1, 2, or 3. In accordance with certain embodiments, d is 1, 2, or 3.

[0126] According to certain embodiments, MaXbis Al2S3, SiS2, FeS, CoS, CoS2, NiS, NiS2, Ni3S2, Ni2S4, Ni9S8, Cu2S, CuS, ZnS, GaS, Ga2S3, MoS2, Ag2S, CdS, In2S3, SnS, TaS2, WS2, ReS2, PbS, Ce2S3, Nd2S3, Pr2S3, Tb2S3, Dy2S3, Ti2S3, TiS2, TiS, MnS, Mn2S3, Mn3S4, Cr2S3, CrS, Cr3S4, V2S5, V2S3, VS4, VS2, Nb2S5, Nb2S3, NbS2, NbS4, NbS, ZrS2, HfS2, ZrS, and / or

[0127] 11

[0128] #15087612vlHfS. For example, according to some embodiments, MaXbis FeS, CoS, CoS2, NiS, NiS2, Ni3S2, Ni2S4, Ni9S8, Cu2S, CuS, ZnS, MoS2, Ag2S, TaS2, WS2, Nd2S3, Pr2S3, Tb2S3, and / or Dy2S3.

[0129] In some embodiments, NcYdis Na2SO4, Na2O, MgSO4, MgO, K2SO4, K2O, CaSO4, CaO, SrSO4, SrO, BaSO4, La2O3, La2(SO4)3, Ce2O3, and / or Ce2(SO4)3. For example, in certain cases, NcYdis BaSO4, La2O3, and / or La2(SO4)3.

[0130] According to some embodiments, forming the molten sulfide electrolyte comprises sulfidation of MaXb, NcYd, and an additional compound. In certain instances, the additional compound is N’c’Y’d’, wherein N’ is selected from any option for N disclosed herein, c’ is selected from any option for c disclosed herein, Y’ is selected from any option for Y disclosed herein, and d’ is selected from any option for d disclosed herein. In some instances, N’c’Y’d’ is different than NcYd. For example, in certain embodiments, N’ is different than N, c’ is different than c, Y’ is different than Y, and / or d’ is different than d.

[0131] Certain aspects comprise performing various methods described herein in an electrochemical cell. As would be appreciated by those of ordinary skill in the art, electrochemical cells can be used to perform electrochemical reactions. One example of an electrochemical cell is shown, for example, in FIG. 8B. Other examples of electrochemical cells are show, for example, in FIGs. 33, 59, and 70. In certain embodiments, the electrochemical cell can be used to perform electrolysis (e.g., by being used as an electrolytic cell, also referred to as an electrolysis cell).

[0132] In certain embodiments, the method comprises removing a gas comprising sulfur formed at an anode of the electrochemical cell. In some instances, the gas comprising sulfur comprises S2, S3, S4, S5, S6, S7, and / or S8. In certain cases, the gas comprising sulfur is S2 gas.

[0133] In accordance with some embodiments, the method comprises producing metallic M at a cathode of the electrochemical cell. As used herein, a “metallic” metal is a metal having an oxidation state of zero. For example, pure iron metal in a zero oxidation state would be metallic iron. Iron that is part of iron sulfide salt, however, would not be metallic iron because iron in that form has an oxidation state of +2.

[0134] According to certain embodiments, at least a portion (e.g., at least 10 mol%, at least 25 mol%, at least 50 mol%, at least 75 mol%, at least 90 mol%, or all) of the metallic M is liquid.

[0135] 12

[0136] #15087612vlIn certain embodiments, the method comprises performing electrolysis continuously for a period of time. For example, in some cases, the method comprises performing electrolysis continuously for a period of time by replenishing the molten sulfide electrolyte. In certain instances, the period of time is greater than or equal to 10 minutes, greater than or equal to 20 minutes, greater than or equal to 30 minutes, greater than or equal to 45 minutes, greater than or equal to 1 hour, greater than or equal to 1.5 hours, greater than or equal to 2 hours, greater than or equal to 3 hours, greater than or equal to 4 hours, or greater than or equal to 5 hours. In some instances, the period of time is less than or equal to 12 hours, less than or equal to 11.5 hours, less than or equal to 11 hours, less than or equal to 10.5 hours, less than or equal to 10 hours, less than or equal to 9 hours, less than or equal to 8 hours, less than or equal to 7 hours, less than or equal to 6 hours, less than or equal to 5 hours, less than or equal to 4 hours, or less than or equal to 3 hours. Combinations of these ranges are also possible (e.g., greater than or equal to 10 minutes and less than or equal to 12 hours or greater than or equal to 1 hour and less than or equal to 12 hours).

[0137] In some embodiments, the anode comprises carbon. For example, in certain cases, the anode comprises carbon with a zero oxidation state. For example, in some cases, the anode comprises graphite. In certain instances, the anode comprises steel, such as stainless steel.

[0138] In some embodiments, the cathode comprises carbon, such as carbon with a zero oxidation state. For example, in some cases, the cathode comprises graphite. In certain instances, the cathode comprises steel, such as stainless steel.

[0139] Additional details of certain embodiments can be found in U.S. Provisional Application No. 63 / 779,516, filed March 28, 2025, and entitled “Molten Sulfide Electrolysis;” KAMIMURA et al., Carbosulfidation of Barium Sulfate Assisted by Iron Sulfide. Metall Mater Trans B. 2026 Jan 29;57: 1995-2010. doi: 10.1007 / s11663-025-03930-z; SURYARAO, Iron Production By Molten Sulfide Electrolysis. Masters Thesis.

[0140] Massachusetts Institute of Technology. 2024 June: 171 pages; United States Patent Application Publication No. US 2019 / 0127221 Al, and entitled “Sulfides electrolyte for metal processing and extraction;” which are hereby incorporated herein by reference in their entireties for all purposes.

[0141] In some embodiments, a method produces metal from metal feedstocks (e.g., comprising sulfides and / or oxides) by electrolysis. In certain embodiments, the method comprises using a molten electrolyte, a positive electrode (anode) where gas is evolved, and a

[0142] 13

[0143] #15087612vlnegative electrode where metal is produced. The disclosed method, in some cases, also identifies how to prepare the components of the electrolyte or the feedstock.

[0144] According to certain embodiments, the method allows control of electrical phenomena in the context of metal, sulfur and oxygen mixture, such that sulfur can be removed as an anodic molecules (in various forms) while metal is won at the cathode, in particular in a liquid state.

[0145] According to some embodiments, one purpose of the disclosed method is to sustain electrolysis in a continuous manner, with minimal degradation of the cell components, while more feedstocks are added and more metal is produced continuously.

[0146] In accordance with some embodiments, the disclosure herein solves the issue of how to process sulfides and sulfur-containing feedstock using electricity and electrolysis phenomena, without extensive consumption of the electrolysis cell component, particularly at the anode.

[0147] The disclosure herein shows the range of oxygen that can be tolerated in the electrolyte, in some embodiments, while using a certain anode material such that the material is inert.

[0148] In some instances, electrolysis in molten sulfide is more energetically efficient.

[0149] Therefore, electrolysis supports the production of liquid metal at temperatures that are lower than other molten oxides electrolysis, while supporting the use of readily available container and electrode materials, in accordance with some embodiments. As such, the process supports lower energy consumption as well as lower cost for the electrolysis cell and its components, in certain cases. Finally, the process is compatible with a given range of oxygen input, in accordance with certain embodiments.

[0150] According to some embodiments, the method comprises producing iron. In some cases, molten sulfide electrolysis for metal production, such as iron production, is advantageous for one or more of the following reasons:

[0151] • Thermodynamic feasibility

[0152] o More favorable thermodynamically (e.g., AG FeS - Fe < AG Fe2O3 - Fe) o Energy requirements < molten oxide electrolysis (MOE)

[0153] o Lower energy consumption compared to BF-BOF (e.g., 40% reduction) • Aspect of Sustainability

[0154] o No use of coal as a reductant

[0155] ■ No direct C emissions

[0156] 14

[0157] #15087612vlo Electrifying metals production - use of green electricity

[0158] ■ As affordable, clean energy increases

[0159] o Scrap integration: excess heat can be used to melt scrap

[0160] o Possibility to use low grade ores, such as FeS

[0161] ■ Selective sulfidation of ores

[0162] ■ Excess heat generated in the cell used to melt gangue • Electrochemical aspect and products

[0163] o No multiple valency (only Fe+2) in FeS

[0164] ■ High faradaic efficiency

[0165] ■ No partial reduction as in oxide electrolysis

[0166] o Elemental sulfur regenerated at the anode

[0167] ■ Availability of a cheap, inert anode: Graphite

[0168] ■ Can be used in the sulfidation step

[0169] o Production of molten iron at cathode

[0170] ■ Steel making in the BOF (unsurpassed productivity and refining standards)

[0171] In certain embodiments, molten sulfide electrolysis is a promising alternative for sustainable iron production because it has low energy consumption; no direct carbon emissions - use of coke eliminated; it integrates well with present BOF technology; faradaic efficiency (anodic) > 90% recorded consistently; scale up of electrolyte, electrode materials is relatively easy; and / or metal composition close to hot metal (BOF feedstock) composition.

[0172] In some embodiments, the method comprises producing steel (see FIG. 76). In some cases, the steel is an alloy comprising iron and carbon. In some embodiments, the alloy contains less than 2 wt% carbon. In some embodiments, the alloy contains less than 2 at% (atomic %) carbon.

[0173] According to some embodiments, the method comprises using sulfur (e.g., a compound comprising sulfur). Non-limiting examples of sulfur sources include elemental sulfur deposits; large reserves of sulfur in crude oil, natural gas and sulfide ores (non-ferrous smelters produce H2SO4); produced at petroleum refineries, natural-gas processing plants; sulfur demand associated with phosphate fertilizers; sulfur in gypsum and anhydrite; 600 billion tons contained in coal, oil shale, and shale rich in organic matter. The world sulfur resources is approximately 5 billion tons with 20% in the USA.

[0174] 15

[0175] #15087612vlAccording to some embodiments, the method comprises using barium (e.g., a compound comprising barium). Non-limiting examples of barium sources include barite: sulfate ore, of which there are approximately 740 million tons in the world and 50 million tons in the USA.

[0176] According to some embodiments, the method comprises using lanthanum (e.g., a compound comprising lanthanum). Non-limiting examples of lanthanum sources include ores: monazite (phosphate) and Bastnaesite (fluorocarbonate), which are abundant in the earth’s crust. The world’s reserves of rare-earth oxides are 120 million MIT, with 2.7 million MT in the USA and over 15 million MT in Canada.

[0177] In accordance with some embodiments, the method is energy efficient. For example, in some cases, the method produces steel at a significantly lower energy than conventional steel production.

[0178] In accordance with some embodiments, the method has the ability to handle gangue (e.g., by melting and eliminating them as slag)

[0179] The following is a first Example that was the basis for other experiments with increased oxygen content in the molten sulfide electrolyte. This first Example was determined to be insensitive to oxygen content, as similar experiments with higher oxygen content were also run successfully. All of the materials and conditions described in this first Example could be used in combination with other embodiments disclosed herein, such as with higher oxygen content, such as higher oxygen contents disclosed herein (e.g., greater than 0.5 wt%).

[0180] Electrolytic design / setup

[0181] High temperature processes impose many functional requirements for the various components of the electrolytic cell, especially in terms of thermal and chemical stability. The following section describes the design of the electrolytic cell including functional and material requirements of the cathode, anode and the electrolyte.

[0182] Cathode

[0183] The cathode is where the ions get reduced, in this case, reduction of Fe+3cations to liquid Fe metal. Usually, the cathode material is conductive (if not, a conductive material is used as a current collector, preferably metals) while being less reactive than the metal to be deposited to avoid a metallothermic reaction. Some alloying of the cathode substrate and the metal product generally occurs due to dissolution of the depositing elements (minimum 16

[0184] #15087612vlsolubility). This may also change the composition of the metal product (alloyed with the cathode substrate) and / or the surface of the cathode substrate thereby altering the deposition thermodynamics of the target metal. With respect to thermal stability and electrical and thermal conductivity, carbon is a suitable candidate. However, C is a reductant for iron oxide and will spontaneously react (carbothermic reduction) with the electrolyte to form Fe-C alloy. Generally, the whole purpose of steel production without C is lost as a spontaneous carbothermic reaction occurs when the cell is not polarized. Usually (as followed in the Al industry), a pool of liquid metal (same as the target metal to be deposited) is maintained as the cathode, while having a solid conductive bottom plate as the current collector.

[0185] Electrolyte

[0186] The electrolyte serves as the solvent for the target metal oxide (in this case iron oxide). Factors such as conductivity, density, composition, viscosity, stability are generally important to be considered. Since it serves as a medium of transport for the ions between the two electrodes, it is supposed to be a good ionic conductor (dissociate to ions). This is generally an important characteristic since low ionic conductivity would lead to very high ohmic drop and too high cell voltage. Electronic conductivity of the electrolyte would reduce the efficiency of the electrolytic process as charge would flow directly as electrons rather than ions participating in faradaic reactions.

[0187] Most oxides (including oxides of iron) are electrically conducting at high temperatures, and above their melting points they exhibit high ionic conductivity. With the addition of certain other oxides like calcia, alumina, silica, etc., their ionic character further increases. These oxides exhibit high melting points and transition-metal oxide solubility, and a large range of thermal stability. Thus, these oxides are key components of the electrolyte in MOE. These are good ionic conductors as compared to silicates (covalent due to the network structure) or the semi-conducting transition metal oxides. The latter show a higher share of electronic conductivity than ionic conductivity. FeO cannot generally be directly used as an electrolyte due to such high electronic conductivity. This is because upon melting, all charge gets transferred as electrons and not as ions; molten oxide mixtures rich in FeO have been shown to exhibit a transport number close to null.

[0188] Other physicochemical properties of the electrolyte are generally important. A low vapor pressure helps reduce molten electrolyte losses. The density of the oxide electrolyte mix, though higher than other molten salt electrolytes, is still less than the Fe (target metal), resulting in sinking of the metal deposit at the bottom, while the electrolyte floats atop. A

[0189] 17

[0190] #15087612vlhigh viscosity may hinder the mass transport of the ions, increasing the electrolyte resistance and limiting the reaction rate. Thus, such processes are operated at high superheats, with an optimized melt composition to control the viscosity as reactions proceed. Additionally, the electrolyte should generally not contain more noble species than Fe (elements having a less negative reduction potential than Fe), as they may compete for reduction.

[0191] The high melting and corrosive / solubilizing oxide melt can be held in a container formed by freezing the electrolyte at the sides forming a frozen / solid wall (similar to the Hall-Heroult process).

[0192] Anode

[0193] The anode is the other electrode in the electrochemical circuit and hence should also generally be electrically conductive, chemically inert to electrolyte, mechanically robust and electrochemically stable. It should generally be thermally stable (operating temperatures higher than 1600 °C) and resistant to attack by the oxygen generation at the anode. This is because the reaction at the anode encompasses the release of oxygen gas (oxidation of O2to O2) thus, any metal used as the anode is subjected to get consumed due to corrosion under anodic polarization. The oxide melt used as the electrolyte has a high solubility for other oxides, thus ends up dissolving most passivating oxide layers on the anode surface, exposing virgin metal for further oxidization.

[0194] Iridium initially serves as an inert oxygen evolving anode in iron-bearing aluminosilicate melts. The mechanism through which the anode material sustained without significant consumption was surface reconstruction by chemically combining with an electrolyte constituent to form an intermediate compound. This was referred to as oxide mediated inert anode. Parameters like temperature, electrolyte and alloy composition can be tuned to adjust the rate of the reactions (dissolution, decomposition, etc.) that deplete the oxide layer over the anode. For example, the consumption of the anode is about 20 times higher in a basic electrolyte (high calcia), than in an acidic electrolyte containing more silica. This is because electrolytes with higher basic (calcia) content dissociate easily into ions providing a path for iridium to transfer between the electrodes (resulting in anode consumption), as compared to those with high silica content which polymerize to form a network to allow direct transfer of electrons. It is for this reason that considering physicochemical properties, a basic electrolyte composition is desirable in terms of ionic conductivity and low viscosity but challenging to be compatible with the anode.

[0195] Even though the stability of the iridium anode is proved to be good in several

[0196] 18

[0197] #15087612vlelectrolytes, it cannot be deployed on an industrial scale especially given the tight economics of steel production, as it is extremely expensive and rare in abundance (not sufficient to meet the demands of the industry).

[0198] Thus, a new anode material, Cr1-xFexalloy (Fe varying from 0- 30%), was recently developed as a more affordable anode. The Fe-rich limit of the alloy composition is due to the Fe-Cr melting point. The stability of this material as an anode is attributed to the in-situ formation of a stable, electronically conductive oxide layer and the limited internal alloy oxidation. The oxide layer is a solid solution of Al, not present in the alloy but coming from the electrolyte, and Cr oxides in a corundum structure. Use of the Cr alloy as anode exhibits similar performance in terms of faradaic efficiencies when compared to the iridium anodes, resulting in a pure liquid Fe with very little dissolved Cr and O.

[0199] Electrochemical reduction of sulfides

[0200] As discussed in the previous sections, there have been some challenges with electrolytic processes resulting in their limited adoption. The foremost one being competing against the well-established and mature carbon-based extractive technology in terms of its cost competitiveness, productivity, leveraged by the operational economies of scale. The level of technology refinement in the BF-BOF steel making process does not allow any operational inefficiency in other steel making alternatives. With MOE, there are challenges with respect to an affordable inert anode being used on commercial scale (though Cr alloys are shown are sustain in lab scale). This is where technology involving the electrochemical reduction of sulfides is promising, because it involves the release of S2 gas in place of O2. If not compromising the stability of the anode, the released S2 gas can be condensed and stored for future use, e.g., energy generation or pretreatment processes like sulfidation.

[0201] The concept of electrolytic reduction of sulfides has been discussed in literature since 1906, although for non-ferrous Chalcolithic ores like Pb, Cu, etc. This is because conventional metal extraction from sulfides has been one of the largest emitters of SO2 and these minerals cannot be reduced carbo-thermically due to the higher stability of sulfides over CS2. Furthermore, conventional smelting incurs handling difficulties with the increased concentrations of impurity elements like As, Sb and Bi in the Cu deposits. Due to its release into gas and condensed streams, it requires more stringent management of the hazardous waste generated. Electrolysis of sulfides not only reduces the energy consumption (sulfides have lower thermodynamic decomposition voltages as compared to oxides and chlorides) but

[0202] 19

[0203] #15087612vlalso could support the selective recovery (as in any other electrolytic process) of multiple elements present in the ore at various sequences of electrolysis. Additionally, electrifying metals production offers avenues for integrating it with renewable energy and reducing the cost of production; lesser number of unit operations, hence low capital. The following sections discuss the sulfide electrolysis in halide and sulfide electrolytes.

[0204] Towards sulfide electrolysis in sulfide-based electrolytes

[0205] To solve some of the challenges of previous approaches, especially the solubility restrictions of sulfides in chlorides(directly impacts the productivity) and their vulnerability to impurities in the melt, alkali and alkaline-earth metal sulfides were studied as potential candidates for electrolytes. Due to the large electronegativity difference of these elements with sulfur, they are ionically bonded with each other. Adding these sulfides suppressed the electronic conducting behavior of transition metal sulfides, stipulating their use as supporting electrolytes. BaS was chosen due to its large bandgap (3.92eV) as compared to CU2S (1.21 eV), resulting in an electrical conductivity (0.01 / ohm-cm) 7000 times smaller than Cu2, its partially ionic nature and significantly higher decomposition voltage as compared to the target metal sulfide. The BaS-Cu2S binary phase diagram was studied to choose the electrolyte composition to result in a homogeneous liquid. Thus, at a temperature of 1105 °C (higher than melting point of Cu), high purity liquid Cu was electrolytically extracted from a sulfide electrolyte composing 43.2 wt.% CU2S and 56.8 wt.% BaS, while simultaneously evolving S2 gas at the graphite anode. The liquid Cu produced was ascertained to be a product of electrolytic reduction as no spontaneous decomposition occurred in the absence of electrochemical signals. Though the measured cathode efficiency was limited to <30% due to challenges in metal dispersions on the small cathode used (mm size cathode), this served as a successful demonstration of sulfide electrolytes being used for faradaic applications. To maintain the efficiency of the electrolysis process over time, continuous feeding of CU2S was envisioned to prevent a BaS rich phase from precipitating around the cathode (due to depletion of feedstock) that obstructs further progress.

[0206] Sahu et.al demonstrated increased faradaic efficiency of 59% for the same process by modifying the electrolyte composition by the addition of La2S3contributing to the increased ionic conductivity. The addition of the rare-earth metal sulfide reduced the vapor pressure of the electrolyte, further increasing the operational temperature range. Condensed elemental sulfur was obtained as the anodic product, along with gas evolution evidence also

[0207] 20

[0208] #15087612vlobserved in the form of voids in the electrolyte. The unavailability of a ternary phase diagram was a limiting factor in the empirical choice of the electrolyte composition. Moreover, due to the unknown solution effects of molten sulfides, any thermodynamic assumption in a multicomponent system may not be valid.

[0209] Recently, Cu and Fe were selectively recovered from chalcopyrite by molten sulfide electrolysis (with preferential Fe deposition). Though CU2S demonstrates high solubility in BaS, to keep the electronic conductivity minimal, only 10 wt.% of chalcopyrite was dissolved in 90% of the supporting electrolyte (55.8 wt.% BaS, 34.2 wt.% La2S3). The electrolysis sequentially produced molten iron and liquid Cu. The productivity of the process exceeded industrial standards with current densities as high as 1.5 A / cm2being recorded. The interaction between Cu-Fe-C37ternary and the unique composition of the supporting electrolyte facilitated the separation of iron and copper due to differential solvation resulting in reduced activity of CU2S. Once the local depletion of Fe occurred due to its deposition, the Cu rich phases started precipitating, following the reaction mentioned below:

[0210] CuFeS2Fe ( / ) + S2(g) + Cu2S (1) C

[0211]

[0212] u2S Cu ( / ) + S2(g) (2) To ensure that the supporting electrolyte is not degraded due to formation of any oxysulfides, and can be reused, the feedstock was treated in a sulfidation reaction to convert any oxygen containing species to sulfides. Before this “sulfidised” feedstock was fed to the cell, the immiscible residual gangue phases were separated.

[0213] Along with the reduction of GHG (greenhouse gas) emissions from metal production, MSE also shows promise in reducing the production of low-value slags as demonstrated by the co-reduction of Cu and Fe from chalcopyrite as well as the reduction of Mo and Re36. However, due to the massive production scales and capital investment of these legacy extractive industries, the emerging technology needs to be economically competitive with the current paradigm.

[0214] Iron sulfide electrolysis

[0215] FeS feedstock is fed into MSE cells (up to 10 wt.%) containing a molten sulfide electrolyte, a mixture of barium sulfide and lanthanum sulfide, to support faradaic reactions. The high decomposition potentials of lanthanum and barium sulfides as compared to FeS leads to the electrolytic decomposition of only the dissolved FeS. As a result, Fe is deposited at the cathode and elemental sulfur gas evolved at the anode according to the final reaction:

[0216] FeS ( / ) Fe ( / ) + ¥1 S2 (g) (in case of pure Fe production) (3)

[0217] 21

[0218] #15087612vlAH°16OO°C = 691.47 kWh / ton Fe; AG°*2 / / °c = 294.91 kWh / ton Fe In comparison, the enthalpy, and Gibbs energy change for reduction of iron oxide directly to metallic Fe, as is carried out in the molten oxide electrolysis process, is higher than reduction from sulfides. Moreover, energetically, due to the stability of Fe+2ion, FeS offers fundamental advantages for the electrification of iron production.

[0219] Fe2O3 (s) 2 Fe (1) + 3 / 2 O2(g) (in case of pure liquid Fe production) (4) AH°i6oo °c = 2077.9 kWh / ton liquid Fe; AG°i6oo°c = 865.6 kWh / ton liquid Fe

[0220] FeS ( / ) + x C Fe - 4.3 w

[0221]

[0222] t.% C (1) + S2(g); molten iron (5) AH°i3oo°c = 776.99 kWh / ton molten iron; AG°i3oo°c = 316.61 kWh / ton molten iron

[0223] The sulfur released is recirculated into the sulfidation circuit, forming a closed loop system. Since the S2released in the MSE process is captured and fed back into the system, an additional of only 0.4 tons will be needed by the system in consecutive cycles in comparison to the 0.96 ton of sulfur required for the initial sulfidation of iron oxide. FIG. 25 is a process flow chart depicting the mass balance to produce 1 metric ton molten iron. As shown in FIG.

[0224] 25, the process generates at least 0.1 tons of gangue from the iron ore (considering high quality iron ore with 65% Fe).

[0225] At an operating temperature of 1300 °C, the practical heat efficiency for a hypothetical electrolytic reactor is assumed to be 66.8%. The logarithmic relation to calculate the heat efficiency is derived by fitting practical industrial data for different electrolytic processes in operation.

[0226] Heat efficiency = 26.064 ln(T in K) - 125 (6) The model considers the current technological thermal limitations in industrial electrolytic metal extraction. However, for the case of making realistic assumptions of heat loss, a 40% loss of the heat (as in industrial Al electrolysis) resulting from the ohmic (IR) drop (Joule heat generated) is considered in the calculations below. It should be noted that the loss of heat should be accounted only for the heat generated by the IR drop, as is done in the calculations presented below.

[0227] Thermodynamic Framework for Iron Sulfide Electrolysis

[0228] Enthalpy and energy balance for pure liquid Fe production at 1600 °C

[0229] 22

[0230] #15087612v1The following calculation depicts the enthalpy and energy balance to produce pure liquid Fe at 1600 °C. For reference, the melting point of Fe is 1539 °C, so at 1600 °C, we can expect a superheated liquid iron product, ensuring ability to cast the product out of the cell.

[0231] Enthalpy ofFeS (s at 25 °C) = -104560 J / mol (7) Enthalpy ofFeS (I at 1600 °C)=29990.9 J / mol (8) AH° FeS (s at 25 °C)^FeS (I at 1600 °C)= 134550.9 J / mol (9) AH rxn @1600°C FeS (l)—> Fe (l)+ S2(g)=139014.1 J / mol Fe (10) kWh

[0232] Total enthalpy needed per ton of liquid Fe produced-1360.74ton Fe(11) U cell — Uchem + ohmic + over voltage (12) AG rxn @1600°C =59295.1 J / mol Fe (13) _

[0233] Uchem = -; F=Faraday’s constant, n=no. of electrons transferred (14) Uchem = -0.307V (15) Amount of chemical work = AG° rxn @i6oo°c =294.94 kWh / ton Fe (16) AG = AH - TAS (17) Amount of energy supplied to provide heat = AH - AG = 1065.8 kWh / ton Fe (18) Uohmic = = -1.11 V (19)

[0234]

[0235] nF Uceii = -1.417V (20) Total electrical energy (for chemical work and heat) = 1360.74 kWh / ton Fe (21) Electrical energy required including 40% heat loss = 1787.06 kWh / ton Fe (22)

[0236] FIG. 26 is a plot depicting the variation of enthalpy and Gibbs energy change with the temperature for sulfidising iron oxide to FeS to electrolytically produce one ton of Fe. The change of slope indicates the fusion of FeS at 1194 °C. FIG. 27 is a plot depicting the variation of enthalpy change and Gibb’s free energy change with the temperature for sulfidising iron oxide to FeS in kWh / ton of Fe produced. The sulfidation is performed at a 1000 °C to overcome any kinetic and mass transport limitations, while also keeping the enthalpy requirements low.

[0237] Enthalpy and energy balance for pure Fe production at 1300°C

[0238] The following depicts the enthalpy and energy balance to produce pure Fe at 1300 °C. However, this calculation is also meant for comparison of the effect of temperature and composition, as at 1300 °C, without any alloying element, the product expected is pure solid iron.

[0239] 23

[0240] #15087612vlEnthalpy ofFeS (I at 1300 °C) = 11225.6 J / mol (23) \H" FeS (s at 25 °C) FeS (I at 1300 °C) = 115785.6 J / mol (24) AH°rxn @1300°C FeS (l) ^ Fe (s)+ ±S2(g) = 124955.9 J / mol Fe (25)

[0241]

[0242] Total enthalpy needed per ton ofFe produced = 1197.47 kWh / ton Fe (26) AG rxn @1300°C = 70207.4 J / mol Fe (27) Uchem = -0.364 V (28) Uohmic = -0.88 V (29) Uceii = -1.24 V (30) Total electrical energy (for chemical work and heat) = 1197.47 kWh / ton Fe (31) Electrical energy required including 40% heat loss = 1536.77 kWh / ton Fe (32)

[0243] Enthalpy and energy balance for molten iron (Fe - 4.3 wt.% C) production at 1300 °C

[0244] The following is the enthalpy and energy balance to produce molten iron at 1300 °C. For reference, the melting point of molten iron of the given composition (Fe - 4.3 wt.% C) is 1147 °C, so at 1300 °C, we can expect a superheated molten iron product, with very good flowability. The advantage of producing molten iron comes with significantly reduced processing temperature, which decreases the complexity of heat management and prevents excessive refractory wear. Apart from reducing the melting point of iron by alloying, carbon also behaves as a powerful agent to control the chemistry of iron, e.g., the solubility of O in Fe increases as C decreases. This is especially crucial for refining the molten iron in the BOF for steel making.

[0245] Enthalpy ofC (s at 25 °C) = 0 (33) Enthalpy of C (s at 1300 °C) = 24977.8^ (34) \H" C (s at 25 °C) ^ C (s at 1300 °C) = 24977.8 — (35) XHmix@i3oo°c 95.7 g Fe (s) + 4.3 g C (s) 100g Fe - 4.3 wt.% C (I) = 33778.5 J (36) Total enthalpy per ton of molten iron = 1264.44 kWh / ton molten iron (37) \G"rxn @1300°CFeS (I) Fe (I) + |s2(g) = 70207.4 J / mol Fe (38) 5Gmix@i3oo°c 95.7 gFe (s) + 4.3 g C(s) 100g Fe - 4.3 wt.% C (I) = -6299.17 J (39) 5GrXn@1300°C FeS + 4.3 wt.% C —>■ Fe - 4.3 wt.% C + - S2(g) = 316.7 - — - (40)

[0246]

[0247] 2 ton molten iron Uchem = -0.345 V (41)

[0248] #15087612vlU ohmic = -1.03 V (42) U ceil = -1.376V (43) h Total electrical energy (for chemical work and heat) = 1264.44 - (44)

[0249]

[0250] ton molten iron kWh

[0251] Electrical energy required including 40% heat loss= 1643.53 - - - (45)

[0252] ton molten iron Electrical energy required including 40% heat loss = 1717.38 kWh / ton Fe (46)

[0253] It can be observed that addition of carbon does not lead to any reduction in the energy required, i.e., producing solid pure Fe at 1300 °C requires less energy than producing molten iron (Fe-C alloy) production. This is because the enthalpy of mixing Fe and C is not sufficient enough to offset the enthalpy of fusion of Fe and that required to heat up the carbon to reaction temperature. Nevertheless, direct production of molten iron is better than solid Fe production, as this allows semi-continuous metal recovery from the cell and avoids an additional facility to melt the solid product. Additionally, with respect to integration within the existing steel making facilities, production of molten iron is the most suitable.

[0254] FIG. 28 is a plot depicting the variation of enthalpy change and Gibbs energy change with the temperature for the electrolysis reaction, in kWh / ton of molten iron produced.

[0255] Enthalpy and energy balance for molten iron (Fe - 4.3 wt.% C) production at 1600°C

[0256] The following depicts the enthalpy and energy balance to produce molten iron (Fe -4.3 wt.% C) at 1600 °C. Given, the melting point of molten iron of the given composition (Fe - 4.3 wt.% C) is 1147°C, we can expect at 1600°C a highly superheated molten iron product. However, this calculation is also meant only for comparison of the effect of temperature and composition, as at 1600°C, the molten iron has a super heat higher than 400°C, which is completely unnecessary and a waste of energy.

[0257] Enthalpy of C (s at 1600 °C) = 32323.7^ (47) \H° C (s at 25 °C) C (s at 1600 °C) = 32323.7— (48) mol 2sHmix@i6oo°c 95.7 g Fe (s) + 4.3 g C (s) 100g Fe - 4.3 wt. % C (I) = 10702.4 J (49)

[0258] kWh Total enthalpy to produce molten iron at 1600°C = 1363.88 - - - (50)

[0259] ton molten iron AG rxn @1600°C FeS (1) Fe (I) + | S2(g) = 59288.7 J / mol Fe (51)

[0260]

[0261] XG@1600°C 95.7 g Fe (s) + 4.3 g C(s) 100g Fe - 4.3 wt. % C (I) = -13,159.7 J (52)

[0262] 25

[0263] #15087612vl1 k lA7h ^Grxn@i600°c FeS + 4.3 wt. % C —>■ Fe - 4.3 wt. % C + -S2(g) = 245. - — - (53)

[0264]

[0265] 2 ton molten iron Uchem = -0.267 V (54) Uohmic = -1.217 V (55) (7ce / / = -1.484 V (56)

[0266] / c h

[0267] Total electrical energy (for chemical work and heat) = 1363.88 - (57) ton molten iron

[0268] kWh

[0269] Electrical energy required including 40% heat loss = 1811.164 - - - (58)

[0270] ton molten iron Electrical energy required including 40% heat loss = 1892.54 kWh / ton Fe (59)

[0271] Summary

[0272] Using a thermodynamic framework, including the mass and energy balance for the MSE process, allows one to understand the expected influence of temperature and composition of the liquid metal on the minimum energy requirements. Above, detailed calculations for both, pure liquid iron and molten iron (alloy of Fe and 4.3 wt.% carbon) are presented. It is found that MSE has lower enthalpy and electrical energy requirement as compared to other electrolytic processes for steel production, specifically molten oxide electrolysis, and is an especially attractive option for molten iron production. This is because of the lower temperature requirements for molten iron production (~ 1300°C), as compared to pure liquid Fe production at ~ 1600°C (minimum requirement). The wide difference in processing temperature facilitates better heat management of the sulfide melt, reduces refractory wear, and allows more flexibility in the process.

[0273] Moreover, the faradaic efficiency of the MSE process, as seen from previous studies with copper production show a 90% efficiency, hence the following calculations use a conservative figure of 90%. Thus, including 40% of heat losses in the MSE reactor, the practical energy requirements of MSE are about 52.8 % less, where molten iron production at 1300°C consumes 1717.38 kW / ton Fe; producing 1.045-ton of molten iron compared to molten oxide electrolysis at around 3640 kWh / ton liquid Fe. FIG. 29 is a bar chart depicting the energy requirements for steel making via BF-BOF, MOE and MSE processes.

[0274] Experimental Methods

[0275] This section discusses the experimental methods to confirm the feasibility of the new process methodology proposed. It explains in detail the experimental preparation and setup for the various sets of experiments carried out using lab equipment. A series of solid-gas

[0276] 26

[0277] #15087612vlreactions were conducted to demonstrate the sulfidation of iron oxide to iron (II) sulfide and results of these experiments were interpreted to identify the working parameters (e.g., minimum temperature required, critical gas flow rate, bed height and porosity, etc.). The electrolysis experiments were conducted in a Thermal Imaging furnace and the experimental preparation, including fabrication of the electrodes, electrolyte synthesis is mentioned in detail here. This is then followed by the procedure for conducting MSE experiments while measuring the faradaic efficiency of the process by simple mass loss calculations.

[0278] Additionally, the parameters affecting the faradaic efficiency are identified and some of them are studied as a part of this work, namely effect of current density and factors affecting the impedance of the system. A lot of experimental samples were generated in both the sulfidation and MSE experiments, and these required characterizing them to analyze and interpret the results. Thus, towards the end of the chapter, the various characterization tools used are discussed.

[0279] Feedstock preparation

[0280] Pre-treatment of the ore - Sulfidation

[0281] The process of preparing the feedstock for the MSE cells, by conducting a solid - gas reaction of an oxide material with sulfur gas, is termed as the sulfidation reaction. This is because iron ore mined for steel production is in the form of oxides in the earth’s crust. The most widely used oxide-based iron ore is hematite, which contains up to 65% Fe (-98% Fe2O3 - rich grade of iron ore). This mined oxide ore cannot be directly fed into the MSE cells, as oxides are not soluble in sulfides. In order to be processed in the MSE cells, the oxide needs to be converted to sulfide. The reaction governing the pretreatment of the ore is as follows:

[0282] 4 Fe2Os (s) + 782(g) — > 8 FeS (s) + 6 SO2 (g) (60) To demonstrate the process of preparing the feedstock for the MSE cells, sulfidation reactions were carried out using pure lab grade Fe2O3 (99.85+ % metal basis, Alfa Aesar). The pure iron (III) oxide was heated in a sulfur-rich atmosphere in a Split Mellen vertical tube furnace (Mellen, PS400-120-20CLT-C2778-R-OT). This methodology can be found in detail in the work by other authors, with relevant details for sulfidation of iron (III) oxide discussed in this section.

[0283] 5 grams of pure lab grade Fe2O3 (99.85+ % metal basis, Alfa Aesar) were crushed using an agate mortar and pestle. It was subsequently sieved using sieving meshes to obtain a particle size between 90 microns and 106 microns. This particle size distribution was chosen

[0284] 27

[0285] #15087612vlto emulate the size distribution of the feed used in the blast - furnace. This crushed iron (III) oxide powder was filled in an in-house machined alumina crucible (50 mm OD, 41 mm ID, 34 mm depth, machinable alumina) up to a bed height of 5 mm. The custom-made alumina crucible, had approximately 100 evenly spaced holes drilled at the bottom (1.5 mm OD) as well as at the sides (6.3 mm OD) for the sulfur gas to react across the entire bed height. To prevent the loss of Fe^O? from the bottom during loading the sample, a kimwipe is placed at the bottom of the crucible.

[0286] The machined alumina crucible was soaked in with deionized water and dried at 200°C in a vacuum oven overnight before use. An alumina crucible was chosen compared to a graphite crucible to avoid any interaction of the oxide with carbon, to prevent carbothermic reduction of the oxide at the high temperature of operation. Additionally, the graphite crucible was not preferred to avoid any carbo-thermically driven sulfur reflux (CDSR) reactions, to give an accurate measure of the partial pressure of S2 to SO2 in the reactor.

[0287] Using an alumina support tube (29 cm length, 5.08cm OD, 4.445 cm ID), the iron (III) oxide containing crucible was positioned in the hot zone of a vertical tube furnace in an alumina tube (600 mm length, 25mm OD, 21mm ID, Advalue). Approximately 100 grams of elemental sulfur powder (99.5%, sublimed, Acros Organics) were carefully melted using a heat gun in a quartz crucible (28 mm OD, 101.6 mm depth). This crucible containing solidified and dense sulfur was loaded through the bottom of the furnace. Throughout the sulfidation, argon (Airgas, ultra- high purity) was flowed at a constant rate of 1500 seem through the chamber, behaving as a carrier gas for S2 and SO2 (g). The critical gas flow rate was calculated by performing mass - balance calculations as described in the methodology in Stinn and Allanore (2021).

[0288] At the set temperature (reaction temperature), to ensure the sulfur partial pressure in the chamber is maintained at absolute pressure of 0.1 atm, the sulfur containing crucible was raised at 0.5 cm per minute using a stainless-steel tube (6.25 mm OD), also used for the Ar inlet, introduced at the bottom of the furnace port. As the reaction progressed, generating more SO2 (g), the sulfur containing crucible was raised gradually and periodically over a certain calculated length, increasing the partial pressure of S2 provided to the system.

[0289] The thermal profile for the sulfidation was as follows: The furnace was heated to a set temperature of 1000°C at a rate of 3°C / min, was held at this temperature (1000°C) for 45 mins and later cooled down to room temperature at a rate of 3°C / min. The ramp up rates were maintained at 3 °C / min to avoid damage to the alumina tubes due to thermal shock. To

[0290] 28

[0291] #15087612vlcapture the unreacted solidified sulfur particles in an in-house constructed gravity separator, a 1” diameter quartz tube was used for the outgas line. FIG. 30 depicts a schematic of the vertical tube furnace used for the sulfidation experiments.

[0292] Post - sulfidation observations

[0293] Once the reactor cools down, the alumina crucible was carefully unloaded. Mass measurements were done to calculate the extent of the reaction (conversion rate) using the difference in mass measurements before and after the reaction, assuming there is no mass loss during loading and unloading. This was done by performing stoichiometric calculations for the expected mass of the product formed upon complete reaction with S2 (g). After the mass measurement, the sample was immediately stored in a glove box to avoid the sulfidised product from getting oxidized by the environment. The powder bed (or pellet) was divided into approximately 4 equal parts and a cross-section across the bed was visually observed to check if the reactant was sulfidised completely up to the reactor bed core. In case it had not, further analysis at higher magnification was required to investigate the cause for incomplete reaction.

[0294] One part of the sample was mounted on epoxy and polished to be observed under the optical microscope for any distinct features using the cross-polarization mode. In case of incomplete reaction, the reaction front could be observed and studied at higher magnification to determine the cause. The same polished sample was observed under the SEM to visualize the features of the sulfide grains and carry out EDS analysis to account for the elemental distribution in the sample. Another part of the initial sample was used for XRD analysis to confirm the chemistry of the product.

[0295] Electrochemical experiments - Molten Sulfide Electrolysis

[0296] Once the production of the feedstock to be fed into the MSE cells was successfully demonstrated via a sulfidation reaction, electrolysis experiments were performed at a scale of 300 mA (0.2 g electrolyte and a cathodic surface area of 0.176 cm2), to investigate the electrochemical deposition of iron, as described below.

[0297] These experiments were performed in a Thermal Imaging furnace (TIF, TX- 12000-1-MIT-VPO-PC, Crystal Systems Corp.). This is a 12kW lamps-based furnace, powered by four 3 kW Xenon lamps. Upon illumination, these with the help of ellipsoidal mirrors create a hot zone of one cubic centimeter volume at the focus. The hot zone can attain temperatures up to 3000°C, in comparison to typical furnaces which are limited to 1500°C. An external quartz tube, inside which the testing sample is placed, is used to maintain different types of

[0298] 29

[0299] #15087612vlatmospheres. This allows melting samples without the use of a container, by suspending an ingot type sample using a probe from the top fitting. Electrodes or thermocouples can be introduced inside the quartz tube from either shaft. Modifications made to the TIF for making electrochemical measurements are described by Nakanishi and Allanore. The TIF has cameras installed on the front and side panels, which allow in-situ visual observations such as bubbling due to gas formation during electrochemical reactions.

[0300] To prepare the electrolyte for the 300 mA experiments, pure lab grade chemicals were used with 10 wt.% of iron (II) sulfide dissolved in a supporting electrolyte. The composition of the electrolyte was as follows: 55.8 wt.% BaS, 34.2 wt.% Ea2Ss and 10 wt.% FeS. Since the faradaic efficiency calculations were to be made using mass loss measurements, it was important to decouple the effect of thermal decomposition / thermal loss (if any) from the effect of the electrochemical reaction, on the observed mass loss. Hence, thermal decomposition trials were carried out in the TIF to observe the mass loss due to thermal decomposition.

[0301] Post-experiment observation and characterization techniques used to analyze the samples were employed. The samples were analyzed using optical microscopy and SEM-EDS to understand the micro structure of the electrolytic deposition and obtain the elemental composition of the electrolyte at various locations, respectively. Further analysis, such as ICP-MS I and WDS II were conducted to give more accurate information about the light element composition in the electrolyte sample, specifically to investigate sulfur depletion after electrolysis.

[0302] The procedure to prepare the iron (II) sulfide feedstock iron (III) oxide for the MSE cells is described above (Pre-treatment of the ore - Sulfidation). However, for the 300 mA electrochemical experiments carried out, pure lab grade iron (II) sulfide was used.

[0303] Electrolyte synthesis

[0304] Previously, electrolytic deposition of Cu and Fe was demonstrated by using a supporting electrolyte of composition 55.8 wt.% BaS (99.7% metals basis, Alfa Aesar), 34.2 wt.% Ea2Ss (99.0% metals basis, Alfa Aesar), (binary system characterized by Boury and Allanore) with 10 wt.% sulfidized chalcopyrite, CuFeS2. This was based on work by Stinn et al., who demonstrated the high solubility of copper sulfide in barium sulfide. Sokhanvaran et al. demonstrated that the supporting electrolyte supported ionic transport. The high decomposition potential of BaS and Ea2Ss as compared to FeS, allows the selective decomposition of FeS into Fe depositing at the cathode and S2 gas evolved at the anode, upon

[0305] 30

[0306] #15087612vlelectrolysis.

[0307] To alleviate the electronic conductivity in the supporting electrolyte for the MSE of FeS, lab grade iron (II) sulfide (99.9 % metal basis, Alfa Aesar) was dissolved at 10 wt.%, while keeping the remaining composition unchanged at 55.8 wt.% BaS, 34.2 wt.% La2S3. However, this proportion could be improved once a relationship of the electrolyte conductivity with composition and temperature and the mechanisms governing electrochemical decomposition are established. The constituent sulfides were ground using a mortar and pestle in a controlled atmosphere inside the glove box. 200 mg portions of the electrolyte powder mixture were loaded into 7 mm OD holes machined in graphite circular puck (53 mm OD, graphitestore). These droplets were pre - melted in the Mellen furnace (Mellen, PS400-120-20CLT-C2778-R-OT), at a set temperature of 1350°C to form a consolidated solid of the powder mixture for ease of handling. The thermal profile of the premelting procedure was as follows: the furnace was heated up to 1350°C (set temperature) at a rate of 5°C / min, held at 1350°C for 60 mins and then cooled down to room temperature at a rate of 5°C / min. Argon (Airgas, ultra-high purity) was flowed through the chamber at a rate of 450 seem to maintain a controlled atmosphere within the reactor during the pre-melting. In case of presence of any residual oxygen in the graphite pucker, there are chances of carbothermic reduction occurring during the electrolyte synthesis. To rule out this possibility, characterization studies (optical microscopy and scanning electron microscopy) were carried out to observe if any metallic phase existed in the pre-melt. SEM image and EDS scan results of the electrolyte after pre - melting were produced. To confirm the elemental composition, the electrolyte was sent for ICP - MS I and LECO.

[0308] Fabrication of the electrodes

[0309] Graphite has been found to be stable and inert during electrolysis of molten sulfides at temperatures greater than 1200°C. Hence, the electrodes, i.e., cathode and anode were machined using EDM quality graphite rods (fine extruded, 6.3mm OD, 305mm length, graphitestore). The height of the cathode stands (shown in FIG. 31 A) was 12 mm with a hemispherical depression / divet at the top to hold the electrolyte droplet. The cathodic surface area of approximately 0.19 cm2(6 mm ID, 0.7 mm depth) was machined using a dremel. The anode, as depicted in FIG. 31B, was machined to be tapered at one end into a fine tip using a belt sander. The height of the anode in total was 16 mm, with the tip (1.3 mm OD) being 11 mm long. On the opposite end of the electrodes, holes 3.06 mm ID) were drilled using a drill press and threaded, to be securely held by a molybdenum rod (>99.97%, 3.2mm OD, 600mm

[0310] 31

[0311] #15087612vllength, Ed Fagan) sheathed in an alumina tube when placed in the TIF. Before use, the graphite electrodes were sonicated for 15 mins using ethanol and were air dried.

[0312] Electrolytic reduction on a graphite cathode

[0313] The electrolytic experiments were carried out in same TIF as described above. The electrode probes, made up of molybdenum rods sheathed in an alumina tube (>99.8%, 6.35mm OD, 4mm ID, Coorstek) were introduced from the top and bottom ports (sealed using Ultratorr fittings, securely holding the anode and cathode respectively. FIG. 32 is a schematic of the 2-electrode experimental set up used in the thermal imaging furnace (TIF) used for MSE experiments. Before the entire assembly was sheltered in a controlled atmosphere inside a quartz tube (customized, Technical Glass Products, Inc.) sealed using Viton O-rings, the cathode, anode, and electrolyte were individually weighed to milligram precision. Any loss of mass from the cathode or anode would indicate presence of oxygen inside the reactor, as mass loss was attributed to carbon getting burnt away as carbon dioxide. The atmosphere inside the tube was maintained under argon as described in the thermal-decomposition trials section.

[0314] To precisely position the electrodes in the hot zone of the furnace, up to 0.1 mm precision, monitored using a camera (EOS Rebel T5i DSLR, Canon Inc.), the probes were controlled by stepper motors. Once oxygen was scrubbed off from the system using the gettering furnace housing a Ti charge, and a constant flow of argon gas was ensured, the lamps were switched on and were powered to 4% power. The electrolyte was heated for about 2-3 mins till a stable molten droplet was formed, by rotating the bottom probe consisting of the cathode constantly at a rate of 10 rpm. To complete the circuit, the rotation was stopped, and the anode tip was moved down to contact the molten droplet, as shown in FIG. 33. Galvanostatic electrolysis at a cathodic current density of 0.85 A / cm2and 1.7 A / cm2(anodic current density 10.64 A / cm2) were carried out. A current of 0.15 A and 0.3 A was applied for varying time lengths of 20, 40,60, 90 and 120 seconds using the Gamry Reference 3000 potentiostat, to account for the faradaic efficiency. Results of the above experiments are shown in the tables in the following sections. The current values were chosen such that the applied current density was similar to the aluminum industry standards (around 1 A / cm2), and double of that to test if the system was stable at higher current density conditions. Additionally, galvanostatic measurements were also carried out to identify the current range where the decomposition potential was achieved.

[0315] Once the experiment was done (usual duration 5-6 mins), the lamps were switched

[0316] 32

[0317] #15087612vloff to quench the sample at a rate greater than 100°C / sec.

[0318] Apart from the electrolysis experiments, a systematic study of the factors affecting the impedance of the system was conducted. The following factors were studied: variation of the impedance with distance between the electrodes and the potential applied. For all the above studies, the counter electrode was used as a reference electrode (2-electrode set up) to record the cell voltages.

[0319] The sample was carefully taken out from the furnace, 30 mins after the lamps were switched off. To prevent the sample from getting contaminated in the ambient atmosphere, mass calculations of the cathode, anode, and electrolyte were done immediately before the sample was mounted in epoxy resin for polishing for further characterization studies. It was then stored in the glove box in an inert atmosphere.

[0320] Sulfidation

[0321] An overall reaction for the sulfidation process can be written as follows:

[0322] 4 Fe2O3(s) + 7 S2(g) → 8 FeS (s) + 6 SO2(g)

[0323] From past sulfidation experimental evidence and preliminary experiments it was observed that the sulfidation process follows the shrinking core reaction mechanism, reacting from the outer surface to the inner core.

[0324] Optical microscopy images of the sulfidized product at magnifications of 10X, 20X, 50X were produced. The sulfide particles were observed to be shiny and porous under the optical microscope as compared to the oxides (Fe2O3or Fe3O4), which are dull and dense. These micrographs are clear evidence of sintering to have occurred during the process, restricting further mass transport of S2 gas inward or SO2 outward. Since the bed height was limited in this case, it resulted in complete conversion of the oxide to sulfide or else a reaction front could be observed with some part of the core staying unreacted or partially reacted.

[0325] Scanning electron micrographs of the sulfidized samples along with their EDS results were produced. From the atomic concentrations of iron and sulfur, as obtained from the EDS results, the stoichiometry of the product compound formed upon sulfidation was estimated, as shown in Table 1 below.

[0326] Table 1: Approximate compositional analysis of the product after EDS analysis. Note that the O is believed to be present due to the use of epoxy.

[0327] Element Concentration (wt.%)

[0328]

[0329] 33

[0330] #15087612vlFe 57.99

[0331] S 37.54

[0332] 0 4.47

[0333]

[0334] XRD

[0335] The product was analyzed for its phase composition using XRD. The peaks as seen in the XRD results corresponded to FeS (iron (II) sulfide) with insignificant traces of any residual oxides. FIG. 34 shows an XRD scan of the product upon sulfidation of iron (III) oxide.

[0336] Electrolytic reduction for iron production

[0337] Electrolyte synthesis

[0338] Based on experimental evidence and literature studies, the percentage of FeS to be dissolved in the supporting electrolyte is governed by the following factors:

[0339] 1. Dependence of conductivity of the electrolyte2 and the current density on the concentration of FeS in the supporting electrolyte: too high an electronic conductivity is not desirable but the process needs high ionic conductivity to ensure low ohmic drop

[0340] 2. Variation of the melting point of the supporting electrolyte with the concentration of FeS in the electrolyte: the electrolyte needs to be molten state, and since the desired product is molten iron (melts at 1147° C, with 4.3 wt.% carbon), the melting point of the electrolyte should be in the range of 1200 o C (+ 100 o C as superheat to ensure flowability)

[0341] 3. Ideally, if a high percentage of FeS shall be soluble, it is advantageous to operate with a high FeS content assuming the previous 2 conditions are satisfied: this allows for a larger amount of molten iron to be processed per electrolysis cell at any time.

[0342] The electrolyte used for the electrochemical experiments was pre-melted in the furnace, resulting in solid droplets of about 200 mg, around 180 mm3. To account for any mass loss during the pre-melting process and / or oxygen contamination, weight measurements were carried before and after the pre-melt procedure. Table 4 below shows the mass loss recorded for each electrolyte droplet. A consistent mass loss of less than 1 mg. This insignificant mass loss indicates the absence of oxygen contamination that would have led to

[0343] 34

[0344] #15087612vlsulfates or oxysulfide. The pre-melt was analyzed using LECOIII to confirm the carbon or oxygen content (if any) resulting from contamination from the graphite pucker or furnace atmosphere respectively, the results of which are shown in Table 3.

[0345] Shown in FIGs. 35A-35B, FIG. 36, and FIGs. 37A-37B are the optical and SEM micrographs of the electrolyte after pre-melting. No compound formation was observed in the pre-melt during SEM characterization. At least 2 phases can be observed - Ba rich phase (dark colored) and a La rich phase (light colored). The approximate composition of these phases (as recorded after EDS analysis) is shown in Table 2.

[0346] Table 2: Approximate phase composition of the observed phases in the electrolyte pre-melt Phase Ba La Fe S Dark colored 54.61 wt.% 25.44 wt.% 4.21 wt.% 15.75 wt.% Light colored 34.56 wt.% 39.18 wt.% 5.59 wt.% 20.67 wt.%

[0347]

[0348] To get an exact composition of Fe, S, C and O in the pre-melt, the electrolyte was sent for ICP-MS and LECO, results of which are shown below:

[0349] Table 3: Results of the ICP-MS I and LECO analysis III of the electrolyte pre-melt

[0350] Fe (ICP) S (ICP)x C (LECO) 0 (LECO) 6.64 wt.% 15.65 wt.% 0.03 wt.% 0.038 wt.%

[0351]

[0352] There is no significant oxygen or carbon contamination in the electrolyte sample, thus eliminating any possibility of oxysulfide formation or carbothermic reactions.

[0353] Electrolytic reduction on a graphite cathode

[0354] Electrolytic reduction was performed at varying current densities of 1.7 A / cm2and 0.85 A / cm2by setting the current at 300 mA or 150 mA, using the droplets of Table 4 as starting electrolytes. FIG. 38 depicts gas bubbles observed during electrolysis, plausibly indicating S2 evolution.

[0355] 35

[0356] #15087612vlTable 4: Results of the mass loss measurements in pre-melt synthesis and following thermal trials

[0357] Mass before Pre-melt at Loss -1 Duration of Mass after Loss -2 pre-melt (g) 1350°C (g) droplet thermal thermal (g)

[0358] (g) treatment (mins) treatment (g)

[0359] 0.2002 0.1997 0.0005 5 0.199 0.0007

[0360] 0.1984 0.1974 0.001 10 0.1958 0.0016

[0361] 0.1986 0.1981 0.0005 20 0.1971 0.001

[0362] 0.1985 0.1979 0.0006 30 0.1968 0.0011

[0363] 0.1998 0.199 0.0008 40 0.1986 0.0004

[0364] 0.1984 0.1956 0.0028 50 0.1946 0.001

[0365] 0.1995 0.1988 0.0007 60 0.1973 0.0015

[0366]

[0367] Electrochemical signals

[0368] With evidence from the electrochemical signals and optical and scanning electron micrographs showing depletion of Fe in the electrolyte, it can be concluded that the Fe deposition occurring is a result of electrolysis.

[0369] The choice of the current to be applied to perform electrolytic decomposition of FeS was arrived at by the following method. Galvanostatic measurements were performed at different currents ranging from 0.0125 A to 0.55 A. The corresponding measured cell voltage variation with time are plotted in FIG. 39. The plot in FIG. 39 depicts the actual recorded cell voltage and is not corrected by the IR drop, R being the ohmic resistance measured between the electrodes at OCP using EIS prior to electrolysis. Three ranges of current can be distinguished from the observed trends in cell voltage:

[0370] (i) Below 0.1 A, the cell voltage is less than the thermodynamic minimum decomposition potential

[0371] (ii) Between 0.1 A and 0.4 A, the cell voltage is between the thermodynamic minimum and 0.9 V [decomposition potential range including the IR drop and over

[0372] 36

[0373] #15087612vlpotential]

[0374] (iii) Above 0.4 A, the cell voltage is greater than 1 V.

[0375] Thus, these measurements served as a good reference for deciding the range of current to be applied to perform electrolysis, (0.1 A - 0.4 A). Thus, 2 values of 0.15 A and 0.3 A were chosen by considering the resulting current density, to study as they yielded cell voltages in the range of the minimum thermodynamic decomposition potential.

[0376] Accounting for the overpotential was out of scope of this work due to the absence of a reference electrode. Hence, for the sake of simplicity, the effect of overpotential was ignored. However, this assumption is reasonable due to the following reasons:

[0377] 1. High concentration of the reactant ions in the electrolyte, the activity ratio between the electrode surface and the bulk is assumed to be about one. Additionally, the minimal amount of time the current is passed, is insufficient to fully deplete the region next to the electrode leading in a concentration gradient. This leads to the overpotential due to mass transfer of the ionic species to be negligible as compared to the product gas.

[0378] 2. Since the operating temperature is greater than 1300° C, it can be assumed that the electrolysis is not limited kinetically. Hence, the contribution of the charge overpotential can also be ignored.

[0379] Chronopotentiometry scans (constant current of 0.3 A and 0.150 A) are shown in FIG. 40 and FIG. 41 for experiments done at different current values for various durations. All cell voltages depicted in the plots are the actual recorded cell voltages and are not corrected by the IR drop. The voltage v / s time plots show the deposition voltage for Fe to be around 0.4 -0.5 V, close to the thermodynamic minimum. The cell voltage is observed to increase as the time increases. The following are plausible reasons for the observed phenomena:

[0380] 1. As the reaction progresses, it becomes thermodynamically challenging for the subsequent removal of the sulfur due to the remaining Fe ions binding more strongly with the remaining sulfur.

[0381] 2. Accumulation of S2 gas bubbles around the anode tip, which might be the cause for increased resistance in the system.

[0382] 3. As the electrochemical reaction proceeds leading to deposition of Fe at the cathode, it must be providing an alternate low-resistance path for the electrons to flow. Thus, it plausibly increases the ionic resistance due to generation of a low electronic resistance path. This system could then be proposed as a parallel resistor

[0383] 37

[0384] #15087612vlcircuit with two resistance paths, ionic resistance and electronic resistance.

[0385] 4. Depletion of the feedstock FeS from the system as the reaction proceeds leads to a change in the composition of the electrolyte. It was experimentally observed that after performing the electrolysis for long durations or higher current values (0.5 A), the electrolyte begins to solidify, while maintaining the same furnace power. Thus, since the flow of ions gets highly restricted in a viscous or solid electrolyte, it leads to an increase in the ionic resistance. Thus, if the above stated hypothesis is true, a continuous feeding of the feedstock to maintain the composition and a periodic removal of the generated products (Fe from the cathode and S2 from the anode) is expected to sustain a constant cell voltage. However further investigation is needed to get a deeper insight into the system’ s response to the applied current and the mechanism of electrolytic activity.

[0386] Other observations included negligible variations of the voltage at lower current densities as compared to the higher current density. However, the plausible causes for the voltage fluctuations at higher current density to be higher could most likely be due to the formation, growth, and separation of the gas bubble from the anode. These are depicted by arrows in FIG. 40 and FIG. 41.

[0387] In some plots shown in FIG. 40 and FIG. 41, some sudden discontinuities are observed in the plot of potential as a function of time. They coincide with the slight adjustments of the anode height, resulting in a minute change of distance between the electrodes. Since the resulting change in voltage was significant, further studies about the variation of the impedance with distance between the electrodes and potential were carried out separately in the same operating conditions. Ideally, in an industrial electrolysis cell, due to the deposition of the metal product (since the liquid metal is not continuously, but periodically tapped), the distance between the electrodes decreases with time. Thus, the IR drop (ohmic overpotential) reduces, affecting the heat generated in the cell. To maintain the interstitial distance between the electrodes, the anode position can be constantly adjusted in the vertical direction by moving it upwards or away from the cathode, such as practiced for vertical carbon anodes in aluminum electro winning. Moreover, the hot metal can be tapped more frequently as it is less sensitive to oxygen as compared to aluminum.

[0388] Factors affecting the impedance of the system

[0389] FIGs. 42A-42B and FIG. 43 show the variation of the impedance with the anode vertical position: the distances between the electrodes were controlled by moving the probe

[0390] 38

[0391] #15087612vlattached to the anode, up and down with a precision of 0.1 mm, controlled using stepper motors. As the distance between the electrodes increases, it is observed that the impedance also increases.

[0392] Based on the results as shown above, it can be concluded that the impedance of the system is extremely sensitive to the distance between the electrodes. This dependence is a feature of the electrolyte's resistivity and the experimental cell design, including the geometry of anode and cathode.

[0393] With the appropriate boundary conditions for current conservation, the electric currents physics model was evaluated to calculate the resistance of the system for the varying distance between the electrodes. FIG. 43 compares the values as calculated using a COMSOL model and those experimentally obtained. A good agreement between the two values is observed along with a similar trend between the experimentally obtained values and those calculated using COMSOL. The little deviation could be attributed to the following factors:

[0394] 1. Variation in the electrochemically active area of the anode: The anode used for experiments was not as perfectly machined as compared to the geometrical design used for the COMSOL model.

[0395] 2. Isothermal conditions assumed in the model; highly likely in the TIF (limited hot zone).

[0396] These values were used to estimate the resistivity of the electrolyte, which approximately equates to 3.44 ohm-cm. Prior work on Cu production via molten sulfide electrolysis of chalcopyrite involved similar studies resulting in the resistivity for CuFeS2-BaS-La2S3being 1.5 ohm-cm. The comparatively lower resistivity can very likely be attributed to the presence of CU2S in the system.

[0397] Mass loss and visual observations

[0398] Based on a visual observation of the quartz tube post experiment (see FIG. 44), it can be confirmed that the electrolysis process resulted in sulfur gas produced at the anode.

[0399] After removing the quartz tube from the furnace, the top part of the tube showed a yellowish tinge, suggesting the S2 gas that left the electrolyte condensed on the sides of the tube. This is due to the sides of the tube being cooler than the boiling point of S2 (boils at 444.6°C). Moreover, there was also a residual odor of sulfur from the tube, serving as evidence of S2 evolution.

[0400] The faradaic efficiency (anodic) was calculated based on the mass loss of the

[0401] 39

[0402] #15087612vlelectrolyte. This is because all mass loss was attributed to the loss of sulfur from the system due to electrolysis, with corresponding metal production at the cathode observed by microscopy. In the range of the recorded cell voltage, only FeS is expected to electrolytically decompose while BaS and La2S3remain, their decomposition voltages being comparatively higher than for FeS. To confirm this and observe the elemental composition of the metallic deposits, optical and SEM microscopy was conducted after electrolysis.

[0403] In cases where metal recovery is possible, the cathodic efficiency can be calculated, where the solidified metal mass recovered at the cathode is compared with the amount of Fe as predicted by Faraday’s law for reduction of Fe+2. Since metal recovery was not feasible at such small scales, only anodic efficiencies were calculated. Table 5 below summarizes the results of the various electrolysis experiments depicting the mass change measurements and faradaic efficiency recorded.

[0404] Table 5: Results of the electrolysis experiments carried out depicting the mass change measurements and faradaic efficiency

[0405] Exp. No Current (A) - Initial Final Mass loss (g) Faradaic duration (sec) mass (g) mass efficiency (%)

[0406] (g)

[0407] 01 0.3 - 20 0.197 0.196 0.001 100.3* 02 0.3 - 40 0.1969 0.1951 0.0018 90.3

[0408] 03 0.3 - 60 0.1976 0.1947 0.0029 97.0

[0409] 04 0.3 - 80 0.1971 0.1949 0.0022 55.2

[0410] 04(ii) 0.3 - 2x40 0.1995 0.1955 0.004 100.3* 05 0.3 - 90 0.1964 0.1944 0.002 44.6

[0411] 06 0.15 - 60 0.1975 0.1962 0.0013 86.9

[0412] 07 0.15 - 90 0.1993 0.1956 0.0037 82.5

[0413] 08 0.15 - 120 0.195 0.1931 0.0019 63.5

[0414] NOTE: Estimated faradaic efficiency c epends on t re accuracy of the mass measurements.

[0415]

[0416] 40

[0417] #15087612vlHindrance in the recovery of the electrolyte happens if it sticks to the cathode, while scrapping off can pick up some material from the cathode.

[0418]

[0419] It is important to note that the estimated faradaic efficiency calculated above is contingent on the ability to completely recover the products or accurately measure the mass losses. As observed from the thermal decomposition studies, since the mass loss of the electrolyte was negligible due to thermal effect, in the above calculations, the thermal mass loss was not accounted for. The resulting mass loss is considered entirely due to the effect of electrolysis. The estimated efficiency is also observed to be dependent on the configuration of the electrolytic cell, such as the following:

[0420] 1. Anode design: The efficiency is reduced when the design of the anode is not optimal. For instance, when the anode tip is sharp and narrow, corresponding to a high anodic current density, the gas evolution is high initially, but the electrochemically active anodic area is quickly getting covered by the gas, inhibiting any further electrochemical activity. The cell configuration for an optimized efficiency would be one that balances the requirement of current density at the electrodes to carry out the necessary reaction as well as allows the release of gas to avoid blocking of the anode (reducing the active area) due to gas evolution.

[0421] 2. Distance between the anode and cathode: This is a feature of the resulting resistance of the electrolyte system, that subsequently affects the ohmic drop, and the dynamics of bubbles evolution, which has a large influence on the interpretation. It was observed that the higher the distance between the electrodes in the electrolyte, the higher the resistance (as was also observed by L. Rush4). Notably three different types of observations were made upon electrolysis experiments and their plausible reasons are included below:

[0422] (i) When the electrode gap is small, the cell voltage recorded is close to the minimum thermodynamic potential for FeS deposition. However, no evidence of electrochemical activity (e.g., gas bubbling, Fe deposits; oxidation or reduction) was observed. This correspond to a gap where the anode tip is below the center of the electrolyte droplet, very close to the cathode.

[0423] (ii) When the anode tip is inserted near the center of the electrolyte droplet, i.e., the distance between the electrodes is of the order of the radius of the electrolyte droplet, the cell voltage was transient. It remained close to the thermodynamic minimum for part of the experiment duration while continuously increasing. Hence,

[0424] 41

[0425] #15087612vlelectrochemical activity would have occurred only for the experiment's duration when the increasing cell voltage went beyond the thermodynamic minimum, thus accounting for the ohmic (IR) drop or other sources of overpotential. Thus, this could be one of the reasons for the reduced faradaic efficiency in some cases. For instance, in an electrolysis experiment for 40 secs, in case the cell voltage remained equal to the thermodynamic minimum for about 10 secs. As it continued to increase, it went beyond the minimum only in the latter 30 secs, during which we can assume electrochemical activity to have occurred for only 30 secs as opposed to 40 secs. Thus, this plausibly translates to the reduced faradaic efficiency calculated for the entire 40 secs.

[0426] (iii) Only when the distance between the electrodes was greater than the radius of the droplet (i.e., the anode was inserted up to the center of the droplet), was the faradaic efficiency calculated close or above 90% for each experiment. In this case, the recorded cell voltage was significantly higher than the thermodynamic minimum, 0.7 to 0.8 V.

[0427] This suggests that the IR drop, controlled by the electrode gap, is linked to the recorded cell voltage. The higher the electrode gap, the higher is the IR drop and hence the cell voltage. One plausible hypothesis for the reduced faradaic efficiency at lower distance between the electrodes, could be due to accumulation of S2 bubbles around the anode tip, resulting in ceasing any further electrochemical activity till the bubbles have sufficient time to escape. Hence, even though the recorded cell voltage was close - though greater - than the thermodynamic minimum, there is very high chance of the initially released S2 bubbles to be accumulating around the anode. The electrode gap being too small, there are multiple plausible explanations can be proposed:

[0428] 1. Bubbles cover the anode and cease further electrochemical activity till they get sufficient time to escape. Moreover, since the size of the S2 gas molecules is large, there is an inherent mass transport limitation, i.e. higher resistance to transport through the electrolyte. Ideally, in such a scenario, it is expected that electrochemical activity would resume once the anode is cleared off the bubbles, however since the duration of the experiment is short, we do not observe this taking place. Evidence supporting this hypothesis is presented later. In the case of 0.15 A (lower current density), the effect of distance between the electrodes observed was less as compared to 0.3 A (higher current density). The plausible explanation for this observation could

[0429] 42

[0430] #15087612vlbe the higher reaction rate in the case of higher current density leading to more vigorous bubbling and hence gas accumulation around the anode. So, controlling the electrode distance was a very sensitive parameter affecting the faradaic efficiency in this case. Even a little increase in the distance allowed the S2 molecules to escape, and thus proportionally increasing the distance between the electrodes up to a certain limit, a trade-off with the increasing IR drop, led to higher faradaic efficiency. However, in the case of lower current density, this sensitivity to the electrode distance was comparatively less.

[0431] 2. Back reaction of the released S2 molecules with the Fe deposited to form FeS, as a result of the electrolyte around the anode being solubilized by sulfur. This would result in a lower to no net electrochemical activity observed. Although, the S2 gas encountering the reduced metal can be prevented by a better cell design.

[0432] According to the experimental observations, if the resulting potential is close to the thermodynamic minimum for FeS decomposition, the efficiency is observed to be negligible due to the possibility of the above stated hypotheses. Only when the recorded cell potential is higher than the decomposition potential, the resulting efficiency is close to 90% or above. The chart in FIG. 45 captures the effect of the time and current density on the experiment's faradaic efficiency.

[0433] As observed from in FIG. 45, the efficiencies recorded for a current of 0.15 A are slightly less as compared to experiments where a current of 0.3 A was passed. This is because a higher polarization is induced with a higher current density, leading to an increased driving force for any chemical reaction to occur. The rate of reaction was qualitatively observed to be higher in 0.3 A as compared to 0.15 A. The video recording in-situ electrolysis showed bubble formation earlier as well as more vigorous in the former as compared to the lower current density case.

[0434] There is an interesting feature that was observed while studying the effect of current density in the electrolysis experiments. For experiments in which a current of 0.3 A was passed, the efficiency dropped to almost half after 60 seconds of electrolysis. As seen in FIG.

[0435] 45, for all times less than 60 seconds, the efficiency recorded is 90% and above, except for those experiments with an electrolysis duration greater than 60 seconds (80 seconds and 90 seconds). From a visual observation in-situ electrolysis (recorded using the camera), a vigorous bubbling was observed at the anode during the time the current was passed, indicating a reduction in the electrochemically active area due to the gas bubbles blocking the anodic surface area till sufficient time was allowed for the gas bubbles to escape.

[0436] 43

[0437] #15087612vlAdditionally, due to diffusion of the ionic species from the bulk electrolyte to the electrodes and later of the S2 gas away from the anode, there is a likelihood of a spatial gradient in the electrolyte composition.

[0438] This hypothesis was tested in a separate experiment where the electrolysis duration was done in 2 steps with a 10 sec inactive period between them. This was done for the system to reach equilibrium once the S2 gas bubbles escaped away from the cathode, allowing to remove bubbles surrounding the anode each time the current was passed.

[0439] Electrolysis during a total of 80 secs resulted in an efficiency of 44.6%. With a sequence of 40 secs with current, a 10 secs gap, and 40 additional seconds passing current - totaling 80 secs of electrolysis time, resulted in an efficiency of 100%. This most likely shows the effect of the gas bubbles around the anode, and the importance of the cell configurational design to support efficient gas removal. This was also the reason experiments with lower current density were performed for longer electrolysis duration. To simplify the comparison, the current density was halved to study its effect for double the electrolysis time. According to the theory discussed above, the effect of the gas evolution limitation in a lower current density case is expected to occur at higher electrolysis duration. For instance, the gas evolution limitation was observed at experiment duration greater than 60 secs in the case of 0.3 A of current. Hence, for the same amount of gas to be released, this limitation is expected to occur at double the time (i.e., 120 secs) when operating at half the initial current density value (i.e., 0.15 A). As seen in FIG. 45, the efficiency for time of 120 secs (0.15 A), indeed drops down as compared to 60 secs or 90 secs. Thus, the experimental results observed were in high coherence to the theory discussed.

[0440] When recast as a plot of faradaic efficiency versus charge, as shown in FIG. 46, it becomes clear that this is not a limitation of the process by itself, but of the cell design. For a similar charge (e.g., 1800°C), higher efficiency is recorded at higher current density, clearly indicating that the process is tolerant to high reaction rates. The same can be concluded based on the point corresponding to 2400°C, though the same current density is maintained, just by changing the method of operating the cell, high efficiency was observed.

[0441] With these results, it can be concluded that the efficiency is highly dependent on the cell configuration, which itself dictates the mass transfer conditions for the anodic and cathode reaction. While the effect of time on the efficiency observed is important, it is a feature of this specific set-up and is expected to be managed during the design of industrial scale cells. During the inactive period, the lamps were still powered at the same intensity, to

[0442] 44

[0443] #15087612vlavoid the electrolyte from losing heat and solidifying, thus assuming the operating conditions remained unchanged between the 2 current passage cycles. However, this cannot be replicated in an industrial setup since the cell is engineered to be self-heated by the flow of current through the resistive electrolyte. Switching off the current in between would result in a sudden temperature drop in the electrolyte, resulting in either a change in the operating conditions when the flow of current is re-established (lower temperature, higher viscosity of the electrolyte), or electrolyte solidification in extreme cases. The experimental study of the effect of time was performed only to show the dependence of the estimated efficiency on the cell design.

[0444] Nevertheless, a major takeaway is the high efficiency of the process, resulting from FeS being the sole iron sulfide stable (+2 being the dominant valency for iron in the electrolyte) at high temperature and low partial pressure of oxygen. This is significant as the multi-valency of iron has afflicted previous electrolytic systems dealing with iron.

[0445] Additionally, the high solubility of the feedstock in the electrolyte, up to 10 wt.%, (as compared to in other electrolysis processes, e.g., Al production) also ensures the high faradaic efficiency, along with establishing a molten electrolyte.

[0446] Further characterization studies

[0447] The optical and SEM micrographs of the electrolyte sample are shown in FIGs. 47 A-47C. Tiny droplets of metal are observed at the bottom, where the electrolyte contacts the graphite cathode, confirming electrolytic reduction of FeS. SEM-EDS analysis does not indicate reduction of BaS or La2S3; only FeS is decomposed. Fe metallic deposits are observed with an average size of 20 microns, with a few deposits of up to 50-micron size.

[0448] However, due to surface tension effects (usually observed at such small scales) and as a result of solidification, the Fe metallic deposits produced were dispersed at the bottom of the electrolyte, found all along the cathode curvature. Due to this, the metal recovery was hindered, preventing calculation of the cathodic faradaic efficiency.

[0449] Some of the metallic deposits are observed at the center of the electrolyte and detached from the cathode most likely floated up when the electrolyte and the metal were in the liquid state and stayed there when the sample was quenched. When observed under the SEM, three regions are identified, with different sulfide composition, though usually a mixture of 2 phases in addition to possibly the Fe metallic phase: (i) anode is inserted; (ii) middle / bulk electrolyte; (iii) region near the cathode.

[0450] It is important to note here that the phases observed may not necessarily represent the

[0451] 45

[0452] #15087612vlphases present during electrolysis. The phase portioning can also result due to the nonequilibrium cooling and solidification. Though the sample was quenched to retain the phases formed during electrolysis as it is, the graphite cathode holds some heat even after the lamps are switched off, resulting in some phase segregation due to the thermal dissipation. In FIG.

[0453] 48, the light-colored regions contain La rich phases whereas the Ba rich phases appear dark in the SEM micrograph, confirmed using both EDS and WDS analysis. Performing an EDS analysis over an area, it was observed that the region in contact with the cathode, especially in regions next to the deposit, are depleted in Fe.

[0454] Line EDS analysis scans shown in FIG. 49 reveal almost no Fe concentration in the sulfides that surrounds Fe deposit. These scans reveal the simultaneous depleted or absence of signal for other elements such as S, Ba and La as we go closer to or within the Fe deposit.

[0455] Table 6 below shows the relation between the concentration of Fe and the Ba / La in the electrolyte. The values observed are from the SEM-EDS analysis over an area providing the average value in the different regions of the electrolyte - near the anode insertion, bulk of the electrolyte and surface near the cathode respectively.

[0456] Table 6: Relation between the concentration of Fe and the Ba / La in the electrolyte, similar Ba / La ratios recorded upon WDS analysis II

[0457] Fe Ba / La

[0458] 6.06 wt.% 1.83

[0459] 4.89 wt. % 1.77

[0460] 4.54 wt. % 1.57

[0461] 3.93 wt.% 1.36

[0462] 2-3 wt.% 0.83

[0463] 0.85 -1.77 wt.% 0.6

[0464]

[0465] It is observed that, in the solidified sulfide electrolyte, Fe is found in phases that are rich in Ba, i.e., phases rich in Fe were correspondingly rich in Ba also. The Fe concentration in the electrolyte was studied together with the Ba / La ratio in the phases. The direct proportionality between Fe concentration and the Ba / La was observed in each region, i.e., as the Fe reduced, the Ba / La ratio also decreased. This is interpreted as indication of the role of mass transport during electrolysis.

[0466] The Fe depletion which occurs due to electrolysis, leads to the formation of a Ba rich

[0467] 46

[0468] #15087612vlsulfide phase. This compositional inhomogeneity locally affects the electrical and other transport properties for further electrolysis to occur. Since the temperature is controlled at the eutectic range of Ba and La sulfide, the formation of the Ba rich sulfide phase results in a solid precipitate2, due to increase in melting point as the concentration of BaS increases, thus inhibiting further electrolysis even if there is any remaining FeS. The solidification of the electrolyte was experimentally observed at longer durations of electrolysis or higher current densities (~ 2.83 A / cm2, 0.5 A), while the furnace power was kept the same. FIG. 50 depicts solidification of an electrolyte after 120 secs of passing 0.5 A. It can be hypothesized that these conditions lead to a depletion of FeS in the electrolyte and hence may have led to the composition of the electrolyte drifting from the eutectic point.

[0469] This is an artefact of the present experimental setup, which is not designed to be continuous. In a continuous reactor, continuous feeding of the feedstock will ensure the composition of the electrolyte is maintained throughout the electrolysis. However, further work needs to be done to quantitatively determine the minimum level of FeS which needs to be present in the electrolyte at all times to keep the electrolysis functioning continuously.

[0470] This will help determine the feeding rate of FeS. This work will also require the FeS-BaS-La2S3 ternary phase diagram to determine the eutectic composition range, which is out of scope of this work.

[0471] FIGs. 51A-51C shows the element intensity map of the bottom region of the electrolyte after electrolysis confirming that the metallic deposits seen in the optical and SEM micrographs (analyzed at the same location) are indeed concentrated in Fe.

[0472] For the S concentration, WDS analysis II was conducted, which results are presented in Table 7. The carbon concentration was not estimated using WDS due to practical difficulties. However, since the WDS analysis did not report presence of Ba or La, the carbon concentration can be estimated as 100 - (Fe wt.% +S wt.%).

[0473] Table 7: Results of WDS analysis II of the observed metallic deposit

[0474] Element Weight Concentration

[0475] Fe 94.55 wt.%

[0476] S 0.05 wt.%

[0477]

[0478] The exact composition of the metallic deposit in terms of C concentration was however difficult to perform due to the size of the deposits and difficulty in recovery.

[0479] 47

[0480] #15087612vlTo get a true qualitative estimate of the carbon, the electrolyte sample was etched to reveal the underlying micro structure and compare it with other Fe-C alloy systems. As seen in FIGs. 52A-52B, a few of the Fe deposits, in contact with a graphite cathode, show the presence of carbide-like inclusions.

[0481] From the micrographs depicted in FIGs. 53A-53D, it is clear that the microstructures revealed after etching, look similar to those of an Fe-C alloy system, specifically the pearlite phase. Observing the shape of the inclusions it can be concluded that they are iron carbides (cementite phase). Since, the ferrite phase was not observed, we can conclude that the metallic deposits are highly likely to be a high-carbon alloy, showing the pearlite phase and cementite (FesC). Each of these deposits were in contact with the graphite cathode, thus the formation of these phases occurred as a result of Fe-C equilibrium when Fe was electrolytically deposited over the graphite cathode. FIGs. 52A-52B shows the optical micrographs of the metallic Fe deposits after etching with 4% Nital solution.

[0482] Table 8, shown below, shows the results of the ICP analysis I to confirm that the mass loss in the electrolyte aligns with the depletion of sulfur before and after electrolysis. A mass loss of 0.0018 gms was observed corresponding to a difference of 0.9 wt.%. The ICP I results are thus in alignment (0.85 wt.%) with the fact that mass loss occurs only due to sulfur loss.

[0483] Table 8: Results of ICP I analysis comparing the difference in S content before and after electrolysis

[0484] Sample S content

[0485] Before electrolysis 15.65 wt.%

[0486] After electrolysis for 40 secs 14.80 wt.%

[0487]

[0488] Applications

[0489] A flowsheet of a proposed process for producing steel using MSE integrated with present steel making infrastructure is shown in FIG. 54. Adopting MSE for iron production can serve multiple advantages: elimination of direct carbon emissions, low energy consumption, and the potential to integrate well with the present BOF steel making. In comparison to other electrolytic methods, MSE exhibits a remarkable, greater than 95%

[0490] 48

[0491] #15087612vlanodic efficiency. One of the major reasons for this is the existence of Fe (+2) as the sole valence state as opposed to multiple valence states in oxides (Fe+2and Fe+3). In fact, the energy required for the sulfide reduction drops down by l / 3rd that of oxide reduction. Post sulfidation, the Fe+3in the oxide gets converted to Fe+2in the sulfide. In this process, the reduction corresponding to one electron per Fe atom occurs in an exothermic spontaneous process.

[0492] Furthermore, the high solubility of FeS (10 wt.%) in the supporting electrolyte, high current density and generation of liquid product make it advantageous as compared to aqueous techniques allowing for higher space-time yield and productivity.

[0493] Thus, MSE offers many benefits compared to other electrochemical techniques. The process thus allows molten iron production at small scale facilities where excess pyrite is available or by treating the oxide ores with sulfur. Integrating the MSE cells into the existing steel making facilities allows benefitting from the unsurpassed productivity and refining capabilities of the BOF and continuous casting.

[0494] Initial molten sulfide electrolysis experiment

[0495] Methodology

[0496] The molten sulfide electrolysis experiments were performed as described above. Initial conditions of operation included very high power of the lamps resulting in partial thermal decomposition of the FeS, resulting in very high mass loss. This mass loss was proportional to the duration of the thermal decomposition experiment, with the highest mass loss of ~ 5 mg occurring at about 40 mins (FIG. 57).

[0497] Observations and results

[0498] A cross-section of the electrolyte sample after a thermal decomposition test was analyzed under the SEM (FIG. 55). It revealed a phase segregation occurring at the bottom of the electrolyte in contact with the cathode. A region of Ba rich sulfide, depleted in both Fe and La, was observed at the bottom whereas the central portion (bulk of the electrolyte) was observed to be a mixture of the 3 sulfides. A similar phenomenon occurred upon electrolysis as well, resulting in low faradaic efficiencies, in the range of 50 - 60%. This was attributed to the Ba rich phase segregating at the bottom, preventing further Fe deposition from occurring (see FIG. 54).

[0499] Several trials of the modified compositions (using the electrolyte composition observed in the center as the starting composition assuming it is the stable phase at high

[0500] 49

[0501] #15087612vltemperature) and cathode designs were investigated. From these experiments, it was concluded that the phase segregation that was observed was a result of the heat distribution from the graphite cathode which remained hot (stored heat) for some time after the lamps were switched off. Thus, due the graphite cathode having a fairly high specific heat capacity, even after switching off the lamps, the cross-section did not represent a quenched sample.

[0502] Some phase segregation occurred later as a result of the bottom of the electrolyte still being at a high temperature than the surface or the top. This was concluded from the shape of the Ba rich region following the shape of the cathode divet, suggesting a change of cathode design being a solution. FIG. 56 shows the optical micrograph of the electrolyte crosssection showing phase segregation.

[0503] The issue with the initial design (deeper divet) was that the graphite prevented complete heat from the lamps from reaching the part of the electrolyte submerged within the cathode, leading to uneven heating of the electrolyte. This was because one part was directly exposed to the lamps getting more heat, the other submerged within getting less heat.

[0504] Similar phase segregation was not observed while making the pre-melts. The reason was uniform heating on all sides of the electrolyte (powder was inside the hole drilled in the pucker). Hence two designs were tested - one deeper divet (no electrolyte surface directly exposed to the lamps - required higher power to melt, also restricted in-situ observation due to the cathode blocking the electrolyte from being viewed. The shallow divet design showed better results and solved the phase segregation issue (see FIGs. 58A-58D).

[0505] The Fe deposits were observed in the Ba rich phase region after electrolysis. The formation of perfect spherical Fe deposits indicated the formation of pure Fe liquid. This suggested the operating temperature to be higher than the melting point of Fe.

[0506] Thus, the cathode design was modified to reduce the contact area of the electrolyte and the cathode. This was done by reducing the depth of the divet, allowing the full electrolyte to being exposed to the lamps and avoiding the higher retention of heat by the cathode. Thus, after these modifications, no phase segregation layer was observed, confirming the above theory. Temperature measurements were done in the thermal imaging furnace to reduce the lamp power and operate in the temperature range of the cast iron production. Moreover, the electrolyte was melting at a lower power (4% as compared to 18% previously) since a higher surface area (almost full electrolyte) was subjected to the radiation from the lamps and was absorbing more energy, eliminating the possibility of thermal decomposition.

[0507] 50

[0508] #15087612vlIn some embodiments, the electrolysis comprises molten sulfide electrolysis of FeS (e.g., FIG. 70), such as:

[0509] 2 FeS (1) + C (cathode) 2 Fe - x% C(l) + S2(g) -1300 °C

[0510] Anode: S

[0511]

[0512] 2’ -> S2(g) + 2e

[0513] Cathode: Fe+2+ 2e⁻ → Fe

[0514] In some embodiments, molten FeS behaves like a metallic conductor, but by itself it does not support ionic conduction. In certain embodiments, a supporting electrolyte is used. For example, in some cases, the supporting electrolyte comprises a mixture of barium sulfide and lanthanum sulfide. In certain cases, the mixture of barium sulfide and lanthanum sulfide supports ionic conduction and provides high decomposition potential as compared to FeS. For example, in some cases, 10 wt.% FeS was dissolved.

[0515] In some cases, FeS reserves are readily available, such as pyrite reserves and in the oil and gas industry (source rocks for oil). In certain instances, a sulfidation reaction is in a rotary kiln (see FIG. 71).

[0516] In certain embodiments, the method comprises selective sulfidation of ores, such as iron oxide ores and copper ores (see FIG. 72). Non-limiting examples of ores include iron oxide ores such as hematite and / or magnetite; copper ores such as chalcopyrite (CuFeS2), and / or non-ferrous oreos (Cu, Ni, Zn, Pb) which may occur as sulfides. In some cases, the sulfidation is:

[0517] • exothermic and spontaneous;

[0518] • no net energy consumption in doing 1 / 3 of the reduction;

[0519] • demonstrated for large scale production of BaS - La2S3from their respective minerals;

[0520] • not merely an anion exchange, reduction step: Fe+3(Fe2O3) + S2(g) Fe+2(FeS).

[0521] In certain instances, the method comprises using steel scrap for molten sulfide electrolysis. For example, in the schematic in FIG. 54, in some cases, steel scrap is an input for molten sulfide electrolysis. In certain embodiments, the sulfur for the rotary kiln in FIG.

[0522] 54 comprises and / or is produced from elemental reserves, natural gas, petroleum, and / or metal sulfides. In some instances, the methods disclosed herein (e.g., of molten sulfide electrolysis) could be well integrated into the present steel making infrastructure, as shown in FIG. 54. According to certain embodiments, there is no use of coke as a reductant, which results in a drastic reduction in C emissions.

[0523] 51

[0524] #15087612v1In some embodiments, the methods disclosed herein use an SM furnace (see FIG. 73).

[0525] The following examples are intended to illustrate certain embodiments of the present invention, but do not exemplify the full scope of the invention.

[0526] EXAMPLE 1

[0527] O-S-Fe-Ba-La-S was synthesized. The results are shown in FIG. 3. FIG. 1 is an E-pO2−diagram for the Ba-La-Fe-S system at 1573 K (1300 C). FIG. 2 is a reference range of stability of O-S-Fe-Ba-La-S.

[0528] Molten sulfide electrolysis was performed with the synthesized O-S-Fe-Ba-La-S as the electrolyte. As shown in FIG. 4, it was determined that it was possible to accommodate oxygen in molten sulfide electrolysis while using refractory and anode materials that were expected to react with oxygen. Between 0.01 ppm and 8% of oxygen were tolerated (weight %) using Ba-La-Fe-S (FIG. 4). The anode was stable (FIG. 5), iron was produced (FIG. 6), and sulfur was removed (FIG. 7), despite the presence of oxygen.

[0529] This example shows that oxygen could be tolerated in a method of molten sulfide electrolysis. This example shows the range of oxygen that could be tolerated in the electrolyte while using a certain anode material such that the material is inert.

[0530] EXAMPLE 2

[0531] Carbosulfidation of Barium Sulfate Assisted by Iron Sulfide.

[0532] Molten sulfide electrolysis offers an energy-efficient and low-CO2technique for metal production, and can be operated with a supporting electrolyte composed of barium sulfide and lanthanum sulfide. Barium sulfide may be produced from barium sulfate, and lanthanum sulfide from lanthanum oxide or lanthanum sulfate. In the present example, the carbosulfidation of barium sulfate in the presence of iron (II) sulfide was investigated. At 1473 K, where iron (II) sulfide was molten, barium sulfate was completely sulfidized through the formation of molten Ba-Fe-S phase. When lanthanum oxide or lanthanum sulfate was added to the reacting barium sulfate, both barium and lanthanum were sulfidized, confirming that electrolytes for molten sulfide electrolysis can be produced from sulfate and oxide raw materials.

[0533] Metals may be produced through the electrolysis of molten metal sulfides, for example, at the 1 kg per day scale. The production of metallic copper (Cu) and Fe from copper (I) sulfide

[0534] 52

[0535] #15087612vl(Cu2S) and chalcopyrite (CuFeS2) was demonstrated, along with results on the electrolytic extraction of Cu, molybdenum (Mo), and rhenium (Re) from sulfide mixtures, the electrolytic extraction of Cu and silver (Ag) from Cu2S-Ag2S, the electrolytic extraction of nickel (Ni) and cobalt (Co) from Ni3S2–CoS, and the production of tantalum (Ta) from TaS2. The electrowinning of Fe from molten sulfide was investigated. FIGs. 8A-8B show a possible route to liquid cast iron from iron oxide. In this process, iron oxide was converted to FeS with sulfur (S2) gas (FIG. 8A), and Fe metal was produced by electrolysis of FeS contained in a molten sulfide electrolyte (FIG. 8B). The advantages of this route include, in accordance with some embodiments,: (i) a carbon-based anode inert to electrochemical oxidation reactions of sulfide ions; (ii) a carbon-containing cathode that supports the production of hot metal, an Fe-C melt with a lower melting point than pure Fe, a product suitable for existing converters for liquid state steelmaking.

[0536] In general, many metal sulfides like FeS exhibit electronic conductivity in the molten state. The electrolysis of molten sulfides with high electronic conductivity may result in low current efficiency, yet the presence of highly ionic BaS in the electrolyte mitigates this issue during molten sulfide electrolysis. Addition of La2S3enhanced the ionic conductivity of the electrolyte, enabling electrolysis with high current efficiency. While the supporting electrolyte components BaS and La2S3were not consumed by electrolytic reactions, they can slowly get contaminated by oxygen if exposed to air. BaS, the main component of the electrolyte, does not exist naturally and can be synthesized from Ba compounds. In some cases, it may be desirable to produce BaS from BaSO4, which is a material available on a large scale. The present disclosure investigated a method to produce BaS by sulfidizing BaSO4 with sulfur (S2) gas in the presence of carbon. Specifically, to manufacture electrolytes for electrolytic steelmaking, sulfidation experiments were conducted by adding FeS — which is an electrolyte component — to BaSO4, and the effect of FeS addition on the sulfidation reaction was investigated.

[0537] Review of barium sulfide production from barium sulfate

[0538] BaS may be industrially produced via the carbothermic reduction reaction of BaSO4 at 1273-1473 K, as shown in equation (61):

[0539] BaSO4(s) + 2 C (5) = BaS (5) + 2 CO2(g) (61) ΔG°r,1= − 296.6 kJ·mol−1at 1473 K (62)

[0540] 53

[0541] #15087612v1Here, ΔG°r,nrepresents the standard Gibbs energy change for reaction ( / ?). The thermodynamic data used was obtained from the FactPS database in FactSage 8.0, unless otherwise noted.

[0542] In practice, the reduction reaction by CO gas, shown in equation (63), can proceed concurrently:

[0543] BaSO4(5) + 4 CO (g) = BaS (5) + 4 CO2(g) (63) ΔG°r,3= − 124.1 kJ·mol−1at 1473 K (64)

[0544] Reactions described in equations (61) and (63) may be utilized industrially to produce high-purity BaSO4from crude BaSO4containing impurities like silica and iron oxide. During the process, BaS is produced as an intermediate. The produced BaS may be dissolved in an aqueous solution to separate impurities and residual carbon materials from BaS. A method for producing BaS through the hydrogen reduction of BaSO4is as shown in equation (65):

[0545] BaSO4(5) + 4 H2(g) = BaS (5) + 4 H2O (g) (65) ΔG°r,5= − 167.7 kJ·mol−1at 1473 K (66)

[0546] This reaction proceeds rapidly at relatively low temperature and generates water as a byproduct. In the presence of a Ni catalyst, the reaction proceeds swiftly at temperatures of 973-1123 K. However, hydrogen production requires significant energy consumption and incurs high costs. A method for producing BaS through the reduction of BaSO4using sulfur gas is shown in equation (67):

[0547] BaSO4(5) + S2(g) = BaS (5) + 2 SO2(g) (67) ΔG°r,7= − 12.2 kJ·mol−1at 1473 K (68)

[0548] In this reaction, inexpensive sulfur can be used as a reducing agent. However, reaction (67) has a small thermodynamic driving force, and achieving a high sulfide conversion rate requires higher temperatures (around 1600 K) than carbothermic or hydrogen reduction. In this method, reactions (61), (63), and (67) proceed simultaneously. Additionally, the reaction between carbon and SO2gas, shown in equation (69), helps maintain low partial pressures of

[0549] 54

[0550] #15087612v1SO2and oxygen in the system, thereby accelerating reaction (67):

[0551] 2 SO2(g) + 2 C ( ) = S2(g) + 2 CO2(g) (69) ΔG°r,9= − 284.4 kJ·mol−1at 1473 K (70)

[0552] In the present disclosure, this process is referred to as the “carbosulfidation” reaction, drawing an analogy to the carbochlorination reaction of oxides by chlorine gas in the presence of carbon. While the overall reaction combining reactions (67) and (69) is represented by reaction (61), both reactions (67) and (69) occur between gas-solid phases, suggesting a potential improvement in reaction rate compared to solid-solid reaction (61). BaSO4sintered at 1273 K and 1473 K has been observed to inhibit the progression of the sulfidation reaction inside the sinter. When high-melting-point La2O3powder was added into BaSO4, the conversion rate to sulfides exceeded 90% at 1373 K.

[0553] Thus, various methods for BaS production have been explored. The present disclosure aimed to establish a BaS production process suited for molten sulfide electrolysis. When targeting steel production via molten sulfide electrolysis, iron sulfides (FeS or FeS2) can be added into the BaS production process. As described herein, carbosulfidation experiments of BaSO4with the addition of FeS were conducted to examine its effects. At temperatures above the melting point of FeS (1462 K), molten FeS may have the following impacts on the sulfidation of BaSO4, potentially promoting the reaction progress (FIG. 9):

[0554] (a) molten FeS may infiltrate between BaSO4particles, suppressing the sintering of BaSO4; (b) molten FeS, which has a high mobility for sulfur and can accommodate more than its stoichiometry, may deliver sulfur to the surface of BaSO4;

[0555] (c) produced BaS may dissolve into the FeS melt, preventing the coating of unreacted BaSO4; (d) the dissolution of BaS into the FeS melt or formation of complex Ba-Fe sulfides may decrease the BaS activity in the system, thereby raising the driving force of the sulfidation reaction.

[0556] In this work, carbosulfidation experiments were performed by adding FeS to BaSO4 to investigate the influence of FeS on BaS production.

[0557] Thermodynamic background for sulfidation of barium sulfate Ba-O-S system

[0558] In the following, a thermodynamic analysis of the sulfidation reaction of BaSO4was conducted. FIG. 10A shows the chemical potential diagram of the Ba-O-S system at 1473 K.

[0559] 55

[0560] #15087612vlAs shown in FIG. 10A, the reduction reactions of BaSC to BaS by carbon, CO gas, and hydrogen gas proceeded from a thermodynamic perspective. When using sulfur gas, BaS was produced from BaSO4 according to reaction (67). From the positions of the S2 / SO2 equilibrium point (point A, ps2 = 1 atm and pso2 = 1 atm) and the BaS stability region in FIG. 10A, as well as the value of AG°r,7, the driving force of reaction (67) was however small. Lower partial pressure of SO2 (pso2 (atm)) was needed to increase the driving force for the reaction.

[0561] When carbon is present in the system, SO2 gas may be reduced to S2 gas according to reaction (69), allowing the oxygen partial pressure (po2(atm)) and pso2 in the system to decrease to values corresponding to the C / CO equilibrium (point B, C / CO / S2 equilibrium). In this way, in sulfidation reactions in the presence of carbon, not only did the reactions between BaSO4 and carbon, CO gas, and sulfur gas proceed in parallel, but the driving force of the sulfidation reaction by sulfur gas also increased. It should be noted that in the coexistence of carbon and sulfur, gases such as CS2 and COS can be formed; however, their thermodynamic stability is low, and their formation is kinetically limited. Due to the small formation rates of these gases, they were neglected in this analysis.

[0562] FIG. 10B shows the chemical potential diagram of the Ba-Fe-O-S system at 1473 K. There exists a region where both BaSO4 and FeS were stable. Therefore, if the pO2–pS2potential in the system was within that region, BaSO4 and FeS did not react with each other and could coexist. In the case of carbosulfidation, the potential in the system was fixed at point B (C / CO / S2 equilibrium). While BaSO4 was converted to BaS, FeS remained stable. There was a possibility of forming sulfide solid / liquid solutions and compounds in the Ba-Fe-S system. At temperatures above the melting point of FeS (1462 K), it was believed that BaS might dissolve into the FeS melt. This was also observed in the phase diagrams of MgS-FeS and CaS-FeS in the alkaline-earth metal systems. Following the MgS-FeS and CaS-FeS systems, the saturated phase of the FeS-rich melt in the BaS-FeS system could be BaS. The ternary compounds (BaxFeySz) may also exist at high temperature. The dissolution of BaS into the FeS melt lowered the BaS activity, making the BaS formation reaction more favorable thermodynamically, and also prevented unreacted BaSO4 from being coated by the solid BaS product.

[0563] In the present example, the addition of raw materials for La2S3was considered simultaneously with FeS. FIG. 10C shows the chemical potential diagram of the La-Fe-O-S system at 1473 K. La2O3or La2(SO4)3can be used as feedstocks to produce La2S3. The carbosulfidation reaction of La20s is shown in equation (71):

[0564] 56

[0565] #15087612vlLa2O3(s) + 3 C (5) + 3 / 2 S2(g) = La2S3(5) + 3 CO (g) (71) AG°r,ii = - 347.1 kJ-mol-1at 1473 K (72)

[0566] From a thermodynamic standpoint, La2O3and La2(SO4)3do not thermodynamically coexist with FeS at 1473 K. When these La compounds and FeS are placed in the same system, they can be converted to La2O2S, La2O2SO4, Fe3O4, and / or FeO. On the other hand, under the carbosulfidation condition at point B (C / CO / S2equilibrium), Ba, La, and Fe are all thermodynamically stable as sulfides. Therefore, it was believed that under the condition at point B, BaS–La2S3–FeS electrolytes for molten sulfide electrolysis could be produced from mixtures of BaSO4, La2O3or La2(SO4)3, and FeS.

[0567] FIG. 11 shows the sulfidation apparatus, and FIG. 12A presents an example photograph of the samples. Additionally, Table 9 lists the experimental conditions, and Table 10 provides the sample information. The sulfidation samples were prepared by mixing BaSO4 powder (99%, ALFA AESAR®), La2O3powder (99.9%, ACROS ORGANICS™) or La2(SO4)3powder (99.9%, THERMO FISHER SCIENTIFIC®), and FeS powder (99%, THERMO FISHER SCIENTIFIC®) using an agate mortar and pestle. The mixed sulfidation samples were placed, without being compacted, on PARAFILM® self-sealing thermoplastic film in graphite crucibles (O. D. 19 mm, H. 25 mm, wall thickness 2-3 mm) made in-house from graphite rods, and three crucibles containing a sample each were set into a graphite tray (O. D. 51 mm, H. 29 mm, wall thickness 3 mm). Holes (01.6 mm) were drilled at intervals of 1-2 mm at the bottom of these crucibles and the tray to allow reaction gases to flow from the bottom of the reaction tube through the samples. The graphite tray with the samples was placed on a mullite supporting tube (O. D. 51 mm, I. D. 44 mm, L. 290 mm) and installed inside an alumina reaction tube (O. D. 64 mm, I. D. 57 mm, L. 610 mm) so that the samples were positioned in the hot zone (1473 K) of the vertical electric furnace. At the bottom of the mullite supporting tube, a quartz crucible (O. D. 28 mm, H. 100 mm, wall thickness 1.5 mm) containing sulfur (99%, SIGMA-ALDRICH®), which was premelted using a heat gun and solidified in air, was placed. Argon (Ar) gas (99.999%) was supplied through a gas inlet tube inserted from the bottom of the furnace. After the furnace was heated to a temperature of 1473 K over 4 hours in an Ar atmosphere, the Ar gas inlet tube was lifted by 10 mm towards the center of the electric furnace every 5 minutes, pushing up the quartz crucible containing sulfur. This continuously moved the sulfur into the high-temperature region of the furnace, where the boiled sulfur gas was

[0568] 57

[0569] #15087612v1transported by the Ar carrier gas and supplied to the samples located in the hot zone at 1473 K. Based on the lifting speed of the quartz crucible, the sulfur evaporation rate was estimated to be 1.8 g-min-1. After the predetermined reaction time, the lift-up of the quartz crucible was stopped, and after maintaining the temperature at 1473 K for about 20 minutes, the system was cooled to room temperature over 4 hours under an Ar atmosphere.

[0570] Table 9. Experimental Conditions for Carbosulfidation of BaSO4in the Presence of FeS.

[0571] Exp. No. Temp., Time, t Ar Flow Rate (Std.), Weight of Sulfur T / K I min f Ar / cm3· min−1Evaporated Sulfur, wS / g Partial Pressure, pS2 / atma04 1473 41 740 74 0.5 ± 0.1 05 1473 41 750 74 0.5 ± 0.1 06 1483 44 470 79 0.6 ± 0.1 07 1483 33 470 59 0.6 ± 0.1

[0572]

[0573] 08 1473 41 470 74 0.6 ± 0.1aThe sulfur partial pressure, ps2 (atm), is calculated with the following equations:

[0574] S2

[0575] [formula as shown in image]

[0576] [formula as shown in image]

[0577] nS2(mol·min−1) = [formula] (C).

[0578]

[0579] Table 10. Sample Information for the Carbosulfidation Experiments.

[0580] Weight of sample / g | Initial concentration, C₀,ᵢ (Wt Pct) | Exp. No. | Sample No. | Initial, Wini / g | After, w / g | BaSO₄ | La₂O₃ | La₂(SO₄)₃ | FeS

[0581] 04 04_l 1.0050 0.7255 93.9 6.1 04_2 0.9960 0.7391 88.5 11.5 05 05_l 1.0031 0.8295a68.8 26.7 4.5 05_2 1.0028 0.8315a65.9 25.5 8.6 06 06_l 1.0031 0.7167 93.9 6.1 06_2 1.0006 0.8056 68.8 26.7 4.5 06_3 0.9972 0.8183 65.9 25.5 8.6 07 07_l 0.9992 0.7392 93.9 6.1 07_2 1.0019 0.8006 68.8 26.7 4.5 07_3 1.0030 0.8142 65.9 25.5 8.6 08 08_l 1.0003 0.7897 65.9 25.5 8.6 08_2 1.0009 0.6995 57.5 38.7 3.7

[0582]

[0583] 08_3 1.0006 0.6984 55.5 37.3 7.2aThis sample was put in a graphite crucible that had been used in Exp. No. 04, which led to carbon contamination and resulting weight gain.

[0584] 58

[0585] #15087612vlA portion of the samples was crushed into powder using an agate mortar and pestle, and phase identification of the powder samples was performed using X-ray diffraction (PANALYTICAL® X'Pert MPD diffractometer, MALVERN PANALYTICAL®), scanned over a range of 10 to 90 degrees using Cu-Ka radiation at an X-ray power of 45 kV and 40 mA. Parts of the sample lumps were embedded in epoxy resin, polished with SiC abrasive papers of #240, #320, #400, and #600 grits, and then buff-polished with diamond particles of 6 μm and 1 μm. The polished surfaces were observed using an optical microscope (OM, GX51, OLYMPUS® Corporation) and a scanning electron microscope (SEM, JEOL JSM-6610LV, JEOL®), and compositional analysis was conducted using the energy-dispersive X-ray spectroscopy (EDS, Sirius SD detector, SGX SENSORTECH®). Furthermore, quantitative elemental analysis of the surfaces was performed using an SEM (JEOL JXA-8200 Superprobe, JEOL®) equipped with wavelength-dispersive X-ray spectroscopy (WDS). Total oxygen and sulfur concentrations in some of the samples were quantified using the combustion method (LECO), carried out by a professional analytical company (APPLIED TECHNICAL SERVICES™).

[0586] FIGs. 12B and 12C are example photographs of the samples; the powdered samples shrunk and became lumpy. The weights of the samples after the experiments are shown in Table 10. The decrease in sample weight may suggest that oxygen in the BaSO4 was removed into the gas phase according to reactions (61), (63), and (67). The weight changes of the graphite tray and crucibles were small, suggesting the generation of COS and CS2 gases was limited, as discussed in the thermodynamic consideration section.

[0587] FIG. 13 shows the XRD spectra of the powder samples obtained after crushing the samples. Strong peaks for BaS suggested that most of the BaSO4 was converted to BaS. Small peaks corresponding to a ternary compound Ba2FeS3 were also observed, and these peaks increased in magnitude with larger amounts of FeS added.

[0588] FIGs. 14A-14E show cross-sectional OM and SEM images of the sample prepared by adding 6.1 wt% of FeS to BaSO4 to achieve a Ba / Fe mole ratio of 85 / 15 (Sample no. 04_l). Primarily, granular phases of several ten micrometers in size and other phases surrounding them were observed throughout the entire area. FIG. 14D presents an enlarged SEM image inside the granular phase. According to EDS and WDS analyses, the granular phase was a BaS phase containing little Fe. No unreacted BaSO4 was found inside the BaS phase. This was consistent with the XRD results. FIG. 14E shows an SEM image of the phase surrounding the BaS phase (Sample no. 04_l). According to EDS and WDS analyses, this phase was a Ba-Fe

[0589] 59

[0590] #15087612vlS ternary phase. Given its network-like morphology, characterized by the absence of gaps surrounding the BaS phase, the ternary phase likely formed by the dissolution of BaS produced through sulfidation into the FeS melt.

[0591] FIGs. 15A-15B plot the WDS analysis results on the Ba-Fe-O-S compositional tetrahedral diagram and the Ba-Fe-(S+O) compositional triangular diagram. The contrast differences in the backscattered electron images and the WDS results indicated that the Ba-Fe-S ternary melt comprised of phases with compositions close to Ba2FeSs and BayFeeSw. The presence of the Ba2FeS3matched the XRD results. The melting point of Ba2FeS3was above 1573 K. No liquid phase existed in the Ba-Fe-S system at 1173 K and 973 K. It was considered that this phase existed as a solid at the temperature of 1473 K or crystallized from the sulfide melt during cooling. On the other hand, the BayFeeSu phase identified by WDS analysis was not observed in the XRD results. As shown in FIGs. 15A-15B, when 11.5 wt% of FeS was added to BaSO4 to achieve a Ba / Fe mole ratio of 74 / 26 (Sample no. 04_2), BaS phases and Ba-Fe-S phases were observed. Increasing the amount of FeS did not increase the Fe concentration in the BaS phase. The Ba-Fe-S phase appeared as a solidified liquid phase in the SEM images with formation of Ba2FeS3and additional phases with high Fe and high oxygen concentrations. Herein, it was found that increasing the amount of FeS resulted in residual oxygen in the Fe-containing liquid phase, indicating incomplete sulfidation.

[0592] Additionally, as shown in FIG. 16, unreacted BaSO4 remained in the sample for the shortest sulfidation duration (Sample no. 07_l), confirming that the sulfidation of BaSO4 was incomplete in such condition.

[0593] Table 11. Oxygen and Sulfur Concentrations of the Samples Before and After the Experiments

[0594] Weight of sample Oxygen Sulfur Oxygen Concentration Concentration Removal Rate Exp. Sample Initial, After, w Initial, After, Initial, After, Ri Ri No. No. Wini / g / g Co, ini Co Cs,ini Cs (Pct)c(Pct)d(Wt Pct) (Wt (Wt Pct) (Wt

[0595] Pct)aPct)b

[0596] 04 04_l 1.0050 0.7255 25.7 0.05 15.1 21.5 99.9 99.8 04_2 0.9960 0.7391 24.3 0.62 16.4 21.4 98.1 97.4 05 05_l 1.0031 0.8295 22.8 0.03 11.1 22.9 99.9 99.9

[0597]

[0598] 08 08_2 1.0009 0.6995 28.9 0.10 15.9 16.4 99.8 99.7 " Estimated error is ± 0.01 wt pct.

[0599] 60

[0600] #15087612vlbResults were obtained by extrapolation of the calibration curve. Estimated error is ± 0.9 wt pct.

[0601] cRi = (wini x Co, ini - w x Co) / (wini x Co, ini) x 100; this equation may overestimate the oxygen removal rate due to the loss of samples before and during the experiments.

[0602] dR2= (Co,ini—Co) / Co, ini x 100, this equation may underestimate the oxygen removal rate, and the actual value is expected to be between R\ and R2.

[0603] The EDS / WDS analyses did not provide reliable quantitative measurement of oxygen concentrations in the samples. Instead, Table 11 shows the oxygen and sulfur concentrations in the samples determined by the combustion method. The oxygen concentration in Sample no.

[0604] 04_l was extremely low, less than 0.1 wt%, suggesting that the sulfidation of BaSCU proceeded almost completely. Since no BaSCU phases were observed in Sample no. 04_l during the SEM observation, it is possible that the oxygen in the sample was either dissolved in the sulfide phases or picked up through moisture absorption or hydrolysis. The oxygen concentration in Sample no. 04_2 (11.5 wt% FeS added) was higher than that in Sample no. 04_l (6.1 wt% FeS added). The higher FeS concentration resulted in more oxygen remaining in sulfides, and this result agreed with the qualitative trends seen with EDS / WDS.

[0605] The results demonstrated that the carbosulfidation of BaSCk in the presence of FeS at 1473 K can effectively produce BaS without residual BaSCE.

[0606] Sulfidizing the mixed powders of BaSCk and La20s in the presence of FeS led to shrinkage and consolidation of the samples in the form of a lump. The final volume of the samples was smaller than that in absence of Ea2O3. FIGs. 17A-17B show the XRD spectra of the powder samples. After sulfidation, high peaks attributed to BaS were observed, and no peaks of BaSCU remained. The peaks for Ea20s and Ea2O2S, which is an intermediate product during sulfidation, were not observed, suggesting that most of the Ea20s were also converted to sulfides. Regardless of the amount of FeS added, peaks for the BaEa2S4 and BaEa2FeS5 phases were observed. The sample with a high amount of FeS added (Sample no. 06_3) showed the peaks characteristic of a Ba2FeSs phase.

[0607] FIGs. 18A-18B show cross-sectional SEM images of the sample prepared by adding 4.5 wt% of FeS to BaSC>4-La2O3 (Sample no. 06_2). The results of EDS analysis are reported on the BaS-Ea2S3-FeS compositional triangle diagram shown in FIG. 19A. Similarly to the BaSO4-FeS system, granular BaS phases were observed in the BaSO4-Ea2O3-FeS system after sulfidation. Whereas Fe was not dissolved in the BaS phase, Fa was dissolved at a couple of weight percents. Like the BaSO4-FeS system, other phases surrounding the BaS phases were formed as shown in FIGs. 18A-18B. These phases were either the BaLa2S4 phase, which

[0608] 61

[0609] #15087612vlcontained little Fe, and quaternary Ba-La-Fe-S phases. The BaLa2S4 phase was a solid solution phase with a compositional range, and the phases observed in this work had compositions slightly BaS-rich compared to the stoichiometric composition of BaLa2S4. The quaternary Ba-La-Fe-S phases showed various compositions, as shown in FIG. 19A. Although it was unclear whether these phases were solid or molten at the temperature of 1473 K, phases with high FeS concentrations may have been melted due to low melting point of FeS.

[0610] FIGs. 18C and 18D show the cross-sectional SEM images of the sample prepared by adding 8.6 wt% of FeS to BaSC>4-La2O3 (Sample no. 06_3). The results of EDS analysis are plotted on the compositional triangle diagram in FIG. 19B. The compositions of the observed phases were similar to those found with a smaller amount of FeS added. On the other hand, unlike the BaSO4-La2O3 with small FeS addition, the Ba2FeS3 phase was observed. The Ba2FeS3 phase was likely stabilized by the high FeS content. It was considered that this phase existed as a solid at the temperature of 1473 K, or had crystallized from the sulfide melt during cooling.

[0611] Phases with high concentrations of La, oxygen, and sulfur were observed shown in FIGs. 19A-19B and FIGs. 20A-20G, in a sample with a relatively low amount of FeS added and sulfidized for a shorter time (Sample no. 07_2), or in samples with a large amount of FeS added (Sample no. 05_2, 07_3, and 08_l). These phases were attributed to lanthanum oxysulfide, likely La2O2S, an intermediate product in the sulfidation process of La2O3, with some Ba and Fe also present at several weight percents. The lanthanum oxysulfide phases were distributed unevenly within the samples as shown in FIGs. 20A-20G. It was proposed that lanthanum oxysulfide crystallized from the sulfide melts containing oxygen during cooling. Thus, when the sulfidation time was short or when a large amount of FeS was added, the progress of sulfidation was incomplete, and oxygen remained in the sulfide melt, supporting the formation of lanthanum oxysulfide phases.

[0612] The XRD results of the powdered samples obtained with La2(SO4)3 and FeS added to BaSO4 are shown in FIGs. 21A-21B. Similarly to the BaSO4-La2O3 system, the BaS, BaLa2S4, and BaLa2FeS5 phases were observed. When a large amount of FeS was added (Sample no.

[0613] 08_3), the Ba2FeS3 phase was also observed. FIGs. 22A-22D show the cross-sectional SEM images, and FIG. 23 shows the EDS analysis results. When a small amount of FeS was added (Sample no. 08_2), BaS, BaLa2S4, and Ba-La-Fe-S phases were observed. On the other hand, when a large amount of FeS was added (Sample no. 08_3), the Ba2FeS3 phase and lanthanum oxysulfide phase were observed. These results were similar to the one obtained with La2O3

[0614] 62

[0615] #15087612vladdition; a large amount of FeS addition lead to incomplete sulfidation.

[0616] Table 11 shows the oxygen and sulfur concentrations by the combustion method for Sample no. 05_l and Sample no. 08_2. These samples showed low oxygen concentrations of less than 0.1 wt%, indicating that the co-sulfidation of BaSCU and La2Ch or La2(SO4)3 mixtures proceeded effectively, and that a BaS-La2S3-FeS electrolyte could be directly produced from the raw materials. Based on the results, as shown in FIG. 24, the electrolyte production process investigated in this work can be incorporated into the molten sulfide electrolysis process for Fe. In the process flow of FIG. 24, a portion of the FeS produced by sulfidation of iron ore (main component: Fe2O3) was added to the sulfidation process of BaSO4 and La compounds to produce the electrolyte for molten sulfide electrolysis. The electrolytes produced under appropriate conditions had low oxygen content.

[0617] In the context of developing an ironmaking process through molten sulfide electrolysis, the effect of FeS addition on the carbosulfidation reactions of BaSO4, BaSO4-La2O3, and BaSO4-La2(SO4)3 at 1473 K for the electrolyte synthesis was investigated. It was demonstrated that in the presence of a small amount of FeS, the coarsening and sintering of BaSO4 was inhibited, enabling effective conversion to BaS. Conversely, with a large amount of FeS added in a system, residual phases with high oxygen concentrations were observed. Under sufficiently long sulfidation conditions and with a small amount of FeS added, it was revealed that BaS-La2S3-FeS electrolytes could be directly produced from BaSO4-La2O3-FeS and BaSO4-La2(SO4)3-FeS. On the other hand, under conditions with a high amount of FeS added, it was found that oxygen remained in the samples in the form of lanthanum oxysulfide.

[0618] EXAMPLE 3

[0619] Iron Production by Molten Sulfide Electrolysis

[0620] This work investigated electrolytic production of molten iron using a sulfide route, molten sulfide electrolysis (MSE). The electrolytic decomposition of iron sulfide into iron and elemental sulfur gas in a molten sulfide electrolyte was demonstrated in a two-electrode configuration, at a temperature sufficient to produce iron with about 4wt% C. The proposed sulfide route supports a virtual elimination of GHG emissions from the reduction step. The absence of trivalent iron species (Fe3+) supports the reduction of divalent iron (Fe2+) and indicates possible energy savings of about 50% for iron production by electrolysis, compared to other oxide-based routes. Results for electrolysis are presented herein to verify such findings and discuss the attributes of MSE for iron production in the context of integrated steelmaking.

[0621] 63

[0622] #15087612vlSteelmaking through the Blast Furnace-Basic Oxygen Furnace (BF-BOF) is energy and emission intensive (-1.85 tonnes of CO2 is released per tonne of steel), given its use of carbon both as a fuel and reductant. As of 2022, the annual global steel production was 1.95 billion tonnes, contributing to -9% of global GHG emissions. As nearly 40% of the world’s electricity is generated from low-carbon sources, using electrical energy for primary metal production can decouple metal extraction from emissions.

[0623] Electrochemical reduction of sulfides

[0624] Technology involving the electrochemical reduction of sulfides may offer advantages compared to processing ores as oxides. The electrolytic decomposition of pyrite (FeS) to molten iron and sulfur gas was investigated in a molten sulfide electrolyte, as shown in reaction 73 below.

[0625] FeS (l) → Fe (l) + ½ S2(g) (73)

[0626]

[0627] Materials and Methods

[0628] This example investigated the feasibility of electrochemically decomposing FeS in the supporting electrolyte into Fe and S2, and determined process conditions to perform a mass and energy balance of a possible process. A series of electrolysis experiments were performed. First, to demonstrate that iron oxide (hematite, Fe20s) could be converted to FeS as feedstock to electrolysis, sulfidation was performed, as described below.

[0629] Sulfidation

[0630] As a proof of concept of preparation of the feedstock for MSE, pure lab grade Fe20s (99.85+% metal basis, ALFA AESAR®) was heated in a sulfur-rich atmosphere in a ‘sulfidation’ process using the Split Mellen vertical tube furnace (MELLEN™, PS400-120-20CLT-C2778-R-OT). 5g of pure, lab-grade Fe2O3 (99.85+ % metal basis, ALFA AESAR®) was crushed using an agate mortar and pestle and subsequently sieved to obtain a particle size between +90 microns and 106 microns. This crushed iron (III) oxide powder was filled in an in-house machined alumina crucible (50 mm OD, 41 mm ID, 34 mm depth, machinable alumina) up to a particle bed height of 4 mm. The custom-made alumina crucible had approximately 100 evenly spaced holes drilled at the bottom (1.5 mm OD) as well as about ten holes in the sides (6.3 mm OD) for the sulfur gas to react across the entire reactor bed height. This machined alumina crucible was soaked in deionized water and dried at 200°C in a vacuum

[0631] 64

[0632] #15087612vloven overnight before use. Using an alumina support tube (29 cm length, 5.08cm OD, 4.445 cm ID), the iron (III) oxide containing crucible was positioned in the hot zone of the vertical tube furnace, in an alumina tube (600 mm length, 25mm OD, 21mm ID, ADVALUE™). Approximately 100 grams of elemental sulfur powder (99.5%, sublimed, ACROS ORGANICS™) were carefully melted using a heat gun in a quartz crucible (28 mm OD, 101.6 mm depth). This sulfur-containing crucible was loaded through the bottom of the furnace. The furnace was heated to set temperature of 1000°C at a rate of 3°C / minute, held at 1000°C for 45 minutes, and cooled to room temperature at a rate of 3°C / minute. Throughout the sulfidation, argon (AIRGAS®, ultra- high purity) flowed at a constant rate of 1500 seem through the chamber, as a carrier gas for S2 and SO2(g). The critical gas flow rate was calculated by performing mass balance calculations. At the maximum hold temperature, to ensure the sulfur partial pressure in the chamber was maintained at absolute pressure of 0.1 atm, the sulfur containing crucible was raised at 0.5 cm per minute using a stainless-steel tube (6.25 mm OD), also used for the argon inlet, introduced at the bottom of the furnace port. The unreacted sulfur particles were captured in an in-house constructed gravity separator, with a 1” diameter quartz tube used for the outgas line.

[0633] Molten sulfide electrolysis

[0634] The decomposition potentials of the electrolyte components, BaS and La2S3, were 1.7 and 1.9V, remote from that of iron sulfide at 1350°C (-0.4V), such that BaS-La2S3 was a supporting electrolyte.

[0635] The preparation of the materials for electrolysis are described in the following subsections: electrolyte synthesis, and fabrication of the electrodes. To evaluate possible mass loss due to the spontaneous thermal decomposition or evaporative losses for such electrolyte, thermal decomposition trials were performed. The equipment and procedure for galvanostatic electrolysis are described followed by the techniques for observation and characterization.

[0636] Electrolyte synthesis

[0637] Electrolyte samples were obtained from laboratory grade chemicals by mixing 55.8 wt.% BaS (99.7% metals basis, ALFA AESAR®), 34.2 wt.% La2S3(99.0% metals basis, ALFA AESAR®) and 10 wt.% FeS (99.9 % metal basis, ALFA AESAR®). The electrolyte components were ground using a mortar and pestle in a controlled atmosphere inside the glove box. Portions (weighing 0.2 g) of the electrolyte powder mixture were loaded into 7 mm OD holes machined

[0638] 65

[0639] #15087612vlin graphite circular puck (53 mm OD, GRAPHITESTORE™). These droplets were pre-melted in a vertical tube furnace (MELLEN™, PS400-120-20CLT-C2778-R-OT), at a set temperature of 1350°C to form a consolidated solid of the powder mixture for ease of handling. Argon (AIRGAS®, ultra-high purity) flowed through the tube at a rate of 450 seem. The pre-melt was analyzed using LECO and ICP to quantify the concentration of C, O and Fe and S. No significant oxygen or carbon contamination in the electrolyte sample was observed. No compound formation was observed in the pre-melt during SEM characterization. At least two phases could be observed: Ba-rich (dark colored) and La-rich (light colored). Mass loss during pre-melting was found to be about 0.001 g (<1%).

[0640] Fabrication of the electrodes

[0641] Graphite is stable and inert towards the sulfur gas released at the anode during electrolysis of molten sulfides. Additionally, it is stable towards molten sulfides at temperatures greater than 1200°C, as no spontaneous carbothermic reduction of the sulfides has been observed. Hence, the electrodes were machined using EDM-quality graphite rods (fine extruded, 6.3mm OD, 305mm length, GRAPHITESTORE™). The height of the cathode stands was 12 mm with a hemispherical depression (“divet”) at the top to hold the electrolyte droplet. The cathodic surface area of approximately 0.18 cm (6 mm ID, 0.7 mm depth) was machined using a dremel. The anode was machined to taper into a fine tip at one end using a belt sander. The height of the anode in total was 16 mm, with the tip (1.3 mm OD) being 11 mm long.

[0642] To hold and provide electrical connection to the electrodes, a molybdenum rod (>99.97%, 3.2mm OD, 600mm length, Ed Fagan) sheathed in an alumina tube (>99.8%, 6.35mm OD, 4mm ID, COORSTEK®) was threaded onto the opposite end of the electrode. Before use, the graphite electrodes were sonicated for 15 mins using ethanol and were air dried. The cathode, anode, and electrolyte were individually weighed to milligram precision before each trial.

[0643] Thermal decomposition trials

[0644] A systematic study to determine the mass loss due to thermal decomposition was carried out by conducting several thermal decomposition trials in the thermal imaging furnace (described below) at varying time lengths in the absence of electrolysis were conducted to determine the mass loss due to thermal decomposition. This was performed to decouple the mass loss due to thermal decomposition, if any, from the mass loss due to electrolysis, for similar duration. The electrolyte was heated at 4 % power for 5, 10, 20, 30, 40 and 60 minutes

[0645] 66

[0646] #15087612vlto determine a trend in the mass loss with time, if any. The individual masses of the anode, cathode and the electrolyte were carefully measured to milligram precision. To determine if any metal had reduced due to thermal decomposition, a cross-section of the electrolyte was characterized under the optical and scanning electron microscope.

[0647] Across all samples, mass loss was minimal, confirming the thermal stability of the electrolyte. This also implied that all the mass loss measured during experiments with electrical inputs was of electrochemical origin. The absence of significant mass loss of the electrodes observed in the thermal-only trials confirmed the following: (1) negligible oxygen contamination in the furnace atmosphere, which would have resulted in mass loss of the graphite electrodes due to reaction with oxygen resulting in CO / CO2 and (2) negligible solubility of carbon in the electrolyte solution.

[0648] Electrolytic reduction on a graphite cathode

[0649] The electrochemical experiments were performed in a Thermal Imaging furnace (TIF, TX-12000-I-MIT-VPO-PC, CRYSTAL SYSTEMS CORP.™), a 12kW lamps-based furnace, powered by four 3 kW Xenon lamps. Upon illumination, these with the help of ellipsoidal mirrors, created a hot zone of one cubic centimeter volume at the focus, able to heat materials to temperatures up to 3000°C. An external quartz tube (customized, TECHNICAL GLASS PRODUCTS®, Inc.) sealed with Viton O-rings was used to maintain an inert atmosphere. The electrode probes were introduced from the top and bottom ports (sealed using ULTRA-TORR™ fittings) securely holding the anode and cathode, respectively. Cameras installed on the front and side panels of the TIF allowed in-situ visual observations such as bubbling due to gas formation during electrochemical reactions. These experiments were carried out in a controlled atmosphere with argon (AIRGAS®, ultra-high purity) flowing at 200 seem for the entire duration. To avoid any minor oxygen contamination, a gettering furnace (OXYGON® Industries, Model OG-120), was used during the experiments after the set up evacuated to a negative pressure of 10’3atm and was held overnight. To precisely position the electrodes in the hot zone of the furnace (monitored using a camera - EOS Rebel T5i DSLR, CANON® Inc.), the probes were controlled by stepper motors. Once oxygen was scrubbed from the system and 200 seem flow of argon gas began, the lamps were switched on and were powered to 4% power. The electrolyte was heated for about 2-3 minutes until a stable molten droplet was formed, by rotating the bottom probe comprising of the cathode constantly at a rate of 10 rpm. FIG. 59 shows the schematic of the assembly ready for electrolysis. The rotation of the bottom electrode was stopped, and the tip of the anode was moved down to contact the molten 67

[0650] #15087612vldroplet. Several runs, each with a new droplet, of galvanostatic electrolysis at a cathode current density of -0.85 A / cm2were conducted by applying a current of 0.15 A, for varying time lengths of 60, 90 and 120 seconds using a GAMRY® Reference 3000 Potentiostat. Once a run was complete, the lamps were switched off to cool down the sample at a rate > 100°C / second. Similar experiments were conducted at a different current density (1.7 A / cm2) by increasing the current to 0.3A for time lengths of 20, 40, 60 and 80 seconds. The current values were chosen such that the applied current density was similar to the aluminum industry standards (around 1 A / cm2), as well as 2 A / cm2to test if the system was stable at higher current density conditions.

[0651] Apart from the electrolysis experiments, a systematic measurement of the ohmic drop by impedance between the 2 electrodes was conducted. The thermal imaging furnace described above allowed for precise control of inter-electrode distance to investigate the relationship between impedance and geometry. The reference electrode lead of the potentiostat was the same as the counter electrode, such that the cell voltage was actually monitored. Impedance was measured with the GAMRY® Reference 3000 Potentiostat using the “electrochemical impedance” function. To verify the obtained values and derive a dimensionless resistivity measurement, the primary current distribution was calculated with a COMSOL model (COMSOL Multiphysics Impedance Modelling section). With the appropriate boundary conditions for current conservation, the electric currents physics model was evaluated to calculate the resistance of the system while varying the distance between the electrodes. An initial conductivity estimate was input into the model, and measurements were validated by increasing the inter-electrode distance in the model and comparing to impedance measurements at those inter-electrode distances.

[0652] Post-experiment observations and characterization

[0653] The sample was taken out from the furnace and mass calculations of the cathode, anode, and electrolyte were done before the sample was mounted in epoxy resin for polishing for further characterization studies: optical microscopy and scanning electron microscopy.

[0654] Mass loss calculations

[0655] Each of the assembly parts were weighed to observe the mass change in the anode, cathode, and electrolyte. At the operating temperature, none of the components volatilized from the electrolyte to incur a mass loss. With the mass loss of the electrodes being insignificant, the

[0656] 68

[0657] #15087612vlchange in mass of the electrolyte was then attributed to the amount of sulfur gas released from the electrolyte during the electrolysis experiments. Thus, the anodic faradaic efficiency of the electrolysis experiment was calculated using the following formula:

[0658] η = (ΔmF) / (I·Δt·MS2) x 100 (74)

[0659]

[0660] Mass loss of the electrolyte (Am) is in grams, 4 is the number of electrons exchanged, F is the Faraday’s constant (96485 C / mol e ), I is the current in amperes, At is the electrolysis time in seconds and MS2is the molar mass of sulfur gas (in g / mol).

[0661] Optical microscopy

[0662] For microscopic observations, the electrolyte sample was mounted in epoxy and then polished down to 1 micron using mineral oil as the lubricant. During polishing, the mounted sample was first ground using silicon carbide grit 180 and 240 to reveal the cross-section of the electrolyte, followed by grit sizes 400, 600, 800, 1000, 1200 and diamond paste (6 micron and 1 micron).

[0663] The sample was observed under the optical microscope. To observe the micro structure of the Fe deposited and carbide inclusions (if any), the electrolyte sample was etched in 3% Nital (Nitric acid HNO3 mixed with ethanol in a volume ratio of 3:100) for about 3 secs. This was done to reveal certain microstructural characteristics of the Fe deposit and get a qualitative estimate of the carbon content by comparing it with the micro structure of cast iron, high and low carbon steel. LEICA® LMDM reflected light scope with a ZEISS® Axiocam 306 camera and ZEISS® Zen imaging software was used for the microscopy.

[0664] Scanning electron microscopy

[0665] To confirm the elemental composition of the metallic deposit and observe depletion of iron from the electrolyte, the electrolyte sample was observed under the scanning electron microscope and subjected to EDS (JEOL® JSM - 6610 LV SEM). To study depletion of iron in the electrolyte, compositional analysis was done at several points at various locations in the electrolyte, such as the part in contact with the cathode (electrolyte bottom), center of the electrolyte and the top part. Line scans and element intensity maps were generated at specific locations to observe possible depletion of Fe in the electrolyte.

[0666] 69

[0667] #15087612vlResults from sulfidation of iron oxide are briefly summarized below, followed by results from the galvanostatic electrolysis and impedance experiments. Thermodynamic analysis, proposed process flow sheets for molten cast iron production and implications for steelmaking are provided and discussed.

[0668] Iron oxide sulfidation

[0669] It was observed that the sulfidation process followed a shrinking core reaction mechanism, reacting from the outer surface to the inner core. The mass change, corresponding to the substitution of O with S from the conversion of Fe20s to FeS was estimated at 99.78%. From the EDS results, the stoichiometry of the product compound formed upon sulfidation was estimated. Iron sulfide is known to exist as various off- stoichiometric compounds, which could be either sulfur rich or sulfur deficient depending on the partial pressure ratio. In this case, the stoichiometry of the product was estimated to be around Feo.gS. To get a more accurate chemical signature of the compound, XRD analysis was performed. The peaks as seen in the XRD results corresponded to FeS (iron (II) sulfide) with almost no traces of any residual oxides.

[0670] Molten sulfide electrolysis for iron production

[0671] Electrolytic reduction was performed at two current densities: 1.7 A / cm2and 0.85 A / cm2. Gas bubbles were observed during electrolysis, plausibly indicating S2 evolution. Chronopotentiometry scans (constant current of 0.3 A and 0.150 A) were taken at different current values. All cell voltages depicted in the plots were the actual recorded cell voltages and were not corrected by the IR drop. Ignoring other source of potentials is a first assumption to interpret the electrolytic nature for such 2 electrode electrolysis conditions, further assuming that (i) the activity ratio between the electrode surface and the bulk is close to unity due to the high reactant concentration and (ii) electron transfer is not kinetically limited at such high operating temperatures. Often in such high temperature molten salts, the cathodic metal deposition reaction is fast kinetically. However, the anodic reaction was of the rate-limiting reaction due to the relatively low mobility of the electronegative element (here S2−) and the size of the gaseous product (here S2).

[0672] The resistance of the cell was measured at various anode-cathode distances. To interpret these measurements, the primary current distribution in the electrolyte was calculated by solving the Laplace equation with a finite element model (COMSOL) of the axi-symmetric configuration, assumed isothermal with current conservation at each graphite electrodes. FIG.

[0673] 70

[0674] #15087612vl60 compares the values of the overall resistance as calculated using the model and those obtained, for different anode / cathode distance. Reasonable agreement between the two values can be observed at one position, along with the trend of the resistance with position. Among the many factors that can explain such deviation, the actual non-isothermal conditions of the experiment in the furnace seem likely, as displacing the graphite electrode mass did not affect the radiative surface of interaction with the light focus, as well as the overall path for heat conduction through the top and bottom electrodes. The resistance values from the model were used to estimate the resistivity of the electrolyte at about 3.44 ohm-cm (extrapolating to an anode-cathode distance of zero).

[0675] Anodic faradaic efficiency was calculated by attributing the mass loss after electrolysis to the loss of sulfur, with corresponding metal production at the cathode as observed by microscopy. In the range of the recorded cell voltage, 0.35-0.75, only FeS was expected to electrolytically decompose. FIG. 46 shows Faradaic efficiencies recorded for different electrolysis duration (total charge passed), at two current densities: 0.85 and 1.7 A / cm2. Faradaic efficiency as high as 100% was recorded in two separate repeat trials. In three trials, the efficiency was less than 65%.

[0676] The variation in efficiency can be explained by subtle differences in cell geometry between trials, as well as the accumulation of sulfur bubbles at the anode. The efficiencies recorded for 0.85 A / cm2current density were systematically lower as compared to 1.7 A / cm2, suggesting an overall electrolysis performance at removing sulfur favoring higher polarization and higher current density. For a similar charge (e.g., 18000 C), higher efficiency was recorded at higher current density, indicating that the process was yet tolerant to high reaction rates. However, for longer trials, it was observed that the efficiency decreased (<60% for the trials at 1.7A / cm2over 24000C). A trial with the same operating conditions was performed, but with a 10 second pause halfway through, while maintaining the furnace lamp power at the same intensity. By running electrolysis in two segments with a pause, the efficiency increased back to nearly 100%. Without wishing to be bound by theory, it is believed that sulfur bubbles accumulating at the anode were released during the pause, aided by the relatively low viscosity of the sulfide electrolytes. This suggested a role of the anode and cell design to mitigate such gas accumulation effect, a common challenge in industrial electrolysis.

[0677] The optical and SEM-BEC micrographs and element intensity map (SEM-EDS) of the electrolyte sample at the graphite cathode interface after electrolysis are shown in FIGs.

[0678] 61A-61C. Droplets of iron metal were observed at the bottom, where the electrolyte contacted

[0679] 71

[0680] #15087612vlthe graphite cathode, confirming electrolytic reduction of FeS. SEM-EDS analysis confirmed no reduction of BaS or La2S3. Fe metallic deposits were observed with an average size of 20 microns, with a few deposits of up to 50-micron size. However, due to surface tension effects (commonly observed at small scales) and because of solidification, the produced Fe metallic deposits were dispersed at the bottom of the electrolyte, all along the cathode curvature. Due to this, the metal recovery was hindered, preventing calculation of the cathodic faradaic efficiency. A few metallic deposits observed at the center of the electrolyte likely detached from the cathode and floated up when the sample was molten.

[0681] The SEM-EDS map in FIGs. 61A-61C show the electrolyte solidified in a mixture of 2 phases, in addition to the Fe metallic phase. The light-colored regions contained La-rich phases whereas the Ba-rich phases appeared darker as shown clearly in the WDS micrograph. Performing EDS analysis over the area, it was observed that the electrolyte in contact with the cathode, including regions next to the deposit, was depleted in Fe. In the solidified sulfide electrolyte, Fe was found at greater concentrations in phases that were rich in Ba. The Fe concentration in the electrolyte was studied together with the Ba / La ratio in the phases. A relationship between Fe concentration and the Ba / La ratio was observed in each region, i.e., as the amount of Fe was reduced, the Ba / La ratio also decreased. At the highest Fe concentration of 6wt%, the Ba / La ratio was 1.83, while at the lowest, 0.9-1.8wt%Fe, the Ba / La ratio was 0.6. This was interpreted as indication of the role of mass transport of Fe during electrolysis. The Fe depletion during electrolysis led to the formation of a Ba-rich sulfide phase. This compositional inhomogeneity may locally affect the electrical conduction and other transport properties of the electrolyte. The bulk composition of the electrolyte was at one of the eutectic of BaS and La2S3, while the Ba-rich sulfide phase had a higher melting point. Local solidification due to iron depletion could then inhibit furter mass transport and electrolysis. The solidification of the electrolyte was observed at longer durations of electrolysis or higher current densities (~ 2.83 A / cm2, 0.5 A), while the furnace power was kept the same. Solidification of an electrolyte after 120 seconds of passing 0.5 A was observed. It was believed that these conditions led to a depletion of FeS in the electrolyte and hence may have led to the composition of the electrolyte drifting from the liquidus point. The absence of a ternary phase diagram BaS-La2S3-FeS prevented further verification of this interpretation. This remained however an artefact of the present setup, which was not designed to be continuous, allowing constant feed of FeS as more Fe was produced at the cathode. In a continuous reactor,

[0682] 72

[0683] #15087612vlcontinuous feeding of the feedstock may ensure the composition of the electrolyte is maintained throughout the electrolysis.

[0684] WDS analysis was conducted to estimate the sulfur concentration of the iron deposits, as presented in Table 12. The carbon concentration was not estimated using WDS though the micro structure clearly indicated its presence in the metal. However, since the WDS analysis did not report presence of Ba or La, the carbon concentration was estimated as 100 - (Fe wt.% +S wt.%).

[0685] Table 12: Results of WDS analysis of the observed metallic deposit. Carbon was not measured.

[0686] Element Concentration (wt%)

[0687] Fe 94.55 wt.%

[0688] S 0.05 wt.%

[0689]

[0690] The exact composition of the metallic deposit in terms of carbon concentration was difficult to perform due to the size of the deposits. Thus, to get a qualitative estimate of the carbon, the electrolyte sample was etched to reveal the underlying micro structure and compare it with other Fe-C alloy systems. FIG. 62 shows the optical and SEM micrographs of the metallic Fe deposits after etching with 4% Nital solution. A few of the Fe deposits, in contact with a graphite cathode, showed the presence of carbide-like inclusions. Observing the shape of the inclusions concluded that they were iron carbides (cementite phase). The microstructures revealed after etching showed the pearlite and cementite phase and plausibly ledeburite. This likely indicated the metallic deposits to be a high-carbon Fe alloy.

[0691] Table 13, shown below, shows the results of the ICP analysis to confirm that the mass loss in the electrolyte aligned with the depletion of sulfur before and after electrolysis. A mass loss of 0.0018 g was observed corresponding to a difference of 0.9 wt.%. The ICPresults were thus in alignment (0.85 wt.%) with the fact that mass loss occured due to sulfur loss. Sulfur deposition could also be observed on the quartz containment tube after electrolysis.

[0692] Table 13. Results of ICP analysis comparing the difference in S content before and after electrolysis.

[0693] Sample S content (wt%)

[0694] Before electrolysis 15.65

[0695] After electrolysis for 40 secs 14.80

[0696]

[0697] 73

[0698] #15087612vlProcess flow and thermodynamic analysis

[0699] A possible process flow for electrolytic production of molten iron, in accordance with some embodiments, is comprised of two steps: (1) feedstock preparation through sulfidation of iron oxide ore (unless pyrite is used) and (2) molten sulfide electrolysis to produce molten iron. The thermodynamics of these steps and implications for a process flow for steel production are summarized below.

[0700] Sulfidation of iron oxide ore

[0701] To utilize conventional iron oxide ore, hematite may be treated with sulfur in a sulfidation reaction to selectively convert the Fe2Ch present in the ore into FeS (as shown in equation 75) in a spontaneous exothermic process, at a 1000°C.

[0702] 4 Fe2O3(s) + 7 S2(g) → 8 FeS (l) + 6 SO2(g) (75)

[0703] ΔH°1000°C= −131.9 KWh / ton Fe ; ΔG°1000°C= −192.82 KWh / ton Fe

[0704] The values indicated here correspond for FeS required to produce 1 tonne of Fe.

[0705] The impurities (gangue) remained as un-molten oxides, were lighter than the molten FeS, and could be separated using physical separation methods. The SO2(g) along with the excess amount of heat released from the reactor may be captured and utilized to produce sulfuric acid in an acid plant and generate electricity, a minimum of 210 kWh / tonne of acid. The excess heat in the reactor may also be used to heat the reactants to the reaction temperature.

[0706] Molten sulfide electrolysis of FeS at 1300 °C

[0707] Molten sulfide electrolysis was operated semi-continuously, with periodic feeding of FeS, and a tapping of molten cast iron in a single-electrolytic cell. The operating temperature for the electrolysis process was 1300- 1350 °C, sufficient to maintain a molten supporting electrolyte (eutectic composition of the pseudo-binary system, BaS-La2S3) and produce superheated molten iron following reaction:

[0708] FeS (l) → Fe (l) + ½ S2(g) (in case of pure Fe production) (76)

[0709] ΔH°1600°C= 691.47 kWh / tonne Fe ; ΔG°1600°C= 294.91 kWh / tonne Fe

[0710] 74

[0711] #15087612vlThe values indicated here correspond for FeS required to produce 1 tonne of Fe.

[0712] For which the standard enthalpy and standard Gibbs energy were found using the data from FACTSAGE™ databases FStel, FTmisc and FactPS. If carbon, in the form of graphite (or carbon- saturated iron) was added to the cathode product, where x is the stoichiometric ratio for carbon- saturated molten iron (4.3wt%C), the equation is:

[0713] FeS (l) + x C → Fe − 4.3 wt. % C (l) + ½ S2(g) (molten iron production) (77) kWh kWh

[0714]

[0715] ΔH°1300°C= 776.99 kWh / tonne molten iron ; ΔG°1300°C= 316.61 kWh / tonne molten iron

[0716] The sulfur released (0.55t) was recirculated into the sulfidation circuit, partially fulfilling the sulfur requirements of the consecutive cycles (0.96t total, or 0.4 It additional to the recirculated sulfur).

[0717] Process mass balance

[0718] Following from the reactions above, FIG. 63 depicts the mass balance for 1 metric tonne production of liquid cast iron (Fe - 4.3 wt.% carbon), a composition similar to that of hot metal, the current product of the blast furnace. It assumed perfect physical and chemical separation at each step. Since the S2 released in the MSE process (0.55t) was captured and recirculated into the system, an addition of 0.4 tonnes sulfur was needed for the sulfidation of iron oxide, Fe2O3. Global production of sulfur in 2023 as a by-product of fossil fuel production totaled 81 Mt. Resources of sulfur in natural gas, crude oil, petroleum, tar sands and metal sulfides including elemental sulfur in volcanic deposits totals about 5 billion tonnes.

[0719] Process energy balance

[0720] The energy requirements for 4 different scenarios of varying product compositions and reaction temperatures were calculated to support a comparison with other routes. For the MSE route, electrical energy requirements were calculated from the projected cell potential, Uceii(V), which is the sum of the potential due to the chemical work contribution (Uchem), potential due to Ohmic contribution (Uohmic) and potential to due to overvoltage requirements (Uovervoitage). Uchem is defined in equation (79), where AG° is the standard Gibbs energy for reaction (76) or (77) depending on the modeled scenario, n is the number of electrons transferred, and F is Faraday’s constant. Uohmic is the amount of energy supplied as heat, defined in equation (80), where

[0721]

[0722] is the standard enthalpy and AG° is the standard Gibbs energy for

[0723] 75

[0724] #15087612vlreaction (76) or (77) depending on the modeled scenario. Uovervoltageis assumed to be negligible in this analysis.

[0725] Ucell Uchem 3” 3” 17overvoj(-age(78)

[0726] Uchem= −ΔG° / nF (79)

[0727] Uohmic= −(ΔH°−ΔG°) / nF (80)

[0728] As shown in FIG. 63, the process may generate at least O.lt of gangue from the iron ore, assuming high quality iron ore with 65% Fe. All mass and energy calculations were done assuming pure compounds in the electrolysis. Other simplifying assumptions included:

[0729] 1. Decomposition potential was constant and not affected by interactions with the inert graphite anode, or components of the supporting electrolyte. Interactions between the different ionic species present in the electrolyte were ignored here, and due to the large difference in the decomposition potential of BaS and La2S3relative to FeS, the supporting electrolyte components were assumed to be stable.

[0730] 2. The Faradaic efficiency of the electrolysis process was 90%, consistent with the results above.

[0731] 3. Heat losses due hot off-gases and the effect of conduction / convection and radiation were assumed to be an additional 40% of the Ohmic energy requirement.

[0732] Table 14 summarizes the total electrical energy requirements for different compositions of the product at two temperatures: 1300°C and 1600°C. As a reference, the melting point of pure Fe is 1539°C while that of cast iron (Fe-4.3 wt.% C) is 1147°C.

[0733] Table 14: Summary of the electrical energy requirements for production of pure Fe and molten cast iron (Fe-4.3 wt.% C) at different temperatures

[0734] Total energy Electrical

[0735] U ohm requirement, Production Ucell (V) energy

[0736] (V) including requirement

[0737] additional

[0738]

[0739] 76

[0740] #15087612vl(kWh / tonne 40% heat loss product) (kWh / tonne product)

[0741] Eiquid Fe at 1600

[0742] 1.1 1.4 1361 1787

[0743] °C

[0744] Solid Fe at 1300

[0745] 0.9 1.2 1197 1537

[0746] °C

[0747] Molten iron (Fe- 4.3 wt.% C) at 1.0 1.4 1264 1644

[0748] 1300 °C

[0749] Molten iron (Fe- 4.3 wt.% C) at 1.2 1.5 1363 1811

[0750] 1600 °C

[0751]

[0752] In case 1 at 1600°C, a superheated liquid iron product was expected, ensuring flowability to tap the product out of the cell. In terms of the minimal energy estimates, liquid Fe production via MSE was less than 50% compared to molten oxide electrolysis (3600-4000 kWh / t Fe, including heat losses). The calculation in case 2 was purely for comparison of the effect of temperature, as at 1300°C, without any alloying element, the product was pure solid iron which would not be practical at tonnage scale. It showed that contrary to several claims, electrical energy savings remain marginal (10 to 15%) by operating about 300°C lower than the existing blast-furnace temperature.

[0753] Other advantages came with lower temperature though, and producing molten iron (Fe-4.3 wt.% C) came with significantly reduced processing temperature, which decreased the complexity of heat management and prevented excessive refractory wear. Apart from reducing the melting point of iron by alloying, carbon also behaved as a powerful agent to control the chemistry of iron (e.g., the solubility of O in Fe increased as C decreased). However, it was observed that the addition of carbon did not lead to any reduction in the energy needed, i.e., producing solid pure Fe at 1300°C required less energy than producing molten iron (Fe-C alloy). This was because the enthalpy of mixing Fe and C was not sufficient to offset the enthalpy of fusion of Fe and that needed to heat the carbon to reaction. Nevertheless, direct production of molten iron allowed semi-continuous metal recovery from the cell and avoided an additional facility to melt the solid product. Production of molten iron, similar to pig iron / hot

[0754] 77

[0755] #15087612vlmetal from the blast furnace, was the most suitable for integration within the existing steel making practices and facilities.

[0756] FIG. 64 compares the energy requirement for molten iron production via MSE with other electrochemical approaches and the BF-BOF route. The energy requirements for molten iron production through MSE were significantly less: 52% less compared to MOE, 43% less than alkaline electrowinning and melting in an electric arc furnace, and 64% less as compared to BF-BOF route per tonne of Fe equivalent produced.

[0757] Table 15 compares the reduction from a sulfide feedstock and an oxide feedstock. Two major differences were evident: the number of moles of electrons per mole of Fe, and the difference in the chemical bonding between Fe and O and Fe and S. In Table 15 FeO is included for the sake of comparison to understand the effect of bonding differences (FeO and FeS) and the effect of number of electrons (FeO and Fe2O3). The results in Table 15 highlight the difference in the chemical work needed between cases. In the case of FeO, 2 moles of electrons needed to be reduced per mole of Fe whereas in case of Fe20s, it was 3 moles of electrons. This suggested that the reduction of FeO would need approximately two-thirds of the energy as compared to Fe2O?. This was also validated by the values mentioned above. The differences in the chemical bonding structure between FeO (Fe double bonded to O) and Fe2O? (2 oxygen atoms double bonded individually to each Fe ion, while one oxygen was shared between the 2 Fe atoms) appeared to account for the deviation in the values. By comparing the values of FeS and FeO, it was concluded that the bond of Fe and S was weaker than the Fe-0 bond. Additionally, the entropy of S2 gas was lower than O2 gas due to the large size of the S2 molecules. These reasons explained the significantly lower energy requirements for reduction of FeS as compared to Fe20s.

[0758] Table 15: Thermodynamic values of the reduction of various iron-bearing feedstock at 1300 °C and 1600 °C.

[0759] 1300 °C 1600 °C

[0760] AS AS AG AH AG AH

[0761] Feedstoc (kWh / K (kWh / K (kWh / tonn (kWh / tonn (kWh / tonn (kWh / tonn k -tonne -tonne e Fe) e Fe) e Fe) e Fe)

[0762] Fe-) Fe) FeS 350 620 0.17 295 690 0.212 FeO 810 1300 0.31 740 1210 0.25

[0763]

[0764] 78

[0765] #15087612vlFeiCh 1050 2005 0.606 865 2080 0.647

[0766]

[0767] In summary, promising aspects of molten sulfide electrolysis for molten iron production include:

[0768] 1. The stability of the +2 valence state for Fe as a sulfide at the electrolysis conditions (high temperature and low partial pressure of oxygen). This is believed to avoid the multi- valency issue encountered in other electrolytic processes for Fe production.

[0769] 2. The S2 gas (thermodynamically the most stable form of sulfur at the operating temperatures of MSE) released at the anode can be captured and condensed, while graphite serves as a scalable inert anode material.

[0770] 3. Molten cast iron production at the cathode allows for smooth integration into the present BOFs, utilizing their high refining capabilities, without requiring their premature retirement.

[0771] Due to the high electrical conductivity of molten FeS (1500 / ohm-cm), 10-20wt% FeS was dissolved into the supporting electrolyte (BaS - La2S3) to mitigate the electronic conductivity. The small electrical conductivity of BaS (0.01 / ohm-cm) resulted in the ternary exhibiting partial ionic conduction, as a function of composition and temperature.

[0772] The high decomposition potentials and thermal stability of BaS and La2S3in the supporting electrolyte provided stability during electrolysis at high temperatures. With feedstock pre-treatments, oxygen contamination can be avoided, enabling recirculation of the electrolyte. BaS and La2S3can be sourced from BaSCU (barite mineral) and La20s (23 wt.% of Bastnaesite mineral), with sulfidation for conversion, which has been demonstrated at the kg-scale. These abundant minerals face low supply risk, and La20s is produced in abundance as demand for accompanying, more dilute rare earth elements, neodymium and dysprosium increases.

[0773] The iron feedstock for MSE could be either naturally-occurring pyrite, FeS (abundant in shales, pyrite in igneous rocks, and tailings from non-ferrous mining), or conventional oxide ores (e.g., hematite, magnetite) treated with sulfur in an exothermic sulfidation reaction. This process may thus allow the usage of lean Fe ores or waste Fe-containing sources which have traditionally not been used (e.g. pyrite).

[0774] This disclosure demonstrated an electrolytic approach based on sulfide chemistry. Greater than 95% anodic efficiency was measured over several trials. MSE demonstrated many benefits compared to other techniques. The existence of Fe (+2) as the sole valence state as

[0775] 79

[0776] #15087612vlopposed to multiple valence states in oxides (Fe+2and Fe+3) reduced energy requirements. Preliminary thermodynamic analysis showed that the energy consumption for the process was favorable, with a possible path toward more than 50% less energy than any other route. Moreover, the low operating temperature range of around 1300°C, allowed for easier heat management and refractory lifetime. Furthermore, the high solubility of FeS (here tested at 10 wt.%) in the supporting electrolyte, high current density and generation of liquid product made it advantageous as compared to lower temperature techniques, allowing for higher space-time yield and productivity. This process allowed for molten iron production at small scale facilities where excess pyrite was available or by treating the oxide ores with sulfur, while still delivering a molten iron product that can be processed with the unsurpassed productivity of steelmaking refining and continuous casting.

[0777] EXAMPLE 4

[0778] Molten sulfide electrolytes were made using the experimental setup shown in FIG. 66 A and FIG. 66B with a set temperature of 1410°C and the internal temperature reached 1350 °C. FIG. 65A, FIG. 65B, and FIG. 65C show images of the electrolytes produced.

[0779] FIG. 65D and FIG. 65E show additional analysis thereof. FIG. 67 shows a Ternary Diagram with La2S3, BaS, and FeS.

[0780] Liquid sulfide electrolytes were prepared at 1300 °C with 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3. Samples were fully molten starting at 12 Wt% FeS. The samples had high oxygen content as oxygen entered the system during testing and was present in the powdered samples. FIG. 68A, FIG. 68B, and FIG. 68C show WDS results and FIG. 68C shows the mass percentage of various elements in various phases.

[0781] Liquid sulfide electrolytes were prepared at 1300 °C with 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3. Samples were fully molten starting at 12 Wt% FeS. The samples had low oxygen content, as the furnace system was completely leak tested and the chemicals were newly purchased. FIG. 69A, FIG. 69B, and FIG. 69C show SEM / EDS results, analyzing a liquid sulfide electrolyte and the amount of oxygen therein. FIG. 68C shows the mass percentage of various elements in various phases.

[0782] 80

[0783] #15087612vlEXAMPLE 5

[0784] FIGs. 68A-68C and FIGs. 69A-69C show liquid sulfide electrolyte chemistry and that the actual amount of Oxygen that was found did not show electrolysis nor electrochemical usage. The samples were fully molten starting at 12 Wt% FeS. FIGs. 68A-68B show the WDS results of 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3at 1300°C. FIG. 68C shows the high oxygen content of the phases, where oxygen entered the system during testing and was present in powdered samples. FIGs. 69A-69B show the SEM / EDS results of 20 Wt% FeS, 49.5 wt% BaS, and 30.5 wt% La2S3at 1350°C. FIG. 69C shows the low oxygen content of the phases, where the furnace system was leak tested and new were chemicals purchased.

[0785] EXAMPLE 6

[0786] Molten sulfide electrolysis was successfully used for iron production as shown in FIG. 74. Gas evolution was observed as current was passed. Bubbling was observed, providing evidence of gas evolution at the anode as current was passed. Electrolyte solidifying was observed at the same power of the furnace. This indicated depletion of FeS from the electrolyte leading to an increase in melting point of the system.

[0787] Observations at the anode included: yellow condensate observed on the quartz tube; evidence of S2 gas evolving from the anode; and distinct smell of sulfur when the tube was removed from the furnace.

[0788] Observations at the cathode included: metallic deposition was observed on the cathode; EDS measurements report 99.45 wt.% Fe; Fe deposit size of 50-300 μm. After electrolysis of the samples for 40 seconds, an optical micrograph (20X) of the electrolyte cross-section indicated metallic deposition on the cathode. SEM images of the electrolyte cross-section and element intensity maps of the SEM images were obtained.

[0789] Inductively coupled plasma mass spectroscopy (ICP) analysis was done on pre-melt and after electrolysis samples. Sulfur depletion, eq. to mass loss electrolyte was observed.

[0790] Wavelength dispersive spectroscopy (WDS) analysis was done. No evidence of Ba or La deposition was observed. 94.55 wt.% Fe and 0.05 wt.% S were observed.

[0791] Quantitative carbon concentration was not obtained as EDS was not good for light element, the sample was too small for LECO, and the sample was coated with carbon for WDS.

[0792] 81

[0793] #15087612vlThis example demonstrated: confirmed mass loss due to loss of sulfur; hot metal composition (BOF feedstock); qualitative measure of carbon; comparison of microstructures of the Fe deposited with available steel and cast-iron microstructures; sample etched with 4% Nital. FIG. 75 shows characterization studies after etch results, showing pearlite-like structure and carbide inclusions observed in deposits near cathode.

[0794] Above 90% faradaic efficiency was consistently recorded. Experiments were performed at 2 current densities: 0.85 A / cm2and 1.7 A / cm2. Productivity was high. High efficiency for duration up to 60 seconds. Over 60 seconds, high amount of gas accumulation near anode shielded the anode, as it needed time to escape. Over 60 seconds, efficiency used to drop, possibly due to a back reaction and / or slower kinetics as the product is not removed. Some cell designs facilitated more efficient gas removal.

[0795] The impedance of the system and factors affecting it were studied. The impedance was sensitive to the distance between the electrodes. There was precise control of electrode position in thermal imaging furnace (TIF). This was experimentally observed using electrochemical impedance spectroscopy (EIS) measurements. It was modelled on COMSOL Multiphysics software. At higher distances, deviation in the values was observed, although the trend remained the same (monotonically increasing). There was a small hot zone in the TIF - minor change results in temperature difference. At higher distances, there is a possibility of a drop in the electrolyte temperature, whereas the COMSOL model assumed isothermal conditions.

[0796] It should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific implementations described above. The specific implementations described above are disclosed as examples only.

[0797] While several embodiments of the present invention have been described and illustrated herein, those of ordinary skill in the art will readily envision a variety of other means and / or structures for performing the functions and / or obtaining the results and / or one or more of the advantages described herein, and each of such variations and / or modifications is deemed to be within the scope of the present invention. More generally, those skilled in the art will readily appreciate that all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and / or configurations will depend upon the specific application or applications for which the teachings of the present invention is / are used. Those skilled in the art will

[0798] 82

[0799] #15087612vlrecognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. It is, therefore, to be understood that the foregoing embodiments are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, the invention may be practiced otherwise than as specifically described and claimed. The present invention is directed to each individual feature, system, article, material, and / or method described herein. In addition, any combination of two or more such features, systems, articles, materials, and / or methods, if such features, systems, articles, materials, and / or methods are not mutually inconsistent, is included within the scope of the present invention.

[0800] The indefinite articles “a” and “an,” as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.” The phrase “and / or,” as used herein in the specification and in the claims, should be understood to mean “either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Other elements may optionally be present other than the elements specifically identified by the “and / or” clause, whether related or unrelated to those elements specifically identified unless clearly indicated to the contrary. Thus, as a non-limiting example, a reference to “A and / or B,” when used in conjunction with open-ended language such as “comprising” can refer, in one embodiment, to A without B (optionally including elements other than B); in another embodiment, to B without A (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0801] As used herein in the specification and in the claims, “or” should be understood to have the same meaning as “and / or” as defined above. For example, when separating items in a list, “or” or “and / or” shall be interpreted as being inclusive, i.e., the inclusion of at least one, but also including more than one, of a number or list of elements, and, optionally, additional unlisted items. Only terms clearly indicated to the contrary, such as “only one of’ or “exactly one of,” or, when used in the claims, “consisting of,” will refer to the inclusion of exactly one element of a number or list of elements. In general, the term “or” as used herein shall only be interpreted as indicating exclusive alternatives (i.e. “one or the other but not both”) when preceded by terms of exclusivity, such as “either,” “one of,” “only one of,” or “exactly one of.” “Consisting essentially of,” when used in the claims, shall have its ordinary meaning as used in the field of patent law.

[0802] 83

[0803] #15087612vlAs used herein in the specification and in the claims, the phrase “at least one,” in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase “at least one” refers, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, “at least one of A and B” (or, equivalently, “at least one of A or B,” or, equivalently “at least one of A and / or B”) can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.

[0804] As used herein, “wt%” is an abbreviation of weight percentage.

[0805] Some embodiments may be embodied as a method, of which various examples have been described. The acts performed as part of the methods may be ordered in any suitable way. Accordingly, embodiments may be constructed in which acts are performed in an order different than illustrated, which may include different (e.g., more or less) acts than those that are described, and / or that may involve performing some acts simultaneously, even though the acts are shown as being performed sequentially in the embodiments specifically described above.

[0806] Use of ordinal terms such as “first,” “second,” “third,” etc., in the claims to modify a claim element does not by itself connote any priority, precedence, or order of one claim element over another or the temporal order in which acts of a method are performed, but are used merely as labels to distinguish one claim element having a certain name from another element having a same name (but for use of the ordinal term) to distinguish the claim elements.

[0807] In the claims, as well as in the specification above, all transitional phrases such as “comprising,” “including,” “carrying,” “having,” “containing,” “involving,” “holding,” and the like are to be understood to be open-ended, i.e., to mean including but not limited to. Only the transitional phrases “consisting of’ and “consisting essentially of’ shall be closed or

[0808] 84

[0809] #15087612vlsemi-closed transitional phrases, respectively, as set forth in the United States Patent Office Manual of Patent Examining Procedures, Section 2111.03.

[0810] 85

[0811] #15087612vl

Claims

CLAIMSWhat is claimed is:

1. A method comprising:a. performing electrolysis on a molten sulfide electrolyte in an electrochemical cell, wherein the molten sulfide electrolyte comprises greater than 0.5 wt% oxygen and the molten sulfide electrolyte comprises M, wherein M is a metal;b. removing a gas comprising sulfur formed at an anode of the electrochemical cell; andc. producing metallic M at a cathode of the electrochemical cell.

2. The method of claim 1, further comprising forming the molten sulfide electrolyte, wherein forming the molten sulfide electrolyte comprises sulfidation of MaXband at least one other compound, wherein the at least one other compound comprises NcYd, wherein M and N are each independently a metal, X and Y are each independently an electronegative species, and a, b, c, and d are each independently a number.

3. A method comprising:a. forming a molten sulfide electrolyte, wherein forming the molten sulfide electrolyte comprises sulfidation of MaXband at least one other compound, wherein the at least one other compound comprises NcYd, wherein M and N are each independently a metal, X and Y are each independently an electronegative species, and a, b, c, and d are each independently a number, and wherein the molten sulfide electrolyte comprises greater than 0.5 wt% oxygen;b. performing electrolysis on the molten sulfide electrolyte in an electrochemical cell;c. removing a gas comprising sulfur formed at an anode of the electrochemical cell; andd. producing metallic M at a cathode of the electrochemical cell.

4. The method of any preceding claim, wherein M comprises iron, cobalt, nickel, copper, zinc, molybdenum, silver, tantalum, tungsten, and / or a rare earth element.86#15087612vl5. The method of any one of claims 2-4, wherein X comprises a sulfide anion (S2-).

6. The method of any one of claims 2-5, wherein N comprises barium and / or lanthanum.

7. The method of any one of claims 2-6, wherein Y comprises a sulfate anion (SO42-), a phosphate anion (PO43-), and / or an oxide anion (O2-).

8. The method of any one of claims 2-7, wherein a is 1, 2, 3, or 9.

9. The method of any one of claims 2-8, wherein b is 1, 2, 3, 4, or 8.

10. The method of any one of claims 2-9, wherein c is 1, 2, or 3.

11. The method of any one of claims 2-10, wherein d is 1, 2, or 3.

12. The method of any one of claims 2-11, wherein MaXbis FeS, CoS, CoS2, NiS, NiS2, Ni3S2, Ni2S4, Ni9S8, Cu2S, CuS, ZnS, MoS2, Ag2S, TaS2, WS2, Nd2S3, Pr2S3, Tb2S3, and / or Dy2S3.

13. The method of any one of claims 2-12, wherein NcYdis BaSO4, La2O3, and / or La2(SO4)3.

14. The method of any preceding claim, wherein the gas comprising sulfur is S2 gas.

15. The method of any preceding claim, wherein the molten sulfide electrolyte comprises less than or equal to 12 wt% oxygen.

16. The method of any preceding claim, wherein the molten sulfide electrolyte comprises less than or equal to 8 wt% oxygen.

17. The method of any preceding claim, wherein the molten sulfide electrolyte comprises greater than or equal to 1 wt% oxygen.87#15087612vl18. The method of any preceding claim, wherein at least a portion of the metallic M is liquid.

19. The method of any preceding claim, wherein performing electrolysis on the molten sulfide electrolyte comprises performing electrolysis continuously for a period of time by replenishing the molten sulfide electrolyte.

20. The method of claim 19, wherein the period of time is greater than or equal to 10 minutes and less than or equal to 12 hours.

21. The method of any preceding claim, wherein the anode comprises carbon.

22. A method comprising:a. forming a molten sulfide electrolyte by sulfidation of BaSO4, La2O3, La2(SO4)3, and FeS;b. removing sulfur (S) via electrolysis in an anodic reaction; andc. producing Fe metal at the cathode.

23. An electrolysis cell comprising an anode, cathode, refractory and electrolyte components, and a molten sulfide electrolyte comprising BaSO4, La2O3, La2(SO4)3, and FeS.

24. The electrolysis cell of claim 23, wherein the electrolysis cell supports the removal of sulfur via an anodic reaction and the production of metal at the cathode.88#15087612vl

Images