A Strategic Metal for Green Technology: The Geologic Occurrence and Global Life Cycle of Lithium

As technological demand for materials with specific physical properties has increased, the importance of strategic metals cannot be ignored. Strategic metals are those materials that have high economic value, are used in a variety of applications, and have few or no viable substitutions; supply may face the risk of restriction due to various technical, economic, or social factors. Mainly due to the growing demand for lightweight and powerful batteries, lithium has become such a metal. While supplies of lithium have historically been mined from pegmatites, brine extraction from salars, or salt flats, has become the dominant source for lithium within the past decade. These salars are found in continental, volcanogenic highlands in arid regions of the world with internal drainage basins; the most noticeable exist in the Andean Altiplano. The lithium in the brines originates from the alteration and weathering of volcanic rocks. The South American salars can be classified by their morphology and by their chemistry. More importantly, the origin and evolution of the brines within these salars can be qualitatively analyzed using the concept of the chemical divide, which determines changing solute concentrations as water evaporates. To quantify the stocks and flows of lithium at present, a global life cycle tracking the metal through its anthropogenic life stages (production, fabrication and manufacturing, use, and waste management) was developed. This life cycle appears to be the first of its kind for lithium and will begin to quantify the rates and types of use of lithium, as well as its potential for recycling. The combination of analysis of both the supply and demand sides of lithium gives a comprehensive picture of the current status of lithium, which is vital to understanding the future of this strategic metal. Introduction The incredible advancement in technology over the past several decades has radically changed human demand for metals. Traditionally, industry and consumers required high quantities of commodity metals such as steel, copper, and aluminum. Technological advances made in every major industry increasingly demand rarer metals with specific physical properties. While high volume production of metals like iron is not yet a thing of the past, platinum group metals and rare earth elements are becoming just as important in products that are taken for granted today. The transition from demand for bulk metals to scarcer metals has created some concern in political and industrial circles about adequate, reliable supplies of the rarer elements. In 1981, a cooperative program called the International Strategic Minerals Inventory (ISMI) was formed by government officials from the United States, Canada, the Federal Republic of Germany, the Republic of South Africa, Australia, and the United Kingdom to address these concerns. The purpose of this organization was to publish publicly available and nonproprietary data about the short-term, medium-term, and longterm supplies of specific minerals for policy considerations. Although ISMI acknowledged that technical, financial, and political complications may interfere with supplies of these minerals, their focus was on deposits and on demand. ISMI evaluated minerals such as chromium, manganese, nickel, phosphate, platinum-group metals, cobalt, titanium, graphite, vanadium, tungsten, tin, zirconium, and lithium (Andstett et al. 1990). ISMI referred to the minerals of concern as “strategic.” This term, they acknowledged, was imprecise; they used it to refer to minerals that come largely from foreign sources, that are difficult to replace, and that are important to a nation’s economy (especially in its defense industry): Andstett writes, “Usually, the term implies a nation’s perception of vulnerability to supply disruptions and of a need to safeguard its industries from the repercussions of a loss of supplies” (1990). Describing minerals as strategic therefore implies that modern society or industry has become dependent on them. While there is some disagreement about the specific definition of a strategic material, the general consensus is that these qualities that make a certain metal strategic (Weil et al. 2009): 1) The metal is used in several important industry sectors 2) It may be difficult to find adequate substitutes for the metal in one or more important applications 3) The number of applications is large and is increasing over time 4) The metal is used in applications in which it is dissipated, meaning that recycling potential is limited 5) The metal has high economic value 6) Production and/or reserves of the metal are geographically concentrated There is some recent disagreement as to the difference between strategic and critical metals. Strategic metals are usually associated with national security or military needs, while supply restrictions of critical metals can cause economic damage. Because critical has broader connotations than strategic, all strategic metals are critical, whereas not all critical metals are strategic (Committee on Critical Mineral Impacts on the U.S. Economy 2008). Throughout this paper, the two terms will be used interchangeably. Lithium fulfills all of the requirements to be considered a strategic metal. Lithium is used in a variety of industries, including ceramics and glass, batteries, lubricating greases, air treatment, in the production of primary aluminum, and alloys. Only lithium in batteries and air conditioning systems can be recovered and recycled; the rest of the uses are dissipative. The amount of lithium used in batteries has increased 20% annually over the past few years and shows no signs of slowing down, as lithium-ion batteries may be ideal for use in electric and hybrid vehicles. Substitutes for lithium in most of its major applications have compromised performance (Jaskula 2010). The lithium production industry is extremely concentrated, with four companies producing 90% of the world’s supply. One deposit, the Salar de Atacama, produces over 60% of the world’s lithium, and there are only ten other deposits that are currently being exploited (Ebensperger et al. 2005); few other deposits have been identified (Figure 1). Lithium was traditionally mined from pegmatite deposits, which may be zoned or unzoned. These pegmatite deposits are usually mined from the surface, and generally contain 0.59 to 1.36 percent lithium and contain between one and 50 million metric tons of ore. Zoned deposits contain lithium minerals such as spodumene, petalite, lepidolite, eucryptite, and amblygonite (Table 1). Unzoned deposits contain spodumene throughout the rock, which is homogenous. Spodumene may account for 25% of the rock; unzoned deposits tend to be the most important source of pegmatitic spodumene (Anstett et al. 1990). However, brine extraction has become the dominant source of lithium within the past decade. Lithium is dissolved by chemical weathering like all other alkali metals; these weathering waters can become concentrated into lithium-rich brines if they are in closed basins where evaporation exceeds precipitation (Anstett et al. 1990). These lithium brines commonly occur in the internal drainage basins in continental, volcanogenic highlands with arid climates. In South America, these salt flats are called salars and are an important feature of the Andean Altiplano (Warren 2010). Although some research regarding the morphology and geochemistry of these salars has been done, few scholars have attempted to place these brine deposits within the context of the ever-expanding lithium industry. The predominant purpose of this paper, therefore, is to develop a comprehensive and multidisciplinary understanding of the lithium brine deposits. Towards that end, the geologic occurrence of the salars, the methods of extraction and beneficiation used to produce lithium from them, and the potential geopolitical and environmental consequences of brine mining in South America will be explored. Additionally, this supply-oriented research will be complemented by exploring the demand side of the lithium industry by evaluating the use of lithium in batteries and through the creation of a global anthropogenic life cycle for lithium. This life cycle will track the stocks and flows of lithium worldwide during 2007 through its life cycle stages of production, fabrication and manufacturing, use, and waste management. The goal is to develop a broad yet rigorous understanding of the status and potential future of lithium. Lithium has gotten significant political and media attention in the past few years due to its potential use in electric and hybrid vehicles. Lithium batteries are ideal for vehicles because they have the greatest energy density of all batteries; this combination of power and low weight is vital to the development of electric and hybrid vehicles to replace traditional fossil fuel-driven vehicles. It seems likely that lithium’s application in a variety of “green” products will increase demand throughout the next century. Therefore, it is vital that scientists and design engineers, as well as leaders in government and industry, understand both the supply and demand sides of this strategic metal. Geologic Occurrence Origin of the Altiplano and Tectonic Context Only the Tibetan Plateau is larger and higher than the Andean Altiplano. Stretching 1800 kilometers across northern Peru and Bolivian down through northern Chile and Argentina, the Altiplano exceeds an elevation of three kilometers across a 300 to 400 km wide plateau. What makes this feature even more awe-inspiring is that it did not result from continental collision or accreted terranes; instead, crustal shortening dominated its formation. The original interpretation of the Altiplano was that it originated from the arc magmatism associated with the subduction of the Nazca Plate beneath the South American Plate: the addition of mantle material would be the cause of crustal thickening. However, it is now argued that magmatic processes were not important in creating the Altiplano. Instead, structural shortening of the crust, causing thickening, and thermal thinning of the lithosphere, causing uplift, are the dominant cause (Allmendinger et al. 1997). According to Isacks’ model for plateau development for the Altiplano, the first stage was a widespread, basin-and-range type shortening that occurred during the late Miocene; the second stage was a foreland fold-thrust belt (Isacks 1988). The foreland compresses and thickens the ductile lower crust, which then lifts up the upper crust into a plateau. The lithosphere becomes hotter and weaker due to the high rate of convergence and the low angle of subduction. Using currently available information, 70 – 90% of the thickening is accounted for solely by shortening. The “missing” crust may be due to either insufficient data or from an unknown process of magmatic addition. Possibilities include hydration of upper mantle rocks to crustal velocities or local tectonic underplating (Allmendinger et al. 1997). It should come as no surprise that such a vast feature would be heterogeneous. There are two distinct parts of the plateau: the Bolivian Altiplano and the Puna in Argentina and Chile. While a detailed discussion of their differences in topography, magmatism, and lithospheric structure is beyond the scope of this paper, it is worth noting that the two regions likely underwent the same stages of development at different times. For example, uplift began about 25 Ma in the Bolivian Altiplano compared to 15 – 20 Ma in the Puna; shortening ceased about 6 – 12 Ma in the Altiplano, while it lasted until 1 – 2 Ma in the Puna (Allmendinger et al. 1997). Several notable features in the Altiplano near northern Chile and their topographies are shown in Figure 2. The Pre-Andean Depression is an intramontane basin in Chile that is at an altitude of 2.5 kilometers and is filled with Tertiary to Holocene continental clastic and evaporite sediments. It contains the Salar de Atacama (which should not be confused with the Atacama Desert) and the Punta Negra salar at its south end. The Salar de Atacama is the largest evaporitic basin in Chile at 3000 square kilometers. The western edge of the Salar de Atacama borders the Cordillera de la Sal, which is the remains of a Tertiary salar deformed during Cenozoic tectonism (Risacher et al. 2003). The Western Cordillera is an elevated plateau above four kilometers in elevation. It is Miocene to Holocene and consists of rhyolitic ignimbrites and andesitic stratovolcanoes, which may tower as much as 2.5 kilometers above the rest of the plateau. Most of the smaller salars in Chile occupy the interior drainage basins demarcated by these volcanoes (Risacher et al. 2003). The Bolivian Altiplano is another major intramontane basin; it separates the Western Cordillera from the Eastern Cordillera above 22 °S latitude. The Salar de Uyuni, which is the world’s largest salt flat at 10,000 square kilometers, occupies the central trough of the Bolivian Altiplano (Risacher et al. 2003). It is estimated that the Salar de Uyuni may contain almost half of the world’s reserve base of lithium (Jaskula 2010). While the tectonic definition of the Altiplano is the part of the plateau reaching elevations above three kilometers, the more specific alternative definition is that the Altiplano consists of the internally-draining (endorheic) basins within this plateau (Allmendinger et al. 1997). The second definition is more appropriate for discussion of the salars, which inherently depend on closed basins for their origin and evolution. Evaporite and Brine Deposits Evaporite deposits are formed from the precipitation of salts when water loss is greater than water gain by the basin. The remaining brine becomes more and more concentrated over time. This process is usually driven by solar evaporation; it is worth noting that aridity is not always dependent on temperature because the driving force for evaporite deposition is the water balance. While some of the most famous evaporite deposits occur in hot deserts, evaporite deposits can also occur in cold, arid climates, such as the highlands of the Andes. Although precipitation increases to the west (Figure 3), it is greatly exceeded by evaporation across the entire Altiplano (Risacher et al. 2003). Evaporites are divided by the source of the original brine: they can be thalassic (marine) or athalassic (nonmarine). Lithium brines in the salars of South America are athalassic (Warren 2010). Beyond the requirement of the dominance of evaporation over precipitation, an internal drainage basin is required for salar formation (Risacher and Fritz 2009). While the precipitated solid salts are often the target of extraction in evaporite deposits, lithium brines are becoming the primary feedstock in certain deposits. These include Clayton Valley in Nevada, Salar de Atacama in Chile, Salar de Hombre Muerto in Argentina, and Dabuxum Salt Lake and Zhabuye Salt Lake, both in southwestern China. The lithium in these brines is a result of weathering of volcanic rocks, which also supply potassium, magnesium, and boron as well as more common ions like sodium and calcium (Warren 2010). Because of the concentration of lithium in these salars, the leaching of the rhyolitic rocks in the region likely occurred at temperatures around 400°C. Because lithium salts are highly soluble, lithium ions stay in solution and concentrate over time (Risacher and Fritz 1990). These lithium brines are “relatively unmodified pore brines (relict or connate waters) residing in permeable saline subsurface lithologies” that are found in “suprasealevel saline lacustrine settings in many arid continental high altitude volcanogenic terrains” (Warren 2010). Lithium salts are highly soluble, and so lithium tends to stay in solution, although sorption onto clays (such as onto hectorite in Clayton Playa, Nevada) is possible. The common minerals found in the salars are listed in Table 2; no minerals of lithium have ever been found in the South American salars. Chile, Argentina, and southwestern China currently dominant lithium brine extraction, although Bolivia has the world’s largest reserve base. These locations have been able to develop salars because of their history of aridity, which is required for the formation of evaporite deposits. The climate of the Altiplano has remained arid to semiarid since the Miocene. Large saline lakes were present at various times throughout Chile and Bolivia; in the early Pleistocene, two large lakes in Bolivia were Lake Ballivian and Pre-Minchin. Present day Lake Titicaca is a remnant of the former (Rettig et al. 1980). The latter lake evolved during the two major lacustrine periods in Bolivia: the Minchin, from 35,000 to 20,000 BP, and the Tauca, from 12,000 to 10,000 BP (Risacher and Fritz 1991); it is likely that Chile also experienced lacustrine phases, as well (Risacher et al. 2003). Fossil salt crusts found throughout the Altiplano are remains of these giant, deep saline lakes. The variety of salars can be astounding, but some attempt has been made to classify them on the basis of their morphology and chemistry. Researchers have traditionally separated Bolivian salars from Chilean salars for the purpose of analysis, although the differences between the two groups may be less significant than between members of the same group. Salar Morphology Salars range in size from 0.03 km to 10,000 km (Figure 4); there are four salars greater in area than 2,000 km while the rest are smaller than 400 km (Risacher and Fritz 2009). Risacher proposed a morphology classification scheme in 1990 that has been adopted by most researchers; it divides salars into four main types (Figure 5). However, the vast majority of salars exhibit more than one, if not all, of these physical types; the transition can happen laterally (different types in different places) or temporally (different types at different times during the year). The first type is the saline lake or permanent salt lake. They rarely dry up and consist of relatively deep saline water, up to 10 meters, on top of a substance with low permeability. The precipitation of salts occurs mainly from freeze-out; precipitated solids commonly include natron, mirabilite, and hydrohalite. The second type is the highly porous salt crust, which is often found in the center of the basin. The pores are usually full of interstitial brine, while the dominant solids are gypsum and halite. This type represents a more complete stage of evaporation where fluid inflow is minimal. The third type is the playa, which is found at the central depression of the basin. A confined aquifer is saturated by the reduced interstitial brine, commonly a few meters deep. A very shallow pool of water, spring-fed and on the order of a few centimeters to a few decimeters deep, often lies on top of the muddy lacustrine sediments. This surface pool usually dries up annually during the dry season and is not directly related to the brine underneath. If it is above the water table, the brine is affected by oxidation and evaporation. Capillary action draws up the brine into the sediments, and can precipitate gypsum, mirabilite, and ulexite; these minerals can often be found in irregular, unzoned lenses. The fourth type is an exposed unconfined aquifer. These occur when unconsolidated deposits of gravel and sand fill a basin and the water table is at or higher than the topographic surface; a very shallow lake is produced. Salts freeze out or evaporate. Salar Chemistry There are two classes of solutes in the Andean salars: those derived from alternation of volcanic rocks, which produce dilute inflows, and those resulting from brine recycling, which produce brackish water. These waters have historically been referred to as ALT (alteration) and EVA (evaporite leaching) waters, respectively (Risacher and Fritz 1990). Risacher and Fritz (1990, 2009) compared the composition of these dilute inflows to the composition of water in North America that was affected only by the weathering of andesitic to rhyolitic rocks, similar to the igneous petrology of the Andes (White et al. 1963, 1980 as qtd. in Risacher and Fritz 1990, 2009). The similarity of the Bolivian dilute inflows to the North American waters, as illustrated in Table 3, suggests that the Bolivian inflows result only from the alteration of volcanic rocks, without the involvement of hydrothermal activity or evaporite dissolution. However, Chilean dilute inflows are more concentrated than their Bolivian counterparts. The increasing proportion of concentration is not the same for all solutes; for example, silica is enriched only 1.7 times more than in Bolivian brines, but sulfate is enriched 8.1 times. The source of the extra sulfate is unlikely to be from the dissolution of gypsum and anhydrite, because calcium would be enriched by approximately the same amount of sulfate. The more likely explanation is the oxidation of sulfur in volcanic glass, as Chilean volcanoes are more sulfuric than their Bolivian counterparts (Risacher and Fritz 2009): S + 1.5O2 + H2O → SO4 + 2H The brackish inflows were once thought to result from the dissolution of halite in ancient salars or from the mixing of waters with brines (Risacher and Fritz 1991). However, it is more likely that brine mixing is the dominant process operating to produce the brackish inflows. While there are large deposits of Neogene evaporites in the Andes that consist mainly of halite and gypsum, the outcrops have very low concentrations of bromine (Pueyo et al. 2001). If the evaporite deposits were the main source of solutes for the brackish waters, the low level of bromine would be preserved in the inflow waters, but this phenomenon is not observed. Instead, the concentration of bromine in the brackish inflows is consistent with dilute inflows mixing with brines. The origin of these unknown, underground brines that mix with dilute waters to produce the brackish inflows is unknown. Two hypotheses seem possible for the origin of the brines: The first explanation is that the source is ancient salars trapped underneath volcanic formations. The climate of the Altiplano has remained arid to semi-arid since the Miocene, concurrent with volcanic activity. Under the Principle of Uniformitarianism, salars probably existed in the Altiplano throughout this period of aridity. These ancient salars could consequently covered by lava and pyroclastic flows. The brines would then be released and would be available to mix with dilute inflows; however, only very large salars could supply the amount of brine required, as smaller volumes would be quickly exhausted (Risacher and Fritz 2009). The other major problem with this explanation is that the high heat from the volcanic activity would have vaporized surface water; only brines protected by sediments could be preserved (Risacher et al. 2003). The more likely source is currently existing salars. Because the bottoms of the vast majority of salars are at least somewhat permeable, the salars are in dynamic equilibrium with their surroundings. Leaked solutes are recycled within the same salar or in adjacent ones; this flow may be driven by high heat flow. Outflux of leaking brine is compensated by the influx of brackish waters, which keeps the concentration of solutes relatively constant (Risacher and Fritz 2009). In this scenario, steady state can be attained in two different scenarios depending upon the permeability of the bottom sediments. If the bottom is impermeable, the concentration of conservative solutes (bromine and lithium) will increase, reducing the brine’s rate of evaporation. At a high enough concentration, the vapor pressure of the brine will balance the relative humidity. However, if the bottom is permeable, steady state is achieved when inflow flux balances outflow flux (Risacher et al. 2003). One way to determine if a lake is at steady state is to take the ratio of a component’s total mass in the lake to its annual input; the result has the dimensions of time. If the age of the last event that could have perturbed equilibrium (be it climatic, geologic, or anthropogenic) is known and is greater than this ratio, then the ratio is the residence time of the component in the lake. Residence time and infiltration rate are inversely related (Risacher et al. 2003). The annual flux and concentration of the solutes in the inflow is equal to the annual leakage of waters and the concentration of solutes in the lake in this salt balance. Ancient trapped salars likely have provided solutes in the past, but their supply has been exhausted. Recycling of brines between and within existing salars accounts for the production of brackish inflows because these brines mix with dilute waters. . Brine Evolution Brine evolution is modeled after Hardie and Eugster (1970, 1978; as qtd. in Risacher et al 2003): during evaporation, the concentration of solutes in the brine increases, and the minerals that precipitate out do so in order of increasing solubility. The concept of the chemical divide arises because the ionic activity product must be equal to the solubility product, so when a mineral precipitates, the concentration of all the solutes cannot increase at once. As evaporation continues, the solution becomes enriched in some solutes and depleted in others depending upon the ratio of solute concentrations at the beginning and the minerals that precipitate. There are two methods to determine brine evolution. The first approach is qualitative; it determines the changing composition of the evaporating solution step-bystep. Its focus is on the pathway, not on the solute composition. The second approach is quantitative; it focuses on the composition of the solution as it evaporates. Risacher and Fritz (2009) use the simulation code EQL/EVP, which is based upon the ion-interaction model and calculates the composition of an evaporating solution step-by-step. Another method is based upon Al-Droubi et al (1980; as qtd. in Risacher et al. 2003), which is valid for solutions with carbonate and silicate species. Throughout the discussion of these methods, the square brackets refer to total concentrations in mol/l or mmol/l. The Chemical Divide Model While there can be dozens of minerals in an evaporating basin, only a handful of them control the evaporative pathways. Therefore, these minerals are the focus of analysis of brine evolution, as shown in the qualitative modeling approach summarized here (after Risacher and Fritz 2009). Sodium and chlorine dominate most brines in terms of concentration, but they do not affect the evaporation pathways because both solutes are conservative until extreme stages of evaporation. Therefore, brine evolution is characterized by calcium, sulfate, and carbonate, even though these ions are less concentrated in solution. This method is not meant to actually predict the evaporative evolution that creates a brine, but is instead meant as an educational tool. The first mineral to precipitate out of a solution is usually calcite, CaCO3: [Ca] * [CO3] = Kcalcite Because the ion activity product must remain constant, calcium and carbonate cannot increase simultaneously; instead, the concentration of one ion will increase and the other will decrease. This is the first chemical divide and it determines two pathways: the alkaline path (carbonate increases while calcium decreases) and the neutral path (calcium increases and carbonate decreases). Which pathway a brine will follow depends on the relative concentrations of calcium and carbonate in the water. If the brine follows the neutral pathway after the first chemical divide, the concentration of calcium increases and the next mineral to precipitate is gypsum, CaSO4·2H2O: [Ca] * [SO4] * [H2O] = Kgypsum Once again, the concentrations of calcium and sulfate ions cannot both increase at the same time. This represents the second chemical divide for the neutral pathway: the two possible pathways are the calcium-rich, sulfate-poor path and the calcium-poor, sulfaterich path. Because some of the initial concentration of calcium was used up in precipitating calcite, the pathway the brine takes after the second chemical divide depends on the concentrations of calcium and sulfate at the beginning of gypsum precipitation and not on the initial concentrations of the brine. The sulfate-rich pathway produces sulfate-rich (Na-SO4-Cl) brines; the calcium-rich pathway produces calciumrich (Na-Ca-Cl) brines. If, however, the brine follows the alkaline pathway after the first chemical divide, pH controls the precipitation of magnesium salts, which can be either carbonates or silicates: Mg + nH4SiO4  Mg-silicates + 2H 2H + CO3  H2O + CO2 Mg + CO3  MgCO3 Because both possibilities utilize carbonate, it is possible to reverse the alkaline path to the neutral path if the decrease of carbonate is sufficient. These magnesium salts control the second chemical divide for the alkaline pathway: one produces Na-CO3-Cl brines and the other produces Na-SO4-Cl brines. The Alkalinity Approach The alkalinity approach (Al-Droubi et al. 1980, as qtd. in Risacher et al. 2003) to brine evolution is more useful than the concept of the chemical divide in predicting pathway, and can be performed without computer modeling. The same general concept is the same as the above method, but the alkalinity approach is more rigorous. The total alkalinity of a solution is: Alkalinity = 2[CO3] + [HCO3] + [OH] + [B(OH)4] – [H] Because all of these terms are interdependent and therefore difficult to manipulate, this operational definition of alkalinity is combined with the electro-neutrality equation. Alkalinity is therefore defined as the difference between the conjugate cations of the strong bases and the conjugate anions of the strong acids: Alkalinity = [Na] + [K] + 2[Ca] + 2[Mg] – [Cl] – 2[SO4] When calcite precipitates, this equation can be rewritten: Alkalinity – 2[Ca] = [Na] + [K] + 2[Mg] – [Cl] – 2[SO4] As long as calcite is the only mineral that precipitates, the concentrations of the solutes on the right hand side of the above equation increase linearly with the concentrating factor, F, of the solution that is evaporating. Using sodium as an example and with the subscript 0 symbolizing the concentration of the initial solution, [Na] = F * [Na]0 Where F = mole number of water in initial solution / mole number of water in solution, such that: F = (H2O)0 / (H20) Therefore, the alkalinity equations for calcite precipitation can be rewritten as: Alkalinity – 2[Ca] = F * {[Na]0 + [K]0 + 2[Mg]0 – [Cl]0 – 2[SO4]0} Which can be simplified to: Alkalinity – 2[Ca] = F * (Alkalinity0 – 2[Ca]0) If the initial solution has more alkalinity than twice its calcium concentration (Alkalinity0 > 2[Ca]0), as the solution evaporates, the difference between alkalinity and calcium concentration will increase and alkalinity will dominate. If the initial solution has less alkalinity than twice its calcium concentration, then calcium will dominate alkalinity as evaporation continues. Comparing alkalinity to twice the calcium concentration will therefore allow determination of what pathway a solution will follow after this first chemical divide. Magnesium silicates or carbonates precipitate in the early stages of evaporation like calcite: if both calcite and magnesium minerals are precipitating, the alkalinity equation is rewritten: Alkalinity – 2[Ca] – 2[Mg] = [Na] + [K] – [Cl] – 2[SO4] Alkalinity – 2[Ca] – 2[Mg] = F * {[Na]0 + [K]0 – [Cl]0 – 2[SO4]0} If the initial solution as more alkalinity than twice its calcium and magnesium concentrations (in other words, if the left hand side of the above equation is positive), then the evaporating water becomes an alkaline brine enriched in carbonates and depleted in calcium and magnesium. If the opposite is true, the solution will follow the neutral pathway and will be depleted in carbonate. If alkalinity is greater than twice the calcium concentration but less than twice the calcium and magnesium concentrations, it will follow the alkaline pathway while calcite is precipitating and will then follow the neutral path once magnesium salts begin to precipitate. If the solution is following the neutral path, it will become more and more enriched with calcium and may precipitate gypsum. The precipitation of gypsum is the second chemical divide in the evolution of a neutral brine; either brines that are calciumrich/sulfate-poor or calcium-poor/sulfate-rich are produced according to the ratio of calcium to sulfate at the beginning of gypsum precipitation, not of the initial solution. Figure 6 provides a schematic to summarize the brine evolution pathways. The brine evolution process results in three major groups of brines: alkaline, sulfate-rich, and calcium-rich. The alkaline brines are Na/HCO3 – CO3 – Cl and follow the alkaline I – IA pathway. The sulfate-rich brines are Na/SO4 – Cl and follow either the sulfate-alkaline I – IIA – III pathway or the sulfate-neutral II – III pathway. The calciumrich brines are Na – Ca/Cl and follow the calcic pathway II – IV. While theoretically possible, the pathways I – IIA – IV and II – (Na/CO3-Cl) have not been observed in any natural evaporating solution (Risacher et al. 2003). It should be apparent from this analysis that even a small variability in calcium, magnesium, or alkalinity in the initial water can change the resulting brine dramatically (Risacher et al 2003). Because the solute concentration in the dilute water is mainly due to rock alteration, the lithology of the region is the dominant control in brine evolution and fate. Moderately mineralized igneous rocks that are weathered tend to produce waters that follow the sulfate alkaline pathway; highly mineralized igneous rocks’ waters follow the sulfate-neutral pathway. Sedimentary rocks, which usually contain high levels of calcium, tend to produce calcic brines (Risacher et al. 2003). Additional Variables in Brine Evolution The brine evolution model described above operates under the assumption that the evaporating basin is essentially a closed system that is not affected by other environmental factors. Risacher et al. explain that the discrepancies between the brines predicted by the model and the brines that are actually observed in the Altiplano are due to these other variables that affect the evaporation pathway. These variables have the most impact if they affect a solution in the early stages of evaporation (Risacher et al. 2003). For example, wind-blown dust and salts may enter the water from the atmosphere, potassium can partake in exchange reactions with clay minerals, and sulfate concentrations can decrease from bacterial reduction. All of these scenarios have the potential to change the evaporation pathway a solution will follow. Additionally, inflow waters that are following different evaporation pathways may mix in a large basin with complex lithology (Risacher et al. 2003). These “disturbing forces” account for the reason why Chile only has one alkaline salar out of more than 200 observed. This unexpected deficiency is due to the high concentrations of native sulfur in the Western Cordillera and the deposition of gypsumrich dust. The sulfur is oxidized and thereby acidifies the water, reducing its carbonate concentration significantly. The gypsum enriches the waters with calcium, but because of calcite’s low solubility, carbonate concentrations decrease. In summary, South American salars occupy the internal drainage basins of the Altiplano where evaporation exceeds precipitation. The lithium in these salars that makes them of economic interest originates from the weathering of volcanic rocks. While the variety of salars in terms of their size, chemistry, and morphology is astounding, some classification is possible. It is also possible to predict the type of brines produced and the evaporation pathway followed by a solution. All the information that is required is the alkalinity and the concentrations of solutes like calcium, magnesium, sulfate, and carbonate; however, environmental contamination may change the composition of a solution significantly and thus change the type of brine it produces. Extraction and Beneficiation Hard Rock In 1978, Averill and Olson argued that extraction technology for pegmatites was well-developed, and further research and development would not be cost-effective; however, researching methods of extraction from clays and, more importantly, brines should be a priority. They also acknowledged that currently available reduction and refining methods may not be suitable for large-scale operations. At the time of their paper, hard rock mining for lithium dominated the supply. Pegmatite ore bodies are the most common type of ore deposit for lithium; the lithium is most commonly found in the minerals lepidolite, petalite, and spodumene. Lithium from pegmatites is recovered and concentrated by froth flotation, hydrometallurgical extraction, and precipitation from aqueous solution (Averill and Olson 1978). It is crushed to finer than 0.3 mm, cleared with caustic and sometimes sodium sulfide, and then conditioned with oleic acid, a collector (Figure 7). To extract lithium from spodumene (which is usually 2% lithium), one of two processes is used. The first process, called the acid process, uses heat to convert spodumene from its alpha to beta structure, which is then ground to 0.15 mm and treated with sulfuric acid. With the addition of heat, the lithium becomes lithium sulfate, which is soluble. Magnesium, calcium, aluminum, and iron are removed before the lithium is precipitated with sodium carbonate. The second process, called the alkaline process, requires heating spodumene or lepidolite with limestone. The lithium silicates become lithium hydroxides after leaching the lithium aluminates. This hydroxide solution is processed with evaporators, which crystallizes lithium hydroxide monohydrate. Lithium chloride is used as the source material for the production of lithium metal in a process similar to the production of sodium: through molten salt electrolysis. It is electrolytically reduced from a fused salt that is a mix of potassium chloride and lithium chloride; this is a low volume and high energy process. A steel box in a refractory-lined fire box is heated to 500 °C, and is typically 3 cubic meters in size. Lithium metal is reduced at the steel cathodes and chlorine is oxidized at the graphite anodes, as shown in the following reactions: Anode reaction: 2Cl → Cl2 + 2e Cathode reaction: Li + e → Li Five kilograms of chlorine gas are produced for each kilogram of lithium metal, which presents an environmental and worker health challenge; it also makes fire control equipment mandatory. This process is also extremely energy-intensive: one kilogram of lithium metal requires about 46 kWh of energy, not including heating. Processing minerals from hard rock sources also has high transport costs and because the deposits tend to be relatively small, hard rock production is not often able to take advantage of economies of scale (Ebensperger et al. 2005). Two decades later, little progress had been made towards reducing the environmental impact and increasing the production of lithium metal, which is used primarily as the anode in lithium ion batteries and in alloys with magnesium and aluminum (Kipouros and Sadoway 1998). The electrolytic reduction process now requires a central mild steel cathode with opposing graphite plates, which serve as the anode. A bell-like structure prevents the mixing of the liquid lithium and chlorine gas. Anhydrous lithium chloride is the source of lithium, and potassium chloride is used as the solvent. Potential ways to reduce the environmental impact of this process include replacing electrode materials; the authors suggest titanium diboride as a replacement for graphite. This material has been proved in the laboratory setting and is now being used in a few industrial applications. There are several emerging technologies that may provide alternatives to the currently-used process, although all are economically unviable currently (Kipouros and Sadoway 1998). Brines While oceans contain 0.2 ppm lithium and some clays, such as hectorite, may contain appreciable amounts of lithium, exploiting these potential sources is both technologically and economically infeasible at present, although increasing demand and value may spur further exploration in this area. Therefore, brines are the only true alternative to hard rock mining. Production of lithium from brines has an inherently smaller environmental impact than hardrock mining, although lithium brines also require extensive processing. In addition to less impact, brine extraction is also less expensive, and is increasingly driving pegmatite producers out of competition (Warren 2010). In 1978, Averill and Olson argued that the two factors that were the dominant controls on lithium production from brines were the grade of the brine and the concentration of calcium and magnesium. Lithium has traditionally been recovered using solar evaporation or flotation, which is ineffective for dilute brines. Averill and Olson suggested ion exchange or liquid-liquid extraction to concentrate lithium from more dilute brines and to reduce concentrations of calcium and magnesium. When concentrations of calcium and magnesium are low, operators have traditionally treated the brine with lime to precipitate magnesium. The brines can be evaporated to recover commodities such as potash, borax, salt cake, soda ash, and dilithium sodium phosphate. Froth flotation is utilized to extract the phosphate from these materials; the dilithium sodium phosphate is treated with sulfuric acid and sodium carbonate to recover the lithium, as was done in the 1970s in Searles Lake, California (Averill and Olson 1978). Brines that contain enough lithium to be economically exploited are mostly found in arid, high altitude, continental volcanogenic terrains (Warren 2010). Beyond concentration of lithium in the brines and the chemical constraints of calcium and magnesium contamination, there is a major physical constraint on brine extraction: the porosity of the salt crust. As depth of the host aquifer increases, effective porosity and permeability approach zero at about 50 m depth. Because of this constraint, there is a limit to economic brine recovery to these shallower regions (Figure 8). Current Brine Operations Different operators use different methods to produce lithium. In the Salar de Hombre Muerto in Argentina, FMC uses a proprietary alumina adsorption system to extract lithium directly from the brine. The Salar de Atacama in Chile produces lithium carbonate from solar evaporation ponds. This facility is able to produce lithium carbonate with 90 ha of evaporation ponds, which is only 1/20 of the area required at Clayton Valley, Nevada due to the extreme aridity of the climate (Warren 2010). Zabuye Salt Lake and Dabuxum/East Taijinier Lakes in China both produce lithium carbonate from extraction and solar evaporation. In the US, the plant at Searles Lake was in operation from 1961 – 1978 and produced lithium carbonate as a byproduct of salt cake and soda ash production, although the source is likely too depleted to become viable again; Clayton Valley in Nevada may be viable (Warren 2010). The Salar de Uyuni in Bolivia, which contains almost half of the world’s reserve base of lithium (Jaskula 2010) may come on line in the next decade. However, production from this salar is currently limited to 40 metric tonnes each month from a state-run pilot plant; it also has very high Mg:Li ratios and therefore would require pretreatment with calcium hydroxide before evaporation. High Mg:Li ratios drive up the price of production (Figure 9). The lithium industry’s moment of truth was in 1997, when Sociedad Quimica y Minera de Chila S.A. (SQM) began to produce lithium chloride from the Salar de Atacama and processed this material into lithium carbonate in Antofagasta. SQM capitalized on the fact that brine extraction was high volume and low cost, and it managed to drive down the market price of lithium by 50% in 1998 (Table 4). Facilities in the US, Russia, Australia, Argentina, and China reduced production or shut down entirely; they produced lithium hydroxide from spodumene, petalite, and lepidolite, which is simply not cost effective when competing with brine extraction (Ebensperger et al. 2005). This is the most recent paper summarizing the state of the industry. Since 2002, the companies producing lithium from brines are SQM in Chile, Chemetall (in Chile and the US), and the FMC Corporation in Argentina. In China, the China Xinjuang Nonferrous Metals Corporation of Mingyuan produces lithium carbonate from brines as well as from domestic and imported ores. Chemetall and SQM accounted for 75% of the market in 2002. It is important to note that while the Salar de Atacama produced 60% of the world’s lithium in 2003, the total value of the industry to Chile is just over 1% of the revenues they receive from copper mining. It is also interesting to note that the industry is highly concentrated, as four companies produce 90% of the world’s supply of lithium (Ebensperger et al. 2005). Chile is likely to continue to dominate the world’s supply of lithium for decades, at least while the Salar de Uyuni remains essentially unexploited. It therefore seems likely that the Chilean government will find that promoting sustainable development will be a priority in this century. The traditional way that governments “share” the benefits of a mining project are through company taxation and royalty payments, which may affect the operations of SQM depending upon the political climate in the country (Ebensperger et al., 2005). The demand for lithium brines will likely be driven by increased use of lithium-ion batteries in technologies such as electric and hybrid vehicles, as well as in consumer electronics (Yaksic and Tilton 2009). Lithium Batteries Introduction As part of the American Recovery and Reinvestment Act of 2009, the US Department of Energy gave lithium battery suppliers and manufacturers a $930 million grant to promote research on lithium-ion battery technology (Jaskula 2010); an additional $10 million was granted to Toxco, the recycling company planning to build the first lithium battery recycling facility in the United States (Hamilton 2009). This staggering amount was a part of a $2.4 billion dollar grant to develop the American capacity for electric drive vehicles through both manufacturing and deployment. This sum, combined with the $25 billion of direct loans as part of the Advanced Technology Vehicles Manufacturing Incentive Program in 2008, illustrates the ever-increasing importance of lithium, which is apparent even to politicians. This governmental commitment is a reflection of recent trends: the amount of lithium consumed for battery manufacturing has increased by 20% each year during the latter half of the decade. In 2008, lithium ion batteries accounted for 70% of the global rechargeable battery market, which was worth $7.4 billion that year (Jaskula 2010). As concerns over America’s dependence on foreign oil and the consequences of climate change grow more serious, consumers, politicians, and scientists are all looking for easy solutions. Lithium-ion batteries, which may be ideal for electric, hybrid electric, and plug-in hybrid vehicles, are becoming an increasingly attractive technology. Types of Lithium Batteries Lithium is suitable for batteries because of its physical properties: it is the most electropositive metal (Table 5). It is also the lightest metal (MW = 6.84 g/mol and  = 0.53 g/cm); therefore, it has the highest energy density (Figure 10). This combination of power and low weight makes it ideal for applications in which portability or mobility is a top priority; an excellent example is electric vehicles. Lithium has been used to make batteries for decades; consequently, there are dozens of types of lithium batteries. However, all lithium batteries fall into three general categories: lithium anode batteries, lithium ion batteries, and lithium air batteries. All batteries have the same basic components: an anode, a cathode, and an electrolyte solution containing dissolved salts. During discharge – when the battery is converting its chemical energy into electrical energy – ions move from the negative electrode (anode) to the positive electrode (cathode). Oxidation occurs at the anode, and reduction occurs at the cathode. The difference between batteries is the type of material that is used in each of these three components and their configuration. Electrolytes can act as a buffer for ion flow or can participate in the reaction (Brain and Bryant 2000). Lithium anode batteries, or lithium batteries, were the earliest lithium-based batteries to be developed and commercialized. They are primary batteries; primary batteries are not rechargeable because one or more of their electrodes is depleted as the battery is used. Lithium batteries utilize lithium metal or lithium compounds as anodes. Because lithium is the most electropositive metal, these batteries produce roughly double the voltage of traditional alkaline batteries. They are often used in smaller consumer devices such as clocks, calculators, and cameras. Lithium batteries have a lifespan of 15 or more years, and thus are often used in implanted medical devices such as pacemakers (Protomatic 2010). The second class of lithium batteries is lithium-ion batteries, which are rechargeable (also called secondary batteries). When they are charging, lithium ions move from the positive to the negative electrode, which is the reverse of the discharge reaction. Lithium-ion batteries have many advantages over older battery technology. Lithium-ion batteries can store the same amount of energy as a lead-acid battery with one-sixth of the mass, they lose only 5% of their charge each month (versus 20% for nickel-metal hydride batteries), and they can withstand hundreds of cycles of charging and discharging. However, many lithium-ion batteries currently on the market have a lifespan of less than five years and have been known to ignite (Brain 2006). The rechargeable nature of lithium-ion batteries is due to intercalation compounds, which let lithium to move into the anode or cathode. Intercalation compounds allow the reversible insertion of a molecule between two other molecules; several inorganic compounds allow alkali metals to react reversibly in this manner (Tarascon and Armand 2001). During discharge, lithium ions move from the anode into the cathode; during charging, lithium ions move from the cathode into the anode. There are many different types of lithium-ion batteries that use different materials for the electrodes and electrolyte. Cathode material is usually a lithium oxide, such as LiCoO2 or LiMn2O4; the anode is often graphite; and the electrolyte consists of lithium salts, such as LiClO4, in an organic solvent. Lithium-ion batteries therefore have extensive material demands beyond that for lithium – for example, batteries account for 25% of global cobalt demand. Using recycled cobalt and nickel in lithium-ion batteries represents a 51.3% savings in natural resources and a 45.3% savings in fossil fuel requirements (Dewulf et al. 2010). Lithium air batteries, or lithium metal-air batteries, are a developing technology that may be feasible within the decade (Luoma 2009). If they are successful, they will have the energy density of gasoline. They utilize a lithium anode and a porous carbonate cathode, in which oxygen molecules are reduced by lithium ions during discharge to form lithium oxide or lithium peroxide. The electrolyte has been demonstrated with a gel polymer (Abraham and Jiang 1996). One of the current research challenges is finding a membrane that allows oxygen in but keeps moisture out, since lithium can ignite when exposed to water. Battery Recycling As lithium batteries become more widely used, they will become a larger proportion and quantity of the waste stream. It is important that waste management practices, whether they be landfilling or recycling, must be safe as well as economical. Fortunately, the lack of metallic lithium in lithium-ion batteries gives recyclers more flexibility in cell disassembly and processing (Lain 2001). It is estimated that up to 98% of lithium can be recovered from battery recycling (Jungst 1999). There are two recycling processes that are currently in use for the recycling of lithium-ion batteries. The first type, the so-called Toxco process, can be used to process any type of lithium waste products. The material is first cooled in liquid nitrogen and is then shredded and mixed with water. This process usually produces lithium hydroxide. The second type, the so-called Sony process, incinerates the cells. While lithium is lost, cobalt can be recovered (Lain 2001). A process that utilized emerging technology is being developed at AEA Technology Batteries in the UK and is effective for lithium-ion batteries that use cobalt. First, the batteries are shredded mechanically without being exposed to water, with which lithium violently reacts. The electrolyte is separated from the solids with a suitable solvent, which is then evaporated away at a reduced pressure. The boiling point at this reduced pressure must be below the decomposition temperature of lithium (80 °C). A different solvent is used to recover pieces of the electrode (Lain 2001). The residual electrode particles are lithium cobalt oxide, which is electrochemically reduced in the