Optimal liberation of waste lithium-ion battery electrode material through attrition milling

This study is focused on the liberation of waste lithium-ion battery electrode material through a physical attrition milling process. High recovery of graphite and cathode material is an important driver for the process economics, thus a robust milling process to liberate these particles is important. To correlate liberation with the attrition milling efficiency, tests were conducted to determine optimal operational parameters, including propeller speed and solid-liquid ratio. Each parameter was tested with three different values, resulting in nine unique testing parameters. Once the tests were completed, the samples underwent size analyses to determine the amount of liberated electrode material. The optimal milling parameters were then determined by the design of experiments and were shown to significantly increase the amount of liberated electrode material compared to size reduction alone.

OPTIMAL LIBERATION OF WASTE LITHIUM-ION BATTERY ELECTRODE MATERIAL THROUGH ATTRITION MILLING By Ruben Ochoa A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Metallurgical Engineering Approved: York Smith Michael Simpson Thesis Faculty Supervisor Chair, Department of Metallurgical Engineering Kevin Perry Sylvia D. Torti Honors Faculty Advisor Dean, Honors College TABLE OF CONTENTS ABSTRACT ii 1 INTODUCTION 1 1.1 History of Lithium-Ion Batteries 2 1.2 Lithium-Ion Battery Chemistry and Manufacturing 3 1.3 Global Production and Waste 6 1.4 Primary Sources for LIB materials 7 1.4.1 Cobalt Mining 7 1.4.2 Aluminum Mining 8 1.4.3 Nickel Mining 9 1.4.4 Copper Mining 10 1.5 Secondary Sources for LIB materials 11 1.5.1 Secondary Smelting 11 1.5.2 Secondary Leaching 12 2 METHODS 12 3 RESULTS AND DISCUSSION 16 4 CONCLUSIONS 21 5 REFERENCES 23 6 APPENDIXES 26 i ABSTRACT This study is focused on the liberation of waste lithium-ion battery electrode material through a physical attrition milling process. High recovery of graphite and cathode material is an important driver for the process economics, thus a robust milling process to liberate these particles is important. To correlate liberation with the attrition milling efficiency, tests were conducted to determine optimal operational parameters, including propeller speed and solid-liquid ratio. Each parameter was tested with three different values, resulting in nine unique testing parameters. Once the tests were completed, the samples underwent size analyses to determine the amount of liberated electrode material. The optimal milling parameters were then determined by the design of experiments and were shown to significantly increase the amount of liberated electrode material compared to size reduction alone. ii 1 1 INTRODUCTION Ever since the Information Age began in the 1970’s with the invention of the personal computer and internet, the market for consumer electronics has experienced exponential growth. It was estimated by The Consumer Electronics Association that in 1998, the United States’ annual sales for consumer electronics exceeded $80 billion. This accounted for 1.6 billion units and averaged about six devices per person [1]. Currently, there are roughly 7.2 billion electronic devices globally, with most being powered by a lithium-ion battery (LIB), due to their high energy density, no memory effect, and a high discharge cycle [2][3]. With the demand seen with consumer electronics and electric vehicles (EVs), LIB’s market is projected to reach 15.7 million units by 2027 [4]. Since there are few established recycling processes to extract the materials needed for LIB through secondary sources [6], most of the material will have to be extracted from primary sources. Therefore, promoting the potential depletion of natural deposits, environmental damage, and social issues associated with mining. The increased production of LIB is projected to generate about two million tons of a secondary source seen as waste LIBs around the globe by 2030 [5]. If current trends remain, most of these LIBs will be sent to landfills, allowing for further safety and environmental issues. Due to the economic factors, only 5% of waste LIBs are recycled in both the United States and the European Union; with the most practiced methods being settling and leaching [5][6]. However, they both have their environmental concerns due to the chemical processes involved and harmful byproducts. To reduce the concerns associated with the increasing amount of waste LIBs being introduced into landfills as well as the concerns with the continual extraction through 2 mining, an economically feasible and environmentally friendly recovery process must be developed to take advantage of the increasing secondary source. This research will is focused on the third subset of LIB recycling, that being direct recycling or physical processing. A baffled attrition mill was used to liberated shredded LIB cathode material and its liberation efficiency response to operation conditions was examined. The system variables examined were solid to liquid ratio and propeller speed. Through this study, optimal operating parameters were established and consequently improved the efficiency of the recovery process. 1.1 History of Lithium-Ion Batteries The development of the modern LIB has been credited to three different scientists including, John B. Goodenough, M. Stanley Whittingham, and Akria Yoshino. Although each scientist created their respective prototypes with noticeable variations, their innovations contributed to the development of the modern LIB and earned them a shared Nobel Prize [7]. The earliest predecessor being Stanley Whittingham’s prototype in the 1970s. At the time Whittingham was serving as a chemist for Exxon mobile, working on a rechargeable battery to eventually replace fossil fuel energy. His original system consisted of titanium disulfide and lithium metal as the cathode and anode, respectively [7]. The Exxon system, however, experienced many safety issues due to the lithium metal anode and LiClO4 electrolyte. This electrolyte was shock-sensitive and became liable to explode under a strong enough shock condition [8]. 3 Ten years later in the 1980s, John B. Goodenough at the University of Oxford in England was able to double the battery’s energy potential. Goodenough’s system was similar to Whittingham’s but had a lithium cobalt oxide cathode instead of titanium disulfide. In 1991, Akira Yoshino at Meijo University at Nagoya, Japan, developed the first modern LIB prototype. Yoshino altered Goodenough’s design by replacing lithium metal with a carbonaceous petroleum coke as the anode material. This alteration leads to a safer battery with a more stable performance [7]. Through the innovations that Goodenough and Yoshio implemented on Whittingham’s original system, they created a stable rechargeable battery. By altering the electrode materials, they were able to improve Whittingham’s prototype by reducing safety concerns and promoting a high energy density with no memory effect. 1.2 Lithium-Ion Battery Chemistry and Manufacturing All batteries operate through by transferring elections from their anode to the cathode through an electrochemical reaction. For non-rechargeable batteries, such as alkaline batteries, the electrochemical reaction is irreversible. This is due to its cathode corroding the anode. As the elections are transferred from anode to cathode, the anode will oxidize and transfer ions to the cathode through the electrolyte medium and plate itself onto the cathode. Rechargeable batteries, such as LIB, can reverse this process and send ions back to the anode by supplying a surplus of electrons during the charging process. This process is successful due to the electrode structures within the LIB. Unlike a nonrechargeable battery, these electrodes, lithium metal oxide and graphite, have a structure 4 that promotes the intercalation of lithium ions, rather than plating. For example, when a LIB is being discharged, the Li-ion will dislodge from the graphite layers in the anode and travel into the metal oxide structure in the cathode creating lithium metal oxide. Since a graphite structure is left in the anode and the Li-ion can easily dislodge from the metal oxide structure, the reaction can easily be reversed [9]. Although the principle reaction is the same throughout all LIB, there are variations within the cathode materials. In the 1980s through a partnership between Oxford and the University of Texas at Austin, Goodenough discovered three classes of oxide cathode structures that would allow for the intercalation of lithium ions. Examples of each structure are shown in Figure 1. All cathodes remain in the lithium metal oxide family with varying metal oxides; layered, spinel, and polyanion will contain cobalt, manganese, and iron respectively [10]. Figure 1: Examples of three classes of cathode structures used as cathode materials in lithium-ion batteries [10] 5 Each of these cathodes have different benefits and faults associated with them and selection dependents on its end-use. For example, the polyanion structure is not as conductive as its spinel and layered counterparts and requires to be coated with carbon. However, the polyanion cathode has higher thermal stability and can sustain higher charge/discharge rates. These characteristics make them desirable for renewable energy grid storage. Due to their structures, spinel and layered cathodes are observed to have higher density, resulting in a higher energy density. Layered oxides are usually more preferred over spinel due to its capability to stabilize more positive metal ions. However, Spinel cathodes are still able to reach a high operating voltage of 4.7 volts [10]. Although providing different benefits, having this chemical variation in the cathode is a major hurdle in developing recycling separation process. Regardless of what cathode type a LIB is unitizing the construction process remains the same. The anode is made by coating copper foil with a layer of graphite and the cathode is made by coating a lithium metal oxide film on an aluminum electrode. After the electrodes are dried, they are cut to their appropriate sizes and assembled with an electrolyte and separator. The foils are then rolled or folded into the desired size and placed into their casing. An example of a LIB cross-section can be seen below in Figure 2 [11]. The cross-section shows the typical layered foil structure that allows for a high surface area and shows the circuit board that regulates the current. 6 Figure 2: Cross Section of LIB [11] 1.3 Global Production and Waste An increase in electric vehicles (EV) is expected and along with it an increase in LIB. The US Energy Information Administration estimated that the United States will sell between 499,000 to 1,146,000 EV units in 2022. Where the global market is expected to reach 6,900,000 units [12]. It is predicted that between 2015 and 2040, 0.33 to 4 million metric tons of LIBs will be produced for EVs alone. Which would lead to about 203,000 metric tons of waste LIB by 2025 [13]. Within this waste stream, it is estimated to contain 42 percent, by weight, materials with established recycling processes such as aluminum, cobalt, copper, nickel, and steel. The waste stream also contains 10 percent valuable metals without an established recycling process, including lithium and manganese [14]. Leaving roughly 85,000 and 20,000 metric tons of recyclable and valuable material, left in landfills in 2025. Which could account for 106,000 tons of material by 2025, that does not need to be extracted through primary sources and can avoid being processed through the landfill. 7 1.4 Primary Sources for LIB materials This section will discuss the extraction processes through primary sources of common LIB materials such as cobalt, aluminum, nickel, and copper. This section will also go over the environmental and social issues associated with mining. 1.4.1 Cobalt Mining About 60 percent of the world’s cobalt is extracted in the Katanga copper belt in the Democratic Republic of the Congo [15]. The mine can vary from open pit to underground depending on the ore grade, size, and surface type. Once the ore is extracted it is introduced to one of the various processes including, leaching, smelting, and/or electrorefining. The leaching process involves milled ore to be mixed with sulfuric acid under pressure and heated to 50-80 degrees Celsius for 90 minutes. This allows for the cobalt and nickel to be converted into sulfate salts, which are then washed through a circuit and from a cobalt-nickel solution. This solution is then leached again with high-pressure oxygen to separate the nickel and cobalt [16]. Smelting will produce a relativity pure matte cobalt carbonate or cobalt oxide. This is obtained by introducing the ore and flux, usually lime, into a furnace. Separation is achieved through the manipulation of their properties including density and melting point. The undesirable materials will likely have a lower density and will be removed with a partition along with the slag. The cobalt carbonate/oxide will then be cast into anodes and purified by electrorefining. In this process, the cobalt anode is introduced into an acidic 8 electrolyte solution alongside a steel cathode. The cobalt ions plate on the cathode, creating a pure cobalt layer [16]. Each of these processes has its respective environmental concerns due to the pressurized acid, fumes emitted, and energy required. The extraction process also has health risks because of uranium-rich dust [15]. This dust is known to cause vision problems, vomiting, heart problems, and Thyroid damage [16]. However, the social issues with cobalt mining out weight these. Anti-slavery economist Siddharth Kara has accused major American tech companies of aiding in the deaths of children miners working in the Congo’s cobalt mines. Families in the Congo also state that their children have been working in mines owned by Glencoe, a UK based mining company [17]. Although environmental risks are concerning, a secondary process for cobalt must be developed to minimize the damage being done in the Congo. 1.4.2 Aluminum Mining Roughly about 70 percent of the world’s aluminum is created by the refinement of alumina that is created through the Bayer process. This process requires bauxite that can be found in the topsoil of tropical and subtropical regions. The estimated bauxite reserves are said to be between 55 to 75 billion metric tons [18]. The reserves are observed to be spread across Africa, Oceania, South American, and Asia, each containing roughly 32, 23, 21, and 18 percent of the global reserves, respectively [18]. In the Bayer process, the incoming bauxite is washed and crushed down to a particle size of 193 – 90 μm. This increases the surface area and promotes the digestion process. Next, the bauxite is submerged into a sodium hydroxide solution raised to, 150 to 200 9 degrees Celsius, leaching the aluminum bearing minerals such as gibbsite, boehmite, or diaspore, creating a pregnant liquor. Depending on the content of each aluminum bearing material, the digestor will operate between 140 – 240 degrees Celsius. After leaching, the bauxite tailings are separated from the pregnant liquor. The liquor is then introduced to a precipitation tank, where the solution can cool and form aluminum hydrate precipitates. These crystals then undergo size classifications to determine if they are large enough for calcination or small enough to serve as precipitation seeds. The crystals suitable for calcination are then be roasted at 1100 degrees Celsius, to remove free moisture and produce alumina [19]. The refinement of alumina into aluminum is similar to cobalt’s, using smelting and electrorefining techniques. 1.4.3 Nickel Mining It is estimated that the world’s nickel reserves are at around 300 million metric tons [20]. This is accounting for both laterite and sulfide type deposits. Roughly 75 percent of laterite ore is accounted for through nine countries being Indonesia, The Philippines, Brazil, Cuba, and New Caledonia. Sulfide ores are more commonly found in South Africa, Russia, and Canada; with Australia having the largest deposits of both types. It is also important to note that more than 290 million tons of nickel are extracted through deep-sea mining operations [20]. After the extraction process, sulfide ores will undergo various milling circuits to ensure a proper particle size of roughly 65 Pm. The ore is then introduced into floatation cells, where reagents will selectively float out the nickel-bearing mineral. The tailings of the initial cell undergo another floatation process to recover the remaining nickel-bearing mineral. This ore concentrate is then smelted at 1350 degrees Celsius to remove both the 10 iron and sulfides present. This process produces a matte consisting of 70 to 75 percent nickel. To reach the desired purity, the matte is cast into anodes and is sent to electrorefining [21]. Laterite ores have been known to be difficult to concentrate due to the uniform distribution of nickel throughout the mineral’s lattice. As a result, concentrating methods such as floatation or gravity separation are futile. In response, most laterite operations process the ore in its entirety [22]. Being an oxide ore, large amounts of moisture is expected with the ore, around 35 to 40 percent water [21]. To remove the moisture from the ore, it is introduced to rotary kiln furnaces. The refinement steps then follow those of a sulfide ore except that the laterite smelter operates at 1360 to 1610 degrees Celsius. The matte created is then sent to electrorefining [21]. Sulfide ores have been historically notorious regarding environmental concerns. Norilsk, Russia was rated one of the most polluted cities in the world in 2016, due to the 350,000 tons of sulfur dioxide that is emitted annually by the Norilsk Nickel mine. This mining company has also been responsible for polluting the city’s river with oxidized nickel causing it to turn red. Due to related health and environmental concerns, 17 nickel mines in The Philippines have been shut down [23]. 1.4.4 Copper Mining Copper has an estimated global reserve of 2.1 billion metric tons [24]. Roughly about 50 percent of the world’s copper originates from South America, South Central Asia, Indochina, or North America. However, it is important to note that copper has been mined in nearly every continent [25]. 11 Much like nickel, copper is also found as both sulfide and oxide ores. The refinement process for copper sulfide ore is similar to that of nickel sulfide. Both creating a concentrate through floatation, which is then introduced to smelting, and finally electrorefining. Copper oxide ore processing, however, does vary. Copper oxide ores are leached with sulfuric acid to create a pregnant solution. The copper is extracted through a solvent extraction process and sent to electrowinning [26]. Contrary to electrorefining, in electrowinning, instead of the copper serving as an anode, it is placed in a solution with an inert anode. The current passed through the electrodes causes the copper ions to plate onto the cathode without any contamination from the anode material. 1.5 Secondary Sources for LIB materials This section discusses the advantages and disadvantages of secondary processes being implemented for LIB such as smelting and leaching. 1.5.1 Secondary Smelting The smelting process for LIB recovery usually does not require any size reduction processes. Waste LIBs are introduced to the smelter at temperatures around 1,450 degrees Celsius, to produce the reduction of cobalt and nickel oxides to metals. However, both aluminum and lithium oxides are transferred to the slag phase and treated as a waste stream. Smelting is capable of recovering up to 70 percent of the cathode value, depending on the amount of cobalt is present in the cathode chemistry [27]. Lithium could be recovered during the smelting, but this would require a prior size reduction. Although this would also allow for the recovery of copper and aluminum foils, it would also increase cost and remove 12 the energy associated with aluminum’s oxidization from the smelter [27]. Although this process can recover a high percentage of cathode value, the high energy cost and cathode chemical variation has made it economically unpredictable. 1.5.2 Secondary Leaching Leaching requires size reduction, allowing for the separation of the electrode foils and electrode powder. This powder consists of graphite and a lithium metal oxide variant, depending on the type of battery cell. The powder is then leached with a strong acid creating a pregnant liquor of various metal ions, due to the complexities of the cathode chemistry. The cobalt and nickel ions can be collected through solvent extraction. However, due to their similar properties, it is difficult to separate these ions. In most cathodes, the remaining ions have a high enough value that their extraction is economically feasible. However, lithium iron phosphate and lithium manganese oxide cathode materials have low value and do not justify leaching [27]. Much like smelting, the variation of cathode chemistry makes the economic feasibility of leaching unstable. 2 Methods The batteries used throughout these experiments were collected from the University of Utah’s Waste Center. These LIBs had a lithium cobalt oxide cathode material and included three pouched cells per unit. The LIBs were manually deconstructed to access and discharge each pouched cell. This process consisted of hooking up each cell to a series with two, two-ohm resistors, and left overnight. Once the cells were determined to be discharged 13 with a voltmeter, the unit was reconstructed to simulate a realist input for the size reduction process. The prepared units are then introduced to a shredder, seen in Figure 3, operating at 40 RPM. The material was then classified by particle size through a shake table and sieves. The sieve sizes used were: 8 mm, 4.75 mm, 2 mm, 1.18 mm, and 500 μm. The seizing process was performed over three sets for five minutes with 20-second intervals at an amplitude of 3.00 mm. This was then repeated with eight more samples to create an averaged mass fraction baseline per each sieve size. The baseline is utilized to simulate a shredded LIB by combining appropriate amounts per sieve size while meeting the 100, 150, 300 grams per liter, solid to liquid ratio requirement. It is important to note that the bassline was adjusted to exclude particles below 500 μm. This insured that the record particles under 500μm after milling were due to the experiment and not size reduction. 14 Figure 3: Image of Shredding System Once the samples were prepared into 50, 75, and 150 gram sets; they were placed into a baffled attrition mill system with 500mL of deionized water. A diagram of the system is shown in Figure 4. This system is constructed of a one-liter beaker with four baffles made from Plexiglas. The four baffles were placed 90 degrees from each other. The placement of these baffles was designed to promote the creation of vertical vortexes during the milling processes. They were placed with an 0.75-inch gap from the bottom of the beaker to reduce any agglomeration or dead zones. All tests had a run for one hour while the solid to liquid ratio and propeller speed were varied. The tested solid to liquid ratios include 100, 150, and 300 grams per liter. The tested propeller speeds were 600, 1200, 1500 RPM. Testing every combination of operating parameters will then result in nine tests. 15 Figure 4: Diagram of Baffled Attrition Mill Once the hour is completed the system is filtered with a metal strain to catch larger particles, roughly larger than 1.18 mm. These particles were then subject to vacuum filtration to catch any smaller particles above 22 μm. The remaining water and fine powder are then subject to centrifugation to recover the particles under 22 μm. After 30 minutes at 700RPM in the centrifuge, the caked powder was decanted from the solution and filtered using a vacuum filter. Once the filtering process was complete, the recovered samples and the filter papers are placed into a tabletop furnace at 60 degrees Celsius and left overnight to remove any remaining moisture. The dried samples are processed again through particle size classification used to establish the baseline, with the addition of 212 μm and 106 μm sieves. Since the total mass will vary throughout the tests, the mass fraction of particles under 500 μm was used to determine the optimal operating parameters of the attrition mill. 16 3 Results and Discussion The parameters chosen for the experiment were derived from the statistical method, design of experiments. This method allows for the simultaneous testing of multiple variables and illustrates the impact of processing parameters on the efficiency of the system. Prior attrition tests have revealed a solid to liquid ratio of 150 grams per liter and a propeller velocity of 1200 revolutions per minute created sufficient liberation, resulting in a 0.41 mass fraction of liberated electrode material. It was assumed that this was because the solid to liquid ratio introduced enough mass to create friction from interparticle interactions and the propeller velocity was significant enough to create vertical vortexes within the baffles to improve slurry mixing. Due to its past success, these parameters were selected to be the origin for the design of experiments. The remaining parameters were determined by setting the upper and lower boundaries of ‘1’ from the origin. For these experiments, the x and y-axis were assigned to revolutions per minute and solid to liquid ratio, respectively. The boundaries for the y-axis were set at +150 and -50, whereas the x-axis had boundaries at -600 and +300. Setting each of the boundaries to a value of ‘1’ a matrix of parameters is built, as shown in Table 1. Each cell in the matrix carry two values representing x and y, or RPM’s and solid to liquid ratio, with each value increasing along their respected axis. Once all points are tested and a contour plot is created, it is analyzed to determine which point produced the highest efficiency. If ା ା this point is not the origin, additional points will be created at ( ξʹ, 0) and (Ͳǡ ξʹ) to ି ି expand the field of view of the plot. These ‘ξʹ’ points will be proportional compared to 17 the original ‘1’ parameters. An example of what the secondary point values would be for these experiments is shown in Figure 5. This is done to help determine the direction of a positive trend. Once this is determined, the closest known point to the positive trend is assigned as the new origin and the process is repeated until the contour’s maximum is observed within its boundaries. S/L Ratio Table 1: Parameter Matrix of Primary DOE Boundary ( 600, 300 ) ( 600, 150 ) ( 600, 100 ) ( 1200, 300 ) ( 1200, 150 ) ( 1200, 100 ) RPM's ( 1500, 300 ) ( 1500, 150 ) ( 1500, 100 ) Figure 5: Secondary boundaries of DOE ା The primary contour plot with only ͳ boundary for these sets of experiments is ି given in Figure 6. Since the maximum is observed at the origin there is no need for the secondary points at ା ି ξʹ , illustrating that the original parameters were the most optimal. 18 Figure 6: Mass Fraction of Particles Less Than 500μm Per Test For the design of experiments to be a success, there must be fixed variables. In this case, those were the amount of deionized water, run time, and mass fraction per sieve size. Although the total mass added to the system varied with the solid to liquid ratio, the mass fraction per sieve size remained the same. Due to the baseline tests performed before the milling experiments, the desired mass fractions per size were determined and are given in Table 1. It is important to note that since no particles under 500 μm were added to the system, the mass fractions were adjusted to exclude the mass of these particles. The actual mass faction per sieve size and total mass of each experiment are shown in Appendix A. 19 Although it is difficult to recreate perfect samples due to the agglomeration of particles, on ା average each sample only varied by ି ͲǤʹͷΨ from its desired overall mass. Table 2: Mass Fraction and Adjusted Mass Fraction Per Seize Size Seize Size Mass Fraction Adjusted Mass Fraction > 8 mm 0.10 0.11 8mm-4.75mm 0.63 0.67 4.75mm-2mm 0.13 0.14 2mm-1.16mm 0.03 0.03 1.16mm-500μm 0.04 0.04 500 μm< 0.07 The mass and mass loss after the milling experiments are given in Appendix B. Unlike the mass variation seen in Appendix A, Appendix B reveals a high mass loss. On average 9.51% of the mass was lost, ranging from 14.23 to 5.24%. It is assumed that this high mass loss is due to the containers used to store the samples. Throughout the experiments, the samples were known to stick to the walls of the containers and became difficult to remove the sample in its entirety. The mass fraction per seize size after milling is given in Appendix C along with the calculated mass fraction of particles below 500 μm. These values were used to construct the contour plot in Figure 6. The mass fraction of particles below 500 μm was determined by adding the mass fractions of 500-212 μm, 212-106 μm, and < 106 μm. It is assumed that the low mass fraction of liberated particles seen with the 300 grams per liter ratio is due to there being too much mass impeding uniform slurry mixing. Without the vortexes, there was little friction caused by interparticle interactions causing less particle liberation from the metal foils. The positive trend line between the 300 solid to liquid ratio with respect to revolution per minute shows that the increasing RPMs introduce higher 20 mechanical energy which promotes vortex development, resulting in increased interparticle interactions and higher yields. Similar reasoning can be applied to the results for 600 RPM. It was observed at this velocity that the propeller did produce significant vortexes and mixing. This is supported by the positive trend in liberated material as the solid to liquid ratio is decreased. At a solid to liquid ratio of 100, 600 revolutions per minute was able to liberate a mass fraction of 0.39, resulting in the second-highest yield. The mass fraction of < 500 μm sized particles, or yield associated with a 100 solid to liquid ratio were consistently high with yields between 0.38 and 0.39. It is believed that this is due to the low energy required to create the vertical vortexes. The mass fraction of liberated particles only varied by 0.01 throughout the RPM range tested, showing that the energy required to achieve maximum liberation was met at the lowest revolutions per minute. It is assumed that to increase the maximum liberation, more mass is required to promote more interparticle scrubbing. Regarding the results for 1500 RPMs, these yields were lower compared to those from 1200 RPMs. It was observed in both the 100 and 150 solid to liquid ratios that the particles would agglomerate along the baffle walls. Illustrating that the radial flow was great enough to limit vertical movement, causing the buildup and low interparticle interaction. To achieve the maximum liberation of electrode material, the operating conditions must be conducive for sufficient vortex flow without promoting particle entrainment along the baffle walls. The solid to liquid ratio must also be low enough to ensure the energy required for the vortexes is feasible for the propeller but, also needs to have enough mass to create the most amount of friction from interparticle interactions as possible. Through 21 the results seen in both Appendix C and Figure 5, the parameters that meet these requirements are a 150 grams per liter solid to liquid ratio and a propeller speed of 1200 revolutions per minute. These parameters resulted in a mass fraction of 0.41 for particles under 500 μm. These parameters increased the mass faction of liberated particles from 0.07 to 0.41, resulting in nearly a 5x increase in liberated particles. However, it is important to note that these experiments were run with 500 ml, not one liter. Although the mass was halved to meet solid to liquid ratio requirements, the interparticle interactions may vary as the system is scaled up to higher volumes. 4 Conclusions Design of experiments was implemented to determine optimal operating parameters of an attrition mill to liberate electrode material from waste LIBs. Testing parameters include solid to liquid ratio and propeller speed. To properly replicate shredded LIBs, size reduction, and particle size classification were performed on waste LIBs to determine average baseline mass fractions to ensure the change in particle size distribution is the result of milling instead of size reduction. This study determined the optimal operating solid to liquid ratio at 150 grams per litter with a propeller speed of 1200 RPMs, resulting in a liberated mass fraction of 0.41. It is believed that these parameters created the highest yield due to the ununiform mixing caused by vortex formations and the high friction created by interparticle interactions. These parameters increased the mass fraction of liberated electrode material by nearly five times compared to its original mass fraction after size reduction of 0.07. This study 22 revealed attrition milling’s capacity to liberate electrode material and its promise of being implemented in waste LIB recovery processes. 23 5 References [1] ." Encyclopedia of Communication and Information. . Encyclopedia.com. 30 Sep. 2020 . 6 Oct. 2020, www.encyclopedia.com/media/encyclopedias-almanacs-transcriptsand-maps/consumer-electronics. [2] There Are Officially More Mobile Devices than People in the World. 7 Oct. 2014, www.independent.co.uk/life-style/gadgets-and-tech/news/there-are-officially-moremobile-devices-people-world-9780518.html. [3] Osmanbasic, Edis. What You Need to Know About EV Batteries. www.engineering.com/ElectronicsDesign/ElectronicsDesignArticles/ArticleID/19504/W hat-You-Need-to-Know-About-EV-Batteries.aspx. [4] Research, Transparency Market. “Lithium-Ion Battery Market to Expand at Growth Rate of ~ 11% Compounded Annually from 2019 to 2027 to Reach 15,764.89 Million Units by 2027: Transparency Market Research.” PR Newswire: Press Release Distribution, Targeting, Monitoring and Marketing, 3 Feb. 2020, [5] Jacoby, Mitch. “It's Time to Get Serious about Recycling Lithium-Ion Batteries.” Chemical & Engineering News, American Chemical Society, 15 July 2019, cen.acs.org/materials/energy-storage/time serious-recycling-lithium/97/i28. [6] Chumwei, Liu, and Lin Jiao. “Recycling of Spent Lithium-Ion Batteries in View of Lithium Recovery: A Critical Review.” Journal of Cleaner Productioin, vol. 228, no. 801. [7] Liu, Zhao, et al. “The History of the Lithium-Ion Battery.” Accelerating Microscopy, 11 Oct. 2019, www.thermofisher.com/blog/microscopy/the-history-of-thelithium-ion-battery/. [8] Blomgren, George E. “The Development and Future of Lithium Ion Batteries.” Iopsiecne.org, iopscience.iop.org/article/10.1149/2.0251701jes/pdf. [9] Brain, Marshall. “How Lithium-Ion Batteries Work.” HowStuffWorks, HowStuffWorks, 14 Nov. 2006, electronics.howstuffworks.com/everyday-tech/lithiumion-battery.htm. [10] Manthiram, Arumugam. “A Reflection on Lithium-Ion Battery Cathode Chemistry.” Nature News, Nature Publishing Group, 25 Mar. 2020, www.nature.com/articles/s41467-020-15355-0. [11] “Full Page Reload.” IEEE Spectrum: Technology, Engineering, and Science News, spectrum.ieee.org/semiconductors/design/how-to-build-a-safer-more-energydenselithiumion-battery. [12] “Annual Energy Outlook 2012.” U.S. Energy Information Administration, May 2012. 24 [13] Winslow, Kevin M., et al. “A Review on the Growing Concern and Potential Management Strategies of Waste Lithium-Ion Batteries.” Resources, Conservation and Recycling, vol. 129, 2018, pp. 263–277., doi:10.1016/j.resconrec.2017.11.001. [14] Richa, Kirti, et al. “A Future Perspective on Lithium-Ion Battery Waste Flows from Electric Vehicles.” Resources, Conservation and Recycling, Feb. 2014. [15] Blijweert, Dirk. “Scientists Reveal the Hidden Costs of Cobalt Mining in DR Congo.” Phys.org, Phys.org, 21 Sept. 2018, phys.org/news/2018-09-scientists-revealhidden-cobalt-dr.html. [16] Farjana, Shahjadi Hisan, et al. “Life Cycle Assessment of Cobalt Extraction Process.” Journal of Sustainable Mining, vol. 18, no. 3, 2019, pp. 150–161., doi:10.1016/j.jsm.2019.03.002. [17] Kelly, Annie. “Apple and Google Named in US Lawsuit over Congolese Child Cobalt Mining Deaths.” The Guardian, Guardian News and Media, 16 Dec. 2019, www.theguardian.com/global-development/2019/dec/16/apple-and-google-named-in-uslawsuit-over-congolese-child-cobalt-mining-deaths. [18] “Bauxite.” Bauxite | The Aluminum Association ,www.aluminum.org/industries/production/bauxite. [19] AG, interstruct. “Alumina Is Used for the Production of Aluminium Metal, through the Hall–Héroult Electrochemical Smelting Process.” Mining and Refining – Process, bauxite.world-aluminium.org/refining/process/. [20] “The Life of Ni.” Nickel Institute, nikelinstitute.org. [21] Wise, Edmund Merriman, and John Campbell Taylor. “Extraction and Refining.” Encyclopædia Britannica, Encyclopædia Britannica, Inc., 5 Sept. 2013, www.britannica.com/technology/nickel-processing/Extraction-and-refining. [22] Pintowantoro, Sungging, and Fakhreza Abdul. “Selective Reduction of Laterite Nickel Ore.” MATERIALS TRANSACTIONS, The Japan Institute of Metals and Materials, 25 Oct. 2019, www.jstage.jst.go.jp/article/matertrans/60/11/60_MT-M2019101/_html/char/ja. [23] Opray, Max. “Nickel Mining: the Hidden Environmental Cost of Electric Cars.” The Guardian, Guardian News and Media, 24 Aug. 2017, www.theguardian.com/sustainable-business/2017/aug/24/nickel-mining-hiddenenvironmental-cost-electric-cars-batteries. [24] How Much Copper Has Been Found in the World?, www.usgs.gov/faqs/howmuch-copper-has-been-found-world?qt-news_science_products=0. [25] “Geology of Copper.” Provident Metals, 15 Sept. 2019, www.providentmetals.com/knowledge-center/precious-metals-resources/geology-ofcopper.html. 25 [26] says:, Bryan Whipple, et al. “Types of Copper Deposits in the World: INN.” Investing News Network, 9 Sept. 2020, investingnews.com/daily/resource-investing/basemetals-investing/copper-investing/types-copper-deposits-world/. [27] Gaines, Linda. “Lithium-Ion Battery Recycling Processes: Research towards a Sustainable Course.” Sustainable Materials and Technologies, Elsevier, 7 July 2018, www.osti.gov/servlets/purl/1558994.