@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Materials Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Fukuda, Hiroki"@en ; dcterms:issued "2019-07-19T15:47:58Z"@en, "2019"@en ; vivo:relatedDegree "Master of Applied Science - MASc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description "Lithium is an essential metal for our society. Notably, increasing energy storage system will necessitate much more lithium in the future. This study focused on brine deposit while lithium exists in hard rocks as well. Conventionally, solar evaporation has been used to concentrate lithium from brine, but it takes more than one year. Thus, a more rapid process is desired for the accelerating demand. Here, two types of adsorbent, ion exchange (IX) resin and heterosite ferric phosphate (FP), were studied in order to extract lithium selectively from brine rapidly. First, more than thirty IX resins were tested in lithium chloride solution. Out of the thirty, sulfonate, iminodiacetate and aminomethylphosphonate resins succeeded in extracting lithium with the value of 16.3–32.9 mg-Li/g. However, no resins could adsorb lithium from a mixed brine solution which contains other interfering cations like sodium. An aluminum loaded resin was also tested since some past studies had reported lithium selectivity with this material. Its adsorption density was 6.6 mg-Li/g and was higher than any other resins tested for the mixed brine in this study. Nevertheless, the overall results showed that the IX resins were not so suitable for lithium extraction from a mixed brine. Then, heterosite FP was investigated as an alternative adsorbent. The FP can adsorb lithium selectively with the addition of a reducing agent to form lithium iron phosphate. This study used thiosulfate (TS) and sulfite (SF) individually as a reducing agent. The maximum adsorption density was 45.9 mg-Li/g by SF reduction at 65 °C, which is almost the same as the theoretical value of 46.0 mg-Li/g. The maximum selectivity over sodium was 2541 by SF reduction at 45 °C. Additionally, it was confirmed that the FP could be recycled by persulfate oxidation without degradation. Finally, the kinetics was studied and fit using pseudo first-order and shrinking sphere model. The two models fit the experimental results and indicated that the lithium extraction reaction was chemical reaction controlled. Since the FP method was found to be promising, it is highly recommended that it should be developed further by using natural brine sources."@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/71058?expand=metadata"@en ; skos:note " LITHIUM EXTRACTION FROM BRINE WITH ION EXCHANGE RESIN AND FERRIC PHOSPHATE by HIROKI FUKUDA B.Eng., Waseda University, Japan, 2017 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Materials Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) July 2019 © Hiroki Fukuda, 2019 - ii - The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled: Lithium extraction from brine with ion exchange resin and ferric phosphate submitted by Hiroki Fukuda in partial fulfillment of the requirements for the degree of Master of Applied Science in Materials Engineering Examining Committee: David Dreisinger, Materials Engineering Supervisor David Dixon, Materials Engineering Supervisory Committee Member Berend Wassink, Materials Engineering Supervisory Committee Member Daan Maijer (Chair), Materials Engineering Additional Examiner Additional Supervisory Committee Members: Supervisory Committee Member Supervisory Committee Member - iii - Abstract Lithium is an essential metal for our society. Notably, increasing energy storage system will necessitate much more lithium in the future. This study focused on brine deposit while lithium exists in hard rocks as well. Conventionally, solar evaporation has been used to concentrate lithium from brine, but it takes more than one year. Thus, a more rapid process is desired for the accelerating demand. Here, two types of adsorbent, ion exchange (IX) resin and heterosite ferric phosphate (FP), were studied in order to extract lithium selectively from brine rapidly. First, more than thirty IX resins were tested in lithium chloride solution. Out of the thirty, sulfonate, iminodiacetate and aminomethylphosphonate resins succeeded in extracting lithium with the value of 16.3–32.9 mg-Li/g. However, no resins could adsorb lithium from a mixed brine solution which contains other interfering cations like sodium. An aluminum loaded resin was also tested since some past studies had reported lithium selectivity with this material. Its adsorption density was 6.6 mg-Li/g and was higher than any other resins tested for the mixed brine in this study. Nevertheless, the overall results showed that the IX resins were not so suitable for lithium extraction from a mixed brine. Then, heterosite FP was investigated as an alternative adsorbent. The FP can adsorb lithium selectively with the addition of a reducing agent to form lithium iron phosphate. This study used thiosulfate (TS) and sulfite (SF) individually as a reducing agent. The maximum adsorption density was 45.9 mg-Li/g by SF reduction at 65 °C, which is almost the same as the theoretical value of 46.0 mg-Li/g. The maximum selectivity over sodium was 2541 by SF reduction at 45 °C. Additionally, it was confirmed that the FP could be recycled by persulfate oxidation without degradation. Finally, the kinetics was studied and fit using pseudo first-order and shrinking sphere model. The two models fit the experimental results and indicated that the lithium extraction - iv - reaction was chemical reaction controlled. Since the FP method was found to be promising, it is highly recommended that it should be developed further by using natural brine sources. - v - Lay Summary Mobile phones and electric cars always have a battery. Some kinds of battery contain lithium (Li) and are called lithium batteries. As we buy a new mobile phone with better quality and buy an electric car for an eco-friendly lifestyle, Li demand has been increasing. However, with a conventional method using solar heat, it may take more than one year to concentrate lithium from brines, which are saltwater and primary lithium sources. More rapid technologies are required to meet the rapidly increasing demand. This work investigated two extraction techniques using a solid adsorbent with experimental results for speedy lithium recovery. The extraction of lithium was demonstrated as a potential pathway to rapid lithium recovery from saltwater solution. - vi - Preface This thesis is original, unpublished, independent work by the author, Hiroki Fukuda. - vii - Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ................................................................................................................................ v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ vii List of Tables ................................................................................................................................. xi List of Figures ............................................................................................................................. xiii Nomenclature ............................................................................................................................. xxi Acknowledgements .................................................................................................................. xxiv Chapter 1. Introduction .......................................................................................................... 1 Chapter 2. Literature Review ................................................................................................. 2 2.1. Lithium Chemistry ............................................................................................................ 2 2.1.1. Electrochemical Properties of Lithium ...................................................................... 2 2.1.2. Lithium Ion Battery.................................................................................................... 3 2.1.3. Eh-pH Diagram .......................................................................................................... 4 2.2. Lithium Market ................................................................................................................. 8 2.2.1. Supply/Demand and Price .......................................................................................... 8 2.2.2. Products and Applications ....................................................................................... 10 2.3. Lithium Deposits ............................................................................................................. 13 2.3.1. Overview .................................................................................................................. 13 2.3.2. Ore Deposits............................................................................................................. 16 2.3.3. Brine Deposits .......................................................................................................... 17 - viii - 2.4. Processes for Lithium Production ................................................................................... 19 2.4.1. Overview .................................................................................................................. 19 2.4.2. Process of Ore Deposits ........................................................................................... 20 2.4.3. Process of Brine Deposits ........................................................................................ 21 2.4.4. Process for Lithium Hydroxide Production ............................................................. 31 2.5. Lithium Extraction Technologies .................................................................................... 33 2.5.1. Overview .................................................................................................................. 33 2.5.2. Phosphate Precipitation ............................................................................................ 35 2.5.3. Ion Exchange Resin ................................................................................................. 36 2.5.4. Al/Mn/Ti-based Adsorbent ...................................................................................... 42 2.5.5. Solvent Extraction .................................................................................................... 43 2.5.6. Membrane Technologies .......................................................................................... 44 2.5.7. Ferric Phosphate Reduction ..................................................................................... 45 2.6. Summary and Objectives ................................................................................................ 56 Chapter 3. Experimental Methods ....................................................................................... 59 3.1. Preparation ...................................................................................................................... 59 3.1.1. Materials .................................................................................................................. 59 3.1.2. Apparatus ................................................................................................................. 63 3.2. Procedures for Lithium Extraction with Ion Exchange Resin ........................................ 64 3.2.1. Preparation of Ion Exchange Resin .......................................................................... 64 3.2.2. Lithium Loading Test .............................................................................................. 67 3.2.3. Lithium Stripping Test ............................................................................................. 68 3.3. Procedures for Lithium Extraction with Ferric Phosphate ............................................. 69 3.3.1. Preparation of Ferric Phosphate ............................................................................... 69 - ix - 3.3.2. Lithium Loading Test .............................................................................................. 70 3.3.3. Lithium Stripping Test ............................................................................................. 71 3.3.4. Acid Digestion ......................................................................................................... 71 3.4. Analysis Procedures ........................................................................................................ 72 3.4.1. Atomic Absorption Spectroscopy (AAS) ................................................................ 72 3.4.2. Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) .............. 72 3.4.3. X-ray Diffraction (XRD) ......................................................................................... 73 Chapter 4. Lithium Extraction with Ion Exchange Resins ................................................ 74 4.1. Lithium Recovery from LiCl Solution ............................................................................ 74 4.2. Lithium Recovery from Brine ......................................................................................... 76 4.3. Summary ......................................................................................................................... 78 Chapter 5. Lithium Extraction with Ferric Phosphate ...................................................... 80 5.1. Preparation of Ferric Phosphate ...................................................................................... 80 5.1.1. Ferric Phosphate Dihydrate...................................................................................... 80 5.1.2. Lithium Iron Phosphate (Reagent Grade) ................................................................ 81 5.1.3. Lithium Iron Phosphate (Battery Grade) ................................................................. 82 5.2. Selection of Reducing Agents ......................................................................................... 83 5.3. Products after Lithium Loading ...................................................................................... 85 5.3.1. Thiosulfate Reduction .............................................................................................. 85 5.3.2. Sulfite Reduction ..................................................................................................... 89 5.4. Capacity and Selectivity ................................................................................................. 93 5.4.1. Thiosulfate Reduction .............................................................................................. 93 5.4.2. Sulfite Reduction ................................................................................................... 105 5.5. Effect of pH .................................................................................................................... 116 - x - 5.6. Recycling of Ferric Phosphate ...................................................................................... 120 5.7. Kinetics ......................................................................................................................... 124 5.7.1. Kinetic Models ....................................................................................................... 124 5.7.2. Thiosulfate Reduction ............................................................................................ 126 5.7.3. Sulfite Reduction ................................................................................................... 131 5.8. Summary ....................................................................................................................... 136 Chapter 6. Conclusion ......................................................................................................... 137 References .................................................................................................................................. 140 - xi - List of Tables Table 2.1 Properties of possible anode materials.35 .................................................................... 2 Table 2.2 Amount of lithium used in various devices.35 ............................................................ 11 Table 2.3 Lithium compounds and applications.35 .................................................................... 11 Table 2.4 Lithium mining sites and resources in tonnes (ores).40 ........................................... 15 Table 2.5 Lithium mining sites and resources in tonnes (brines).40 ........................................ 16 Table 2.6 Lithium minerals.40..................................................................................................... 17 Table 2.7 Analytical values of continental brines.40 ................................................................. 18 Table 2.8 Solubility of chloride salts at 25 °C.44 ....................................................................... 23 Table 2.9 Solubility of hydroxide salts at 25 °C.44 .................................................................... 23 Table 2.10 Solubility of carbonate salts at 25 °C.44 .................................................................. 24 Table 2.11 Technologies for lithium extraction from brine and their mechanism. ............... 33 Table 2.12 Solubility of phosphate salts at 25 °C.44 .................................................................. 36 Table 2.13 List of functional groups of IX resin studied in this thesis. .................................. 39 Table 3.1 List of resins tested in this study. .............................................................................. 61 Table 3.2 Actual density and swelling ratio of resins. .............................................................. 65 Table 3.3 Metal ion concentration in synthetic LiCl solution and brine. ............................... 67 Table 3.4 Metal ion concentration in eight synthetic solutions. .............................................. 70 Table 3.5 Wavelength for the measured elements. ................................................................... 73 Table 5.1 Molarity, molality and mean ionic activity of chloride salts and water activity in B1 solution at 25 °C, calculated by Meissner’s method71. ............................................................. 95 Table 5.2 k value and activation energy calculated from kinetics in the case of thiosulfate - xii - reduction. ................................................................................................................................... 129 Table 5.3 k value and activation energy calculated from kinetics in the case of sulfite reduction. ................................................................................................................................... 134 - xiii - List of Figures Figure 2.1 Schematic illustration of charge-discharge in lithium ion batteries, Me= Co, Fe+P, Mn, Ni+Mn+Co, Ni+Co+Al, Ti, S, etc.2 ....................................................................................... 3 Figure 2.2 pH-Eh diagram for Li-H2O system at 25 °C, 1 atm, (a) [Li]=1 M or (b) [Li]=0.072 M (Drawn by HSC). ...................................................................................................................... 5 Figure 2.3 pH-Eh diagram for Li-C-H2O system at 25 °C, 1 atm and (a) [Li]=[C]=1 M or (b) [Li]=[C]=0.072 M (Drawn by HSC). ........................................................................................... 7 Figure 2.4 Lithium supply/demand and price and their estimates.38 ....................................... 8 Figure 2.5 Annual use of lithium in tonnes in each of the primary lithium usage industries from 2002 to 2020.39 .................................................................................................................... 10 Figure 2.6 Lithium value chain.35 .............................................................................................. 12 Figure 2.7 (a) Resources and (b) Reserves in terms of deposit types in thousands metric tonnes Li metal equivalent.2 ....................................................................................................... 13 Figure 2.8 (a) Resources and (b) Reserves in terms of countries in thousands metric tonnes Li metal equivalent.3 ................................................................................................................... 14 Figure 2.9 An example of flow chart of Li2CO3 production from ores.41 ............................... 20 Figure 2.10 A simplified flow chart of Li2CO3 production from ores.42 ................................. 21 Figure 2.11 A schematic illustration of lithium production from brines.43 ............................ 22 Figure 2.12 A representative flow of lithium carbonate production from brines.40 .............. 22 Figure 2.13 Process flow of Salar de Atacama, Chile.42 ........................................................... 25 Figure 2.14 Process flow of Salar de Hombre Muerto, Argentina.42 ...................................... 26 Figure 2.15 Process flow of Salar de Oraroz, Argentina.42 ...................................................... 27 - xiv - Figure 2.16 Process flowsheet for lithium recovery in China & Tibet.42 ............................... 28 Figure 2.17 Process flow of CITIC method.42 ........................................................................... 29 Figure 2.18 Process flow of POSCO method.42 ........................................................................ 30 Figure 2.19 Process flow of Bateman method.42 ....................................................................... 31 Figure 2.20 Chemical structure of SDB copolymer. ................................................................. 37 Figure 2.21 Chemical structure of resin with (a) sulfonate, (b) quaternary ammonium. .... 38 Figure 2.22 A flow sheet of lithium production using solvent extraction by Tenova.9 .......... 44 Figure 2.23 Schematic illustrations of heterosite FP structure: (a) oblique view, (b) view along the c-axis (Drawn by VESTA). ................................................................................................... 46 Figure 2.24 Schematic illustrations of triphylite structure: (a) oblique view, (b) view along the c-axis (Drawn by VESTA). ................................................................................................... 48 Figure 2.25 pH-Eh diagram for TS-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC). Sulfate species (SO42- and HSO4-) were excluded in this diagram. ......................................... 50 Figure 2.26 pH-Eh diagram for PS-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC)........................................................................................................................................................ 51 Figure 2.27 pH-Eh diagram for I-H2O system at 25 °C, 1 atm, [I]=1 M (Drawn by HSC). . 53 Figure 2.28 pH-Eh diagram for SF-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC)........................................................................................................................................................ 55 Figure 4.1 Lithium adsorption density of resins in LiCl solution in mg-Li/g-dry resin. ...... 75 Figure 4.2 Correlation between results of loading and stripping tests in the case of lithium chloride solution (A1 solution). .................................................................................................. 76 Figure 4.3 Lithium adsorption of resins in brine in mg-Li/g-dry resin. No Data = No experiment was conducted with brine because of its low capacity in pure LiCl solution. ... 77 Figure 4.4 Correlation between results of loading and stripping tests in the case of brine - xv - solution (A2 solution). ................................................................................................................. 78 Figure 4.5 Lithium adsorption density of the resins which showed higher selectivity in LiCl solution. ........................................................................................................................................ 79 Figure 5.1 XRD patterns of ferric phosphate dihydrate calcined at 300, 500 and 600 °C. .. 80 Figure 5.2 XRD patterns of ferric phosphate obtained by calcination of ferric phosphate dihydrate at 600 °C and samples after lithium loading tests at 20 °C and 65 °C. ................ 81 Figure 5.3 XRD patterns of reagent-grade LFP and delithiated and lithiated ones at 65 °C in comparison with calcined FPD. ................................................................................................. 82 Figure 5.4 XRD patterns of battery-grade (BG) LFP and delithiated FP by PS oxidation. 83 Figure 5.5 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by cuprous (Cu+) at room temperature (RoC), by cuprous (Cu+) at 65 °C (65C), by thiosulfate (TS) at 65 °C (65C), by sulfite (SF) at room temperature (RoC) and by sulfite (SF) at 65 °C (65C). All the loading experiments used B1 brine solution. ........................................................................................................................ 84 Figure 5.6 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 25 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 86 Figure 5.7 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 45 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 87 Figure 5.8 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 65 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 88 Figure 5.9 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS - xvi - oxidation), and products after loading experiments by sulfite (SF) at 25 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 90 Figure 5.10 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by sulfite (SF) at 45 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 91 Figure 5.11 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by sulfite (SF) at 65 °C. The loading experiments used B1–B8 brine solution. ................................................................................... 92 Figure 5.12 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 25 °C (B1–B8 brine solution). ......................................... 94 Figure 5.13 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 25 °C (B1–B3 brine solution). ....................................................................................... 96 Figure 5.14 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 25 °C (B1–B3 brine solution). .................................................................... 97 Figure 5.15 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 45 °C (B1–B8 brine solution). ......................................... 98 Figure 5.16 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 45 °C (B1–B3 brine solution). ....................................................................................... 99 Figure 5.17 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 45 °C (B1–B3 brine solution). .................................................................. 100 Figure 5.18 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 65 °C (B1–B8 brine solution). ....................................... 101 Figure 5.19 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 65 °C (B1–B3 brine solution). ..................................................................................... 102 - xvii - Figure 5.20 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 65 °C (B1–B3 brine solution). .................................................................. 103 Figure 5.21 Adsorption density vs. temperature in the case of thiosulfate (TS) reduction (B1–B4 brine solution). The theoretical maximum of the adsorption density is 46.0 mg/g. ...... 104 Figure 5.22 Weight percentage of lithium iron phosphate in solid vs. temperature in the case of thiosulfate (TS) reduction (B1–B4 brine solution). ............................................................ 105 Figure 5.23 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 25 °C (B2–B7 brine solution). ............................................... 106 Figure 5.24 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 25 °C (B2 and B3 brine solution). ............................................................................... 107 Figure 5.25 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 25 °C (B2 and B3 brine solution). ............................................................................... 108 Figure 5.26 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 45 °C (B2–B7 brine solution). ............................................... 109 Figure 5.27 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 45 °C (B2 and B3 brine solution). ................................................................................ 110 Figure 5.28 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 45 °C (B2 and B3 brine solution). ................................................................................ 111 Figure 5.29 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 65 °C (B2–B7 brine solution). ................................................ 112 Figure 5.30 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 65 °C (B2 and B3 brine solution). ................................................................................ 113 Figure 5.31 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 65 °C (B2 and B3 brine solution). ................................................................................ 114 - xviii - Figure 5.32 Adsorption density vs. temperature in the case of sulfite (SF) reduction (B2–B4 brine solution). The theoretical maximum of the adsorption density is 46.0 mg/g. ............. 115 Figure 5.33 Weight percentage of lithium iron phosphate in solid vs. temperature in the case of sulfite (SF) reduction (B2–B4 brine solution). .................................................................... 116 Figure 5.34 Iron (Fe) dissolution and final pH in the case of thiosulfate (TS) reduction (B1–B8 brine solution). ...................................................................................................................... 117 Figure 5.35 Iron (Fe) dissolution and final pH in the case of sulfite (SF) reduction (B1–B8 brine solution)............................................................................................................................. 119 Figure 5.36 Adsorption density and iron (Fe) dissolution at pH 4 (uncontrolled) and 7 (controlled with sodium hydroxide) in the case of thiosulfate (TS) reduction (B1 brine solution). ..................................................................................................................................... 120 Figure 5.37 Schematic flowsheet of the ferric phosphate method. Flow of FP and LFP solid is in orange and flow of brine solution is in blue. FP and LFP are recycled by a continuous loading and stripping cycle. ..................................................................................................... 121 Figure 5.38 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after two cycles of loading and stripping experiments. The loading experiments used B1 brine solution at pH 7 controlled by sodium hydroxide and lithium was loaded by thiosulfate (TS) at 65 °C. The stripping experiments used persulfate to oxidize the loaded materials. ....................................................................................................................... 122 Figure 5.39 Concentration of metal ions in the initial brine solution (B1 brine solution) and solution after stripping in the first and second cycle of the loading and stripping experiments. The loading experiments used B1 brine solution at pH 7 controlled by sodium hydroxide and lithium was loaded by thiosulfate (TS) at 65 °C. The stripping experiments used persulfate to oxidize the loaded materials and it was carried out with a solid concentration of 10 wt%. - xix - In addition to the metal concentration of the stripped ones in the graph, potassium from K2S2O8 was also in the solution. ............................................................................................... 123 Figure 5.40 x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). .................................................................................................. 126 Figure 5.41 Calculation of Pseudo first-order model: “ln (1-α) vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of thiosulfate (TS) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature...................................................................................................................................................... 127 Figure 5.42 Calculation of shrinking sphere model: “1-(1-α)1/3 vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of thiosulfate (TS) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature...................................................................................................................................................... 128 Figure 5.43 Comparison of experimental data and calculated values by Pseudo first-order model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). .................................................................................................. 130 Figure 5.44 Comparison of experimental data and calculated values by shrinking sphere model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). .................................................................................................. 130 Figure 5.45 x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution). .................................................................................................. 131 Figure 5.46 Calculation of Pseudo first-order model: “ln (1-α) vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of sulfite (SF) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. . 132 Figure 5.47 Calculation of shrinking sphere model: “1-(1-α)1/3 vs. time at 25 °C, 45 °C and - xx - 65 °C” and “ln k vs. -1000/RT” in the case of sulfite (SF) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. . 133 Figure 5.48 Comparison of experimental data and calculated values by Pseudo first-order model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution)..................................................................................................................... 135 Figure 5.49 Comparison of experimental data and calculated values by shrinking sphere model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution)..................................................................................................................... 135 - xxi - Nomenclature AAS Atomic adsorption spectroscopy ADB Acrylic-divinylbenzene AEM Anion exchange membrane Ah Ampere hour BG Battery grade Ce Equilibrium concentration in solution Co Initial concentration in solution CR Resin concentration in solution da Actual denisty DI Deionized dry R Dry resin DVB Divinylbenzene E Electrode potential E° Standard electrode potential ESS Energy storage system EV Electric vehicle FP Ferric phosphate FPD Ferric phosphate dihydrate HSC HSC Chemistry 6.0 ICP-OES Inductively coupled plasma - optical emission spectroscopy IX Ion exchange - xxii - KLi Selectivety of lithium to a metal KLi, solid Molar ratio of lithium to a metal in solid KLi, solution Molar ratio of lithium to a metal in solution kWh kilowatt hour LCE Lithium carbonate equivalent LFP Lithium ferrous (iron) phosphate LIB Lithium ion battery mAh Milli ampere hour mFP Weight of ferric phosphate mM Weight of an element in solid NF Nanofiltration ppm Parts per million PS Persulfate qe Adsorption density on resin Qe Adsorption density on ferric phosphate Rpm Revolutions per minute rs Swelling ratio SAC Strong acid cation SDB Styrene-divinylbenzene SF Sulfite SFP Sodium ferrous (iron) phosphate SHE Standard hydrogen electrode SX Solvent extraction TS Thiosulfate - xxiii - V Volt VESTA Visualization for Electronic and STructual Analysis 31 wet R wet resin XRD X-ray diffraction ΔE Electrode potential difference ΔE° Standard electrode potential difference - xxiv - Acknowledgements I would like to express my special thanks to my supervisor, Dr. David Dreisinger, for his kind guidance and support throughout the Master’s program. All of his words lead me to learn a lot about hydrometallurgy and how to do meaningful research works. I would also like to sincerely thank Dr. Berend Wassink for his kind assistance in my lab work. He always helped me with preparing chemical reagents and experimental equipment and keeping lab work safe. I also thank Jacob Kabel, Eunike Heli for their support of XRD analysis and Kresimir Ljubetic for his support of ICP analysis. This study couldn’t be completed without material support from Dow, Purolite, Lanxess, Mitsubishi, Samyang and software support by HSC and VESTA. I would also like to express my gratitude to these groups. I really appreciated the warm support from my group members: Dr. Jianming Lu, Brighty Dutta, Chih Wei Chao, Chulho Song, Eunso Shin, Fei Wang, Jackie Zhou, Junichi Ito, Mariam Melashvili, Maryam Mohammadi, Maryam Rasouli, Roselyn Yeboah. My research life was fruitful and enjoyable. This thesis was supported by my thesis committee of Dr. David Dixon, Dr. Berend Wassink, and Dr. David Dreisinger and the chairperson Dr. Daan Maijer. I would like to thank the committee for their kind review. Finally, I would like to express my special thanks to my family and friends in Canada and Japan. Thanks to all the support from you, I was able to complete the Master’s program. - 1 - Chapter 1. Introduction Lithium is one of the critical metals in today’s world, and the demand for lithium is now sharply increasing due to the accelerating use of energy storage systems containing lithium ion batteries. While lithium is obtained from mineral ores or brines, lithium from brines is now getting more attention since brines are known to have more reserves than ores.2 Brines containing lithium exist mainly in the “lithium triangle” of Argentina, Bolivia and Chile in South America.3,4 There are also a lot of medium size or small size of brines deposits around the world including oilfield brines and geothermal brines. Since brines which contains a higher concentration of lithium have been preferentially mined in the past decades, conventional solar evaporation processes, which require vast areas and an enormous amount of time, have been utilized as a major method to concentrate and recover lithium. However, new processes now need to be developed in order to meet the accelerating demand by mining brines with a lower concentration of lithium. To recover lithium efficiently from low concentration resources, new processes should be more rapid, lithium-selective, and small-scale than the solar evaporation process. Therefore, some researchers have proposed new rapid processes such as ion exchange resin5–8, solvent extraction9,10, inorganic absorbent such as aluminum oxide11–16, manganese oxide17–20, titanium oxide21,22 and ferric phosphate23–25, phosphate precipitation26,27, nanofiltration28–30, membrane electrolysis31–34. Some of them are now used commercially or under a pilot test. However, there is no established process other than solar evaporation. This thesis focused on two processes out of the available options: ion exchange resin and ferric phosphate adsorption. - 2 - Chapter 2. Literature Review 2.1. Lithium Chemistry 2.1.1. Electrochemical Properties of Lithium Lithium is one of the alkali metals with an atomic number of 3 and an atomic weight of 6.941. It is reactive as with the other alkali metals and is generally dealt with in the form of salts or compounds such as lithium carbonate Li2CO3. From the electrochemical point of view, lithium has the lowest oxidation-reduction potential with the standard electrode potential of -3.05 V as shown in Equation (2.1) and Table 2.1. -Li e Li E 3.05 V vs. SHE+ + = = − (2.1) Table 2.1 Properties of possible anode materials.35 Standard electrode potential E° (vs. SHE) [V] Density [g/cm3] Capacity density [Ah/kg] [Ah/dm3] Li -3.05 0.53 3860 2060 Na -2.71 0.97 1170 1130 Al -1.66 2.70 2980 8050 Zn -0.76 7.14 820 5860 Fe -0.44 7.85 960 7550 Cd -0.40 8.65 480 4120 Pb -0.13 11.4 260 2940 Lithium also has a very low density with 0.53 g/cm3, which enables batteries to have higher - 3 - capacity per unit weight. For the above two reasons, lithium is very useful for battery development. 2.1.2. Lithium Ion Battery Lithium ion battery (LIB) is one of the most important lithium products. This thesis doesn’t focus on the chemistry of lithium ion batteries, but it is essential to know this topic as a “lithium researcher”. Therefore, this section only summarizes chemical reactions at the anode and cathode when the charge and discharge occur. Figure 2.1 Schematic illustration of charge-discharge in lithium ion batteries, Me= Co, Fe+P, Mn, Ni+Mn+Co, Ni+Co+Al, Ti, S, etc.2 When a battery is charged, the following reactions in Equation (2.2) occur. Lithium metal oxide at the anode provides carbon at the cathode with electrons via lithium ion electrolyte. -1-x-6 x 6Cathode: LiMeO Li MeO xLi xeAnode: C xLi xe Li C++→ + ++ + → (2.2) When a battery is discharged, the reactions in Equation (2.3) occur. Carbon at the cathode supplies Anode Cathode Anode CathodeMetal oxide layers Graphite layersLi+ ion- 4 - electrons to lithium metal oxide at the anode. Lithium moves into and out of the electrode materials. -1-x-x 6 6Cathode: Li MeO xLi xe LiMeOAnode: Li C xLi C xe+++ + →→ + + (2.3) There have been various metals used in lithium ion batteries such as Co, Fe+P, Mn, Ni+Mn+Co, Ni+Co+Al, Ti. Moreover, lithium sulfur (Li+S) batteries are currently being focused on by many researchers.36 The cathode reaction of LiS batteries is Equation (2.4). SLi 2e S 2Li 2- →+++ (2.4) Although the electronic conductivity of sulfur is lower than other cathode materials, it has a higher theoretical capacity (1675 mAh/g) than that of other available cathode materials for lithium ion batteries such as lithium cobalt oxide (–140 mAh/g). There are various problems for practical use, but this is why lithium sulfur batteries are now eliciting much interest. 2.1.3. Eh-pH Diagram To understand the aqueous chemistry of lithium, an Eh-pH diagram is one of the essential tools. Figure 2.2 is an Eh-pH diagram for the Li-H2O system when (a) [Li+]=1 M and (b) [Li+]=0.072 M. Almost all the experiments in this thesis used the condition of [Li+]=0.072 M(=500 mg/L) because this value is close to the concentration in South American brines. These figures show that lithium exists as lithium ion (Li+) in almost all the range of pH and Eh regardless of the concentration. - 5 - Figure 2.2 pH-Eh diagram for Li-H2O system at 25 °C, 1 atm, (a) [Li]=1 M or (b) [Li]=0.072 M (Drawn by HSC). 1412108642043210-1-2-3-4Li - H2O - System at 25.00 CC:\\HSC6\\EpH\\Li25.iep pHEh (Volts)LiLi(+a)ELEMENTS Molality PressureLi 7.200E-02 1.000E+001412108642043210-1-2-3-4Li - H2O - System at 25.00 CC:\\HSC6\\EpH\\Li25.iep pHEh (Volts)LiLi(+a)ELEMENTS Molality PressureLi 1.000E+00 1.000E+00(a)(b)- 6 - From Figure 2.3 (a), it is clear that lithium can be recovered from aqueous solution as Li2CO3, which is the standard commercial product of lithium for the battery market. However, when the concentration is reduced to 500 mg/L or 0.072 M, lithium carbonate is not precipitated unless an over-addition of carbonate is used (Figure 2.3 (b)). This result confirms that it is necessary to pre-concentrate lithium in brines prior to precipitation of lithium carbonate. - 7 - Figure 2.3 pH-Eh diagram for Li-C-H2O system at 25 °C, 1 atm and (a) [Li]=[C]=1 M or (b) [Li]=[C]=0.072 M (Drawn by HSC). 1412108642043210-1-2-3-4Li - C - H2O - System at 25.00 CC:\\HSC6\\EpH\\LiC25.iep pHEh (Volts)LiLi(+a)ELEMENTS Molality PressureLi 7.200E-02 1.000E+00C 7.200E-02 1.000E+001412108642043210-1-2-3-4Li - C - H2O - System at 25.00 CC:\\HSC6\\EpH\\LiC25.iep pHEh (Volts)LiLi2CO3Li(+a)LiHCO3(a)ELEMENTS Molality PressureLi 1.000E+00 1.000E+00C 1.000E+00 1.000E+00(a)(b)- 8 - 2.2. Lithium Market 2.2.1. Supply/Demand and Price The lithium market is expanding. It is a new market without a clear history unlike the copper market. It is therefore more difficult to predict supply/demand and price of lithium in a today’s rapidly fluctuating market.37 While there are several reports from some companies, the data by BMO Capital Markets is cited here in Figure 2.4. Figure 2.4 Lithium supply/demand and price and their estimates.38 According to the graph, there won’t be a supply deficit by 2025, although, around 2015, there was a concern of lithium shortage. The price of lithium carbonate and hydroxide are expected to drop and then increase gradually over time. At first glance, it seems that there is no problem in the - 9 - lithium market because, according to the current estimates, a sufficient amount of lithium would be supplied to the lithium market and the price would be stable. However, this topic is controversial since the lithium market is unique, and it’s not easy to predict future demand. First of all, it takes some time for lithium to be supplied into the market, especially from brine deposits. Since the process used for brine deposits is mainly solar evaporation, which normally takes at least one year to concentrate lithium, the market can be affected by the supply situation in the past few years. In other words, a certain amount of lithium should be prepared beforehand to meet the demand in the next few years. On the other hand, the demand side of the market is rapidly changing, and the future demand is unclear at least in the current situation. Some automobile companies announced a sales target for EVs and the lithium market reacted so as to prepare for those targets, but there has been less demand because EVs are not yet common among consumers due to the restriction of infrastructure and available models. It is more complicated than expected to keep the lithium market in balance by predicting supply and demand. This is one of the reasons that the price of lithium compounds hit the highest point in 2018. In the early 2010s, the market was thought to become short of lithium soon due to EVs production, and the price rose rapidly though the supply finally managed to meet the demand. This means there is a possibility that lithium demand will catch and surpass its supply in the near future while the current estimates show the opposite situation. For example, China, which is one of the biggest markets for EVs in the world, has started to head toward becoming an EV society by making laws to restrict the emission of greenhouse gases. It is uncertain how fast EVs will spread to the world. To maintain the supply of lithium, a process is needed to make lithium compounds quickly. The slow response time for the use of long-term solar evaporation of brines is not sufficient to meet rapidly escalating demand for lithium chemicals. This is the reason that a lot of researchers are working on rapid lithium recovery technologies. - 10 - 2.2.2. Products and Applications Lithium is utilized in various ways. Figure 2.5 shows the changes in lithium uses from 2003 to 2020. According to the graph, lithium use has been increasing because the demand for lithium batteries has been dramatically increasing. Almost half of the lithium produced was consumed for lithium batteries in 2019. The rapid increase in demand in the current period (Figure 2.4) due to lithium battery production is expected to continue. Figure 2.5 Annual use of lithium in tonnes in each of the primary lithium usage industries from 2002 to 2020.39 Rockwood Holdings, Inc. report about 23 kg-LCE (lithium carbonate equivalent) of lithium is needed for one EV while only about 3 g is needed for one cellphone as shown in Table 2.2.35 For instance, Tesla, Inc. made a plant, called Tesla Gigafactory 1, to make lithium ion battery EVs in a larger scale than ever before. - 11 - Table 2.2 Amount of lithium used in various devices.35 Usage Lithium used in LCE Cellphone 3 g Laptop 30 g Plug-in hybrid car (16 kWh) 9 kg EV (25 kWh) 23 kg Other than lithium batteries, ceramics and the lubricant industry are the primary lithium consumers. Aluminum production and pharmaceuticals also use lithium as one of their critical components though the total consumption is not so large. For those industries, various kinds of lithium compounds are prepared. Table 2.3 shows key products and key applications of lithium, and Figure 2.6 illustrates a lithium value chain.35 Table 2.3 Lithium compounds and applications.35 Key product Key applications Lithium carbonate Lithium ion batteries, Glass ceramics, Cement, Aluminum Lithium hydroxide Lithium ion batteries, Grease, CO2 adsorption, Mining Lithium metal Lithium batteries, Pharmaceuticals, Aluminum alloy Butyl-lithium Elastomers, Pharmaceuticals, Agrochemicals Lithium carbonate and hydroxide are the main lithium products used in lithium ion batteries, glass ceramics and lubricants, which, in total, occupy more than 50 percent of the lithium use. Lithium carbonate is used for most glass products because it decreases the viscosity and melting point of glass. In the ceramics industry, lithium hydroxide is used for lithium grease. Lithium greases - 12 - adhere well to metal over a wide range of temperature. Lithium carbonate is also useful for aluminum production since lithium can lower the melting point of the molten electrolyte of aluminum and increase the electrical conductivity in the electrolysis cell. This helps decrease costs and energy for aluminum production. When welding steel alloys, lithium may be used to reduce the melting temperature and lower the surface tension. Some other lithium compounds also enhance electrical conductivity in some metal processes. Interestingly, as some environmental concerns are increasing, lithium hydroxide is applied to adsorption of one of the greenhouse gases, carbon dioxide. Lithium metal and lithium organic compounds like butyl-lithium are also critical products for life science such as pharmaceuticals. Figure 2.6 Lithium value chain.35 Though more explanation will be provided later, it should be noted here that lithium carbonate is firstly produced from ores and brines in most cases, from which other compounds are produced - 13 - such as lithium hydroxide and lithium metal. 2.3. Lithium Deposits 2.3.1. Overview It is vital to review lithium deposit types before discussing the separation science and technology of lithium because separation technology is dependent on lithium deposit type. Figure 2.7 (a) Resources and (b) Reserves in terms of deposit types in thousands metric tonnes Li metal equivalent.2 Figure 2.7 is the estimate of the global lithium resources and reserves based on data summarized in 2015.2 According to the graph, lithium deposit types are divided into two. Brines constitute about 58% of global resources and 82% as reserves while ore minerals about 37% and 17% respectively. Brine (Salt lakes), 23572Brine (Oil fields), 1497Brine (Geothermal), 1000Ore, 16735Others, 2397(a) ResourcesBrine (Salt lakes), 10003Ore, 2089Others, 107(b) Reserves- 14 - Figure 2.8 (a) Resources and (b) Reserves in terms of countries in thousands metric tonnes Li metal equivalent.3 Figure 2.8 shows that more than half of the resources are in the lithium triangle of Argentina, Bolivia and Chile in South America and almost 80 percent of the estimated reserves are in the triangle as well.3 These data indicate that brines would be expected to be a more abundant lithium source in the future if technologies are developed which can extract lithium from brine more rapidly and economically. Table 2.4 and Table 2.5 show some ore-type and brine-type mining sites respectively.40 Ore-type deposits mainly exist in the US, Canada, Australia and China. Some deposits are in Africa and Europe, but they are not major ones. Australia and China are still expanding the operation of lithium production from pegmatites. Argentina, 9800Bolivia, 9000Chile, 8400China, 7000United States, 6800Australia, 5000Canada, 7350(a) ResourcesArgentina, 2000Chile, 7500China, 3200United States, 35Australia, 2700Canada, 131(b) Reserves- 15 - Table 2.4 Lithium mining sites and resources in tonnes (ores).40 Deposit types Mining Sites Li [t] Ores Pegmatites North Carolina Former operations 230,000 North Carolina Undeveloped 2,600,000 Barraute, Quebec 90,000 Bernic Lake, Manitoba 18,600 Others, Canada 147,000 Bikita, Zimbabwe 56,700 Manono, Zaire 2,300,000 Greenbushes, Western Australia 1,500,000 Mount Marion, Western Australia 19,800 Mount Catlin, Western Australia 20,000 Koralpa, Austria 100,000 Larritta, Finland 14,000 Various, Russia 1,000,000 Brazil, Minas Gerais & Ceara 85,000 China 750,000 Hectorites McDermitt Caldera Oregon/Nevada 2,000,000 Jadarite Jadar, Serbia 850,000 Of course, brine-type deposits are mainly found in the lithium triangle as shown in Table 2.5. It is also known that some geothermal and oilfield brines in the US, such as the Smackover formation, have the potential for lithium production. That is why there are now many interests in the methods - 16 - of how to extract lithium from oilfield brines. Table 2.5 Lithium mining sites and resources in tonnes (brines).40 Deposit types Mining Sites Li [t] Brines Continental Brines Silver Peak, Nevada 40,000 Salar de Uyuni, Bolivia 8,900,000 Salar de Hombre Muerto, Argentina 850,000 Salar de Rincon, Argentina 1,400,000 Salar de Oraroz, Argentina 300,000 Salar de Atacama, Chile 6,900,000 Salar de Maricunga,Chile 200,000 China & Tibet 2,600,000 Geothermal Brines Brawley, Southern California 1,000,000 Oilfield Brines Smackover Formation USA 750,000 2.3.2. Ore Deposits As shown in Table 2.4, pegmatite is the most common lithium ore type, which contains some lithium minerals such as spodumene, while there are some other ores such as hectorite. Table 2.6 summarizes some lithium minerals with the theoretical and practical amount of lithium contained. - 17 - Table 2.6 Lithium minerals.40 Mineral Composition Li2O [wt%] Theoretical Practical Petalite LiAlSi4O10 4.9 3.0–4.5 Spodumene LiAlSi2O6 8 1.5–7.0 Amblygonite (Li,Na)AlPO4(F,OH) 10 8–9 Lepidolite K(Li,Al)3(Al,Si,Rb)4O10(F,OH)2 4.1 3–4 Eucryptite LiAlSiO4 11.9 5 Hectorite Na0.3(Mg,Li)3Si4O10(OH)2 - 0.3–0.7 Jadarite LiNaSiB3O7OH - - Igneous rock - - 1–35 ppm Carbonate rocks - - 8 ppm Shales and clay - - 70 ppm Seawater - - 0.18 ppm Many pure minerals contain several percent of lithium, while the amount of lithium found in natural minerals is smaller because the minerals are not pure. Usually, the concentration in hard rock deposits is higher than brine deposits. 2.3.3. Brine Deposits As shown in Table 2.5, continental brines in South America are the main brine-type deposits that are exploited, while a few geothermal brines or oilfield brines are also commercialized. Continental brines contain at most 0.05–2% of lithium. Some analytical data of - 18 - brines are provided below in Table 2.7.40 Continental brines generally contain 5–12% of sodium, and some other ions such as magnesium, sulfate, boron, potassium as shown in the table. Calcium is also found in continental brines. Sodium and potassium may be partially recovered via evaporation as chloride salts before lithium recovery. Magnesium and calcium are precipitated as carbonates by soda ash addition before the lithium concentration gets high enough to be coprecipitated as a carbonate. It is necessary to remove sulfate and boron because they disturb lithium production in the following process. Sulfate precipitates with sodium as sodium sulfate when brine temperature is low. Boron is an unfavorable impurity in lithium products and must be removed. Table 2.7 Analytical values of continental brines.40 Brines Evaporation [mm/y] Li [%] Mg [%] (Mg/Li) SO4 [%] (SO4/Li) B [%] (B/Li) K [%] (K/Li) Salars de Atacama (Chile, SQM) 3200 0.15 0.96 (6.4) 1.65 (11.0) 0.05 (0.3) 2.36 (15.7) Salars de Atacama (Chile, Chemetall) 3200 0.16 1.02 (6.4) 1.76 (11.0) 0.05 (0.3) 1.97 (12.3) Salars del Hombre Muerto (Argentina, FMC) 2400 0.069 0.097 (1.4) 0.952 (13.8) 0.034 (0.5) 0.61 (8.8) Great Salt Lake (Utah, USA) - 0.004 - 2 (500.0) 0.006 (1.5) 0.65 (162.5) SilverPeak (Nevada, USA, Chemetall) 760–1200 0.016 0.028 (1.4) 0.494 (30.9) 0.008 (0.5) 0.53 (33.1) - 19 - Table 2.7 Analytical values of continental brines,40 cont’d. Brines Evaporation [mm/y] Li [%] Mg [%] (Mg/Li) SO4 [%] (SO4/Li) B [%] (B/Li) K [%] (K/Li) Dead Sea (Israel/Jordan) - 0.002 - 0.07 (35.0) 0.003 (1.5) 0.6 (300.0) Rincon (Argentina, Admiralty) 1500–2100 0.04 0.344 (8.6) 1.228 (30.7) 0.12 (3.0) 0.75 (18.8) Uyuni (Bolivia, Gerve−) 1200–2200 0.035 0.651 (18.6) 0.854 (24.4) 0.08 (2.3) 1.6 (45.7) West Taijnar (China, Qinhai Citic Group) 2600–3100 0.025 1.538 (61.5) 3.45 (138.0) 0.038 (1.5) 0.84 (33.6) East Taijnar (China, Qinhai Lithium) 2600–3100 0.05 1.87 (37.4) 2.2 (44.0) 0.022 (0.4) 0.38 (7.6) Zhabuye (China, Zhabuye Lithium) 2300 0.13 0.03 (0.2) 2.228 (17.1) - - Dongxiung Cuo (China, Sterling) 2300 0.035 0.008 (0.2) 0.571 (16.3) - - 2.4. Processes for Lithium Production 2.4.1. Overview As explained in 2.3, ore and brine deposits are the primary resources of lithium. In most cases, when making products, especially for lithium ion batteries, those deposits are processed and refined into lithium carbonate. If necessary, lithium carbonate is then converted into lithium - 20 - hydroxide. In this section, therefore, the following three processes are explained with some practical examples. 1) Process from ore deposits to lithium carbonate 2) Process from brine deposits to lithium carbonate 3) Process from lithium carbonate to lithium hydroxide 2.4.2. Process of Ore Deposits Ores firstly go through mineral processing like grinding and sieving to get lithium mineral concentrates. The lithium concentration of the concentrates is slightly higher than brine but is still low. In order to extract lithium from the concentrates, there is a common method decomposing ores by using sulfuric acid. Figure 2.9 and Figure 2.10 present the process schematically and in a simplified flow chart. Figure 2.9 An example of flow chart of Li2CO3 production from ores.41 - 21 - Figure 2.10 A simplified flow chart of Li2CO3 production from ores.42 First, the concentrates are roasted and reacted with sulfuric acid in a rotary kiln at high temperature (–1100 °C). Lithium sulfate (Li2SO4) is extracted into water and the solution is recovered by filtration followed by neutralization for impurity rejection. Li2CO3 can be obtained by adding soda ash. 2.4.3. Process of Brine Deposits A schematic illustration of the brine process is shown in Figure 2.11, and a representative flow chart is in Figure 2.12, which is drawn based on a report by JOGMEC, 201040. 1100 °C (a→b)ConcentratesPhase conversionRoasting/DissolutionWater ExtractionFiltrationLi2SO4 solutionNeutralizationH2SO4Li2CO3TailingsTailingsSoda Ash (Na2CO3)- 22 - Figure 2.11 A schematic illustration of lithium production from brines.43 Figure 2.12 A representative flow of lithium carbonate production from brines.40 BrinesLi 6.0%, Mg 2.0%, B 0.8%Solar HeatConcentration Evaporation Pond (About 1 year)Halite (NaCl), Na2SO4Sylvinite (NaCl+KCl)Bischofite (MgCl2・6H2O)Li 0.2%Li 0.9%Li 6%LiCl solutionDensity 1.25–1.30B removal SX (Load: Isooctanol , Strip: HCl)Mg removalLime (Ca(OH)2 or CaO)Mg(OH)2Li precipitationSoda Ash (Na2CO3)Li2CO3Drying 170°C- 23 - First, brine containing around 0.2% of lithium is pumped up from wells. Then in evaporation ponds, brine is concentrated by solar heat for about one year so that lithium concentration finally gets to 4–6%. During the evaporation, some sodium and potassium are removed as chloride salts by making use of a solubility different as shown in Table 2.8. Table 2.8 Solubility of chloride salts at 25 °C.44 Compounds Solubility [g/100 g-H2O] LiCl 84.5 NaCl 36.0 KCl 35.5 MgCl2 56.0 CaCl2 81.3 Table 2.9 Solubility of hydroxide salts at 25 °C.44 Compounds Solubility [g/100 g-H2O] LiOH 12.5 NaOH 100 KOH 120.8 Mg(OH)2 0.00069 Ca(OH)2 0.160 - 24 - Boron and magnesium removal follow the evaporation. Magnesium is rejected by adding a base to precipitate magnesium hydroxide. During this step, magnesium and calcium precipitates and are removed from brine as hydroxides. Other monovalent cations remain in solution because of the difference of the solubility of their hydroxide in Table 2.9. Lithium carbonate is finally precipitated by adding soda ash Na2CO3 and then dried. As magnesium and calcium are removed beforehand, only lithium precipitates as lithium carbonate. Some sodium and potassium ions remain in solution, but these species can be ignored due to the solubility difference of their carbonate salts in Table 2.10. This lithium carbonate is a product itself and can be a source of other lithium compounds like lithium hydroxide LiOH. Table 2.10 Solubility of carbonate salts at 25 °C.44 Compounds Solubility [g/100 g-H2O] Li2CO3 1.3 Na2CO3 30.7 K2CO3 111.4 MgCO3 0.18 CaCO3 0.00066 This section introduces seven practical cases below. Figure 2.13 is the process flow of Salar de Atacama, Chile. After the evaporation step, boron is removed by SX and magnesium is precipitated by lime. - 25 - Figure 2.13 Process flow of Salar de Atacama, Chile.42 Figure 2.14 is the process flow of Salar de Hombre Muerto, Argentina. This process uses an adsorption method onto alumina at a certain pH and temperature to enhance lithium concentration before evaporation because of low concentration in brines. Brines: 0.16%LiEvaporation0.9%LiB removalNaCl, KClLi2CO3SXSoda Ash (Na2CO3)Salar de Atacama, ChileLiCl solution: Max 6%LiMgCl2Mg removalLime (CaO)- 26 - Figure 2.14 Process flow of Salar de Hombre Muerto, Argentina.42 Figure 2.15 is the process flow of Salar de Oraroz, Argentina. In this process, magnesium removal is performed in the first place to increase the lithium concentration before evaporation because of low concentration in brines. Besides, crude lithium carbonate is first recovered and then redissolved and refined to remove impurities efficiently. Brines: 0.09%LipH&temp adjustmentLi adsorptionLi2CO3AluminaSoda Ash (Na2CO3)Salar de Hombre Muerto, ArgentinaLiCl solution: 1%LiImpurity removalEvaporationLiCl solution: 3%Li- 27 - Figure 2.15 Process flow of Salar de Oraroz, Argentina.42 Figure 2.16 shows the process flowsheet for lithium recovery in China and Tibet. This method firstly removes sulfate by natural cooling because sulfate inhibits evaporation and lithium cannot be concentrated sufficiently as a result. Brines: 0.08%LiEvaporationLiCl solution: 0.84%LiB removalNaCl, KClCrude Li2CO3IXSoda Ash (Na2CO3)Salar de Oraroz, ArgentinaMg removalLime (CaO)RedissolutionCO2HeatingLi2CO3- 28 - Figure 2.16 Process flowsheet for lithium recovery in China & Tibet.42 Figure 2.17 is the process flow of CITIC method (CITIC Group Corporation Ltd., formerly the China International Trust Investment Corporation, is a state-owned investment company of the People's Republic of China). This method is similar to the process of Salar de Atacama, Chile (Figure 2.13), but it removes boron by hydrochloric acid treatment, whose process is not clearly explained in the literature. Spray drying and kiln processing are used to concentrate lithium more efficiently. Brines: 0.05%LiNatural coolingNa2SO4, MgSO4Li2CO3Soda Ash (Na2CO3)China & TibetLiCl solutionNaCl, KClMg removalLime (CaO)Evaporation- 29 - Figure 2.17 Process flow of CITIC method.42 Figure 2.18 is the process flow of POSCO method (POSCO, formerly Pohang Iron and Steel Company, is a multinational steel-making company headquartered in Pohang, South Korea). This process does not use evaporation in an ideal case and precipitates and removes magnesium hydroxide and calcium carbonate by adding sodium hydroxide and sodium carbonate. After the removal, lithium phosphate is precipitated by adding sodium phosphate, and the precipitates are redissolved by phosphoric acid. Membrane electrolysis makes lithium hydroxide from the lithium BrinesLi2CO3Soda Ash (Na2CO3)CITIC methodLiCl solutionNaCl, KClB removalEvaporationHCl treatmentSpray dryingKiln processRedissolutionMg removalLime (CaO)- 30 - phosphate solution and CO2 addition forms lithium carbonate form. Figure 2.18 Process flow of POSCO method.42 Figure 2.19 is the process flow of Bateman method. This process doesn’t use evaporation either in an ideal case. Membrane electrolysis and solvent extraction are used to enhance lithium concentration in lithium chloride solution. BrinesLi2CO3CO2POSCO methodLi3PO4 solutionMg(OH)2Membrane electrolysisLiOHLi3PO4 precipitateH3PO4NaOHCaCO3Na2CO3Na3PO4- 31 - Figure 2.19 Process flow of Bateman method.42 As mentioned above, there are different optimum methods to recover lithium from brine with different characteristics. Additionally, it is generally said that favorable conditions for lithium extraction are (1) High Li concentration, (2) Low Mg and B concentration, (3) Low SO4 concentration, and (4) High K concentration. Conditions (1) and (2) directly lead to lower cost. Condition (3) makes it easier to separate lithium from brines while (4) increases profit with greater amounts of byproducts. 2.4.4. Process for Lithium Hydroxide Production Lithium hydroxide is an essential product for the production of lithium ion batteries. Although there are some methods proposed to make LiOH directly from brines, this section only BrinesLiOHBateman methodMg(OH)2Membrane electrolysisLiCl solutionNaOHCaCO3Na2CO3Solvent extractionLi2CO3Na2CO3 Membrane electrolysis- 32 - mentions a method using lithium carbonate because there are few direct processes which are feasible. Lithium hydroxide solution is firstly produced by Equation (2.5). 2LiOH CaCO Ca(OH) COLi 3232 +→+ (2.5) Then, lithium hydroxide monohydrate LiOH・H2O is crystallized from the mother solution by evaporation as shown in Equation (2.6). →+ OHLiOH OH LiOH 22 (2.6) In order to precipitate and crystallize higher grade of lithium hydroxide monohydrate for lithium ion batteries, it is important to prepare a higher concentration of lithium hydroxide solution with a higher grade of lithium carbonate.45 There are also various processes proposed using membrane electrodialysis to get lithium hydroxide solution. - 33 - 2.5. Lithium Extraction Technologies 2.5.1. Overview For the past several decades, a combination of solar evaporation and carbonation has been mainly utilized to produce pure lithium compounds. However, this combination can be applied only to brine that has a relatively high concentration of lithium, roughly more than 500 mg/L. The evaporation method also takes approximately one year to concentrate lithium and subsequently precipitate lithium carbonate. Thus, other rapid processes have been considered and studied recently as shown in Table 2.11.46 This section reviews those technologies by considering some practical examples. Table 2.11 Technologies for lithium extraction from brine and their mechanism. Technology Mechanism Developer Solar evaporation Lithium-containing solutions in ponds are concentrated by solar heating so that lithium carbonate can precipitate after adding soda ash (sodium carbonate). 2 3 2 32LiCl Na CO Li CO 2NaCl+ → + Conventional Phosphate precipitation In place of soda ash, phosphoric acid is used to precipitate lithium. 3 4 3 43LiCl H PO Li PO 3HCl+ → + POSCO26,27 - 34 - Table 2.11 Technologies for lithium extraction from brine and their mechanism, cont’d. Technology Mechanism Developer Ion exchange resin Lithium ions are absorbed into aluminum hydroxide’s layers formed in ion exchange resins. 3 23 2LiCl+NaCl 2Al(OH) nH ONaCl+LiCl 2Al(OH) nH O →   Dow5–7 Aluminum based adsorbent Lithium ions are absorbed by the same mechanism as the ion exchange resin adsorption. The adsorbent in the form of powder or granulated beads. FMC11–13 Simbol14,15 Eramet16 Manganese based adsorbent Lithium ions are adsorbed into layers of manganese oxide such as H1.6Mn1.6O4 and λ-MnO2. JOGMEC20 Titanium based adsorbent Lithium ions are adsorbed into layers of titanium oxide such as H2TiO3. Neometals22 - 35 - Table 2.11 Technologies for lithium extraction from brine and their mechanism, cont’d. Technology Mechanism Developer Solvent extraction Lithium ions are extracted to an oil phase from a water phase. sol aq sol aqR-H +LiCl R-Li +HCl→ Tenova9,10 Nanofiltration Lithium ions are concentrated by making use of the differences in ion rejection ratios and water flow rejection through a membrane MGX29 Membrane electrolysis Magnesium and calcium ions are removed as hydroxide. 2 222222H O 2e H 2OHMg 2OH Mg(OH)Ca 2OH Ca(OH)− −+ −+ −+ → ++ →+ → Research underway34 Ferric phosphate reduction Lithium ions are selectively extracted to form lithium ferrous phosphate by ferric phosphate reduction by thiosulfate. 4 2 2 34 2 4 62FePO +2LiCl 2Na S O2LiFePO Na S O 2NaCl+→ + + Research underway23–25 2.5.2. Phosphate Precipitation Phosphate precipitation is a newly invented technology to replace the carbonation step in the solar evaporation process as shown in Figure 2.18. Phosphoric acid is added to precipitate lithium from brine as Equation (2.7). OrganicAqueous- 36 - 3 4 3 43LiCl H PO Li PO 3HCl+ → + (2.7) While this method needs extra processes to convert lithium phosphate into lithium carbonate or hydroxide,47 it has succeeded in shortening the solar evaporation step by precipitating lithium phosphate whose solubility, 0.027 g per 100g of H2O, is much lower than lithium carbonate’s solubility, 1.3 g per 100g of H2O, at ambient temperature (Table 2.12). Table 2.12 Solubility of phosphate salts at 25 °C.44 Compounds Solubility [g/100 g-H2O] Li3PO4 0.027 Na3PO4 14.4 K3PO4 105.8 Mg3(PO4)2 - Ca3(PO4)2 0.00012 POSCO currently has projects to recover lithium from brine or battery recycling with this method.26,27 However, it is doubtful that phosphate precipitation can be applied to low-concentration brines because this method still partially uses solar evaporation in practical cases. 2.5.3. Ion Exchange Resin IX resin is a widely used material for water softening, water purification and metal separation. IX resins commonly consist of a polymer matrix and functional groups introduced into - 37 - the matrix. Most commercial IX resins use styrene-divinylbenzene (SDB) as their matrix. SDB forms a three-dimensional structure by styrene polymer chains and divinylbenzene (DVB) crosslinks (Figure 2.20). Here, DVB is called a crosslinking agent. Figure 2.20 Chemical structure of SDB copolymer. The number of crosslinks partially determines how small the micropores in resin are and how hard resin is. The more crosslinking, the smaller the micropores and the harder the resin. The percentage of DVB amount is usually calculated as an index of the degree of crosslinking with Equation (2.8). Mass of DVBCrosslinkage DVB [%] 100Mass of total monomer= =  (2.8) Also, a porous-type resin can be manufactured by using a special polymerization method, while otherwise, SDB resin is usually gel-type. As for functional groups, sulfonate (R-SO3-H+) and quaternary ammonium (R-NR3+OH-) are commonly used as a strong acid - and strong base - exchanger, respectively. They are introduced on the copolymer matrix by chemical reaction as illustrated in Figure 2.21. In the figure, R-SO3- and R-NR3+ are called fixed ion because they cannot move freely in a resin, and H+ and OH- are called counter ion, which makes a neutral pair with the fixed ion. CH2 CH CH2 CHCH2 CHCH2 CHCH2 CH CH2 CHCH2 CHCH2 CH CH2 CH CH2 CHCH2 CHCH2 CHCH2 CHCH2 CH CH2 CHCH2 CH CH2 CHCH2 CH- 38 - Figure 2.21 Chemical structure of resin with (a) sulfonate, (b) quaternary ammonium. However, there are many other types of resins. Some have special functional groups on them, and others contain additional chemical compounds on functional groups. Keeping this in mind, IX resin for lithium extraction is next reviewed. Although IX resins are conventional adsorbents to recover metal ions from solutions, lithium is much more difficult to adsorb selectively than other metal ions such as copper ions. The reason is that, in brine, a much higher concentration of sodium, potassium, calcium and magnesium ions are present (sometimes 100 times as high as lithium concentration in mol/L), and these ions have a higher affinity to cation exchange resins. Despite the unfavorable conditions for ion exchange resins, Dow Chemical in the US invented aluminum loaded resins which can take lithium selectively from brine by the reaction in Equation (2.9),5–7 although the approach remains challenging for practical use. 3 2 3 2LiCl+NaCl 2Al(OH) nH O NaCl+LiCl 2Al(OH) nH O  →   (2.9) Some other investigations have looked at solvent impregnated resins that extract lithium into resins loaded with an organic phase. S. Nishihama et al., for example, studied 1-phenyl-1,3-tetradecanedione (C11phβDK) / tri-n-octylphosphine oxide (TOPO) impregnated resin to separate lithium ions from sodium and potassium ions.8 Solvent impregnated resin method is one of the (a)CH2 CHCH2 CHCH2 CHSO3-H+CH2 CHCH2 CH CH2 CHSO3-H+CH2 CHSO3-H+CH2 CHSO3-H+(b)CH2 CHNR3+OH-CH2 CHNR3+OH-CH2 CHNR3+OH-CH2 CHCH2 CHCH2 CHNR3+OH-CH2 CHCH2 CH- 39 - most promising ways to recover lithium because of its high selectivity, but the gradual extraction of solvent from the resin, which makes the resin more difficult to be reused, remains problematic especially in oil and gas wastewater which can contain some organic phases. The two special types of ion exchangers above are not believed to be in commercial use. However, some commercial IX resins with unique functional groups may be candidates for lithium extraction as well. Table 2.13 lists the functional groups of cation exchange resins and chelating resins which were surveyed in this study. It was not sure that these functional groups are effective for selective lithium extraction, but commercial resins of these functional groups were tested for lithium recovery and the results are discussed in Chapter 4. Table 2.13 List of functional groups of IX resin studied in this thesis. Name Chemical structure Note Sulfonate Commonly used as cation exchange resin Phosphonate + Sulfonate Selective for antimony, bismuth and iron CHCH2 SOOO-M+CHCH2 POO-O-CHCH2 SOOO-M3+- 40 - Table 2.13 List of functional groups of IX resin studied in this thesis, cont’d. Name Chemical structure Note Iminodiacetate Selective for copper, cobalt, nickel and zinc Aminomethylphosphonate (Aminophosphonate) Selective for lead, copper, zinc, nickel, iron, antimony and bismuth Bispicolylamine Selective for copper and nickel Hydroxypropylpicolylamine Selective for copper and nickel CHCH2 CH2 NCH2CH2 COO-COO-M2+CHCH2 CH2 NHCH2 POO-O-M2+CHCH2 CH2 NCH2CH2NNM2+CHCH2 CH2 NCH2NM2+CH2CHCH3O-- 41 - Table 2.13 List of functional groups of IX resin studied in this thesis, cont’d. Name Chemical structure Note Thiouronium (Isothiouronium) Selective for mercury and precious metals Thiourea Selective for mercury and precious metals Thiol Selective for mercury and precious metals N-methylglucamine Selective for borate (boron) Amidoxime Selective for uranyl (uranium) and precious metals CHCH2 CH2 SCNH2NHM2+CHCH2 CH2 NHCNH2SM2+CHCH2 S-CHCH2 S- M2+BOOHOOCHCH2 CH2 NCH3CH2CHCHCHCHCH2OOOOHOHCHCH2 CH2 CNH2N O- CHCH2 CH2 CNNH2O- M2+- 42 - Table 2.13 List of functional groups of IX resin studied in this thesis, cont’d. Name Chemical structure Note Polyamine Selective for divalent heavy metals such as mercury, iron, copper, zinc Amphoteric (Quaternary amine + Carboxylate) Effective for salts removal 2.5.4. Al/Mn/Ti-based Adsorbent Some metal oxides/hydroxides also can adsorb lithium. They have a higher selectivity for lithium than ion exchange resins although it takes more time than ion exchange resins to complete extraction because lithium needs to be intercalated into the layers of the metal oxides/hydroxides. Aluminum-based adsorbents are typically LiCl/Al(OH)3 compounds, which have a similarity to the aluminum-loaded resin. The loading mechanism is the same as well. Since the aluminum-loaded resin only has adsorption sites on the surface of the resin, and the aluminum-based adsorbent has adsorption sites along its whole surface, the latter generally has a higher capacity. Some companies, FMC, Simbol and Eramet, have their patents for aluminum adsorbents.11–16 A manganese-based adsorbent was first reported by K. Ooi et al.18,19,48 and had high selectivity for lithium, raising interest among many researchers. Some papers were published after the first report; however, the adsorption mechanism is essentially the same as the aluminum adsorbents. The manganese-based adsorbent is now being tested with the Uyuni salt lake in Bolivia CHCH2 CH2 NHCH2CH2NHNHNH2MCHCH2 CH2 N+X-CH3CH3CH2 COO-M+- 43 - by Japan Oil, Gas and Metals National Corporation (JOGMEC).20 A titanium-based adsorbent also has comparable capacity for lithium recovery according to the study by R. Chitrakar et al.,21 and Neometals used a similar kind of titanium-based absorbent to extract lithium directly from brine.22 Furthermore, the three adsorbents are granulated for practical operation in columns, and they are currently being studied in terms of their optimal binder and the method of granulation. In spite of the potential of the three adsorbents, there are still few practical operations; only a few pilot tests have been carried out. 2.5.5. Solvent Extraction Solvent extraction is another promising method for lithium recovery.49 Lithium ion is extracted and stripped by the reaction between aqueous and organic phase (Equation (2.10)). orgaqaqorgorgaqorgaqHL Li H LiL :StrippingLiL H HL Li :Loading+→++→+++++ (2.10) The symbol L refers to the lithium extractant. Several extractants have been developed to satisfy each condition for each metal. For example, hydroxyoxime extractants from BASF and Solvay are very famous for copper SX from dilute leach solutions. Solvent extraction technology shows higher selectivity for lithium over other monovalent ions such as sodium and potassium ions. On the other hand, other divalent ions, such as magnesium and calcium ions, should be removed and lithium should be pre-concentrated before the solvent extraction step to maintain the efficiency of the process. As for a practical case, Tenova, an Italian company, is now testing an SX process in Israel.9,10 Figure 2.22 is a flow sheet of the process. This process uses SX to extract lithium and separate sodium and potassium after calcium and magnesium removal. The extracted lithium is - 44 - directly conveyed to electrolysis to make lithium hydroxide solution from lithium sulfate. Figure 2.22 A flow sheet of lithium production using solvent extraction by Tenova.9 2.5.6. Membrane Technologies Membrane technology is also expected to play an essential role in lithium extraction. Somrani et al.28 studied nanofiltration and low pressure reverse osmosis for lithium ion separation from brine. Although temperature and pressure control are required, the membrane process can be applied to low concentrations of lithium, such as lithium in oil and gas wastewater. Furthermore, while fouling is a huge concern, as discussed in other membrane applications, the membrane process can be applied to a variety of brines or wastewaters when conditions are adequately determined and controlled. MGX has reported success in lithium extraction from oilfield produced wastewater.29,30 Besides nanofiltration, some researchers have investigated membrane electrolysis. - 45 - Nieto et al.34 studied the removal of magnesium and calcium by making hydroxide ions by membrane electrolysis. 2 22H O 2e H 2OH− −+ → + (2.11) With an anode of a titanium mesh coated with an iridium-based mixed metal oxide (IrO2/TiO2; 65/35%), a cathode of a stainless steel wire mesh with a stainless steel current collector and an in-between anion exchange membrane, hydroxide ions are produced on the cathode side by Equation (2.11). 2222Mg 2OH Mg(OH)Ca 2OH Ca(OH)+ −+ −+ →+ → (2.12) Then, with the reactions in Equation (2.12), the produced hydroxides precipitate magnesium and calcium ions in brine. This technology is novel because it does not need to use either base like sodium hydroxide or carbonate salts like soda ash in order to remove both ions from brine. 2.5.7. Ferric Phosphate Reduction This technology uses the redox couple of ferric phosphate (FP) and lithium ferrous (or iron) phosphate (LFP), which is inspired by the mechanism of a rechargeable battery cell of lithium iron phosphate. LFP is one of the most popular cathode materials for LIBs because of low cost, performance and safety. The practical capacity is typically 140–160 mAh/g while the theoretical capacity is 170 mAh/g. When FP is reduced in aqueous solution by a reducing agent or an external potential, lithium ions in the solution are selectively adsorbed into FP with the formation of LFP. According to the existing researches, it is known that only the reversible couple of “heterosite-structure FP” and “olivine-structure LFP” can take up and release lithium selectively due to their unique structure. Therefore, the structures of heterosite FP and olivine LFP are firstly reviewed - 46 - here before getting into the detail of this technology. FP has several structures. The most common form adopts α-quartz structure, and two orthorhombic phases and one monoclinic phase are also known. One of the orthorhombic FPs is heterosite FP. This structure is named after heterosite (Fe3+,Mn3+)PO4. Heterosite FP is in orthorhombic lattice system, Pbnm (No. 62) space group with lattice parameters of a=4.76 Å, b=9.68 Å, c=5.82 Å.50 Figure 2.23 is the structure of heterosite FP drawn by VESTA. The figures clarify that, among PO4 tetrahedra along the c-axis, there are octahedral spaces for some cations being inserted. Figure 2.23 Schematic illustrations of heterosite FP structure: (a) oblique view, (b) view along the c-axis (Drawn by VESTA). LFP is generally known with the mineral name of triphylite and has only one structure, (a) (b)PO4 tetrahedra FeO6 octahedra- 47 - which is the olivine structure. This structure is named after olivine (Mg2+,Fe2+)2SiO4, a common mineral on the earth. Like olivine, triphylite or LFP is in orthorhombic lattice system, Pbnm (No. 62) space group with lattice parameters of a=4.71 Å, b=10.38 Å, c=6.05 Å.51 This means that heterosite FP and olivine LFP have almost the same structure. With the parameters, the illustrations of triphylite structure can be drawn by VESTA in Figure 2.24. By comparing Figure 2.24 with Figure 2.23, it is shown that olivine LFP is structured by the octahedral spaces of heterosite FP being occupied by lithium ions. When the transformation between FP and LFP occurs by lithium insertion and desertion, the unit-cell volume expands only by 6.5%.52 This stability is an advantage of LFP as a battery material. On the other hand, it also has a structural drawback. Since there are only one-dimensional pathways for lithium ion migration (c-axis direction), LFP/FP has lower conductivity than other cathode materials with two- or three-dimensional pathways. This phenomenon was reported and discussed by some researchers in terms of diffusion coefficient, activation energy and migration energy.53–56 When the structure is reviewed in detail, the strong P-O covalent bonds of PO4 tetrahedra keep oxygen from being released, which contributes to high resistance to thermal ignition.57 This is also why the FP/LFP couple is preferable as a cathode material of LIB in terms of safety. The P-O covalent bonds additionally weaken covalent bonds of the iron ion and lower the Fe3+/Fe2+ redox potential, which makes the Fe3+/Fe2+ couple easier to be oxidized and reduced even in the solid phase of FP or LFP.58 Thanks to those unique structures, the FP/LFP redox couple can absorb and release lithium selectively and repeatedly. When it comes to methods to make LFP powders, they can be synthesized by both solid-state and solution-based methods59. Standard solid-state methods are mechano-chemical activation, carbo-thermal reduction and microwave heating. They are simple ways to obtain well-crystallized structure, but typically consume so much time and energy and often result in poor purity. On the other hand, solution-based methods can usually obtain high purity and can control particle sizes - 48 - for the homogeneous property. Solution-based methods include hydrothermal synthesis, sol-gel synthesis, spray pyrolysis, co-precipitation and microemulsion drying. Figure 2.24 Schematic illustrations of triphylite structure: (a) oblique view, (b) view along the c-axis (Drawn by VESTA). Coming back to the lithium extraction technology, the ferric phosphate reduction method is not used in practical sites currently, but its research has been underway in recent years. While LFP is widely used in rechargeable batteries, it is only recently that the FP/LFP redox couple has started to be used to recover lithium selectively from brine. X. Liu et al. firstly proposed an electrochemical cell with FP/LFP electrodes for lithium extraction from brine.31,32 With an external potential applied, lithium ions are extracted from catholyte brine, and lithium ions are stripped (a) (b)PO4 tetrahedra FeO6 octahedra Li+ ion- 49 - from anode LFP into anolyte like NaCl solution as shown in Equation (2.13). 4 44 4Cathode: FePO +Li e LiFePOAnode: LiFePO FePO +Li e+ −+ −+ →→ + (2.13) The electrochemical cell consumes electrical energy. As an alternative, Intaranont et al.23–25 used thiosulfate (S2O32-, TS) as a reducing agent to promote the redox reaction of FP/LFP. As in Equation (2.14), thiosulfate is oxidized to tetrathionate by reducing FP particles into LFP particles. The reduction potential of FP/LFP is 3.45 V vs. Li/Li+ 60 and 0.40 V vs. SHE23, while S4O62-/S2O32- reduction potential is -0.015 V vs. SHE from HSC database. It is clear that the overall reaction is favorable because the potential difference is positive. 04 42 2 02 3 4 62 2 04 2 3 4 4 6Cathodic: FePO +Li e LiFePO E 0.40V vs.SHEAnodic : 2S O S O 2e E 0.015V vs.SHEOverall : 2FePO +2Li 2S O 2LiFePO S O E 0.42V vs.SHE+ −− − −+ − −+ → =→ + = −+ → +  = (2.14) Figure 2.25 is an Eh-pH diagram for the TS-H2O system. As there is no potent oxidizing agent in the experimental system of this research, the diagram considers the gentle oxidation into tetrathionate (S4O62-) and possible acid decomposition of Equation (2.15). - 50 - Figure 2.25 pH-Eh diagram for TS-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC). Sulfate species (SO42- and HSO4-) were excluded in this diagram. 2 2 3 2 2Na S O 2HCl 2NaCl S SO H O+ → +  +  + (2.15) Conversely, Intaranont et al.23–25 prepared FP for the lithium extraction by the delithiation of LFP with persulfate (S2O82-, PS) in Equation (2.16). The reduction potentials of S2O82-/SO42- is 1.94 V vs. SHE from HSC database. This reaction is also favorable from a thermodynamic point of view. 2 2 02 8 404 42 2 04 2 8 4 4Cathodic: S O 2e 2SO E 1.94V vs.SHEAnodic : LiFePO FePO +Li e E 0.40V vs.SHEOverall : 2LiFePO S O 2FePO +2Li 2SO E 1.54V vs.SHE− − −+ −− + −+ → =→ + =+ → +  = (2.16) Figure 2.26 is an Eh-pH diagram for the PS(persulfate)-H2O system. Both the reaction equation 141210864202.01.51.00.50.0-0.5-1.0-1.5-2.0TS - H2O - System at 25.00 CC:\\HSC6\\EpH\\S25.IEP pHEh (Volts)SSO2(g)H2S(a)HS(-a)S(-2a)S2O3(-2a)S4O6(-2a)ELEMENTS Molality PressureS 1.000E+00 1.000E+00FPLFP 0.42 V- 51 - and the diagram prove that the reduction potential of persulfate is high enough to oxidize LFP to FP. Figure 2.26 pH-Eh diagram for PS-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC). With the reducing and oxidizing reactions, Intaranont et al.23–25 studied the lithium extraction from brine and reported the following points. 1) Heterosite-FePO4 was chosen as FP adsorbent due to its unique structure, where Li+ can be inserted to form olivine-structured LFP.61 2) Heterosite-FePO4 can be obtained by the delithiation of LFP with the use of potassium PS (K2S2O8) as an oxidizing agent.62 3) The delithiation rate was dependent on the molar ratio of PS:LFP, while it took 2 141210864202.01.51.00.50.0-0.5-1.0-1.5-2.0PS - H2O - System at 25.00 CC:\\HSC6\\EpH\\S25.IEP pHEh (Volts)SH2S(a)HS(-a)HSO4(-a)S(-2a)SO4(-2a)S2O8(-2a)ELEMENTS Molality PressureS 1.000E+00 1.000E+00FPLFP1.54 V- 52 - hours at most to convert 99% of LFP to FP in the case of the minimum theoretical reaction ratio of 1:2. 4) Sodium TS (Na2S2O3) completed the lithiation of FP into LFP. 5) The kinetics of the lithiation could be explained by a surface-reaction limited model. 6) The lithiation rate was first-order proportional to both lithium and thiosulfate concentrations, and an overall second order rate constant was around 0.03 min-1 M-2. 7) The maximum adsorption capacity was 46 mg-Li/g-solid, which is comparable with that of any other types of adsorbents except some manganese-based adsorbents with the capacity of 38–40 mg-Li/g-solid63–66. 8) The extractions of other ions such as sodium, potassium and magnesium from a brine solution were less than 4 mg/g-solid. In place of thiosulfate, other reducing agents have been studied by some researchers. Iodide (I-) is one of them. Kuss et al.67 successfully converted FP into LFP with lithium iodide in acetonitrile, which is an organic solvent. Intaranont tried to use the same reaction of iodide in aqueous solution as shown in Equation (2.17). The potential of I2/I- couple is from HSC database. 04 40204 4 2Cathodic: FePO +Li e LiFePO E 0.40V vs.SHEAnodic : 2I I 2e E 0.62V vs.SHEOverall : 2FePO +2Li 2I 2LiFePO I E 0.22V vs.SHE+ −− −+ −+ → =→ + =+ → +  = − (2.17) Since the overall reaction is thermodynamically unfavorable, the reaction was aided with the zinc-iodine reaction of Equation (2.18).24 2 2I Zn ZnI+ → (2.18) However, the author found that the reaction was incomplete although a high concentration of iodide was used and iodine was removed from aqueous phase to a solid phase of zinc iodide in - 53 - order to promote the overall reaction in Equation (2.17) by following Le Chatelier’s principle. Figure 2.27 pH-Eh diagram for I-H2O system at 25 °C, 1 atm, [I]=1 M (Drawn by HSC). Figure 2.27 is an Eh-pH diagram for I-H2O system. Triiodide (I3-) is a byproduct of the I2/I- couple formed by the following reaction. Regardless of the presence of triiodide, the diagram clearly shows that the reduction potential of iodine/iodide couple is above that of FP/LFP couple. 2 3I I I− −+ → (2.19) The same author studied sodium nitrite (NaNO2), formaldehyde (CH2O), formic acid (HCOOH), and sodium sulfite (Na2SO3) as alternative reducing agents and found that sulfite was the most suitable. It is reported that sulfite fully converted FP to LFP.24 Equation (2.20) shows the 141210864202.01.51.00.50.0-0.5-1.0-1.5-2.0I - H2O - System at 25.00 CC:\\HSC6\\EpH\\I25.iep pHEh (Volts)HIO3(a)I2(a)I(-a)I3(-a)IO3(-a)ELEMENTS Molality PressureI 1.000E+00 1.000E+00FPLFP-0.22 V- 54 - reactions which are considered to occur. The potential of the sulfate/sulfite couple at each pH is calculated from data in HSC. 04 423 2 402 23 2 4024 3 2 4 40Cathodic: FePO +Li e LiFePO E 0.40V vs.SHEAnodic :HSO H O SO 3H 2eE 0.34, 0.52V vs.SHE (pH 5,7)SO H O SO 2H 2eE 0.64V vs.SHE (pH 9)Overall :2FePO +2Li HSO H O 2LiFePO SO 3HE 0+ −− − + −− − + −+ − − ++ → =+ → + += − − =+ → + += − =+ + → + + =2 24 3 2 4 40.74, 0.92V vs.SHE (pH 5,7)2FePO +2Li SO H O 2LiFePO SO 2HE 1.04V vs.SHE (pH 9)+ − − +=+ + → + + = = (2.20) The advantage of sulfite over thiosulfate as a reducing agent is that sulfite needs an only half dosage of thiosulfate since SF reacts with FP with the ratio of one to two while TS does with the ratio of one to one. Figure 2.28 is an Eh-pH diagram for SF-H2O system. The diagram considers the acid decomposition of sulfite into sulfur dioxide in Equation (2.21). 2 3 2 2Na SO 2HCl 2NaCl SO H O+ → +  + (2.21) - 55 - Figure 2.28 pH-Eh diagram for SF-H2O system at 25 °C, 1 atm, [S]=1 M (Drawn by HSC). According to the equation and diagram, the potential difference is dependent on pH and, in the range of neutral pH, it is around 1 V. This difference is more than twice that of the thiosulfate case, ΔE°=0.42 V. At the same time, it should be noted that, under an uncontrolled pH condition, the potential difference decreases as the reaction produces protons. As for an oxidizing agent, hydrogen peroxide is an alternative to persulfate. LFP would be oxidized by the following reaction in (2.22). The potential of hydrogen peroxide is from HSC database. The potential difference ΔE°=1.44 V is comparable to that of persulfate ΔE°=1.54 V. However, it’s not promising because it can only be used as an oxidizing agent in acid solution,24 where ferric phosphate is partially dissolved into solution. 141210864202.01.51.00.50.0-0.5-1.0-1.5-2.0SF - H2O - System at 25.00 CC:\\HSC6\\EpH\\S25.IEP pHEh (Volts)SO2(g) HSO3(-a)HSO4(-a)SO3(-2a)SO4(-2a)S2O8(-2a)ELEMENTS Molality PressureS 1.000E+00 1.000E+00FPLFP1.04 V0.92 V0.74 V- 56 - 02 2 204 404 2 2 4 2Cathodic: H O 2H 2e 2H O E 1.84V vs.SHEAnodic : LiFePO FePO +Li e E 0.40V vs.SHEOverall : 2LiFePO H O 2H 2FePO +2Li 2H O E 1.44V vs.SHE+ −+ −+ ++ + → =→ + =+ + → +  = (2.22) 2.6. Summary and Objectives This chapter firstly focused on lithium chemistry from a scientific point of view. Second, the lithium market was reviewed in order to grasp supply and demand in the future. Lithium production was then reviewed from types of deposit to processes and technologies. Out of the technologies reviewed, this thesis focused on ion exchange resin and ferric phosphate reduction. The following summarizes the development of each technology and objectives for this research. Ion exchange resin probably has the longest history as a rapid lithium recovery process. Although this method is a conventional method to recover metal ions from solutions, lithium is much more challenging to adsorb selectively than other metal ions. The reason is that, in brine, a much higher concentration of sodium, potassium, calcium and magnesium ions are present (sometimes 100 times as high as lithium concentration in mol/L), and these ions have a stronger affinity than lithium to general cation exchange resins. Despite the unfavorable conditions for ion exchange resins, in the 1980s, Dow Chemical in the US invented aluminum-loaded resins which can adsorb lithium selectively from brine.5–7 However, as the invention needs some special conditions to be carried out and the selectivity is not so high, it has not yet been applied to a practical situation. Since then, there are fewer promising reports about lithium-selective ion exchange resins for brine while a lot of new types of resins have been developed like chelating resins. Therefore, this thesis tested a wide variety of commercial resins from various producers and summarized the current situation of ion exchange resins around lithium extraction from brine. Ferric phosphate is a relatively new adsorbent for lithium extraction from brine. While - 57 - lithium iron phosphate is widely used in rechargeable batteries, it is only recently that the ferric phosphate (FP)/lithium iron phosphate (LFP) redox couple is used to recover lithium selectively from brine. X. Liu et al. used an electrochemical cell with FP/LFP electrodes for lithium extraction from brine.31,32 Furthermore, as an electrochemical cell consumes electrical energy, Intaranont et al. used sodium thiosulfate as a reducing agent to promote the redox reaction of FP/LFP to load lithium from a mixed salt brine.23–25 To be sure, this method using a reducing agent showed excellent lithium selectivity over other cations such as Na+, K+, Mg2+, but more studies should be carried out and it is necessary to develop a scientific understanding in order to apply the process into a practical site. Therefore, in this thesis, lithium extraction by FP/LFP redox reaction promoted by reducing agents was studied from the following points of view. 1) What kind of material is the best to make heterosite FP? 2) Which reducing agent is favorable? 3) How much are the capacity and selectivity? 4) How do temperature and pH have an impact on the process? 5) Is it possible to recycle ferric phosphate? 6) How long does it take to extract lithium? (Kinetics) Overall, this study investigated the possibility of the use of ion exchange resin and ferric phosphate method for lithium extraction from brine. Especially, the points which distinguish this work from the other works are as follows. a) Many kinds of commercial IX resins were screened for lithium extraction from brine b) Some precursor materials were compared for the formation of heterosite FP - 58 - c) Some reducing agents were investigated and compared for the FP reduction into LFP d) Lithium extraction by the sulfite reduction was investigated in detail for the first time e) Effects of temperature and pH on the FP method were well investigated for the first time f) Simplified kinetic models were developed for the FP method at different temperature This thesis consists of 6 chapters. After an explanation of the experimental methods, Chapter 4 and Chapter 5 discuss the experimental results of lithium extraction by ion exchange resin and ferric phosphate respectively. Finally, Chapter 6 concludes all the discussions and mentions challenges and outlook for future works. - 59 - Chapter 3. Experimental Methods 3.1. Preparation This work required preparation of synthetic solutions from various salts, measurement of pH and the testing of many ion exchange materials and ferric phosphate substrates. The chemical materials, sources of ion exchange materials obtained, and experimental and analytical equipment used are summarized below 3.1.1. Materials All the experiments in this study were carried out with the following materials unless otherwise noted. IX resins are listed in Table 3.1 with their characteristics.  Buffer, Reference Standard pH 10.00 ± 0.01 at 25 °C (Color Coded Blue), VWR Analytical BDH®  Buffer, Reference Standard pH 4.00 ± 0.01 at 25 °C, VWR Analytical BDH®  Buffer, Reference Standard pH 7.00 ± 0.01 at 25 °C (Color Coded Yellow), VWR Chemical BDH®  Calcium chloride hexahydrate, BioUltra, ≥99.0% (calc. based on dry substance, KT), Sigma Aldrich  Calcium Standard for AAS, TraceCERT®, 1000 mg/L Ca in nitric acid, Sigma Aldrich  Copper(I) chloride, reagent grade, 97%, Sigma Aldrich  Hydrochloric acid 1.0 N, VWR Chemical BDH®  Hydrochloric acid 36.5 - 38.0% ACS, VWR Chemicals BDH®  Iron(III) phosphate dihydrate, Fe 29%, Sigma Aldrich - 60 -  Lithium chloride, anhydrous, free-flowing, Redi-Dri™, ACS reagent, ≥99%, Sigma Aldrich  Lithium iron phosphate (LiFePO4) powder, 500g/bag, LIB-LFP-500G, MSE supplies  Lithium iron(II) phosphate, powder, <5 μm particle size (BET), >97% (XRF), Sigma Aldrich  Lithium Standard for AAS, TraceCERT®, 1000 mg/L Li in nitric acid, Sigma Aldrich  Magnesium chloride hexahydrate ACS, Amresco  Magnesium Standard for AAS, TraceCERT®, 1000 mg/L Mg in nitric acid, Sigma Aldrich  Nitric acid 68 - 70%, ARISTAR® ACS, VWR Chemicals BDH®  Phosphorus Standard for AAS, TraceCERT®, 1000 mg/L P in H2O, Sigma Aldrich  Potassium chloride 99.0-100.5% ACS, VWR Chemicals BDH®  Potassium Persulfate (Crystalline Powder/Certified), Fisher Chemical  Potassium Standard for AAS, TraceCERT®, 1000 mg/L K in nitric acid, Sigma Aldrich  Sodium chloride ≥99.0% ACS, VWR Chemicals BDH®  Sodium hydroxide 1.0 N, VWR Chemical BDH®  Sodium Standard for AAS, TraceCERT®, 1000 mg/L Na in nitric acid, Sigma Aldrich  Sodium Sulfite Anhydrous (Crystalline/Certified ACS), Fisher Chemical  Sodium Thiosulfate Anhydrous (Certified), Fisher Chemical - 61 - Table 3.1 List of resins tested in this study. No. Manufacturer Name Structure Matrix Functional group Ionic form 1 Dow XE 832 Gel SDB Aluminum loaded - 2 Dow AMBERLITE IR120 plus Gel SDB Sulfonate Na+ 3 Lanxess LEWATIT MonoPlus S 108 Gel SDB Sulfonic acid Na+ 4 Purolite PUROLITE C100 Gel SDB Sulfonic acid Na+ 5 Mitsubishi DIAION SK1B Gel SDB Sulfonic acid Na+ 6 Samyang TRILITE SCR-B Gel SDB Sulfonate Na+ 7 Lanxess LEWATIT MonoPlus SP 112 Macroporous SDB Sulfonic acid Na+ 8 Purolite PUROLITE C150 Macroporous SDB Sulfonic acid Na+ 9 Purolite PUROLITE C160 Macroporous SDB Sulfonic acid Na+ 10 Purolite PUROMET MTC1500 Macroporous SDB Sulfonic acid Na+ 11 Mitsubishi DIAION PK216 Macroporous SDB Sulfonic acid Na+ 12 Mitsubishi DIAION PK228 Macroporous SDB Sulfonic acid Na+ 13 Samyang TRILITE CMP28 Macroporous SDB Sulfonate Na+ 14 Purolite PUROMET MTS9570 Macroporous SDB Phosphonic acid and Sulfonic acid Na+ 15 Dow AMBERLITE IRC748i Macroporous SDB Iminodiacetic acid Na+ - 62 - Table 3.1 List of resins tested in this study, cont’d. No. Manufacturer Name Structure Matrix Functional group Ionic form 16 Lanxess LEWATIT MonoPlus TP 207 Macroporous SDB Iminodiacetic acid Na+ 17 Lanxess LEWATIT MonoPlus TP 208 Macroporous SDB Iminodiacetic acid Na+ 18 Purolite PUROMET MTS9300 Macroporous SDB Iminodiacetic acid Na+ 19 Mitsubishi DIAION CR11 Macroporous SDB Iminodiacetate Na+ 20 Dow AMBERLITE IRC747 Macroporous SDB Aminomethylphosphonate Na+ 21 Lanxess LEWATIT MonoPlus TP 260 Macroporous SDB Aminomethylphosphonic acid Na+ 22 Purolite PUROMET MTS9500 Macroporous SDB Aminophosphonic acid Na+ 23 Lanxess LEWATIT MDS TP 220 Macroporous SDB Bis-picolylamine H2SO4 salt 24 Purolite PUROMET MTS9600 Macroporous SDB Bispicolylamine FB/SO4 25 Dow DOW XUS 43605 Macroporous SDB Hydroxypropylpicolylamine - 26 Dow DOW XUS 43600 Macroporous SDB Thiouronium - - 63 - Table 3.1 List of resins tested in this study, cont’d. No. Manufacturer Name Structure Matrix Functional group Ionic form 27 Lanxess LEWATIT MonoPlus TP 214 Macroporous SDB Thiourea - 28 Purolite PUROMET MTS9140 Macroporous SDB Thiourea - 29 Purolite PUROMET MTS9200 Macroporous SDB Isothiouronium H+ 30 Dow AMBERLITE IRA743 Macroporous Polystyrene N-methylglucamine Free Base 31 Mitsubishi DIAION CRB05 Macroporous SDB N-methyl glucamine Free Base 32 Purolite PUROMET MTS9100 Macroporous ADB Amidoxime Free Base 33 Purolite PUROMET MTS9240 Macroporous SDB Thiol H+ 34 Mitsubishi DIAION CR20 Macroporous SDB Polyamine Free Base 35 Mitsubishi DIAION AMP03 Gel SDB Quaternary amine and Carboxylate Inner Salt 3.1.2. Apparatus All the experiments in this study were carried out with the following apparatus unless otherwise noted. - 64 -  Electric furnace: Exciton  pH meter: Oakton ION 700  Shaker: Lab-Line Orbit Environ-Shaker 3527  Water bath shaker: Thermo Scientific SWB 25  AAS: Varian AA240  ICP-OES: Agilent ICP-OES 5110  XRD: Rigaku Multiflex 2kW 3.2. Procedures for Lithium Extraction with Ion Exchange Resin 3.2.1. Preparation of Ion Exchange Resin Before a series of lithium extraction experiments with IX resins, all the resins obtained from the companies were pretreated. The resins were washed with deionized (DI) water several times to remove impurities on their surface and immersed into DI water so that they were fully hydrated. Then, they were vacuum filtered for just 30 seconds. These resins were directly used. After washing and before experiments, actual density and swelling ratio of the resins were measured for later calculation of the experimental results. Actual density da was calculated from the volume and mass of the resins as follows. aMass of wet resins [g-wet R]Actual density d [g-wet R/mL-wet R]=Volume of wet resins [mL-wet R] (3.1) The volume was measured by a 10 mL graduated cylinder. The resins were put into the cylinder with DI water and tapped so that they were fully packed. After vacuum filtration for 30 seconds, all the packed resins were weighed using an electronic scale. Next, the swelling ratio rs was - 65 - calculated with wet volume and dry mass in Equation (3.2). sVolume of wet resins [mL-wet R]Swelling ratio r [mL-wet R/g-dry R]=Mass of dry resins [g-dry R] (3.2) Dry mass was obtained after the weighed wet resins were dried at 60–65 °C in an oven for 12 hours. These measurements were repeated three times, and the average was used for subsequent experiments and calculation. The values of each resin are listed in Table 3.2. Table 3.2 Actual density and swelling ratio of resins. No. Resin da [g-wet R/L-wet R] rs [mL-wet R/g-dry R] 1 XE832 825 1.89 2 AMBERLITE IR120 plus 854 2.24 3 LEWATIT MonoPlus S 108 841 2.05 4 PUROLITE C100 814 2.21 5 DIAION SK1B 848 2.13 6 TRILITE SCR-B 829 2.18 7 LEWATIT MonoPlus SP 112 810 2.61 8 PUROLITE C150 800 2.27 9 PUROLITE C160 839 1.98 10 PUROMET MTC1500 799 2.50 11 DIAION PK216 818 2.29 12 DIAION PK228 826 1.90 13 TRILITE CMP28 841 1.83 14 PUROMET MTS9570 751 3.62 - 66 - Table 3.2 Actual density and swelling ratio of resins, cont’d. No. Resin da [g-wet R/L-wet R] rs [mL-wet R/g-dry R] 15 AMBERLITE IRC748i 720 3.99 16 LEWATIT MonoPlus TP 207 740 3.41 17 LEWATIT MonoPlus TP 208 749 3.49 18 PUROMET MTS9300 763 2.93 19 DIAION CR11 713 3.89 20 AMBERLITE IRC747 745 3.97 21 LEWATIT MonoPlus TP 260 768 3.40 22 PUROMET MTS9500 749 3.63 23 LEWATIT MDS TP 220 735 3.02 24 PUROMET MTS9600 707 2.98 25 DOW XUS 43605 696 3.22 26 DOW XUS 43600 725 2.23 27 LEWATIT MonoPlus TP 214 735 3.25 28 PUROMET MTS9140 686 3.19 29 PUROMET MTS9200 718 2.93 30 AMBERLITE IRA743 699 2.90 31 DIAION CRB05 730 2.71 32 PUROMET MTS9100 791 2.14 33 PUROMET MTS9240 730 2.62 34 DIAION CR20 668 3.35 35 DIAION AMP03 703 2.84 - 67 - 3.2.2. Lithium Loading Test Lithium loading test was performed to see how much lithium could be extracted from a brine solution. First of all, two types of synthetic solution were prepared with DI water and reagents. The first was a lithium chloride (LiCl) solution and the other was a brine containing chloride of lithium, sodium, potassium, magnesium and calcium. Table 3.3 shows the concentration of each metal ion in both solutions and they were labeled as A1 and A2 solution respectively. The concentration was determined roughly by referring to that of saline lakes in South America.68–70 Table 3.3 Metal ion concentration in synthetic LiCl solution and brine. Species LiCl (A1) Brine (A2) mg/L mol/L mg/L mol/L Li 500 0.07 500 0.07 Na - - 90000 3.91 K - - 6000 0.15 Mg - - 2500 0.10 Ca - - 2000 0.05 The washed wet resins were contacted with both solutions in batch tests. A 300 mL volume of LiCl solution or brine was prepared in an Erlenmeyer flask. Initial pH before putting IX resins was always around 6 at ambient temperature. Before adding the resin, 0.1 mL of the solution was taken by a syringe filter for AAS analysis of initial concentration. Then, 1.5 mL of the wet resins were added to the solution, which gave the dosage of 5 mL-wet R/L. Since it was hard to measure such - 68 - a small volume of resins with a graduated cylinder, the volume was converted into weight by the actual density prepared in the last section so that 1.5 mL of resins could be taken by weight on an electronic balance. With the flask covered with parafilm, it was shaken for 24 hours on the shaker at 220 rpm at an ambient temperature of around 20 °C. After 24 hours, 0.1 mL of the solution was taken again for AAS analysis of final concentration. Both samples for AAS analysis were diluted 200 times in solution with 1% of hydrochloric acid (HCl) and 3.81 g/L of potassium chloride (KCl). After AAS analysis, adsorption density qe [mg-Li/g-dry R] was calculated as follows: o eeRsC [mg-Li/L] C [mg-Li/L]Adsorption density q [mg-Li/g-dry R]=C [mL-wet R/L]r [mL-wet R/g-dry R]− (3.3) where Co: initial lithium concentration in solution before loading, Ce: equilibrium (or final) lithium concentration in solution after loading, CR: resin concentration in solution and rs: swelling ratio. 3.2.3. Lithium Stripping Test Lithium stripping test was carried out after the loading test to see how much lithium could be recovered from resins back into solution. First, all the resins after 24-hour loading were washed with DI water three times. The washed Li-loaded resins were used for a batch-style stripping test. They were put into 20 mL of 1 M HCl and shaken for 12 hours on the shaker at 220 rpm and ambient temperature. After 12 hours, 0.1 mL of the solution was taken for AAS analysis. The sample for AAS analysis was treated in the same way as described for the loading experiments. Adsorption density qe [mg-Li/g-dry R] was calculated as follows: eeRsC [mg-Li/L]Adsorption density q [mg-Li/g-dry R]=C [mL-wet R/L]r [mL-wet R/g-dry R] (3.4) where Ce: equilibrium lithium concentration in solution after stripping, CR: resin concentration in - 69 - solution and rs: swelling ratio. 3.3. Procedures for Lithium Extraction with Ferric Phosphate 3.3.1. Preparation of Ferric Phosphate Ferric phosphate powder was prepared in two different ways before a series of lithium extraction experiments because there is no commercial reagent of ferric phosphate. One was dehydration of ferric phosphate dihydrate (FePO4·2H2O, FPD) and the other was oxidation of battery-grade lithium iron phosphate (LiFePO4, LFP). First, the dehydration of FPD was done by calcination. FPD powder was put into a crucible and calcined for 2 hours at 300–600 °C in the electric furnace. Second, oxidation of LFP was treated chemically by potassium persulfate (K2S2O8) in aqueous solution. Equation (3.5) is the oxidation reaction and LFP and PS react in the ratio of 2:1. 4 2 2 8 4 2 4 2 42LiFePO K S O 2FePO +Li SO +K SO+ → (3.5) Here in this experiment, in order to oxidize LFP completely, LFP and PS were reacted in the ratio of 1:1. Specifically, 370 mL of 0.2 M PS solution was prepared in an Erlenmeyer flask and 0.2 M of LFP powder (11.67 g/370 mL) was put into the solution. Then, it was shaken on the shaker for 24 hours at 220 rpm and ambient temperature. After 24 hours, the powder was vacuum filtered and dried at 60–65 °C for further experiments and XRD analysis and acid digestion as well. The solution was also recovered and diluted 100 times in solution with 2% of nitric acid (HNO3) for ICP analysis - 70 - 3.3.2. Lithium Loading Test A lithium loading test was performed to see how much lithium could be extracted from solution onto ferric phosphate. First, eight kinds of synthetic solution were prepared with DI water and reagents as shown in Table 3.4. In order to study the effects of other ions on lithium extraction, solutions with each metal ion, Li+, Na+, K+, Mg2+ and Ca2+, were separately prepared in addition to brines. Each of the solutions was labeled as B1 to B8 respectively. Table 3.4 Metal ion concentration in eight synthetic solutions. Species Brine (B1) Brine w/o Ca (B2) Brine w/o Na Ca (B3) LiCl (B4) NaCl (B5) KCl (B6) MgCl2 (B7) CaCl2 (B8) mg/L mol/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Li 694 0.10 694 694 694 - - - - Na 45980 2.00 45980 -* - 45980 - - - K 5865 0.15 5865 5865 - - 5865 - - Mg 2431 0.10 2431 2431 - - - 2431 - Ca 2004 0.05 - - - - - - 2004 *There was still some concentration of sodium from reducing agents (Na2S2O3 and Na2SO3). The prepared FP was tested for these eight solutions by batch testing. A 100 mL volume of each solution was prepared in a plastic bottle. Before adding FP powder, 0.1 mL of the solution was taken by a syringe filter for ICP analysis of initial concentration. The initial pH before adding a reducing agent and LFP was always around 6 at ambient temperature. A reducing agent was - 71 - added and dissolved into the solution. The dosage of sodium thiosulfate and sodium sulfite was 15.8 g/L (=0.1 mol/L) and 12.6 g/L (=0.1 mol/L) respectively. Then, 0.75 g of FP was added, which gave the dosage of 7.5 g/L or 0.05 mol/L. A lid was placed on the bottle and it was shaken for 24 hours on the water bath shaker at 100 rpm at 25, 45 or 65 °C. The pH was uncontrolled or controlled at 5, 7 or 9 with sodium hydroxide. During the 24 hour experiment, 0.1 mL samples of the solution were taken several times at a certain time for ICP analysis. All the samples for ICP analysis were diluted 1000 times in solution with 2% of nitric acid (HNO3). After 24 hours, the powder was washed with DI water three times, vacuum filtered and dried at 60–65 °C for stripping experiments or for XRD analysis or acid digestion. 3.3.3. Lithium Stripping Test A lithium stripping test was carried out after each loading test to see how much lithium could be recovered from LFP back into solution. The dried powder after loading was used for batch-style stripping test. A stripping method is same as that of preparation of FP, but the volume of solution was varied. A mass of 0.63 g of LFP was put into 20 mL of 0.2 M K2S2O8 (LFP:PS=1:1) and shaken for 24 hours on the shaker at 220 rpm and ambient temperature. After 24 hours, the powder and solution were recovered and treated for XRD and digestion/ICP analysis in the same way above. 3.3.4. Acid Digestion In order to analyze an exact ratio of elements in the solid, FP or LFP powder was digested with concentrated HCl. In a small beaker, 0.01 g of FP or LFP was reacted with 5 mL of 38% HCl and 5 mL of DI water for an hour under a fume hood. After one hour, all the solution was filtered, - 72 - collected in a 100 mL volumetric flask and diluted in 2% HNO3 for ICP analysis. Based on the analysis results of acid digestion, adsorption density Qe [mg/g-FP] was calculated as follows for each element: MeFPm [mg]Adsorption density Q [mg/g-FP]=m [g] (3.6) where mM: weight of an element in solid after loading, mFP: weight of ferric phosphate used for loading. The selectivity of lithium over other ions KLi [-] was also calculated as follows: Li,solidLiLi,solutionK [mol-Li/mol-M]Selectivity K [-]=K [mol-Li/mol-M] (3.7) where KLi,solid: molar ratio of lithium over each metal in solid after loading, KLi,solution: molar ratio of lithium over each metal in solution after loading. 3.4. Analysis Procedures 3.4.1. Atomic Absorption Spectroscopy (AAS) The aqueous samples were diluted with a solution of 1% HCl and 3.81 g/L KCl and were directly analyzed by AAS. Since AAS can measure only one element at the same time, in this study, AAS was used to measure the concentration of lithium (Li) only. The measurement was carried out at a wavelength of 670.9 nm, which is the most sensitive peak of Li. 3.4.2. Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) The aqueous samples were diluted with 2% HNO3 and were directly analyzed by ICP. ICP was used to measure the concentration of lithium (Li), sodium (Na), potassium (K), magnesium (Mg), calcium (Ca), iron (Fe), phosphorus (P) at the same time. The wavelength for each element - 73 - was set as follows in Table 3.5. Table 3.5 Wavelength for the measured elements. Element Wavelength [nm] Li 670.78 Na 589.59 K 766.49 Mg 280.27 Ca 396.85 Fe 259.94 P 213.62 3.4.3. X-ray Diffraction (XRD) The dried solid samples were analyzed by XRD to determine their crystal structure. Samples were ground in a mortar and pestle if required. The samples were mounted and smoothed on a glass plate and analyzed by x-ray diffraction at 40 kV and 40 mA with Cu-Kα. - 74 - Chapter 4. Lithium Extraction with Ion Exchange Resins 4.1. Lithium Recovery from LiCl Solution First of all, all the resins listed in Table 3.1 were tested in lithium chloride solution (A1 solution) in order to check the capacity of each resin. Lithium capacity of each resin is summarized in Figure 4.1. Overall, functional groups of sulfonate, iminodiacetate and aminophosphonate showed better performance than the others with the capacity of around 20–30 mg/g. Only resins with relatively high capacity from XE 832 to PUROMET MTS9500 were tested in the next experiment with brine solution (A2 brine solution). - 75 - Figure 4.1 Lithium adsorption density of resins in LiCl solution in mg-Li/g-dry resin. Figure 4.2 is a graph of the weight of lithium stripped vs. that of lithium loaded. There is a good correlation between the results from the loading and the stripping. This is evidence of the accuracy of AAS measurement of the solution after both loading and stripping. 2.9422.021.120.521.320.622.021.318.420.521.318.918.016.324.927.532.928.022.130.630.627.23.380.2381.543.290.7780.8663.161.160.2177.010.9422.810.6820 10 20 30 40XE 832AMBERLITE IR120 plusLEWATIT MonoPlus S 108PUROLITE C100DIAION SK1BTRILITE SCR-BLEWATIT MonoPlus SP 112PUROLITE C150PUROLITE C160PUROMET MTC1500DIAION PK216DIAION PK228TRILITE CMP28PUROMET MTS9570AMBERLITE IRC748iLEWATIT MonoPlus TP 207LEWATIT MonoPlus TP 208PUROMET MTS9300DIAION CR11AMBERLITE IRC747LEWATIT MonoPlus TP 260PUROMET MTS9500LEWATIT MDS TP 220PUROMET MTS9600DOW XUS 43605DOW XUS 43600LEWATIT MonoPlus TP 214PUROMET MTS9140PUROMET MTS9200AMBERLITE IRA743DIAION CRB05PUROMET MTS9100PUROMET MTS9240DIAION CR20DIAION AMP03Capacity [mg-Li/g-dry R]- 76 - Figure 4.2 Correlation between results of loading and stripping tests in the case of lithium chloride solution (A1 solution). 4.2. Lithium Recovery from Brine Lithium capacity of each resin is summarized in Figure 4.3. The resins generally have poor selectivity (and lithium loading) when exposed to brine solution with high salinity. Only the aluminum loaded resin (XE 832) maintained a reasonable lithium capacity. Furthermore, the aluminum loaded resin showed a higher capacity in the brine compared to the simple solution of lithium chloride. It is indicated that the high concentration of other ions increases the activity of lithium ions and promotes lithium loading (calculation of activity is carried out in Chapter 5). 051015200 5 10 15 20Lithium stripped [mg]Lithium loaded [mg]- 77 - Figure 4.3 Lithium adsorption of resins in brine in mg-Li/g-dry resin. No Data = No experiment was conducted with brine because of its low capacity in pure LiCl solution. Figure 4.4 is a graph of the weight of lithium stripped vs. that of lithium loaded in the case of brine solution (A2 solution). There is no correlation between the results from the loading and the stripping. That is because lithium uptake was so small that measurement error would overlap the 6.6390.3360.3170.2790.3080.3360.3430.3180.3470.2930.2830.3130.3450.2090.0860.1550.1420.1660.0690.1360.0840.1320 10 20 30 40XE 832AMBERLITE IR120 plusLEWATIT MonoPlus S 108PUROLITE C100DIAION SK1BTRILITE SCR-BLEWATIT MonoPlus SP 112PUROLITE C150PUROLITE C160PUROMET MTC1500DIAION PK216DIAION PK228TRILITE CMP28PUROMET MTS9570AMBERLITE IRC748iLEWATIT MonoPlus TP 207LEWATIT MonoPlus TP 208PUROMET MTS9300DIAION CR11AMBERLITE IRC747LEWATIT MonoPlus TP 260PUROMET MTS9500LEWATIT MDS TP 220PUROMET MTS9600DOW XUS 43600LEWATIT MonoPlus TP 214PUROMET MTS9140PUROMET MTS9100PUROMET MTS9200PUROMET MTS9240DIAION CR20DIAION CRB05DOW XUS 43605AMBERLITE IRA743DIAION AMP03Capacity [mg-Li/g-dry R]No DataNo DataNo DataNo DataNo DataNo DataNo DataNo DataNo DataNo DataNo DataNo DataNo Data- 78 - difference between initial and final concentration in the loading experiment (samples were diluted 200 times when analyzed). Therefore, the adsorption density shown in Figure 4.3 was calculated from the results of the stripping experiment. Figure 4.4 Correlation between results of loading and stripping tests in the case of brine solution (A2 solution). 4.3. Summary Out of the resins listed in Table 3.1, the resins with functional groups of sulfonate, iminodiacetate and aminophosphonate had better performance in contact with lithium chloride solution (A1 solution). However, these resins could not recover lithium selectively from a brine 02468100 2 4 6 8 10Lithium stripped [mg]Lithium loaded [mg]XE 832(Aluminum loaded resin)- 79 - solution (A2 solution) except the aluminum loaded resin (XE 832). While XE 832 is worth investigating furthermore under various condition, the results indicated that the use of ion exchange materials is not so suited to recover lithium from brine solutions. Therefore, the attention of this study turned to iron phosphate based adsorbents. Figure 4.5 Lithium adsorption density of the resins which showed higher selectivity in LiCl solution. 0 10 20 30 40XE 832AMBERLITE IR120 plusLEWATIT MonoPlus S 108PUROLITE C100DIAION SK1BTRILITE SCR-BLEWATIT MonoPlus SP 112PUROLITE C150PUROLITE C160PUROMET MTC1500DIAION PK216DIAION PK228TRILITE CMP28PUROMET MTS9570AMBERLITE IRC748iLEWATIT MonoPlus TP 207LEWATIT MonoPlus TP 208PUROMET MTS9300DIAION CR11AMBERLITE IRC747LEWATIT MonoPlus TP 260PUROMET MTS9500Capacity [mg-Li/g-dry R]0 10 20 30 40XE 832AMBERLITE IR120 plusLEWATIT MonoPlus S 108PUROLITE C100DIAION SK1BTRILITE SCR-BLEWATIT MonoPlus SP 112PUROLITE C150PUROLITE C160PUROMET MTC1500DIAION PK216DIAION PK228TRILITE CMP28PUROMET MTS9570AMBERLITE IRC748iLEWATIT MonoPlus TP 207LEWATIT MonoPlus TP 208PUROMET MTS9300DIAION CR11AMBERLITE IRC747LEWATIT MonoPlus TP 260PUROMET MTS9500Capacity [mg-Li/g-dry R]LiClaq BrineSulfonateAl loadedIminodiacetateSO3HCH2CH2NCH2 COOHCOOHAminomethylphosphonateCH2 NH CH2 POOHOH- 80 - Chapter 5. Lithium Extraction with Ferric Phosphate 5.1. Preparation of Ferric Phosphate 5.1.1. Ferric Phosphate Dihydrate Ferric phosphate dihydrate (FPD) was calcined at temperatures of 300, 500 and 600 °C. The as-received FPD was amorphous. After calcining at 300 and 500 °C, there was no crystalline material left; just amorphous FPD. Crystal ferric phosphate (FP) was obtained at 600 °C. However, the structure was different from heterosite FP (suitable for Li extraction). The calcine had another structure of ferric phosphate: α-quartz type. Figure 5.1 XRD patterns of ferric phosphate dihydrate calcined at 300, 500 and 600 °C. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FePO4·2H2O (Sigma Aldrich)Calcined at 600°CCalcined at 500°CCalcined at 300°CFerric phosphate (α-quartz type)- 81 - After the loading test, there was no change in the FP structure, and lithium was not extracted. Figure 5.2 XRD patterns of ferric phosphate obtained by calcination of ferric phosphate dihydrate at 600 °C and samples after lithium loading tests at 20 °C and 65 °C. 5.1.2. Lithium Iron Phosphate (Reagent Grade) Reagent grade lithium iron phosphate was studied by delithiation by oxidation with potassium persulfate and then attempted relithiation by reduction with thiosulfate. It was confirmed by AAS analysis that lithium could be wholly extracted from LFP by PS oxidation. However, according to XRD results in Figure 5.3, the structure of oxidized FP was not well-crystallized heterosite FP. In addition, the delithiated FP couldn’t extract lithium even from LiCl solution when reduced by TS. This is mainly because the reagent grade LFP has low electronic conductivity59 and could not carry the electronic current during the reduction reaction. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FePO4・2H2O_calcined at 600°CLi loading at 20°CLi loading at 65°C- 82 - Figure 5.3 XRD patterns of reagent-grade LFP and delithiated and lithiated ones at 65 °C in comparison with calcined FPD. 5.1.3. Lithium Iron Phosphate (Battery Grade) LFP battery grade was converted to heterosite FP by PS oxidation. The complete extraction of lithium was confirmed by AAS analysis of solution after delithiation (100% of the lithium from LFP was extracted into solution). Typically, in order to get carbon coated and obtain high electronic conductivity, battery-grade LFP is synthesized in the presence of carbon black whose surface area is high59. Therefore, it is indicated that the successful result is due to the carbon coating. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)Reagent grade LiFePO4Delithiated at 65°CHeterosite FePO4 (reference)Heterosite ferric phosphate- 83 - Figure 5.4 XRD patterns of battery-grade (BG) LFP and delithiated FP by PS oxidation. In summary, it was not possible to prepare ferric phosphate from FPD and reagent-grade LFP while persulfate oxidation of battery-grade LFP successfully made heterosite FP, which can be used for lithium extraction. Therefore, ferric phosphate used as a lithium adsorbent in the following experiments was prepared from battery-grade LFP by persulfate oxidation. 5.2. Selection of Reducing Agents Some experiments were carried out to compare a selection of reducing agents to determine which is preferable for lithium extraction. Three agents were tested: cuprous (Cu+), thiosulfate (TS) and sulfite (SF). Cuprous was experimentally studied because no literature uses cuprous as a reducing agent for the FP method. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)DelithiatedBattery-grade FPHeterosite ferric phosphate- 84 - Figure 5.5 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by cuprous (Cu+) at room temperature (RoC), by cuprous (Cu+) at 65 °C (65C), by thiosulfate (TS) at 65 °C (65C), by sulfite (SF) at room temperature (RoC) and by sulfite (SF) at 65 °C (65C). All the loading experiments used B1 brine solution. First, reduction by cuprous was not successful while it dissolves well in brine with a high concentration of chloride as chloride complex. XRD patterns in Figure 5.5 show that it could not 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGCu+ RoCCu+ 65CTS 65CSF RoCSF 65CBotallackite (Cu2(OH)3Cl) Hannebachite (CaSO3·0.5H2O)- 85 - reduce FP into LFP completely. Also, it left an unfavorable material called Botallackite (Cu2(OH)3Cl). This result indicates that cuprous ions were oxidized immediately in solution by air before reacting with FP. Second, thiosulfate successfully reduced FP into LFP. There is an XRD pattern of LFP produced by thiosulfate reduction at 65 °C only in Figure 5.5 and it cannot promote the reaction entirely at ambient temperature, but it surely had the potential for lithium recovery. Third, sulfite is also a promising reducing agent. While it is a problem that sulfite ions make unfavorable salts with calcium called Hannebachite (CaSO3·0.5H2O), this reagent can recover lithium efficiently even at ambient temperature if there is no calcium in solution. In Figure 5.5, XRD patterns of LFP produced by sulfite reduction clarified that it could make LFP at room temperature and the higher temperature. Therefore, this study investigated thiosulfate and sulfite as a reducing agent for the reaction of lithium extraction in the following discussion. 5.3. Products after Lithium Loading The work reported above confirmed that heterosite FP could be made using thiosulfate and sulfite as reducing agents. This section starts to discuss details of the ferric phosphate based lithium extraction method. As a first step, the capacity and selectivity of this method were studied by contacting FP with solutions containing metal ions other than lithium by checking XRD patterns in the case of thiosulfate and sulfite separately. 5.3.1. Thiosulfate Reduction Figure 5.6 shows XRD patterns after thiosulfate reduction at 25 °C in B1–B8 brine solution. There is a bit of change of FP into LFP in the case of B1–B3 solution. However, there is - 86 - no change in the case of the lithium chloride solution (B4). This may be because higher salinity may be required to facilitate the reduction reaction of FP. The experiments with FP in contact with sodium, potassium, magnesium and calcium ions at 25 °C did not show evidence of change in structure. Figure 5.6 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 25 °C. The loading experiments used B1–B8 brine solution. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGTS Brine (B1)TS Brine w/o Ca (B2)TS Brine w/o Na, Ca (B3)TS Li (B4)TS Na (B5)TS K (B6)TS Mg (B7)TS Ca (B8)- 87 - Figure 5.7 shows XRD patterns after thiosulfate reduction at 45 °C in B1–B8 brine solution. In the case of B1–B3 solution, FP was almost wholly converted to LFP. In the case of lithium chloride solution (B4), there is a more substantial change in the pattern at 45 °C than 25 °C, but some FP was still in the final product. Again, a new material is not formed from FP in contact with sodium, potassium, magnesium and calcium ions at 45 °C. Figure 5.7 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 45 °C. The loading experiments used B1–B8 brine solution. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGTS Brine (B1)TS Brine w/o Ca (B2)TS Brine w/o Na, Ca (B3)TS Li (B4)TS Na (B5)TS K (B6)TS Mg (B7)TS Ca (B8)- 88 - Figure 5.8 shows XRD patterns after thiosulfate reduction at 65 °C in B1–B8 brine solution. FP was utterly reduced to LFP in B1–B4 solution while sodium iron phosphate (NaFePO4, SFP) was produced from FP and sodium ions at 65 °C. Potassium, magnesium and calcium ions did not make any compounds with FP even at 65 °C. Figure 5.8 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by thiosulfate (TS) at 65 °C. The loading experiments used B1–B8 brine solution. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGTS Brine (B1)TS Brine w/o Ca (B2)TS Brine w/o Na, Ca (B3)TS Li (B4)TS Na (B5)TS K (B6)TS Mg (B7)TS Ca (B8)NaFePO4- 89 - 5.3.2. Sulfite Reduction Figure 5.9 shows XRD patterns after sulfite reduction at 25 °C in B1–B8 brine solution. Firstly, B1 and B8 brine solution, which contains calcium ions, produced the salts of calcium and sulfite. This means that it is difficult to separate lithium from calcium by sulfite reduction. That’s why this study won’t discuss sulfite reduction in B1 and B8 solution from the next section. There are more significant LFP patterns than those observed for thiosulfate at 25 °C in the case of B2–B4 solution, but the pattern of FP remains. There was no significant change in the pattern when FP was in contact with sodium, potassium and magnesium ions in salt solutions at 25 °C. - 90 - Figure 5.9 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by sulfite (SF) at 25 °C. The loading experiments used B1–B8 brine solution. Figure 5.10 shows XRD patterns after sulfite reduction at 45 °C in B1–B8 brine solution. Hannebachite was also formed at 45 °C in B1 and B8 brine solution. In the case of B2–B4 solution, LFP was completely produced without FP remaining. Sodium, potassium and magnesium ions do not interact with FP even at 45 °C. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGSF Brine (B1)SF Brine w/o Ca (B2)SF Brine w/o Na, Ca (B3)SF Li (B4)SF Na (B5)SF K (B6)SF Mg (B7)SF Ca (B8)Hannebachite (CaSO3·0.5H2O)- 91 - Figure 5.10 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by sulfite (SF) at 45 °C. The loading experiments used B1–B8 brine solution. Figure 5.11 shows XRD patterns after sulfite reduction at 65 °C in B1–B8 brine solution. The calcium salt was formed again at 65 °C in B1 and B8 brine solution. In the case of B2–B4 solution, LFP was produced entirely without FP remaining. Sodium iron phosphate (SFP) was formed in B5 solution only. Solutions containing potassium and magnesium ions at 65 °C did not 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGSF Brine (B1)SF Brine w/o Ca (B2)SF Brine w/o Na, Ca (B3)SF Li (B4)SF Na (B5)SF K (B6)SF Mg (B7)SF Ca (B8)Hannebachite (CaSO3·0.5H2O)- 92 - change the structure of the solid. Figure 5.11 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after loading experiments by sulfite (SF) at 65 °C. The loading experiments used B1–B8 brine solution. 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFP-BGSF Brine (B1)SF Brine w/o Ca (B2)SF Brine w/o Na, Ca (B3)SF Li (B4)SF Na (B5)SF K (B6)SF Mg (B7)SF Ca (B8)NaFePO4Hannebachite (CaSO3·0.5H2O)- 93 - 5.4. Capacity and Selectivity The previous section reported on lithium extraction by checking XRD patterns of the products. It is confirmed that FP was appropriately reduced into LFP under most of the conditions, while there were some differences due to temperature and a reducing agent. Notably, it turned out that it was not useful to add sulfite as a reducing agent in B1 and B8 solution containing calcium ions. Based on the discussion in the last section, this section will discuss capacity and selectivity. 5.4.1. Thiosulfate Reduction Figure 5.12 shows the adsorption density of each metal on FP after thiosulfate reduction at 25 °C in B1–B8 brine solution. In the case of B2 brine solution, the adsorption density of lithium was 12.5 mg/g and the other metals were also adsorbed with the density of several milligrams per gram of FP. - 94 - Figure 5.12 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 25 °C (B1–B8 brine solution). When lithium adsorption density is compared between B1 and B4 solution, the adsorption density was much higher in B1 case than B4 case, although they contained the same molarity or molality of lithium. The reason why this phenomenon was observed is probably an effect of the activity of lithium. Since B1 solution was composed of a high concentration of chloride salts of other cations, the activity of each chloride salts must be different from its molality. To evaluate the difference, the mean ionic activity of each chloride salts was calculated by Meissner’s method71. Table 5.1 shows the calculated values of the mean ionic activity. It is found that activity of lithium chloride, 0.432 mol/kg-H2O, is four times as high as its molality, 0.100 mol/kg-H2O. Therefore, it’s reasonable to mention that the adsorption density was higher because lithium activity was increased by the high concentration of other chloride salts and the reaction of lithium insertion got 01020304050607080TS Brine(B1)TS Brinew/o Ca(B2)TS Brinew/o Na Ca(B3)TS Li(B4)TS Na(B5)TS K (B6) TS Mg(B7)TS Ca(B8)Adsorption density [mg/g-FePO4]LiNaKMgCa- 95 - easier to proceed. Table 5.1 Molarity, molality and mean ionic activity of chloride salts and water activity in B1 solution at 25 °C, calculated by Meissner’s method71. Species Molarity [mol/L] Molality [mol/kg-H2O] Mean ionic activity [mol/kg-H2O] LiCl 0.100 0.100 0.432 NaCl 2.00 2.09 1.60 KCl 0.150 0.152 0.358 MgCl2 0.100 0.101 0.246 CaCl2 0.0500 0.0503 0.193 Water activity [-] 0.875 Figure 5.13 shows the selectivity of lithium over the other metals after thiosulfate reduction at 25 °C in B1–B3 brine solution. Selectivity over sodium was 428 in B2 brine solution while 18 in B3 brine solution because lithium was not adsorbed well in B3 solution and because the initial concentration of sodium was not so high in B3 solution. Selectivity over potassium, magnesium and calcium was lower than that over sodium because the initial concentration of those ions was not so high as that of sodium ions. - 96 - Figure 5.13 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 25 °C (B1–B3 brine solution). Figure 5.14 shows the weight percentage of FP in solid after thiosulfate reduction at 25 °C in B1–B3 brine solution when assuming each metal would make metal iron phosphate. FP was not all changed to LFP at 25 °C in 24 hours. Even in solution B2, the percentage was 27.8%. The amount of LFP produced in B3 solution, which contains lower total dissolved solids (TDS) than B1 or B2, was smaller. This indicates that high TDS may promote the reaction of lithium recovery in some ways. 050010001500200025003000TS Brine (B1) TS Brine w/o Ca (B2) TS Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 97 - Figure 5.14 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 25 °C (B1–B3 brine solution). Figure 5.15 shows the adsorption density of each metal on FP after thiosulfate reduction at 45 °C in B1–B8 brine solution. Adsorption density of lithium at 45 °C was more than three times as high as that at 25 °C with the value of 44.5 mg/g in the case of B2 solution. The adsorption density of the other metals also got a little bit larger. 0% 20% 40% 60% 80% 100%TS Brine (B1)TS Brine w/o Ca(B2)TS Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 98 - Figure 5.15 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 45 °C (B1–B8 brine solution). Figure 5.16 shows the selectivity of lithium over the other metals after thiosulfate reduction at 45 °C in B1–B3 brine solution. Selectivity over sodium was 1430 in B2, which is also more than three times as high as that at 25 °C. 01020304050607080TS Brine(B1)TS Brinew/o Ca(B2)TS Brinew/o Na Ca(B3)TS Li(B4)TS Na(B5)TS K (B6) TS Mg(B7)TS Ca(B8)Adsorption density [mg/g-FePO4]LiNaKMgCa- 99 - Figure 5.16 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 45 °C (B1–B3 brine solution). Figure 5.17 shows the weight percentage of FP in solid after thiosulfate reduction at 45 °C in B1–B3 brine solution. The weight percentage was 90.5% in the case of B2 solution. Although FP did not remain in the case of B1 and B2 solution, less than 10% of the products were impurities. In B3 solution, FP was still left even at 45 °C. 050010001500200025003000TS Brine (B1) TS Brine w/o Ca (B2) TS Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 100 - Figure 5.17 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 45 °C (B1–B3 brine solution). Figure 5.18 shows the adsorption density of each metal on FP after thiosulfate reduction at 65 °C in B1–B8 brine solution. Adsorption density of lithium at 65 °C was 44.6 mg/g in the case of B2 solution. This value is almost the same as the value at 45 °C because it is so close to a theoretical value of 46.0 mg/g. Additionally, as it is indicated in the last section that sodium iron phosphate was produced at 65 °C, the adsorption density of sodium was much higher than that at 25 °C and 45 °C, especially in B5 solution. 0% 20% 40% 60% 80% 100%TS Brine (B1)TS Brine w/o Ca(B2)TS Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 101 - Figure 5.18 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by thiosulfate (TS) at 65 °C (B1–B8 brine solution). Figure 5.19 shows the selectivity of lithium over the other metals after thiosulfate reduction at 65 °C in B1–B3 brine solution. Selectivity over sodium was 570 in B2, which is much lower than that at 45 °C because of SFP formation. In the case of B3 solution, there was no sodium adsorbed and so a small amount of potassium adsorbed. That is why the selectivity over sodium and potassium was respectively N/A and 1159. 01020304050607080TS Brine(B1)TS Brinew/o Ca(B2)TS Brinew/o Na Ca(B3)TS Li(B4)TS Na(B5)TS K (B6) TS Mg(B7)TS Ca(B8)Adsorption density [mg/g-FePO4]LiNaKMgCa- 102 - Figure 5.19 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by thiosulfate (TS) at 65 °C (B1–B3 brine solution). Figure 5.20 shows the weight percentage of various forms of FP in solid after thiosulfate reduction at 65 °C in B1–B3 brine solution. The weight percentage was 89.6% LFP in the case of B2 solution, which is almost the same value as that at 45 °C. No FP was remaining in every case. Although there were impurities more than 10% of the products in B1 and B2 cases, impurities in B3 solution were only 3.3%. That is because there was no calcium ion and few sodium ions from sodium thiosulfate in B3 solution. 050010001500200025003000TS Brine (B1) TS Brine w/o Ca (B2) TS Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 103 - Figure 5.20 Weight percentage of iron phosphate in solid after loading experiments by thiosulfate (TS) at 65 °C (B1–B3 brine solution). Figure 5.21 is a graph of adsorption density vs. temperature in the case of thiosulfate reduction in B1–B4 brine solution. According to the graph, adsorption density is higher at a higher temperature. Though the experimental values are almost the same as the theoretical value at 65 °C, they are lower at 25 °C and 45 °C. 0% 20% 40% 60% 80% 100%TS Brine (B1)TS Brine w/o Ca(B2)TS Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 104 - Figure 5.21 Adsorption density vs. temperature in the case of thiosulfate (TS) reduction (B1–B4 brine solution). The theoretical maximum of the adsorption density is 46.0 mg/g. Figure 5.22 is a graph of weight percentage of lithium iron phosphate in solid vs. temperature in the case of thiosulfate reduction in B1–B4 brine solution. The weight percentage is higher at a higher temperature though the experimental values are around 90% even at 65 °C because of the presence of impurities like sodium and calcium. 0510152025303540455020 30 40 50 60 70Adsorption density [mg/g-FePO4]Temperature [°C]TS Brine (B1) TS Brine w/o Ca (B2) TS Brine w/o Na Ca (B3) TS Li (B4)Theoretical- 105 - Figure 5.22 Weight percentage of lithium iron phosphate in solid vs. temperature in the case of thiosulfate (TS) reduction (B1–B4 brine solution). 5.4.2. Sulfite Reduction Figure 5.23 shows the adsorption density of each metal on FP after sulfite reduction at 25 °C in B2–B7 brine solution. In the case of B2 brine solution, the adsorption density of lithium was 28.6 mg/g. This value is more than twice as that of thiosulfate case at 25 °C. In contrast to the case of thiosulfate, lithium adsorption density is not lower in low TDS solution like B3 and B4. This fact suggests that thiosulfate is not an effective reductant for lithium loading under low TDS condition. 02040608010020 30 40 50 60 70LiFePO4produced [wt%]Temperature [°C]TS Brine (B1) TS Brine w/o Ca (B2) TS Brine w/o Na Ca (B3) TS Li (B4)- 106 - Figure 5.23 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 25 °C (B2–B7 brine solution). Figure 5.24 shows the selectivity of lithium over the other metals after sulfite reduction at 25 °C in B2 and B3 brine solution. The selectivity of lithium over sodium was 1503 in the B2 brine solution, which is more than three times that achieved with thiosulfate as the reductant. This means that sulfite is better than thiosulfate at 25 °C when calcium doesn’t exist. 01020304050607080SF Brine w/oCa (B2)SF Brine w/oNa Ca (B3)SF Li (B4) SF Na (B5) SF K (B6) SF Mg (B7)Adsorption density [mg/g-FePO4]LiNaKMgCa- 107 - Figure 5.24 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 25 °C (B2 and B3 brine solution). Figure 5.25 shows the weight percentage of FP in solid after sulfite reduction at 25 °C in the B2 and B3 brine solution. It is clear that FP was not all changed to LFP at 25 °C, but the value is 62.6% in the B2 solution case and more extensive than that of the thiosulfate case. 050010001500200025003000SF Brine w/o Ca (B2) SF Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 108 - Figure 5.25 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 25 °C (B2 and B3 brine solution). Figure 5.26 shows the adsorption density of each metal on FP after sulfite reduction at 45 °C in B2–B7 brine solution. In B2 case, adsorption density of lithium at 45 °C was 45.2 mg/g, which is so close to the theoretical value. Again, in contrast to the case of thiosulfate, lithium adsorption density is not lower in B3 and B4 case than in B2 case, which indicates a difference between thiosulfate and sulfite. 0% 20% 40% 60% 80% 100%SF Brine w/o Ca(B2)SF Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 109 - Figure 5.26 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 45 °C (B2–B7 brine solution). Figure 5.27 shows the selectivity of lithium over the other metals after sulfite reduction at 45 °C in B2 and B3 brine solution. Selectivity over sodium was 2541 in B2, which is close to twice as that at 25 °C and that of the thiosulfate case at 45 °C. 01020304050607080SF Brine w/oCa (B2)SF Brine w/oNa Ca (B3)SF Li (B4) SF Na (B5) SF K (B6) SF Mg (B7)Adsorption density [mg/g-FePO4]LiNaKMgCa- 110 - Figure 5.27 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 45 °C (B2 and B3 brine solution). Figure 5.28 shows the weight percentage of FP in solid after sulfite reduction at 45 °C in B2 and B3 brine solution. The weight percentage was 93.3% in the case of B2 solution. Although FP didn’t remain in the case of B1 and B2 solution, less than 10% of the products were impurities. 050010001500200025003000SF Brine w/o Ca (B2) SF Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 111 - Figure 5.28 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 45 °C (B2 and B3 brine solution). Figure 5.29 shows the adsorption density of each metal on FP after sulfite reduction at 65 °C in B2–B7 brine solution. Adsorption density of lithium at 65 °C was 45.9 mg/g in the case of B2 solution. Due to SFP formation, the adsorption density of sodium was much higher at 65 °C in B5 solution. 0% 20% 40% 60% 80% 100%SF Brine w/o Ca(B2)SF Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 112 - Figure 5.29 Adsorption density of Li, Na, K, Mg, Ca on ferric phosphate after loading experiments by sulfite (SF) at 65 °C (B2–B7 brine solution). Figure 5.30 shows the selectivity of lithium over the other metals after sulfite reduction at 65 °C in B2 and B3 brine solution. Selectivity over sodium is 1056 in B2, which is lower than that at 45 °C because of SFP formation. On the other hand, selectivity over potassium is higher than that at 45 °C. This indicates that SFP formation may be disturbed adsorption of potassium. 01020304050607080SF Brine w/oCa (B2)SF Brine w/oNa Ca (B3)SF Li (B4) SF Na (B5) SF K (B6) SF Mg (B7)Adsorption density [mg/g-FePO4]LiNaKMgCa- 113 - Figure 5.30 Selectivity of lithium over Na, K, Mg, Ca after loading experiments by sulfite (SF) at 65 °C (B2 and B3 brine solution). Figure 5.31 shows the weight percentage of FP in solid after sulfite reduction at 65 °C in B2 and B3 brine solution. The weight percentage was 92.6% in the case of B2 solution. Impurities were less than 10% in both cases, and both products have a higher purity than the products after thiosulfate reduction. 050010001500200025003000SF Brine w/o Ca (B2) SF Brine w/o Na Ca (B3)Lithium selectivity [-]Li/NaLi/KLi/MgLi/Ca- 114 - Figure 5.31 Weight percentage of iron phosphate in solid after loading experiments by sulfite (SF) at 65 °C (B2 and B3 brine solution). Figure 5.32 is a graph of adsorption density vs. temperature in the case of sulfite reduction in B2–B4 brine solution. According to the graph, adsorption density is higher at a higher temperature. The experimental values reached the theoretical value at 45 °C, which means sulfite is better than thiosulfate in the absence of calcium. 0% 20% 40% 60% 80% 100%SF Brine w/o Ca(B2)SF Brine w/o Na Ca(B3)Mineral contained [wt%]LiFePO4 NaFePO4 KFePO4 Mg0.5FePO4 Ca0.5FePO4 FePO4- 115 - Figure 5.32 Adsorption density vs. temperature in the case of sulfite (SF) reduction (B2–B4 brine solution). The theoretical maximum of the adsorption density is 46.0 mg/g. Figure 5.33 is a graph of weight percentage of iron phosphate in solid vs. temperature in the case of sulfite reduction in B2–B4 brine solution. The weight percentage is higher at a higher temperature. Though the experimental values are larger than 90% even at 45 °C and of course at 65 °C, which again means sulfite can reduce FP to LFP better compared to thiosulfate in the absence of calcium. 0510152025303540455020 30 40 50 60 70Adsorption density [mg/g-FePO4]Temperature [°C]SF Brine w/o Ca (B2) SF Brine w/o Na Ca (B3) SF Li (B4)Theoretical- 116 - Figure 5.33 Weight percentage of lithium iron phosphate in solid vs. temperature in the case of sulfite (SF) reduction (B2–B4 brine solution). 5.5. Effect of pH This section will investigate whether pH influences the ferric phosphate method. The relation between pH change and iron dissolution is firstly discussed, and then the prevention of iron dissolution by controlling pH and the pH effect on lithium capacity of ferric phosphate are discussed. This section focuses on only the case of 65 °C because pH change and iron dissolution were observed at 65 °C and it was confirmed that almost no iron was extracted at 25 °C and 45 °C by either method of reduction. Figure 5.34 shows iron dissolution from FP and the final pH of a solution in the case of thiosulfate reduction at 65 °C. 02040608010020 30 40 50 60 70LiFePO4produced [wt%]Temperature [°C]SF Brine w/o Ca (B2) SF Brine w/o Na Ca (B3) SF Li (B4)- 117 - Figure 5.34 Iron (Fe) dissolution and final pH in the case of thiosulfate (TS) reduction (B1–B8 brine solution). According to the graph, almost 40% of iron extracted from FP and the pH went down to 4–5 in B1–B4 solution while sodium thiosulfate solution itself is neutral at around 7. 4 2 2 3 4 2 4 62FePO +2LiCl 2Na S O 2LiFePO Na S O 2NaCl+ → + + (5.1) However, as shown in Equation (5.1), there is no acid produced to lower pH and no iron dissolution by the thiosulfate reduction. Ferric phosphate would not dissolve even at 65 °C and at pH 4. Therefore, it is appropriate to guess another reaction which produces proton and dissolves ferric phosphate. The reaction should be involved in thiosulfate because this phenomenon occurred in the thiosulfate case only. Based on these considerations, it is assumed that ferric phosphate is reduced to ferrous sulfate by tetrathionate with the following reaction. 2 2 3 24 4 6 2 4 414FePO S O 10H O 14Fe 14PO 4SO 20H− + − − ++ + → + + + (5.2) 02468101214020406080100TS Brine(B1)TS Brinew/o Ca(B2)TS Brinew/o Na Ca(B3)TS Li (B4) TS Na(B5)TS K (B6) TS Mg(B7)TS Ca(B8)pHFe dissolution [%]Fe dissolutionFinal pH- 118 - At the same time, there is a possibility that oxygen oxidizes tetrathionate. 2 24 6 2 2 42S O 7O 6H O 8SO 12H− − ++ + → + (5.3) Both oxygen and ferric can be an oxidant, and it depends on a condition how much oxygen or ferric or both contribute to the oxidation. For example, Druschel et al.72 suggested, based on their experimental results, that the overall reaction of tetrathionate oxidation at low pH was as follows. 2 3 2 24 6 2 2 4S O 3Fe 2.75O 4.5H O 4SO 3Fe 9H− + − + ++ + + → + + (5.4) On the other hand, when ferric phosphate dissolves into solution as ferrous sulfate, phosphate ions are neutralized by water and release hydroxide ions, which increase pH. This reaction contradicts the pH decrease observed while it may be because the amount of protons released from the reaction above is much more than that of hydroxides. 3 24 2 4PO H O HPO OH− − −+ → + (5.5) For these reasons, the reaction system of the FP method is so complicated. That’s why, to investigate the reactions above more in detail, it is necessary to analyze phosphate ions by ICP and polythionate species by ion chromatography (IC). Figure 5.35 shows iron dissolution from FP and the final pH of a solution in the case of sulfite reduction at 65 °C. The graph shows that the pH decreased to 4–5 in the sulfite case while sodium sulfite solution itself is a little bit alkaline at around 9. There is much less iron dissolution than observed in the thiosulfate experiments. 4 2 3 2 4 2 42FePO +2LiCl Na SO H O 2LiFePO Na SO 2HCl+ + → + + (5.6) The decrease in pH can be explained by Equation (5.6). The reduction by sulfite produces proton as a byproduct, and the small amount of iron dissolution is attributed to the pH decline. - 119 - Figure 5.35 Iron (Fe) dissolution and final pH in the case of sulfite (SF) reduction (B1–B8 brine solution). In order to prevent iron dissolution in the thiosulfate case, another experiment was conducted with B1 solution by thiosulfate reduction at controlled pH 7. Figure 5.36 compares lithium adsorption density and iron dissolution at pH 4 with that at pH 7. Setting pH at 7 was effective to keep iron phosphate from dissolving. At the same time, controlling pH didn’t have any effects on the adsorption density of lithium. 02468101214020406080100TS Brine(B1)TS Brinew/o Ca(B2)TS Brinew/o Na Ca(B3)TS Li (B4) TS Na(B5)TS K (B6) TS Mg(B7)TS Ca(B8)pHFe dissolution [%]Fe dissolutionFinal pH- 120 - Figure 5.36 Adsorption density and iron (Fe) dissolution at pH 4 (uncontrolled) and 7 (controlled with sodium hydroxide) in the case of thiosulfate (TS) reduction (B1 brine solution). When it comes to delithiation of LFP, the oxidation reaction in Equation (5.7) occurs. It is readily found out that the reaction does not dissolve ferric phosphate. 4 2 2 8 4 2 4 2 42LiFePO K S O 2FePO +Li SO K SO+ → + (5.7) While potassium persulfate solution itself is acidic at around 3.9, it is confirmed by ICP analysis that almost no iron eluted. 5.6. Recycling of Ferric Phosphate The previous sections discussed details of the ferric phosphate recovery method by using FP obtained from new battery-grade LFP. However, FP and LFP should be recycled to use this 010203040506070809010001020304050pH 4 pH 7Fe dissolution [%]Adsorption density [mg-Li/g-FP]Adsorption density Fe dissolution- 121 - method more efficiently in practical situations as shown in Figure 5.37. Figure 5.37 Schematic flowsheet of the ferric phosphate method. Flow of FP and LFP solid is in orange and flow of brine solution is in blue. FP and LFP are recycled by a continuous loading and stripping cycle. Figure 5.38 shows XRD patterns of solid after the loading and stripping in the first and second cycle by thiosulfate reduction at 65 °C and at pH 7 in the B1 brine solution. In both the first and the second cycle, well crystallized LFP and FP were obtained after the loading and stripping respectively. New BG-LFPStripping(Oxidation by PS)Loading(Reduction by TS or SF)FPLFPLi pre-concentratedsolutionBrineStripped brine Further extractionorDisposalConcentrationRefiningPrecipitationSalt lakeSeawaterGeothermal waterOil & gas produced waterLi2CO3, LiOH- 122 - Figure 5.38 XRD patterns of battery-grade (BG-) LFP and BG-FP (delithiated LFP by PS oxidation), and products after two cycles of loading and stripping experiments. The loading experiments used B1 brine solution at pH 7 controlled by sodium hydroxide and lithium was loaded by thiosulfate (TS) at 65 °C. The stripping experiments used persulfate to oxidize the loaded materials. Figure 5.39 shows metal concentration in solution after stripping in the first and second cycle by persulfate oxidation with a solid concentration of 10 wt%. Lithium concentration in 10 20 30 40 50Relative intensity [-]2θ [deg] (CuKα)FP-BGLFPBG1st loading1st stripping2nd loading2nd stripping- 123 - solution after stripping was seven times as the concentration in the initial brine in both the first and the second cycle. Sodium was almost completely rejected. Figure 5.39 Concentration of metal ions in the initial brine solution (B1 brine solution) and solution after stripping in the first and second cycle of the loading and stripping experiments. The loading experiments used B1 brine solution at pH 7 controlled by sodium hydroxide and lithium was loaded by thiosulfate (TS) at 65 °C. The stripping experiments used persulfate to oxidize the loaded materials and it was carried out with a solid concentration of 10 wt%. In addition to the metal concentration of the stripped ones in the graph, potassium from K2S2O8 was also in the solution. Potassium concentration is a little bit higher than the concentration in the initial brine because of potassium persulfate. However, as Table 2.10 shows, the solubility of potassium carbonate is almost 100 times as that of lithium carbonate and they can be separated by carbonate 05000100001500020000250003000035000400004500050000Initial brine Stripped (1st cycle) Stripped (2nd cycle)Concentration in solution [mg/L]LiNaKMgCa05001000150020002500300035004000Initial brine Stripped (1st cycle) Stripped (2nd cycle)- 124 - precipitation. Alternatively, further concentration and refining would help them separate. Therefore, the ferric phosphate method can recycle the adsorbent (FP) without degradation. 5.7. Kinetics Finally, this section discusses the kinetics of the reactions by both thiosulfate and sulfite. Two models are utilized here. One is the pseudo first-order model and the other is shrinking sphere model. While, as mentioned, the reactions in the system of the FP method are so complicated, kinetic models are simplified here. Only one parameter α, reacted fraction of FP, is included in the kinetic models, and temperature is considered by Arrhenius formula. However, the other factors, such as the concentration of lithium, thiosulfate, sulfite and particle size distribution of FP powder and pH, are ignored for simplification. Furthermore, while the two models assume that the reactions of the FP method are normal chemical reactions, they are also electrochemical redox reactions. That’s why it is also possible to consider kinetics by the Butler-Volmer equation and the Tafel equations if the reactions are electron transfer controlled. Here, this study hands over the electrochemical consideration of kinetics to future works. 5.7.1. Kinetic Models To begin with, the theory of the two models are reviewed in the following.72 Pseudo first-order model is described as Equation (5.8). ktdkdt1 e−a= − aa = − (5.8) where α is fraction reacted, k is reaction rate constant, and t is time. By taking the natural logarithm, the equation changes to Equation (5.9) and k is obtained as a slope of plots of ln(1-α) vs. time. - 125 - ln(1 ) kt−a = − (5.9) Furthermore, k is described by Arrhenius formula as shown in Equation (5.10). aERTak AeEln k ln ART−== − + (5.10) where A: frequency factor, Ea: activation energy, R: gas constant, T: absolute temperature. By taking the natural logarithm again, Ea is obtained as a slope of plots of ln k vs. -1000/RT. The shrinking sphere model is described as Equation (5.11). 22 22 23 33 323dVkS k(4 r )dtdV 3V 3k(4 ) k(4 ) V k 'Vdt 4 4dVk 'Vdt− = =    − =  =  =       − = (5.11) where V: volume of a particle, t: time, k: reaction rate constant, S: surface area of the particle, r: radius of the particle, k’=k(4π)(3/4π)2/3. In the calculation of experimental results, initial radius ro was set at 1.5 μm, which is D50 of the battery-grade LFP. The following equations are derived by integrating the reaction above. ( )1 13 3o131o 3o13k ' tV V3V k '1 tV3V1 1 k '' t− = − =  − −a = (5.12) where Vo: initial volume of the particle, α: fraction reacted (=1-V/Vo), k’’=k’/(3Vo1/3). From the last equation, k’’ is obtained as a slope of plots of 1-(1-α)1/3 vs. time. By calculating back to Equation (5.13), k can be obtained. For k in shrinking sphere model, the Arrhenius formula can - 126 - be applied as well. aERTak AeEln k ln ART−== − + (5.13) Lastly, it is known that kinetics is diffusion controlled when Ea is less than 20 kJ/mol and chemical reaction controlled when more than 40 kJ/mol. 5.7.2. Thiosulfate Reduction Figure 5.40 shows x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate reduction in B2 brine. At 65 °C, the reaction was completed in 2 hours or so. Figure 5.40 x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). 00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp.45°C_Exp.65°C_Exp.- 127 - Figure 5.41 shows graphs for the calculation of Pseudo first-order model. Rate constant k and activation energy Ea are obtained as a slope of those graphs. While the correlation coefficient is low at 25 °C, the other three graphs have good correlation. Figure 5.41 Calculation of Pseudo first-order model: “ln (1-α) vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of thiosulfate (TS) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. 25°C 45°C65°Cln k vs. -1000/RTy = -0.0111xR² = 0.7885-0.4-0.35-0.3-0.25-0.2-0.15-0.1-0.0500 10 20 30ln (1-α)t [hour]y = -0.0932xR² = 0.9747-1.8-1.6-1.4-1.2-1-0.8-0.6-0.4-0.200 5 10 15 20ln (1-α)t [hour]y = -1.4209xR² = 0.9748-3.5-3-2.5-2-1.5-1-0.500 0.5 1 1.5 2 2.5ln (1-α)t [hour]y = 99.285x + 35.53R² = 0.9807-5-4-3-2-101-0.41 -0.4 -0.39 -0.38 -0.37 -0.36 -0.35ln k-1000/RT- 128 - Figure 5.42 shows graphs for the calculation of the shrinking sphere model. Again, the graph at 25 °C has a worse correlation but the others have a better one. Figure 5.42 Calculation of shrinking sphere model: “1-(1-α)1/3 vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of thiosulfate (TS) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. 25°C 45°C65°C ln k vs. -1000/RTy = 0.0035xR² = 0.801200.020.040.060.080.10.120 10 20 301-(1-α)1/3t [hour]y = 0.0246xR² = 0.960700.050.10.150.20.250.30.350.40.450.50 5 10 15 201-(1-α)1/3t [hour]y = 0.3147xR² = 0.999900.10.20.30.40.50.60.70 0.5 1 1.5 2 2.51-(1-α)1/3t [hour]y = 90.818x + 31.375R² = 0.9789-6-5-4-3-2-10-0.41 -0.4 -0.39 -0.38 -0.37 -0.36 -0.35ln k-1000/RT- 129 - Table 5.2 is a summary of k values and activation energy of the thiosulfate reduction calculated from the two models. The rate constant of each model gets larger along with temperature. As activation energy is more than 40 kJ/mol in both cases, it is found that the lithium extraction reaction by thiosulfate reduction is chemical reaction controlled. Table 5.2 k value and activation energy calculated from kinetics in the case of thiosulfate reduction. Values Temperature Pseudo first-order model [hour-1] Shrinking sphere model [μm hour-1] k 25 °C 0.01 0.01 45 °C 0.09 0.04 65 °C 1.50 0.47 Activation energy [kJ/mol] 99.3 90.8 Using the calculated k values, the two models are compared with the experimental results in Figure 5.43 and Figure 5.44, respectively. While the models ignored some complicated reactions and particle size distribution and so on to simplify the discussion, they fit the experimental data well. - 130 - Figure 5.43 Comparison of experimental data and calculated values by Pseudo first-order model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). Figure 5.44 Comparison of experimental data and calculated values by shrinking sphere model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of thiosulfate (TS) reduction (B2 brine solution). 00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp. 45°C_Exp. 65°C_Exp.25°C_Calc. 45°C_Calc. 65°C_Calc.00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp. 45°C_Exp. 65°C_Exp.25°C_Calc. 45°C_Calc. 65°C_Calc.- 131 - 5.7.3. Sulfite Reduction Figure 5.45 shows x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite reduction in B2 brine. The reaction was completed in 12 hours and 2 hours at 45 °C and 65 °C respectively. Figure 5.45 x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution). Figure 5.46 shows graphs for the calculation of Pseudo first-order model. 00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp.45°C_Exp.65°C_Exp.- 132 - Figure 5.46 Calculation of Pseudo first-order model: “ln (1-α) vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of sulfite (SF) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. Figure 5.47 shows graphs for the calculation of the shrinking sphere model. At 45 °C, the correlation coefficient of the graph is a little bit smaller than the others. 25°C 45°C65°C ln k vs. -1000/RTy = -0.037xR² = 0.9791-1-0.9-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.100 10 20 30ln (1-α)t [hour]y = -2.4801xR² = 0.9881-6-5-4-3-2-100 0.5 1 1.5 2 2.5ln (1-α)t [hour]y = -0.2598xR² = 0.9862-5-4-3-2-100 5 10 15 20ln (1-α)t [hour]y = 88.32x + 32.243R² = 0.99-4-3-2-1012-0.41 -0.4 -0.39 -0.38 -0.37 -0.36 -0.35ln k-1000/RT- 133 - Figure 5.47 Calculation of shrinking sphere model: “1-(1-α)1/3 vs. time at 25 °C, 45 °C and 65 °C” and “ln k vs. -1000/RT” in the case of sulfite (SF) reduction (B2 brine solution). α: fraction of reacted, k: reaction rate constant, R: gas constant, T: absolute temperature. Table 5.3 is a summary of k values and activation energy of the sulfite reduction calculated from the two models. Like the thiosulfate case, the rate constant is more significant at a higher temperature. The extraction reaction of lithium by sulfite reduction is also chemical 25°C 45°C65°C ln k vs. -1000/RTy = 0.0109xR² = 0.984300.050.10.150.20.250.30 10 20 301-(1-α)1/3t [hour]y = 0.4293xR² = 0.977700.10.20.30.40.50.60.70.80.910 0.5 1 1.5 2 2.51-(1-α)1/3t [hour]y = 0.1347xR² = 0.985400.050.10.150.20.250.30 0.5 1 1.5 2 2.51-(1-α)1/3t [hour]y = 76.69x + 27R² = 0.9703-5-4-3-2-10-0.41 -0.4 -0.39 -0.38 -0.37 -0.36 -0.35ln k-1000/RT- 134 - reaction controlled since activation energy is more than 40 kJ/mol. Table 5.3 k value and activation energy calculated from kinetics in the case of sulfite reduction. Values Temperature Pseudo first-order model [hour-1] Shrinking sphere model [μm hour-1] k 25 °C 0.04 0.02 45 °C 0.24 0.19 65 °C 2.57 0.61 Activation energy [kJ/mol] 88.3 76.7 Using the calculated k values, the two models are compared with the experimental results in Figure 5.48 and Figure 5.49, respectively. While the models ignored some complicated reactions and particle size distribution and so on to simplify the discussion, they fit the experimental data well. - 135 - Figure 5.48 Comparison of experimental data and calculated values by Pseudo first-order model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution). Figure 5.49 Comparison of experimental data and calculated values by shrinking sphere model. x in LixFePO4 vs. time at 25 °C, 45 °C and 65 °C in the case of sulfite (SF) reduction (B2 brine solution). 00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp. 45°C_Exp. 65°C_Exp.25°C_Calc. 45°C_Calc. 65°C_Calc.00.20.40.60.810 6 12 18 24x in Li xFePO4[-]t [hour]25°C_Exp. 45°C_Exp. 65°C_Exp.25°C_Calc. 45°C_Calc. 65°C_Calc.- 136 - 5.8. Summary The preparation of FP as an adsorbent was studied. It was challenging to prepare ferric phosphate from FPD and reagent-grade LFP, but persulfate oxidation of battery-grade LFP successfully made heterosite FP, which can be used for lithium extraction. Next, a variety of reducing agents was studied. Out of Cu+, TS and SF, TS and SF seemed appropriate for the lithium recovery reaction. The cuprous reduction experiments were not successful. With the adsorbent and reducing agents, all the experiments were carried out. By checking XRD patterns, it was found that FP was successfully reduced to LFP in most cases with a significant dependence on temperature. When sulfite is used for brine solution containing calcium ions, calcium sulfite salts were produced. At 65 °C, sodium iron phosphate was formed in the absence of lithium. Experimental results showed the lithium adsorption density could be the same as the theoretical value of 46.0 mg/g. For example, the value was 45.9 mg/g in the case of sulfite reduction at 65 °C. Lithium selectivity was also high. The selectivity over sodium, for instance, 2541 in the case of sulfite reduction at 45 °C. There were pH changes during experiments, but they didn’t have much impact on adsorption density and iron dissolution. However, 40% of iron was extracted when thiosulfate was used at 65 °C. The dissolution of iron can be prevented by pH control to ~ 7. When it comes to practical operation, FP and LFP should be recycled. The cycle of the loading and stripping test was repeated, and it was confirmed that FP and LFP could be recycled without degradation. In the last, kinetics was studied by using Pseudo first-order model and shrinking sphere model. These models were well fitted to experimental results and found that the lithium extraction reaction was chemical reaction controlled in both thiosulfate and sulfite cases. - 137 - Chapter 6. Conclusion Lithium is now getting more and more attention from around the world as an essential material for various energy storage systems, especially for electric vehicles, and demand for lithium has been increasing rapidly. To meet the demand, lithium has been mined from ore minerals and brines. Since it is known that brines have more lithium reserves than ore minerals in the world, many researchers are now studying how to recovery lithium effectively from brine. The difficulty is that brine contains a high concentration of sodium, potassium, magnesium and calcium, which limit selective extraction of lithium. Conventionally, solar evaporation has been used to reject those cations into a solid phase and concentrate lithium in brine. However, as the evaporation process takes more than one year and is highly dependent on climatological conditions, a more rapid process is strongly desired for the accelerating demand. Therefore, many studies have investigated new rapid lithium recovery methods using ion exchange resin, solvent extraction, inorganic absorbents such as manganese oxide, titanium oxide and ferric phosphate, phosphate precipitation, nanofiltration, membrane electrolysis and so forth. This thesis focuses on two types of adsorbents: ion exchange (IX) resin and heterosite-type ferric phosphate (FP). First, IX resin was studied. More than twenty cation exchange resins were tested in both lithium chloride solution without any other metal cations and a mixed saline solution containing lithium, sodium, potassium, magnesium and calcium chloride. Sulfonate, iminodiacetate and aminomethylphosphonate resins successfully extracted lithium from the lithium chloride solution. The resins recovered 16.3–32.9 mg of lithium per one gram of dried resins. However, no resins could adsorb lithium strongly or selectively from the saline solution with less than 1 mg of lithium per one gram of dried resins in loading. This behavior was due to the lack of selectivity for lithium - 138 - extraction. Sodium, potassium, magnesium and calcium ions compete with the lithium adsorption. Therefore, aluminum loaded resin, which some past studies reported that had relatively good selectivity for lithium, was also tested. This resin extracts lithium on the surface in the form of LiCl·2Al(OH)3·nH2O. It showed higher capacity and selectivity than any other resins tested in this study in the saline solution with the value of 6.6 mg/g. Nevertheless, overall, it was found that IX resins poorly suited for lithium recovery. A second study focused on FP as an adsorbent. FP can adsorb lithium ion from solution by a reduction reaction and turns to lithium iron phosphate (LFP). As a second step, an oxidizing agent can strip lithium and regenerate FP. In this project, sodium thiosulfate (TS) or sodium sulfite (SF) was used as a reducing agent. By analyzing X-ray diffraction patterns of the products after the reduction, it was found that FP was successfully reduced to LFP in most cases with faster reaction at a higher temperature. However, when SF was used as a reducing agent for lithium loading from a brine solution containing calcium ions, calcium sulfite salts were produced. In addition, at 65 °C, sodium iron phosphate was formed in addition to LFP. The maximum of lithium adsorption capacity was almost the same value as the theoretical value of 46.0 mg-Li/g-FP. For example, the value was 45.9 mg/g in the case of SF reduction at 65 °C. Lithium selectivity was also high. The selectivity over sodium, for instance, 2541 in the case of SF reduction at 45 °C. When TS and SF are compared as a reducing agent, TS is better when a brine solution contains calcium because SF produces calcium sulfite. On the other hand, SF needs an only half dosage of TS since SF reacts with FP with the ratio of one to two while TS does with the ratio of one to one. There was a pH decrease from around 7 to 4 during experiments, but pH did not have much impact on adsorption capacity. However, 40% of iron was dissolved when TS was used for reduction at 65 °C. This problem of iron dissolution can be prevented by pH control in the neutral region of –7. When it comes to practical operation, FP should be recycled. The cycle of the loading and - 139 - stripping test was repeated, and it was confirmed that FP could be recycled without degradation. In the last part of the study, the kinetics of lithium extraction was also studied and fit using both Pseudo first-order model and shrinking sphere model. These models fit the experimental results and indicated that the lithium extraction reaction was chemical reaction controlled in both the TS and SF cases. In conclusion, as this study found out that the ferric phosphate method is a promising technology, it is recommended that this technology be developed further by using natural brine sources. - 140 - References (1) Momma, K.; Izumi, F. VESTA 3 for Three-Dimensional Visualization of Crystal, Volumetric and Morphology Data. J. Appl. Crystallogr. 2011, 44 (6), 1272–1276. https://doi.org/10.1107/S0021889811038970. (2) Chagnes, A.; Swiatowska, J. Lithium Process Chemistry: Resources, Extraction, Batteries, and Recycling; Elsevier: Amsterdam, Netherlands, 2015. https://doi.org/10.1016/B978-0-12-801417-2.00007-4. (3) United States Geological Survey. Mineral Commodity Summaries 2018; 2018. (4) Vikström, H.; Davidsson, S.; Höök, M. Lithium Availability and Future Production Outlooks. Appl. Energy 2013. https://doi.org/10.1016/j.apenergy.2013.04.005. (5) Lee, J. M.; Bauman, W. C. Recovery of Lithium from Brines. Patent US4159311A, 1979. (6) Lee, J. M.; Bauman, W. C. Recovery of Lithium from Brines. Patent US4347327A, 1982. (7) John L. Burba, I. Method of Making Crystalline 2-Layer Lithium Aluminates in Ion Exchange Resins. Patent US4461714A, 1984. (8) Nishihama, S.; Onishi, K.; Yoshizuka, K. Selective Recovery Process of Lithium from Seawater Using Integrated Ion Exchange Methods. Solvent Extr. Ion Exch. 2011, 29 (3), 421–431. https://doi.org/10.1080/07366299.2011.573435. (9) Tenova. Tenova Lithium Recover Process. (10) Tenova. Lithium Processing https://www.tenova.com/product/lithium-processing/ (accessed Nov 29, 2018). (11) Bauman, W. C.; John L. Burba, I. Recovery of Lithium Values from Brines. Patent US5389349A, 1995. - 141 - (12) Bauman, W. C.; John L. Burba, I. Recovery of Lithium Values from Brines. Patent US5599516A, 1997. (13) Bauman, W. C.; John L. Burba, I. Composition for the Recovery of Lithium Values from Brine and Process of Making/Using Said Composition. Patent US6280693B1, 2001. (14) Harrison, S.; Sharma, C. V. K.; Viani, B. E.; Peykova, D. Lithium Extraction Composition and Method of Preparation Thereof. Patent US8637428B1, 2014. (15) John L. Burba, I.; Stewart, R. F.; Viani, B. E.; Harrison, S.; Vogdes, C. E.; Lahlouh, J. G. S. Sorbent for Lithium Extraction. Patent US8753594B1, 2014. (16) Boualleg, M.; Lafon, O.; Burdet, F. A. P.; Soulairol, R. C. J. R. Method of Preparing an Adsorbent Material Shaped in the Absence of Binder and Method of Extracting Lithium from Saline Solutions Using Said Material. Patent US20160317998A1, 2016. (17) Ooi, K.; Miyai, Y.; Katoh, S. Recovery of Lithium from Seawater by Manganese Oxide Adsorbent. Sep. Sci. Technol. 1986. https://doi.org/10.1080/01496398608056148. (18) Miyai, Y.; Ooi, K.; Kato, S. Adsorbent for Lithium and a Method for the Preparation Thereof. Patent US4665049A, 1987. (19) Chitrakar, R.; Kanoh, H.; Miyai, Y.; Ooi, K. Recovery of Lithium from Seawater Using Manganese Oxide Adsorbent (H1.6Mn1.6O4) Derived from Li1.6Mn1.6O4. Ind. Eng. Chem. Res. 2001, 40 (9), 2054–2058. https://doi.org/10.1021/ie000911h. (20) Chitrakar, R.; Makita, Y.; Ooi, K.; Sonoda, A. Magnesium-Doped Manganese Oxide with Lithium Ion-Sieve Property: Lithium Adsorption from Salt Lake Brine. Bull. Chem. Soc. Jpn. 2013, 86 (7), 850–855. https://doi.org/10.1246/bcsj.20130019. (21) Chitrakar, R.; Makita, Y.; Ooi, K.; Sonoda, A. Lithium Recovery from Salt Lake Brine by H2TiO3. Dalt. Trans. 2014. https://doi.org/10.1039/c4dt00467a. (22) Neometals. Successful ‘Proof of Concept’ testing of Direct Lithium and Potassium - 142 - Extraction from Brines https://www.neometals.com.au/reports/749-DX01062017.pdf (accessed Nov 29, 2018). (23) Intaranont, N.; Garcia-Araez, N.; Hector, A. L.; Milton, J. A.; Owen, J. R. Selective Lithium Extraction from Brines by Chemical Reaction with Battery Materials. J. Mater. Chem. A 2014. https://doi.org/10.1039/c4ta01101e. (24) Intaranont, N. Selective Lithium Extraction from Salt Solutions by Chemical Reaction with FePO4, PhD Thesis, University of Southampton, 2015. (25) Owen, J. R.; Garcia-Araez, N.; Intaranont, N. Sequestration of Lithium. WO2015121684A1, 2015. (26) POSCO. POSCO Begins Lithium Production First Time in Korea http://www.posco.com/homepage/docs/eng5/jsp/prcenter/news/s91c1010025v.jsp?idx=2701&onPage=1 (accessed Nov 29, 2018). (27) Jamasmie, C. POSCO to Buy Lithium Mining Rights in Argentina from Galaxy http://www.mining.com/posco-buy-lithium-mining-rights-argentina-galaxy/ (accessed Nov 29, 2018). (28) Somrani, A.; Hamzaoui, A. H.; Pontie, M. Study on Lithium Separation from Salt Lake Brines by Nanofiltration (NF) and Low Pressure Reverse Osmosis (LPRO). Desalination 2013, 317, 184–192. https://doi.org/10.1016/J.DESAL.2013.03.009. (29) Bogner, S. Successful Independent Verification of the MGX Lithium Extraction Technology https://www.rockstone-research.com/images/PDF/MGX20en.pdf (accessed Nov 29, 2018). (30) McEachern, P. Lithium Recovery from Oilfield Produced Water Brine & Wastewater Treatment https://rockstone-research.com/images/pdfs/MGX_TechnicalReportRapidLithiumExtractionProcess.pdf - 143 - (accessed Nov 29, 2018). (31) Liu, X.; Chen, X.; Zhao, Z.; Liang, X. Effect of Na+ on Li Extraction from Brine Using LiFePO4/FePO4 Electrodes. Hydrometallurgy 2014, 146, 24–28. https://doi.org/10.1016/J.HYDROMET.2014.03.010. (32) Liu, X.; Chen, X.; He, L.; Zhao, Z. Study on Extraction of Lithium from Salt Lake Brine by Membrane Electrolysis. Desalination 2015. https://doi.org/10.1016/j.desal.2015.08.013. (33) Romero, V. C. E.; Tagliazucchi, M.; Flexer, V.; Calvo, E. J. Sustainable Electrochemical Extraction of Lithium from Natural Brine for Renewable Energy Storage. J. Electrochem. Soc. 2019, 165 (10), A2294–A2302. (34) Díaz Nieto, C. H.; Palacios, N. A.; Verbeeck, K.; Prévoteau, A.; Rabaey, K.; Flexer, V. Membrane Electrolysis for the Removal of Mg2+ and Ca2+ from Lithium Rich Brines. Water Res. 2019, 154, 117–124. https://doi.org/10.1016/J.WATRES.2019.01.050. (35) Sawada, A. Current Situation of Lithium Resource Development in the World. Bull. Soc. Sea Water Sci. Japan 2012, 66, 2–7. (36) Manthiram, A.; Fu, Y.; Chung, S.-H.; Zu, C.; Su, Y.-S. Rechargeable Lithium–Sulfur Batteries. Chem. Rev. 2014. https://doi.org/10.1021/cr500062v. (37) Martin, G.; Rentsch, L.; Höck, M.; Bertau, M. Lithium Market Research – Global Supply, Future Demand and Price Development. Energy Storage Mater. 2017. https://doi.org/10.1016/j.ensm.2016.11.004. (38) Jamasmie, C. Lithium demand from battery makers to almost double by 2027 http://www.mining.com/lithium-demand-battery-makers-almost-double-2027/ (accessed Sep 12, 2019). (39) Kavanagh, L.; Keohane, J.; Garcia Cabellos, G.; Lloyd, A.; Cleary, J. Global Lithium - 144 - Sources—Industrial Use and Future in the Electric Vehicle Industry: A Review. Resources 2018. https://doi.org/10.3390/resources7030057. (40) JOGMEC. Metal Resources Report, Current Situation of Lithium Resources; 2010. (41) Canada lithium Corp. Objective Capital Rare Earth and Minor Metals Investment Summit: The challenges of developing a lithium project – reopening the Quebec Lithium Project - Peter Secker https://www.slideshare.net/objectivecapital/objective-capital-rare-earth-and-minor-metals-investment-summit-the-challenges-of-developing-a-lithium-project-reopening-the-quebec-lithium-project-peter-secker (accessed Dec 29, 2018). (42) JOGMEC. Introduction of Cases of Metal Recovery and Refining Technology Development from Seawater and Brines; 2016. (43) Rodinia Lithium. Exploring Lithium http://rodinialithium.com/ (accessed Sep 29, 2018). (44) Rumble, J. R.; Lide, D. R.; Bruno, T. J.; Martinsen, D. CRC Handbook of Chemistry and Physics, 99th Print.; CRC Press: Boca Raton, U.S., 2018. (45) Kim, K. J. Recovery of Lithium Hydroxide from Spent Lithium Carbonate Using Crystallizations. Sep. Sci. Technol. 2008. https://doi.org/10.1080/01496390701784088. (46) Kumar, A.; Fukuda, H.; Hatton, T. A.; Lienhard, J. H. Lithium Recovery from Oil and Gas Produced Water: A Need for a Growing Energy Industry. ACS Energy Lett. 2019, 1471–1474. https://doi.org/10.1021/acsenergylett.9b00779. (47) Chon, U.; Lee, I. C.; Kim, K. Y.; Han, G.-C.; Song, C. H.; Jung, S. R. Method for Manufacturing Lithium Hydroxide and Method Using Same for Manufacturing Lithium Carbonate. Patent US9598291B2, 2017. (48) Ooi, K.; Miyai, Y.; Katoh, S. Recovery of Lithium from Seawater by Manganese Oxide Adsorbent. Sep. Sci. Technol. 1986, 21 (8), 755–766. https://doi.org/10.1080/01496398608056148. - 145 - (49) Cohen, L.; McCallum, T.; Tinkler, O.; Szolga, W. Technological Advances, Challenges and Opportunities in Solvent Extraction from Energy Storage Applications. In Extraction 2018 Preceedings; 2018; pp 2033–2045. (50) Bjorling, C. O.; Westgren, A. Minerals of the Varustrask Pegmatite. IX. X-Ray Studies of Triphylite, Varulite and Their Oxidation Products. Geol. Föreningens i Stock. Förhandlingar 1938, 60, 67–72. (51) Losey, A.; Rakovan, J.; Hughes, J. M.; Francis, C. A.; Dyar, M. D. Structural Variation in the Lithiophilite-Triphylite Series and Other Olivine-Group Structures. Can. Mineral. 2004, 42 (4), 1105–1115. https://doi.org/10.2113/gscanmin.42.4.1105. (52) Okubo, M.; Asakura, D.; Mizuno, Y.; Kim, J.-D.; Mizokawa, T.; Kudo, T.; Honma, I. Switching Redox-Active Sites by Valence Tautomerism in Prussian Blue Analogues AxMny[Fe(CN)6]·nH2O (A: K, Rb): Robust Frameworks for Reversible Li Storage. J. Phys. Chem. Lett. 2010, 1 (14), 2063–2071. https://doi.org/10.1021/jz100708b. (53) Islam, M. S.; Daniel J. Driscoll; Fisher, C. A. J.; Slater, P. R. Atomic-Scale Investigation of Defects, Dopants, and Lithium Transport in the LiFePO4 Olivine-Type Battery Material. Chem. Mater. 2005, 17 (20), 5085–5092. https://doi.org/10.1021/CM050999V. (54) Ellis, B. L.; Lee, K. T.; Nazar, L. F. Positive Electrode Materials for Li-Ion and Li-Batteries. Chem. Mater. 2010, 22 (3), 691–714. https://doi.org/10.1021/cm902696j. (55) Morgan, D.; Van der Ven, A.; Ceder, G. Li Conductivity in LixMPO4 (M=Mn, Fe, Co, Ni) Olivine Materials. Electrochem. Solid-State Lett. 2004, 7 (2), A30–A32. https://doi.org/10.1149/1.1633511. (56) Allen, J. L.; Jow, T. R.; Wolfenstine, J. Kinetic Study of the Electrochemical FePO4 to LiFePO4 Phase Transition. Chem. Mater. 2007, 19 (8), 2108–2111. https://doi.org/10.1021/CM062963O. - 146 - (57) Wang, J.; Sun, X. Olivine LiFePO4: The Remaining Challenges for Future Energy Storage. Energy Environ. Sci. 2015, 8 (4), 1110–1138. https://doi.org/10.1039/C4EE04016C. (58) Wu, B.; Ren, Y.; Li, N. LiFePO4 Cathode Material. In Electric Vehicles - The Benefits and Barriers; Soylu, S., Ed.; IntechOpen: London, U.K., 2011; pp 199–216. https://doi.org/10.5772/18995. (59) Satyavani, T. V. S. L.; Srinivas Kumar, A.; Subba Rao, P. S. V. Methods of Synthesis and Performance Improvement of Lithium Iron Phosphate for High Rate Li-Ion Batteries: A Review. Engineering Science and Technology, an International Journal. 2016. https://doi.org/10.1016/j.jestch.2015.06.002. (60) Padhi, A.K.; Nanjundaswamy, K.S.; Masquelier, C.; Okada, S. Goodenough, J. B. Effect of Structure on the Fe3+∕Fe2+ Redox Couple in Iron Phosphates. J. Electrochem. Soc. 1997. https://doi.org/10.1149/1.1837649. (61) Chen, J.; Vacchio, M. J.; Wang, S.; Chernova, N.; Zavalij, P. Y.; Whittingham, M. S. The Hydrothermal Synthesis and Characterization of Olivines and Related Compounds for Electrochemical Applications. Solid State Ionics 2008. https://doi.org/10.1016/j.ssi.2007.10.015. (62) Ramana, C. V.; Mauger, A.; Gendron, F.; Julien, C. M.; Zaghib, K. Study of the Li-Insertion/Extraction Process in LiFePO4/FePO4. J. Power Sources 2009. https://doi.org/10.1016/j.jpowsour.2008.11.042. (63) Chitrakar, R.; Kanoh, H.; Miyai, Y.; Ooi, K. Recovery of Lithium from Seawater Using Manganese Oxide Adsorbent (H 1.6 Mn 1.6 O 4 ) Derived from Li 1.6 Mn 1.6 O 4. Ind. Eng. Chem. Res. 2001. https://doi.org/10.1021/ie000911h. (64) Chitrakar, R.; Makita, Y.; Ooi, K.; Sonoda, A. Selective Uptake of Lithium Ion from - 147 - Brine by H 1.33 Mn 1.67 O 4 and H 1.6 Mn 1.6 O 4 . Chem. Lett. 2012. https://doi.org/10.1246/cl.2012.1647. (65) Yu, Q.; Sasaki, K.; Hirajima, T. Bio-Templated Synthesis of Lithium Manganese Oxide Microtubes and Their Application in Li+ Recovery. J. Hazard. Mater. 2013. https://doi.org/10.1016/j.jhazmat.2013.08.027. (66) Zhang, Q. H.; Li, S. P.; Sun, S. Y.; Yin, X. S.; Yu, J. G. Lithium Selective Adsorption on 1-D MnO2 Nanostructure Ion-Sieve. Adv. Powder Technol. 2009. https://doi.org/10.1016/j.apt.2009.02.008. (67) Kuss, C.; Carmant-Dérival, M.; Trinh, N. D.; Liang, G.; Schougaard, S. B. Kinetics of Heterosite Iron Phosphate Lithiation by Chemical Reduction. J. Phys. Chem. C 2014. https://doi.org/10.1021/jp502346f. (68) Rettig, S. L.; Jones, B. F.; Risacher, F. Geochemical Evolution of Brines in the Salar of Uyuni, Bolivia. Chem. Geol. 1980. https://doi.org/10.1016/0009-2541(80)90116-3. (69) JOGMEC; Toyota Tsusho. Joint Study of Process Development of Lithium Separation and Refining at Salar Brines in Puna Region, Argentina; 2011. (70) Nuibe, Y. Lithium Resources in Argentina and Projects of Exploration Development; 2013. (71) Dixon, D. G. Chloride-Based Leaching of Sulfides: Principles and Practice. (72) Druschel, G. K.; Hamers, R. J.; Banfield, J. F. Kinetics and Mechanism of Polythionate Oxidation to Sulfate at Low PH by O2 and Fe3+. Geochim. Cosmochim. Acta 2003. https://doi.org/10.1016/S0016-7037(03)00388-0. (73) House, J. E. Inorganic Chemistry; Elsevier Inc., 2008. "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2019-09"@en ; edm:isShownAt "10.14288/1.0379929"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Materials Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@* ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Lithium extraction from brine with ion exchange resin and ferric phosphate"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/71058"@en .