WO2020072532A1 - Methods and systems for enhancing smelter grade alumina quality - Google Patents

Abstract

Methods and systems for calcining aluminum hydroxide (e.g. Gibb site, Boehmite) and then reacting the calcined product with hydrogen fluoride (HF) before being recycled as fluorinated alumina to the molten electrolyte of an electrolysis cell. The systems and methods reduce the capital and operating costs for aluminum hydroxide calcination and alumina smelting processes and reduce the environmental impact of these processes.

Description

METHODS AND SYSTEMS FOR ENHANCING SMELTER GRADE ALUMINA

QUALITY

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application No. 62/740,193, filed October 2, 2018, which is incorporated herein by reference. This application, U.S. Patent No. 9,920,442 and U.S. Patent Application Serial No. 15/745,832, which are incorporated herein by reference, are commonly assigned to Bechtel Mining & Metals, Inc.

FIELD OF THE DISCLOSURE

[0002] The following disclosure generally relates to methods and systems for calcining aluminum hydroxide (e.g. Gibbsite, Boehmite) and then reacting the calcined product with hydrogen fluoride (HF) before being recycled as fluorinated alumina to the molten electrolyte of an electrolysis cell. More particularly, the present disclosure relates to reducing the capital and operating costs for aluminum hydroxide calcination and alumina smelting process systems and reducing the environmental impact of these processes using disclosed methods and systems to mitigate the generation and emissions of hydrogen fluoride (HF) and Green House Gases (GHG), such as perfluorocarbons and carbon dioxide, to the environment.

BACKGROUND

[0003] The Bayer Process is the dominant method used to extract alumina from the bauxite ore. Overall, the main process steps consist of grinding, digestion, dissolution, precipitation, and calcination. Mined bauxite is initially washed, ground and dissolved in a hot caustic soda (sodium hydroxide) and lime (calcium oxide) at high pressure and temperature creating a liquor solution that contains sodium aluminate and undissolved bauxite residues iron, silicon, and titanium. Insoluble materials are then separated from the sodium aluminate solution in thickeners and filters. The bauxite residue is then washed, dewatered, and then stored in a land fill; the wash water, containing caustic soda, is recycled back to the process. Aluminum hydroxide is precipitated by cooling the liquid and adding crystal seeds; the precipitate is filtered and washed to remove and recover entrained caustic solution. The final calcination step rapidly heats the aluminum hydroxide to approximately l000°C to drive off free water and water that is chemically combined, leaving smelter grade alumina. This last step of the process may be referred to herein generally as a calcination process. The commercial calcination of aluminum hydroxide to form smelter grade alumina commonly utilizes the circulating fluid bed process as generally described and schematically illustrated in FIG. 1.

[0004] According to FIG. 1, wet aluminum hydroxide sludge enters a filter 50 which then discharges partially dewatered moist aluminum hydroxide via a mechanical screw feeder 54 to an aluminum hydroxide charging station 1 feeding a drier 60 operating at approximately H0°C. Dried aluminum hydroxide 3 is discharged from the drier 60 to a waste gas heat exchanger 4 preheating the dried aluminum hydroxide 3 and waste gases. The dried aluminum hydroxide and waste gases are then conveyed to an electrostatic gas cleaning system 8 that separates waste gases, that are discharged to an exhaust stack (not shown), from the dried aluminum hydroxide 3. At the discharge of the electrostatic gas cleaning system 8 dried aluminum hydroxide 3 is gravity fed to a series of separation cyclones 12 and 18 before the dried aluminum hydroxide 3 is fed to a fluid bed reactor 20. In the fluid bed reactor 20 the dried aluminum hydroxide 3 is rapidly heated to approximately l000°C in the order of 1 minute to 3 minutes thermally decomposing (calcining) the dried aluminum hydroxide 3 forming smelter grade alumina. A series of counter flow coolers 28 and 30 reduce the temperature of the calcined alumina and recover energy for the upstream drier 60. A series of cyclone separators 29, 31 and 33 remove fine calcined alumina particulate from waste gases before entering the calcined alumina cooler 36. Calcined alumina product is discharged from the calcined alumina cooler 36 by a discharge chute 38 at a temperature of approximately 80°C. [0005] Alumina quality, in terms of particle size distribution, under calcined residual hydroxyl content and over calcined alpha content, are a function of the technologies used in the Bayer process and calcination system operating parameters.

[0006] Aluminum metal is produced industrially by electrolysis of smelter grade (or other) alumina in a molten electrolyte, using the well-known Hall-Heroult process. This process may be referred to herein generally as a smelting process. The electrolyte is contained in a pot comprising a steel pot shell, which is lined on the inside with refractory and insulating materials, and a cathodic assembly located at the bottom of the electrolytic cell. Carbon anodes extend into the above referenced electrolyte composed of molten cryolite and dissolved alumina. A direct current, which may reach values of more than 600 kA, flow through the anodes, electrolyte and cathode generating electrochemical reactions that reduce the alumina to aluminum metal, and that heat the electrolyte by the Joule effect to a temperature of approximately 960° C.

[0007] Emissions from the electrolytic cell comprise several gaseous and particulate constituents, also referred to as process gases, such as hydrogen fluoride (Fg), particulate fluoride (Fp), and Green House Gases (GHGs), such as carbon dioxide (CO2) and perfluorocarbons (PFCs).

[0008] The mechanics involved with the generation of hydrogen fluoride by the electrolytic cell include:

i. electrochemical hydrolysis of hydrogen (H) sources that react with the molten electrolyte (at ~960°C); and

ii. thermal hydrolysis of hydrogen (H) sources entering the electrolytic cell that react with electrolyte vapor (~400°C) escaping through the crust.

[0009] Dry adsorption of gaseous fluorides onto the surface of fresh alumina, followed by the recycle of the fluorinated alumina back to the electrolytic cell as the feed material for an aluminum smelting process, is a widely utilized technique for abating fluoride emissions from an electrolytic cell. An injection type dry scrubbing system adsorbs gaseous hydrogen fluoride onto the surface of smelter grade alumina, and then filters the alumina and particulate before releasing scrubbed gases (including residual emissions) to the environment. Depending on the electrolytic cell operating current and operating conditions the temperature of the process gases exhausted from conventional electrolytic cells typically varies between l00°C to l40°C above ambient temperature. Because the temperature of the process gas exhausted from the electrolytic cell varies inversely with ambient air flow entering the electrolytic cell, conventional smelting process systems with significantly reduced ventilation flow can theoretically generate process gas temperatures up to and above 400° C.

[0010] Greenhouse gas emissions arise from the calcination process and process-related conditions in the electrolysis cell, such as the consumption of carbon anodes and by anode effects occurring within a cell. Reduction of greenhouse gas emissions are important to reduce the overall carbon footprint of alumina calcination and primary aluminum production. To date, the reduction of GHG emissions to the environment that are derived from the calcination of aluminum hydroxides and the reduction of alumina forming aluminum metal have been limited to incremental fuel combustion efficiency and electrolysis cell process control system enhancements. GHG formation sources and related species generated by the calcination and electrolysis processes include:

i. fuel combustion for the calcination of aluminum hydroxides resulting in the emission of carbon dioxide (CO2);

ii. fuel combustion, when applicable, for the generation of electricity resulting in the emission of CO2;

iii. electrochemical reactions between alumina and carbon anodes occurring within the molten electrolyte resulting in the emission of CO2; and iv. anode effects, due principally to insufficient alumina being dissolved in the molten electrolyte causing some of the electrolyte and carbon anode to be consumed, resulting in the emission of PFC gases tetrafluoromethane (CF₄) and hexafluoroethane (C2F6).

[0011] Aluminum fluoride compounds are a critical additive to the electrolyte in the aluminum smelting process. When combined with cryolite, aluminum tri-fluoride (AIF3) lowers the melting point of the electrolyte solution, which increases its conductivity reducing the amount of energy needed for electrolysis. The efficient capture and return of fluorides evolved from the molten electrolyte solution is thus important for environmental compliance, the operating efficiency and operating cost of an electrolytic cell.

[0012] The capability of calcined alumina to capture and adsorb HF on its surface is dependent on: i. Calcination temperature and exposure time: Aluminum hydroxide (e.g. Gibbsite, Boehmite) is heated at a high rate to a temperature between 950°C to l050°C typically for 1 to 3 minutes in a modern circulating fluid bed calciner. The product produced is predominately metastable transition phase aluminas (e.g. gamma, gamma-prime, delta, theta); however, the calcined product typically also contains under-calcined phases with excessive residual hydroxyl content (e.g. Gibbsite, Boehmite) located in the core of larger particles, and over-calcined alpha phase alumina (void of hydroxyls) within smaller particles and the outer shell of larger particles. The calcination temperature and alumina phases and thermal decomposition pathways typically experienced by conventional calcination are known to those skilled in the art; ii. The structural change and surface reactivity development during calcination: The transition phase alumina formed, its pore structure and surface chemistry development are intimately linked to the thermal decomposition temperature and heating rate of aluminum hydroxides. As thermal decomposition proceeds with increasing temperature and time, the pore size increases and surface reaction sites evolve within the pores to a maximum, then decline with increasing pore size as higher transition phases are formed. Commercial calcination processes promote the formation of higher transition phases irrespective of surface reactivity development; iii. Pore surface chemistry and reactivity: Aluminum hydroxide surface pH evolves as a function of calcination temperature. The surface basicity density increases with calcination temperature to some maxima occurring at or near 500°C, then (surface basicity) falling to some minima at the maximum calcination temperature of approximately l050°C. The surface acidity density evolves inversely as compared to the surface basicity density with its maxima coinciding with the maximum calcination temperature of l050°C. The capture of HF varies directly with the specific surface basicity;

iv. Sodium migration toward internal pore surfaces and reactivity: Residual sodium (an impurity in the aluminum hydroxide) migrates from the crystalline structure to pore surfaces formed during the thermal decomposition (calcination) under conventional calcination methods. The sodium ions tend to block narrow pores that lead to internal surface reaction sites; and

v. Pore Diffusion Resistance & Pore Blockage: HF adsorption capability is directly proportional to the number of kinetically accessible alumina surface reaction sites. The Specific Surface Area (SSA), a measure of the number of reaction sites, stated on material certificates from the alumina supplier do not properly characterize an alumina’s capability to allow HF molecules (or complexes thereof e.g. hydroxy- fluorides) access to internal reaction sites. Pore openings smaller than 35 Angstroms (A) kinetically restrict or block the pore entrance prohibiting internal reaction sites from adsorbing HF during dry scrubbing.

[0013] Bulk transport of calcined alumina from the refinery to the smelter site, material handling within the smelter facility and finally storage in the electrolytic cell’s superstructure, before being fed to the molten electrolyte, typically detracts from alumina consistency (quality) by way of varying:

i. Temperature: calcined fluorinated alumina fed to the electrolytic cell can vary from ambient temperature to that discharged from the dry scrubber (typically l20°C above ambient). Additional energy is required to heat the feed material to the molten electrolyte temperature of 960°C;

ii. Moisture: calcined alumina readily equilibrates with moisture under ambient conditions. Loosely bound moisture (water) and structural hydroxyls enter the electrolytic cell with the alumina feed material, measured respectively in terms of Moisture on Ignition (MOI) and Loss on Ignition (LOI). Additional energy is required to drive off the excess moisture. Material data sheets issued with each alumina delivery capture the conditions at the exit of the calciner. Depending on bulk material transport method used and local conditions, the total moisture content of alumina can increase by 2wt% to 6wt.% or more before entering the electrolysis cell; and

iii. Time: extended exposure to ambient moisture will cause the reformation of structural hydroxyls in the form of surface crystals (e.g. Gibbsite, Boehmite) that increase alumina structural (LOI) moisture content, change surface pore structure and reduce material flowability. In the extreme, cases have been reported whereby the contents of an entire alumina storage silo have fused together.

[0014] Adsorption is a surface phenomenon. Surface adsorption is the adhesion of atoms, ions or molecules (adsorbate) in the gaseous phase onto the surface of an adsorbent. This process creates a film of adsorbate on the surface of the adsorbent. The forces attracting and retaining HF and ThO molecules (adsorbate) captured on to the surface of the alumina (adsorbent) during dry scrubbing can be categorized as either weak or strong, that being:

i. relatively weak attraction forces (e.g. van der Waals, London, Polar) between molecules is referred to as Physisorption (or physical adsorption). Physisorption increases with the gas pressure gradient and decreases with increasing temperature. Physisorption is in general favored by temperatures close to the boiling point of the adsorbate. Molecules can be physically adsorbed more than those in direct contact with the surface resulting in multilayer adsorption. Equilibrium is established between the absorbent’s surface and the gaseous phase molecules when under steady state conditions. A change in equilibrium conditions for example, reducing the gas pressure gradient and/or increasing the temperature would cause the removal of molecules from the adsorbent’s surface until a new equilibrium is established between the adsorbent and adsorbate. The equilibrium between the alumina’s surface and the gaseous phase molecules (i.e. HF and H2O) varies with changes to process conditions that occur in the dry scrubbing system and when fed to the molten electrolyte solution at elevated temperature;

ii. strong attraction forces between molecules is referred to as Chemisorption (or chemical adsorption). Chemisorption occurs when valence electrons in the outermost electronic“shell” of atoms interact to form chemical compounds that occupy adsorption sites on the absorbent’s surface. Chemisorbed molecules form only a monolayer at reaction sites of kinetically accessible pores. Chemisorption of molecules initially increases with temperature as sufficient energy is being provided for the molecules to reach the activation energy; however, after a certain degree it (chemisorption) decreases as the high-temperature helps in breaking of the bond (dissociative chemisorption) between the adsorbate and the adsorbed molecules; and iii. physisorption and chemisorption mechanisms can simultaneously occur during a reaction between alumina and HF and H2O. Molecules adsorbed by physisorption on to adsorbent’s surface can change to a stable chemisorbed bond with rising temperature.

[0015] Less than optimal fluorinated alumina quality fed to the molten electrolyte will impact the electrolytic cell operating efficiency and the fluoride and GHG emissions to the environment.

[0016] The release of the structural hydroxyl (OH-) component is very important for alumina dispersion and dissolution in to the molten electrolyte during feeding as it (hydroxyls) “blow apart” the alumina planes (layers) of hydrogen-bonded (H+) AIO, crystals stacked along the [001] direction.

[0017] Alpha phase alumina has essentially no residual hydroxyls and is thus difficult to dissolve in the molten electrolyte. Alpha phase alumina is also known to be prone to forming undesirable fines that inhibit bulk material flowability.

[0018] Alternatively, too much loosely bound moisture and/or an excess of hydroxyls can lead to (respectively) excessive HF formation via thermal and electro-chemical hydrolysis and alumina being ejected from the breaker-feeder hole due to the“volcano effect” created by rapid heating of the moisture within alumina particles resulting in the explosion of alumina particles and their expulsion away from the molten electrolyte.

[0019] The outer shell of the calcined alumina is typically alpha phase rich (over-calcined). This condition tends to isolate the internally bound hydroxyls from conditions that break-up alumina particles causing the alumina charged to raft on the molten electrolyte’s surface impeding dispersion and dissolution. It is also known that relatively cool alumina charged onto the surface of molten electrolyte tends to locally freeze the electrolyte forming a shell around the periphery of the alumina charge delaying the dissolution of the alumina until heated and the frozen electrolyte melted. Both excess moisture and heating of charged alumina detract from cell efficiency as more energy is required to maintain electrolyte superheat and are the principal causes for an anode effect to occur.

[0020] Optimal conditions that mitigate HF and GHG formation and that promote efficient HF capture and the return of adsorbed fluorides to the electrolytic melt, as well as, that which promote efficient electrolysis cell operation, include:

i. calcined alumina with a uniform transition phase throughout the particle regardless of particle size variation;

ii. calcined alumina pore structure that promotes the kinetic access of HF to reaction sites within pores;

iii. calcined alumina that maximizes the number of surface reaction sites, expressed as a function of specific surface area (SSA), and reactivity in terms of surface basicity density;

iv. calcined alumina that mitigates the migration of sodium ions to internal pore surfaces blocking HF access to internal reaction sites; v. optimal dry scrubbing conditions that promote HF capture and retention at internal reaction sites;

vi. chemisorbed fluoride byproduct species that are thermodynamically and concentration gradient stable as opposed to being unstable species formed by physisorption;

vii. removal of loosely bound moisture (MOI) associated with the alumina and all residual active phases (i.e. Gibbsite, Boehmite) leaving sufficient structural hydroxyls (LOI) to facilitate breakdown of hydrogen-bonded (H+) AIO, crystals stacked along the [001] direction thereby mitigating sources of variation and promoting dissolution of non-fluorinated and fluorinated alumina into the molten electrolyte; and

viii. Preheating the alumina, ideally to 960°C prior to being fed (dropped by gravity) on to the surface of a cell’s electrolytic melt.

BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present disclosure is described with reference to the accompanying drawings, in which like elements are referenced with like reference numbers, in which:

[0022] FIG. 1 is a schematic diagram illustrating a conventional system for the calcination of aluminum hydroxide.

[0023] FIG. 2 is a plot illustrating a Thermal Gravimetric Analysis (TGA) of non- fluorinated alumina after thermal treatment.

[0024] FIG. 3 is a plot illustrating morphological changes to cumulative (cc/g) and incremental dV(d) pore volume of non-fluorinated alumina with no thermal treatment compared to dry and humid thermal treatment. [0025] FIG. 4 is a plot illustrating morphological changes to cumulative (m²/g) and incremental dS(d) pore surface area of non-fluorinated alumina with no thermal treatment compared to dry and humid thermal treatment.

[0026] FIG. 5 is a plot illustrating cumulative (m²/g) and incremental dS(d) pore surface area distribution over pore width (A) for Gibbsite laboratory calcined 20h at temperatures varying between 300°C to 700°C.

[0027] FIG. 6 is a plot illustrating cumulative pore surface area (m²/g) for Gibbsite laboratory calcined 20h at temperatures varying between 300°C to 700°C.

[0028] FIG. 7 is a plot illustrating cumulative (cc/g) and incremental dV(d) pore volume distribution over pore width (A) for Gibbsite laboratory calcined 20h at temperatures varying between 300°C to 700°C.

[0029] FIG. 8 is a plot illustrating cumulative pore volume (cc/g) for Gibbsite laboratory calcined 20h at temperatures varying between 300°C to 700°C.

[0030] FIG. 9A is a plot illustrating Bulk C SGA reacted with humidified nitrogen carrier gas 1000 ppm HF at l20°C in an experimental fluid bed reactor to 70% saturation.

[0031] FIG. 9B is a plot illustrating Bulk C SGA reacted with humidified nitrogen carrier gas 1000 ppm HF at 400°C in an experimental fluid bed reactor to 70% saturation.

[0032] FIG. 10 is a plot illustrating Bulk C reacted samples thermally post-treated in humidified nitrogen carrier gas (0.0 ppm HF) at 400°C in an experimental fluid bed reactor.

[0033] FIG. 11 is a plot illustrating a TGA of Bulk C SGA reacted, thermally post-treated reference samples at l20°C vs. 400°C.

[0034] FIG. 12 is a plot illustrating nitrogen porosimetry of Bulk C SGA reacted, thermally post-treated reference samples at l20°C vs. 400°C. [0035] FIG. 13A is a plot illustrating Bulk C SGA HF captured by weight up to 70% breakout.

[0036] FIG. 13B is a plot illustrating Bulk C SGA HF released by weight in post thermal treatment at 400°C.

[0037] FIG. 13C is is a plot illustrating Bulk C SGA HF retained by weight in post thermal treatment at 400°C.

[0038] FIG. 13D is is a plot illustrating Bulk C SGA HF retention efficiency after post thermal treatment at 400°C.

[0039] FIG. 14A is a plot illustrating Bulk C SGA reacted at l20°C.

[0040] FIG. 14B is is a plot illustrating Gibbsite thermally treated @ 500°C and reacted at

400°C.

[0041] FIG. 15A is a plot illustrating Bulk C SGA reacted at l20°C and post treated at

400°C.

[0042] FIG. 15B is a plot illustrating Gibbsite thermally treated at 500°C, reacted at 400°C and post treated at 400°C.

[0043] FIG. 16 is a plot illustrating a TGA of Bulk C SGA reacted at l20°C compared to Gibbsite treated at 500°C and reacted at 400°C.

[0044] FIG. 17 is a plot illustrating nitrogen porosimetry of Bulk C reacted at l20°C compared to Gibbsite thermally treated at 500°C and reacted and post treated @ 400°C.

[0045] FIG. 18A is a plot illustrating pretreated Gibbsite HF captured by weight at 400°C up to 70% breakout.

[0046] FIG. 18B is a plot illustrating pretreated Gibbsite HF released by weight in post treatment at 400°C [0047] FIG. 18C is a plot illustrating pretreated Gibbsite HF retained by weight in post treatment at 400°C

[0048] FIG. 18D is a plot illustrating pretreated Gibbsite HF retention efficiency after post treatment at 400°C.

[0049] FIG. 19A is a plot illustrating Bulk C reacted at l20°C.

[0050] FIG. 19B is a plot illustrating Gibbsite thermally treated at 500°C and reacted at l20°C.

[0051] FIG. 20 is nitrogen porosimetry of Bulk C reacted at l20°C compared to Gibbsite thermally treated (laboratory calcined) at 500°C and reacted at l20°C.

[0052] FIG. 21A is a plot illustrating Bulk C HF captured by weight up to 70% breakout.

[0053] FIG. 21B is a plot illustrating Bulk C HF released by weight in post treatment at

400°C

[0054] FIG. 21C is a plot illustrating Bulk C HF retained by weight in post treatment at

400°C

[0055] FIG. 21D is a plot illustrating Bulk C HF retention efficiency after post treatment at

400°C

[0056] FIG. 22 is a system for enhancing smelter grade alumina quality by thermal treatment in an electrolysis cell superstructure.

DETAILED DESCRIPTION OF THE ILLUSTRATIVE EMBODIMENTS

[0057] The subject matter of the present disclosure is described with specificity, however, the description itself is not intended to limit the scope of the disclosure. The subject matter thus, might also be embodied in other ways, to include different structures, steps and/or combinations similar to and/or fewer than those described herein, in conjunction with other present or future technologies. Although the term“step” may be used herein to describe different elements of methods employed, the term should not be interpreted as implying any particular order among or between various steps herein disclosed unless otherwise expressly limited by the description to a particular order. Other features and advantages of the disclosed embodiments will be or will become apparent to one of ordinary skill in the art upon examination of the following figures and detailed description. It is intended that all such additional features and advantages be included within the scope of the disclosed embodiments. Further, the illustrated figures and dimensions described herein are only exemplary and are not intended to assert or imply any limitation with regard to the environment, architecture, design, or process in which different embodiments may be implemented.

[0058] The systems and methods disclosed herein thus, overcome the prior art disadvantages by calcining aluminum hydroxide (e.g. Gibbsite, Boehmite) at temperatures and heating rates below and for durations longer than that commonly used by the prior art, and the calcination of dried aluminum hydroxide in or outside the superstructure of an electrolysis cell. More particularly, the present disclosure relates to the reduction of capital and operating costs for aluminum hydroxide calcination and alumina smelting process systems and reducing the environmental impact of these processes using disclosed methods and systems to mitigate hydrogen fluoride (HF) and Green House Gases (GHG) emissions at the source. Although the following description refers to the aluminum smelting industry, the systems and methods described herein are not limited thereto and may also be applied in other industries and processes to achieve similar results.

[0059] In one embodiment, the present disclosure includes a method for calcining alumina hydroxide, which comprises: i) heating a composition comprising or comprising only alumina and alumina hydroxide inside or outside a superstructure of an electrolysis cell to a temperature above about 400°C; and ii) maintaining the temperature of the composition between the above about 400°C and about 550°C for no greater than about 20 hours.

[0060] In another embodiment, the present disclosure includes a method for capturing and retaining hydrogen fluoride on the surface of calcined alumina, which comprises: i) heating a composition comprising or comprising only the hydrogen fluoride and the calcined alumina at a temperature from greater than 400°C up to about 550°C before entering an electrolyte.

[0061] In yet another embodiment, the present disclosure includes a system for thermal treatment of a feed material in a superstructure of an electrolysis cell, comprising or comprising only alumina hydroxide, calcined smelter grade alumina or a composition comprising or comprising only the alumina hydroxide and the calcined smelter grade alumina, which comprises: i) an adjustable feed material level controller for maintaining a predetermined depth of feed material in a fluid bed of the electrolysis cell; ii) a plurality of feed material conveyors extending along a length of the superstructure and positioned below a gas skirt of the superstructure; iii) a plurality of stand-pipes connected to the plurality of feed material conveyors for delivering thermally treated feed material from the plurality of feed material conveyors to the fluid bed and venting moisture to an exhaust above the fluid bed; and iv) a plurality of fluid bed compartments positioned inside the superstructure for reacting the thermally treated feed material with hydrogen fluoride to form a fully calcined and fluorinated alumina by-product.

[0062] In one example, the present disclosure demonstrates the reduction in moisture content of non-fluorinated alumina resulting from the thermal treatment of alumina comprised of various phases. The residual moisture content, as measured by Thermal Gravimetric Analyses (TGA), of Gibbsite and commercially calcined alumina samples thermally treated (or not) are shown in FIG. 2. The results in FIG. 2 demonstrate: i. a significant reduction in residual moisture content of the Gibbsite alumina thermally treated above 400°C as compared to Gibbsite alumina thermally treated at 400°C;

ii. a significant reduction in residual moisture content of the commercially calcined Bulk C alumina thermally treated at 400°C as compared to Bulk C alumina thermally treated at l20°C or not thermally treated;

iii. elimination of loosely bound moisture of the commercially calcined Bulk C and Gibbsite alumina thermally treated above 400°C as evidenced by the absence of endothermic reactions in DTA measurements;

iv. elimination of residual under-calcined phases of Gibbsite and Boehmite, referred to as structural water (hydroxyls), that were thermally decomposed above 400°C as evidenced by the absence of an endothermic reaction in Gibbsite DTA measurements for samples treated at and above 450°C.

Table 1 Differential moisture reduction to the electrolytic cell post thermal treatment [0063] In addition to the above discoveries regarding the thermal treatment of alumina samples, Table 1 above demonstrates that significant moisture reduction of alumina fed to the molten electrolyte of an electrolysis cell, in terms of loosely bound water (MOI) and structural water (LOI), would be achieved by thermally treating commercially calcined smelter grade alumina at temperature above 400°C for up to 20 hours as compared to the prior art. Reduced moisture entering the electrolytic cell would significantly abate the formation of hydrogen fluoride, due to thermal and electro-chemical hydrolysis, as compared to the prior art.

[0064] In another example, the present disclosure demonstrates changes to commercially calcined smelter grade alumina (SGA) pore structure resulting from thermal treatment using disclosed methods that promote the kinetic access of HF to reaction sites within pores of non- fluorinated alumina and that increases the pore specific surface area as compared to the prior art.

[0065] The pore size distribution depicted in FIG. 3, for example, demonstrates the morphological change to the cumulative (cc/g) and incremental dV(d) pore volume of non- fluorinated alumina, with residual Gibbsite content (Bulk D), for three conditions: i) without thermal treatment (prior art) 3.1 and 3.4 respectively; ii) post thermal treatment (20h @ 400°C) under dry conditions (present disclosure) 3.2 and 3.5 respectively, and; iii) post thermal treatment (20h @ 400°C) under humid conditions (present disclosure) 3.3 and 3.6 respectively. The morphological change due to the thermal decomposition of the residual Gibbsite in the alumina, under dry and humid conditions, dramatically shifts the incremental pore volume dV(d) 3.7 to larger pores by increasing the pore diameters for microporous (defined as < 2qA diameter) to that of mesoporous (defined as 2qA to 5qA) that would reduce the probability of pore blockage at the entrance to a pore and pore attenuation due to blockage forming within the pore. [0066] In yet another example, the pore size distribution depicted in FIG. 4 demonstrates the morphological change to the cumulative (m²/g) and incremental dS(d) pore surface area of non- fluorinated alumina, with residual Gibbsite content (Bulk D), for three conditions: i) without thermal treatment (prior art) 4.1 and 4.4 respectively; ii) post thermal treatment (20h @ 400°C) under dry conditions (present disclosure) 4.2 and 4.5 respectively, and; iii) post thermal treatment (20h @ 400°C) under humid conditions (present disclosure) 4.3 and 4.6 respectively. The morphological change of the alumina of the present disclosure significantly increases both the cumulative pore surface area and incremental pore surface area dS(d), for pore diameters in the range of 35 A to 200A, as compared to alumina entering the electrolysis cell under conditions of the prior art (4.1 and 4.4 respectively). The morphological change due to the calcination of the residual Gibbsite in the alumina under dry and humid conditions dramatically shifts the incremental pore surface area dS(d) 4.7 to larger pores by increasing the pore diameters for microporous (defined as < 2qA diameter) to that of mesoporous (defined as 2qA to 5qA) that would significantly increase the number of internal reaction sites kinetically available for hydrogen fluoride adsorption on to the surface of alumina.

[0067] In yet another example, the present disclosure demonstrates changes to aluminum hydroxide (e.g. Gibbsite, Boehmite) pore structure, resulting from laboratory calcination using disclosed methods that promote the kinetic access of HF to reaction sites within pores of non- fluorinated alumina and that increases the pore specific surface area as compared to the prior art.

[0068] The pore size distribution depicted in FIG. 5 demonstrates the morphological change to the cumulative (m²/g) and incremental dS(d) pore surface area of non-fluorinated Gibbsite when laboratory calcined for 20h at temperatures of: a) 300°C (5.1-5.2; b) 400°C (5.3-5.4); c) 450°C (5.5-5.6); d) 500°C (5.7-5.8); e) 550°C (5.9-5.10); f) 600°C (5.11-5.12), and g) 700°C (5.13-5.14). The morphological change of Gibbsite as it is thermally decomposed (calcined) at increasing temperatures reduces both the cumulative pore surface area (m²/g) and the peak incremental pore surface area dS(d) thus, dramatically shifting the distribution lower (right) to larger pore sizes by increasing the pore diameters from microporous to that defined above as mesoporous sized pores.

[0069] However, as demonstrated in FIG. 6 for pore diameters above 35 A to 200 A the morphological change of the non-fluorinated Gibbsite of the present disclosure significantly increases the net pore surface area (m²/g), a measure of the number of reaction sites, to a maximum in and around 450°C. HF adsorption capability on to the surface of alumina is directly proportional to the number of kinetically accessible pore reaction sites. The net quantity of kinetically accessible reaction sites for hydrogen fluoride adsorption on to the surface of alumina is thus maximized and is in the order of twice that of the prior art (FIG. 4).

[0070] The pore size distribution depicted in FIG. 7 demonstrates the morphological change to the cumulative (cc/g) and incremental dV(d) pore volume of non-fluorinated Gibbsite when laboratory calcined for 20h at temperatures of: a) 300°C (7.1-7.2); b) 400°C (7.3-7.4); c) 450°C (7.5-7.6); d) 500°C (7.7-7.8); e) 550°C (7.9-7.10); f) 600°C (7.11-7.12), and g) 700°C (7.13- 7.14). The thermal decomposition of the Gibbsite in the alumina, under the above specified conditions, dramatically shifts the peak incremental pore volume dV(d) lower (right) to larger pore sizes by increasing the pore diameters for microporous (defined as < 2qA diameter) to that of mesoporous (defined as 2qA to 50A).

[0071] As demonstrated in FIG. 8 the morphological change of the non-fluorinated Gibbsite of the present disclosure significantly increases the cumulative pore volume (cc/g) to a maximum in and around 500°C. Increased pore access effectively reduces the occurrence of pore blockage at the entrance to a pore and pore attenuation due to a blockage forming within the pore. HF molecule kinetic access to internal pore reaction sites is thus maximized and is in the order of one and a half times that of the prior art (FIG. 3).

[0072] In yet another example, the present disclosure demonstrates the comparative change in mass capture (wt.%), mass retention (wt.%) and retention efficiency (%) and of the by-product species formed for HF reactions with commercially calcined smelter grade alumina (SGA) at conditions used by the prior art verses that described in the present disclosure.

[0073] A series of experimental tests using SGA (Bulk C) samples reacted with hydrogen fluoride (HF) in a laboratory scale fluid bed reactor were performed and the samples analyzed to assess changes to pore structure, mass loss when heated, fluoride speciation and fluoride content by (respectively) nitrogen porosimetry, thermogravimetric analysis (TGA) and magic angle spinning solid state nuclear magnetic resonance spectroscopy (¹⁹F SS MAS-NMR). Each experimental run is comprised of sample (a) run that was split and post treated at 400°C as a (b) run.

[0074] Results from the experimentation (presented below) demonstrate that though more HF by mass can be captured at a relatively low reaction temperature (l20°C as per the prior art), the retention efficiency (stability) under post thermal treatment, with varying thermodynamic and concentration gradient conditions, is significantly less as compared to HF captured and retained on to the surface of SGA reacted at a relatively high-temperature (e.g. 400°C).

[0075] The results shown in FIGS. 9A-9B comparing experimental sample run 8a (ER8a) at a low reaction temperature (l20°C) versus ER3 la reacted at a higher temperature of 400°C replicate industry knowledge (prior art) that HF capture is reduced as the reaction temperature increases. However, the thermal post-treatment at 400°C of the same samples indicates that HF retention efficiency as a percentage of HF mass captured dramatically increases as compared to the prior art (l20°C) when the reaction temperature increases to 400°C i.e. 37% retention versus 88% retention respectively per FIG. 10. These results demonstrate that the HF captured on to the alumina’s surface reacted at l20°C is predominately physisorbed and unstable.

[0076] Thermo-gravimetric analysis (TGA) of the sample material (ER8a) also demonstrated that alumina reacted with HF at l20°C readily released adsorbents from the alumina during heating at a rate of 5°C/min from ambient temperature to 1 l00°C resulting in a dramatic loss of mass attributed to relatively unstable physisorbed fluoride and water in the form of hydroxy- fluorides released from the alumina’s surface well before reaching an electrolyte temperature of 960°C. Comparatively, the TGA analysis of the same sample material post reaction and thermal treatment at 400°C (ER8b) indicates minimal mass loss. The remaining byproducts of reaction are thermodynamically stable (FIG. 11). Alternatively, the alumina reacted with HF at a temperature of 400°C (ER3 la) and post treated at 400°C (ER3 lb) lost minimal mass during the TGA test indicating that the majority of HF captured is most likely bonded (chemisorbed) to the alumina’s surface and thermodynamically stable up to and beyond the electrolyte temperature of 960°C. Most of the HF captured would be returned to the electrolyte under these reaction conditions.

[0077] Nitrogen porosimetry testing (shown in FIG. 12) capture the morphological changes to these samples reacted and treated under the above described conditions. The results compare the experimental samples (8 a/b; 31 a/b) and a reference Bulk C SGA sample unreacted and thermally pretreated at 400°C (12.9 - 12.10). The results demonstrate that the unit and cumulative Pore Surface Area (dS(d) & m²/g respectively) of the reacted Bulk C SGA samples (12-1-12.4), representing practices used by the prior art reacted at l20°C, provide evidence that the majority of pore surface area is inaccessible after reaction with HF and its capability to further adsorb HF being limited. Comparatively, the unit and cumulative Pore Surface Area (dS(d) & m²/g respectively) of the reacted Bulk C samples 3 la/b, (12.5-12.8) representing the present disclosure, are similar to thermally treated unreacted reference Bulk C SGA sample (12.9-12.10); samples are both capable of adsorbing more HF with an extended reaction time. Bulk C SGA sample 8a results present what appears to be blocked pore entrances for pore sizes > 35 A up to approximately 60 A in width, and that after thermal post treatment at 400°C (8b) there is evidence that a majority of pore surface area is accessible after thermal post Treatment at 400°C (8a vs 8b) that support above observations that HF were desorbed.

[0078] Results from the fluoride speciation and relative fluoride content by magic angle spinning solid state ¹⁹F nuclear magnetic resonance spectroscopy (¹⁹F SS MAS-NMR) are tabulated below in Table 2. Most of the species formed by reacting Bulk C SGA with HF at l20°C (prior art) are hydroxy-fluorides; no terminal Al-F species were formed. The surface species changed dramatically when samples were thermally post treated at 400°C with retained fluorides forming more stable terminal Al-F short bonds and reducing the quantity of hydroxy -fluorides. The analysis demonstrates that most of the HF captured by sample run 8a were physisorbed to the surface as supported by the above TGA and Porosimetry results for the same samples (8a vs. 8b). The total area under fluoride species peaks also demonstrate the magnitude of fluorides lost when thermal post treatment of the samples was performed. The fluoride loss is attributed to species that were thermodynamically unstable; thermal hydrolysis and hydrogen bond decomposition being the primary loss mechanism. The sodium impurity present in the sample 8a, reacted under conventional conditions, was present as an ionically bonded Na-F species and that after thermal post treatment (8b) at 400°C none was detected. For samples 3 la and 3 lb, reacted and post treated at 400°C, there were no Na-F species detected. The reaction of HF with SGA does not create any terminal or bridged sodium bonds with aluminum such as cryolite (Na3AlF₆) / chiolite (NasAbF 14).

Table 2 F19 SS MAS NMR of SGA Reacted and Post Treated @ 400°C - Total area under peaks & relative contribution by species

[0079] The experimental trends for HF capture, release, retention and retention efficiency over increasing reaction temperatures from l20°C to 400°C with SGA are presented in FIGS. 13A- 13D. The trends demonstrate a marginal increase in HF mass captured (wt. %) and HF retention efficiency (%) (FIG. 13D) for SGA thermally pretreated at temperatures greater than 300°C. These findings support the evidence presented in FIG. 3 and FIG. 4, whereby the thermal pretreatment of the SGA increases both the accessible specific surface area and pore volume (over pore sizes 35 A to 200A). This demonstrates that thermal pre-treatment of SGA at elevated temperature reduces HF diffusion resistance and pore blockage via increased pore size and formation of additional surface reaction sites.

[0080] In yet another example, the present disclosure demonstrates the comparative change in mass capture (wt.%), mass retention (wt.%) and retention efficiency (%) and of the by-product species formed for HF reactions with industrial grade Gibbsite that was thermally pretreated (laboratory calcined), at conditions described in the present disclosure, as compared to the experiments conducted using commercially calcined alumina at conditions used by the prior art.

[0081] A series of experimental tests using laboratory calcined Gibbsite samples reacted with hydrogen fluoride (HF) in a laboratory scale fluid bed reactor were performed and the samples analyzed to assess changes to pore structure, mass loss when heated, fluoride speciation and fluoride content by (respectively) nitrogen porosimetry, thermogravimetric analysis (TGA) and magic angle spinning solid state ¹⁹F nuclear magnetic resonance spectroscopy (¹⁹F SS MAS-NMR). Each experimental run is comprised of sample (a) run that was split and post treated at 400°C (b) run.

[0082] Results from the experimentation (presented below) demonstrate that the net HF mass retained (wt.%) and the retention efficiency (%) under post thermal treatment exceed that of the prior art using SGA at a reaction temperature below l25°C, and that the Gibbsite calcination temperature to achieve optimal HF capture and retention lies between 400°C up to 550°C. These findings correspond closely with that presented in FIG. 6 and FIG. 8, whereby the thermal pretreatment of Gibbsite results in the morphological change of the non-fluorinated Gibbsite that significantly increases (respectively) both the net pore surface area (m²/g), a measure of the number of reaction sites, to a maximum in and around 450°C and the cumulative pore volume (cc/g), a measure of kinetic access to internal pore reaction sites, to a maximum in and around 500°C.

[0083] The results shown in FIGS. 14A-14B comparing experimental sample run 8a (FIG. 14A) reacting HF at low temperature (l20°C) with SGA versus sample run 22a (FIG. 14B) reacting HF at a relatively high-temperature (400°C) with Gibbsite thermally pretreated at 500°C demonstrate that thermally pretreated Gibbsite is significantly more effective in the capture of HF than SGA at elevated temperature as compared to sample run 3 la in FIG. 9B. The total fluoride captured (wt.%) remains below that of SGA reacted at relatively lower temperature of l20°C (8a). However, thermal post-treatment at 400°C of the same samples (8a and 22a) indicates that HF retention efficiency as a percentage of HF mass captured, as with experimental run 22b, dramatically increases as compared to the prior art (l20°C) when the reaction temperature increases to 400°C (i.e. 37% retention versus 88% retention respectively per FIGS. 15A-15B). These results demonstrate that most of the HF captured on to the surface of thermally pretreated Gibbsite at a reaction temperature of 400°C is chemisorbed and would be retained within an electrolytic cell’s operating environment. The thermal post treatment of sample run 22a/b was performed under varying equilibria through multiple step-wise reductions in HF concentrations from 1000 ppm to 500 ppm and then to 0.0 ppm. Whereas, the post treatment of samples 8a/b and 3 la/b were performed under zero HF (0.0 ppm) conditions. The results shown in FIGS. 15A-15B demonstrate that changes to the HF concentration gradient (under constant temperature of 400°C) resulted in the loss of physisorbed fluoride from sample run 22b.

[0084] Thermo-Gravimetric Analysis (TGA) of the reacted samples demonstrate that a majority of the HF captured by the thermally preheated Gibbsite (FIG. 16, 22a vs 22b) was stable, chemisorbed to surface reaction sites, and to a large degree, not impacted by changes to the HF concentration gradient.

[0085] Nitrogen porosimetry testing (shown in FIG. 17) capture the morphological changes to the thermally treated Gibbsite samples (22a/b) reacted with HF and post treated using methods disclosed herein as compared to the prior art (8a/b). As with experimental results for samples 3 la/b reacted and treated at the same temperature in FIG. 12, there is no evidence of pore blockage or pore attenuation. There is also evidence that the cumulative specific surface area of the thermally treated, reacted and post thermal treatment Gibbsite samples is approximately double that for the SGA samples (i.e. -135 m²/g for 22b versus - 60 m²/g for 8b). This demonstrates that the thermally pretreated and reacted Gibbsite has added capability, with extended exposure time, to capture and retain additional HF as compared to the prior art.

[0086] Results from the fluoride speciation and relative fluoride content by magic angle spinning solid state ¹⁹F nuclear magnetic resonance spectroscopy (¹⁹F SS MAS-NMR) are tabulated below in Table 3. The NMR results of Gibbsite samples thermally pretreated at 400°C, 450°C and 500°C, and reacted with HF and post treated under the same conditions at 400°C (samples l7ab, 28ab and 22ab respectively) support the present disclosure in terms of byproduct species formed and respective stability with post thermal treatment. The results demonstrate that a majority of the HF captured by the Gibbsite calcined at relatively low temperature and for an extended period (e.g. 400°C to 550°C for up to 20 hours) and reacted with HF at a relatively high- temperature would be returned to the electrolyte under these reaction conditions. The notable results as compared to the prior art reactions using SGA (i.e. samples 8ab and 3 lab) include:

i. Sodium fluoride (Na-F) reaction byproduct is not formed;

ii. Sodium reaction byproducts formed involve bonds with aluminum and fluoride molecules e.g. cryolite (Na3AlF₆) / chiolite (NasAbF 14) which are relatively stable under post thermal treatment, and;

iii. No Hydroxy-fluorides are formed, mostly terminal Al-F bonds that appear to be stable after thermal post treatment - no water or hydroxyls present.

iiiiiiil

Table 3 F19 SS MAS NMR of SGA & Calcined Gibbsite Reacted and Post Treated @ 400°C - Total area under peaks & relative contribution by species

[0087] In FIGS. 18A-18D, the experimental trends for HF capture, release, retention and retention efficiency are illustrated using Gibbsite thermally pretreated at temperatures varying between 300°C and 700°C for 20 hours and reacted at 400°C under the same conditions as that for the above referenced smelter grade alumina (SGA).

[0088] In yet another example, the present disclosure demonstrates the comparative change in mass capture (wt.%), mass retention (wt.%) and retention efficiency (%) and of the by-product species formed for HF reactions under conditions used by the prior art (i.e. reaction temperature l20°C) with commercially calcined smelter grade alumina (SGA) verses industrial grade Gibbsite that was thermally pretreated (laboratory calcined) at conditions described in the present disclosure.

[0089] As with the above examples, the experiment was conducted using a laboratory calcined Gibbsite sample (thermally treated at 500°C for 20 hours) reacted with HF in the same laboratory scale fluid bed reactor with the samples analyzed to assess changes to pore structure and fluoride speciation and fluoride content by (respectively) nitrogen porosimetry and ¹⁹F SS MAS- NMR spectroscopy. Each experimental run is comprised of sample (a) run that was split and post treated at 400°C (b) run.

[0090] Results from this experiment (presented below) demonstrate that the HF captured (wt.%) by the alumina and the net HF mass retained (wt.%) and the retention efficiency (%) under post thermal treatment exceed that of the prior art using commercially calcined smelter grade alumina (SGA) at a reaction temperature of l20°C.

[0091] For the results shown in FIGS. 19A-19B, the samples were reacted with HF under the same experimental conditions; however, due to the extended HF adsorption time of the thermally treated Gibbsite (sample 49a), the experiment was prematurely stopped at 10% saturation as compared to the SGA sample run 8a that was stopped at 70% saturation. The results for experimental run 8a reacting HF at low temperature (l20°C) with SGA versus 49a reacting HF at the same temperature (l20°C) with Gibbsite (49a) thermally pretreated at 500°C demonstrate that thermally pretreated Gibbsite is significantly more effective in the capture of HF than SGA (8a). Despite the lower saturation level of the Gibbsite sample (49a 10% vs. 70% for 8a), the total fluoride captured (wt.%) is significantly greater than that of SGA.

[0092] FIG. 20 illustrates the morphological changes to the thermally treated Gibbsite samples (49a/b) reacted at l20°C with HF and post treated under the same conditions as compared to the prior art (8a/b). There is evidence of minor pore blockage or pore attenuation for the laboratory calcined Gibbsite; however, it is significantly lower than that (blockage) detected for the SGA sample (8a/b). There is also evidence that the cumulative surface area (SSA) of the thermally treated and reacted Gibbsite samples is nearly double that for the SGA samples. It is evident that the thermally pretreated and reacted Gibbsite has the capability to significantly outperform commercially calcined smelter grade alumina (SGA - prior art) over all reaction temperatures in terms of HF adsorption and retention.

[0093] The NMR results for Gibbsite treated at 500°C (present disclosure) and reacted with HF at l20°C per prior art practices (49a/b) are tabulated below in Table 4. The results support the following discoveries associated with the present disclosure relating to Gibbsite calcination and reaction with relatively low concentration of HF at relatively high temperature (a new operating domain):

i. greater conversion efficiency of hydroxy fluorides to terminal Al-F bonds when thermally post treated at 400°C, and;

ii. all hydroxy fluorides are eliminated post thermal treatment at 400°C reducing the transport of moisture to the electrolysis cell that would mitigate HF formation.

Table 4 F19 SS MAS NMR of SGA & Calcined Gibbsite Reacted a 120 C and Post Treated @ 400°C - Total area under peaks & relative contribution by species

[0094] FIGS. 21A-21D illustrate experimental trends for HF capture, release, retention and retention efficiency using Gibbsite thermally pretreated at 500°C for 20 hours and reacted at l20°C. The results are presented overlaid that for the above referenced smelter grade alumina (SGA) trends presented in FIGS. 13A-13D. The experimental results demonstrate that the HF mass captured (wt.%) and retained (wt.%) and the retention efficiency (%) by Gibbsite treated at 500°C (present disclosure) exceed that of the prior art using SGA at a reaction temperature of l20°C.

[0095] Experimental research demonstrates a new operating domain for the thermal decomposition (calcination) of aluminum hydroxide and the capture (chemisorption) of HF onto the surface of calcined alumina. This disclosure also relates to the formation of calcined alumina properties that enhance the return of stable fluoride species formed on the surface of reacted (fluorinated) alumina and the dissolution of the alumina into the electrolytic melt as compared to the prior art. The methods disclosed herein therefore include:

i. thermally treating (calcinating) alumina with an aluminum hydroxide (e.g. Gibbsite, Boehmite) content ranging between 0.0 wt.% to 100 wt.%, heated in or outside the superstructure of an electrolysis cell at a relatively low rate to, and held at, a temperature above 400°C to 550°C for up to 20 hours or more resulting in the thermal decomposition of the aluminum hydroxides forming meta-stable transition phase aluminas through-out the alumina particle’s crystalline structure regardless of particle size; and

ii. capturing and retaining hydrogen fluoride (HF) onto the surface of calcined alumina that is thermally calcined by methods disclosed herein and/or conventionally and reacted with HF at a temperature greater than 400°C and up to about 550°C before entering the cell’s molten electrolyte.

[0096] As compared to the prior art, the methods disclosed herein include the following benefits that mitigate HF and GHG (e.g. CO2, CF₄, C2F6) formation and promote efficient HF capture and the return of adsorbed fluorides to the electrolytic melt, as well as efficient electrolysis cell operation, specifically:

i. an alumina composition that contains only meta-stable transition phase(s) throughout the alumina particle’s structure regardless of particle size. The calcined alumina composition thus does not contain any under-calcined phases (e.g. Gibbsite, Boehmite) that promote HF formation by way of electro-chemical hydrolysis of residual hydroxides (OH-) or over-calcined phases (e.g. alpha) that impede alumina dissolution in to the molten electrolyte that form perfluorocarbon emissions by way of an anode effect(s);

ii. calcined alumina that is preheated and fluorinated before being fed to the molten electrolyte removing the loosely bound moisture from the alumina thus eliminating the formation of HF due to thermal hydrolysis of HF adsorbed on to the surfaces of fluorinated alumina, and electrolyte present as a liquid droplet, a gaseous vapor or a condensed particulate and reducing the energy required to maintain electrolyte superheat as compared to the prior art;

iii. a calcined alumina pore structure that promotes the kinetic access of HF molecules to reaction sites within pores due to increased pore width (d) and increased pore volume (cc/g) that reduce the probability of pore blockage at the entrance to a pore and pore attenuation due to blockage forming within the pore. The kinetic access of HF molecules to internal pore reaction sites is thus enhanced by approximately one and a half times that of the prior art; iv. a calcined alumina pore structure and reactivity that increases the pore surface area dS(d) for pore sizes 35 A to 200A. The ability to capture and retain additional HF is thus enhanced by approximately twice that of the prior art;

v. calcined alumina that when reacted with HF at a temperature greater than 400°C and up to about 550°C that does not form a hydroxy-fluoride by-product thus eliminating the transport of moisture (hydroxyls and hydrates) associated with the byproduct (e.g. AlFx (OH) -x · nH₂0) to the molten electrolyte eliminating the formation of HF due to thermal hydrolysis of electrolyte present as a liquid droplet, a gaseous vapor or a condensed particulate and the loss of fluoride return to the electrolyte;

vi. a calcined alumina structure that mitigates the migration of residual sodium (Na) from the internal lattice structure of alumina to the surface of pores during thermal decomposition, that tend to block and attenuate narrow pores by sodium molecules, thus enhancing HF molecule access to and capture of at internal reaction sites;

vii. HF adsorption on to the surface of alumina that forms chemisorbed terminal or bridged bonds to aluminum molecules (e.g. Al-F, cryolite (Na3AlF₆), chiolite (NasAbFw)) that are thermodynamically stable and concentration gradient stable (e.g. HF, H2O) and that tend not to form hydroxy-fluoride or sodium fluoride species (e.g. AlF_x (OH)₃-x · nH₂0, Na-F) that are typically physisorbed (not chemisorbed) on to the alumina’s surface. Thus, upwards of 90% of the mass of reaction by-products formed on the alumina’s surface are retained under varying thermodynamic and concentration gradient conditions (such as that when fluorinated alumina is fed to the molten electrolyte of an electrolysis cell) as compared to a low of 40% of the reaction byproduct mass retained by the prior art; viii. HF adsorption on to the surface of calcined alumina that exceed that of the prior art in terms of HF mass captured (wt.%) and mass retained (wt.%) and retention efficiency (%). Thus, the mass of fluoride reaction by-products formed on the alumina’s surface at reaction temperatures up to 550°C that are retained when fluorinated alumina is fed to the molten electrolyte of an electrolysis cell would exceed that of the reaction byproduct mass retained by the prior art; and ix. the thermal decomposition (calcination) of aluminum hydroxide within or outside the electrolysis cell eliminating the variation of calcined alumina quality entering the electrolyte in terms of temperature, moisture content, pore structure (specific surface area and pore volume) and material flowability. Thus, HF capture and retention performance, cell operating efficiency and the reduction of anode effects are enhanced as compared to the prior art.

[0097] In yet another example, the present disclosure includes a system for the thermal treatment (calcination) of dried aluminum hydroxide or commercially calcined smelter grade alumina or a blend of these materials (the feed material) 2206 inside the superstructure 2202 of an electrolysis cell 2204 as illustrated in FIG. 22, which comprises:

i. gravity feed of dried aluminum hydroxide or commercially calcined smelter grade alumina or a blend of these feed materials (the feed material) 2206 to the superstructure 2202 of an electrolysis cell 2204;

ii. a material feed distribution system 2208 that would equally feed a plurality of material conveyors 2210 at a rate based on the electrolysis cell 2204 operating production condition;

iii. an adjustable feed material level control system 2212 that would set and maintain the selected fluid bed 2214 depth for a range of cell operating production conditions; iv. a plurality of material conveyors 2210 horizontally spanning the length of the superstructure 2202 located below the gas skirt 2216 and exposed to hot exhaust gases 2218 and radiant heat from the electrolysis cell’s crust 2220, forming partially calcined material 2224 in this the first stage of a multi-stage thermal treatment (calcination) system;

v. a plurality of stand-pipes 2222 connected to material conveyors 2210 located below the gas skirt 2216 delivering thermally treated (partially calcined) material 2224 from the material conveyors 2210 to the fluid bed 2214 at a rate based on the cell operating production conditions and venting moisture 2226, derived from thermal treatment (calcination) of the feed material 2206 forming partially calcined material 2224, to an exhaust point 2230 above the surface of the fluid bed 2214, venting the moisture 2226 directly to the filter system 2228 and cell exhaust connection 2232; and

vi. a plurality of fluid bed 2214 compartments configured inside an electrolytic cell’s superstructure 2202 wherein simultaneous second stage thermal treatment (calcination) of the partially calcined material 2224 and reaction with HF in the hot exhaust gases 2218 occurs to form a fully calcined and fluorinated alumina byproduct 2234 that is suitable for feeding to the electrolytic cell.

[0098] The methods and systems disclosed herein are intended to be applied to either new electrolysis cell 2204 superstructure 2202 applications or the retrofit of existing conventional prebake electrolysis cell superstructures. Modifications to existing electrolysis cell superstructures to produce fluorinated alumina byproduct of the required quality will vary with electrolysis cell size and design. [0099] For the retrofit of existing conventional prebake electrolysis cell superstructures, the fluorinated alumina byproduct is normally stored in a plurality of static bins located in the upper section of the superstructure. The conversion of an existing superstructure to accommodate the disclosed methods and systems would require the removal of a plurality of existing static bins and the installation of the systems disclosed herein including but not limited to the following structures: 2208, 2210, 2212, 2214, 2222, 2224, 2226, 2228 and 2228 to support a functioning fluid bed 2234.

[0100] Commercial calcination of aluminum hydroxide to form smelter grade alumina commonly utilizes the circulating fluid bed process as generally described and schematically illustrated in FIG. 1. For the fluid bed calcination process shown in FIG. 1, the dried aluminum hydroxide 3 could be extracted at the discharge of the electrostatic gas cleaning system 8 via the discharge chute 9 downstream by introducing a device to divert part of or all of the process stream to a dried aluminum hydroxide storage bin (not shown). Alternatively, the dried aluminum hydroxide 3 could be extracted at the discharge of the series of separation cyclones 12 and 18 before the dried aluminum hydroxide 3 is fed to a fluid bed reactor 20. The extraction location is a function of aluminum hydroxide dryness to meet specified requirements and will vary with specifics of the calcination facility systems and operating practices. Such modifications to existing commercial calcination processes would allow the facility to operate in three modes, converting production mode as product demand changes:

i. 100% smelter grade alumina using existing methods and systems;

ii. 100% dried aluminum hydroxide using existing (or new) methods and systems; iii. Simultaneous production of both smelter grade alumina and dried aluminum hydroxide using existing (or new) methods and systems.

[0101] Other processes and configurations exist for the calcination of aluminum hydroxide to form smelter grade alumina e.g. rotary kilns, flash calcination systems. Modifications for these systems to produce wholly or partial production of dried aluminum hydroxide and smelter grade alumina vary with each process; however, the scope of the modifications would be of similar magnitude as dried aluminum hydroxide is an intermediate process step before calcination at high temperature that is common to all calcination processes.

[0102] As compared to the prior art, the systems disclosed herein therefore include the following benefits:

i. eliminate the need for process steps and related equipment to calcine dried aluminum hydroxide at a commercial process plant, as illustrated in FIG. 1 by the equipment enclosed by the dashed line, including the preheating of dried aluminum hydroxide, separation of the aluminum hydroxide from process gases, high- temperature calcination of the aluminum hydroxide and cooling of the calcined alumina product;

ii. eliminate fuel combustion and electrical power consumption related to the calcination of dried aluminum hydroxide that result in the emission of carbon dioxide (CO2); and

iii. eliminate the bulk transport of calcined alumina from the refinery to the smelter site, material handling within the smelter facility and finally storage in the electrolytic cell’s superstructure. Instead, dried aluminum hydroxide would be delivered to the superstructure of the electrolysis cell improving the consistency (quality) of calcined alumina being fed to the electrolyte of an electrolysis cell.

[0103] While the present disclosure has been described in connection with presently preferred embodiments, it will be understood by those skilled in the art that it is not intended to limit the disclosure to those embodiments. It is therefore, contemplated that various alternative embodiments and modifications may be made to the disclosed embodiments without departing from the spirit and scope of the disclosure defined by the appended claims and equivalents thereof.

Claims

1. A method for calcining alumina hydroxide, which comprises:

heating a composition comprising alumina and alumina hydroxide inside or outside a superstructure of an electrolysis cell to a temperature above about 400°C; and

maintaining the temperature of the composition between the above about 400°C and about 550°C for no greater than about 20 hours.

2. The method of claim 1, wherein the aluminum hydroxide is Gibbsite or Boehmite.

3. The method of claim 1, wherein the calcined alumina hydroxide does not contain any under-calcined phases that promote hydrogen fluoride formation or over-calcined phases that impede alumina dissolution.

4. The method of claim 1, wherein the formation of a hydroxy fluoride byproduct on a surface of the calcined alumina hydroxide, including transporting moisture associated with the hydroxy fluoride byproduct to a molten electrolyte, is eliminated.

5. The method of claim 1, wherein the calcined alumina hydroxide includes a pore structure that enables kinetic access of hydrogen fluoride to internal pore reaction sites in the calcined alumina hydroxide.

6. The method of claim 5, wherein the calcined alumina hydroxide pore structure and reactivity increases a pore surface area for pore sizes from about 35 A to about 200 A.

7. The method of claim 1, wherein the calcined alumina hydroxide mitigates migration of sodium from an internal lattice structure of the alumina to a pore structure of the calcined alumina hydroxide.

8. The method of claim 1, wherein the calcined alumina hydroxide inside the superstructure of the electrolysis cell eliminates a variation of calcined alumina attributes entering an electrolyte in the electrolysis cell, the attributes including at least one of a temperature, a moisture content, a pore structure and a material flow.

9. A method for capturing and retaining hydrogen fluoride on the surface of calcined alumina, which comprises:

heating a composition comprising the hydrogen fluoride and the calcined alumina at a temperature from greater than 400°C up to about 550°C before entering an electrolyte.

10. The method of claim 9, further comprising:

forming stable reaction byproducts on the surface of the calcined alumina after heating the composition; and

retaining about 90% of mass of the reaction byproducts.

11. A system for thermal treatment of a feed material in a superstructure of an electrolysis cell, comprising alumina hydroxide, calcined smelter grade alumina or a composition comprising the alumina hydroxide and the calcined smelter grade alumina, which comprises:

an adjustable feed material level controller for maintaining a predetermined depth of feed material in a fluid bed of the electrolysis cell;

a plurality of feed material conveyors extending along a length of the superstructure and positioned below a gas skirt of the superstructure;

a plurality of stand-pipes connected to the plurality of feed material conveyors for delivering thermally treated feed material from the plurality of feed material conveyors to the fluid bed and venting moisture to an exhaust above the fluid bed; and

a plurality of fluid bed compartments positioned inside the superstructure for reacting the thermally treated feed material with hydrogen fluoride to form a fully calcined and fluorinated alumina by-product.

12. The system of claim 11, wherein the thermally treated feed material and the fluorinated alumina in the fluid bed compartments are dehydrated of loosely bound moisture.

13. The system of claim 11, wherein the thermally treated feed material and the fluorinated alumina in the fluid bed compartments are partially dehydrated of structural hydroxyls.

14. The system of claim 11, further comprising a feed material distributor for equally distributing the feed material from a gravity feed in the superstructure to the plurality of feed material conveyors at a rate based on operating conditions for the electrolysis cell.