HYDROMETALLURGICAL PROCESSES FOR THE RECOVERY OF LITHIUM FROM SILICATES

by
Adrian Griffin
Lithium Australia, Australia

Presenter and corresponding author

Adrian Griffin
adrian.griffin@lithium-au.com

ABSTRACT

Lithium Australia NL has been developing a number of hydrometallurgical process flowsheets for the recovery of lithium from silicates. The most advanced of these, the halogen-based SiLeach™ process, was developed specifically for the digestion of spodumene, a refractory lithium pyroxene previously able to be processed only by roasting, followed by sulphation bake and water leach. The SiLeach™ process, which relies on the reaction of halogens with silicon-oxygen bonds, is widely applicable to the dissolution of all categories of silicate minerals and has found practical application in the processing of lithium micas. SiLeach™ is currently being commercialised, utilising pilot facilities established at ANSTO Minerals (a division of the Australian Nuclear Science and Technology Organisation).

With SiLeach™, the halogens can be added to the process slurry in a number of ways. However, the preferred method is via the addition of ground fluorine minerals (which may include the lithium micas themselves) to the process slurry prior to the addition of sulphuric acid. Due to the kinetics of competing reactions, this sequence allows the momentary generation of fluoride in solution – and its almost instantaneous reaction with the silicates – without any accumulation of hydrogen fluoride (‘HF’) in the slurry. Process plant operations can be accomplished without the hazards often considered a risk with fluorine-based processes.

The SiLeach™ process is capable of recovering a wide range of by-products, as all metals in the target silicate are dissolved. The low energy requirements of the process, as well as by-product credits, enhance the economics, which may result in lithium production costs using SiLeach™ to be among the lowest currently experienced.

SiLeach™ has the potential to provide access to a wide range of plant feed previously not considered viable. Such feed includes low-grade spodumene concentrates and micas as primary feed sources. Furthermore, vast quantities of lithium minerals currently discharged as tailings from non-lithium mining operations create very attractive targets for the application of this type of technology. Utilisation of these materials as lithium sources will greatly enhance the sustainability of the lithium industry.

Keywords: SiLeach™, lithium, spodumene, mica, halogen, fluorine, hydrometallurgy

 INTRODUCTION

There are two historic sources of lithium: hard rock and brine occurrences. Each presently accounts for about half of global lithium demand (USGS, 2017). Lithium-bearing ‘clays’ are also a promising source, with a number currently under investigation for future production (Ausenco, 2016).

Regardless of type, all commercially significant lithium sources have a common provenance – they result from magmas that have accumulated a high concentration of volatile elements. Those volatile elements, including lithium, reach the near-surface, exploitable accumulations by:

  • emplacement of pegmatites;
  • development of greisens;
  • degassing of aqueous components, or
  • generation of hydrothermal fluids.

Very specific tectonic conditions are required to provide an environment in which volatile components and lithium accumulate. It is the chemistry of the last stages of volatile magma emplacement that provide many of the keys to dissolution of silicate minerals under commercial conditions. Indeed, due to the unique circumstances encountered in these intrusions, which invariably solidify at shallow crustal levels or generate violent explosive volcanic events, natural fluxes and solvents accumulate in abundance.

High quantities of water in these melts depress the liquidus and high pressures (relative to atmospheric) prevent the water separating as a single phase – such conditions are beyond the critical point of water. Other accumulated fluxing elements further depress the solidus, in particular lithium (‘Li’), boron and phosphorus. As the magmas rise and pressures reduce in these melts, the distinction between magmatic and hydrothermal fluid blurs.

These final fluid phases are of great interest, since their behaviour demonstrates the possibility of recreating similar conditions within commercial processing systems that can reverse the natural sequence and take minerals back into solution. The SiLeach™ process has achieved this reversal; as such, it can be considered a digestion process rather than a simple leach.

THE PYROMETALLURGICAL/HYDROMETALLURGICAL TRANSITION

In terms of extractive metallurgy, the definitions of ‘pyro’ and ‘hydro’ are quite distinct, but within the bowels of the earth the processes cross – there is no distinct boundary: the extreme pressures can generate a process continuum. However, other factors are also at work. The fluxing characteristic of many of the ‘incompatibles’ results in hydrous melts of specific chemistry with very low melting points. At the extreme, the water-saturated melts approach the operating conditions that can be achieved in mineral processing autoclaves. That being the case, the phase chemistry of these late-stage magmatic products offer some keys as to how the processes may be reversed, exploiting the ability of metals to return to an aqueous phase.

Late-stage magmas, and magmatic fluids accumulate incompatible elements – those that don’t readily fit into the lattice of the common rock-forming minerals. Some are too big to fit conveniently into most silicates and get a ‘free ride’ into the late-stage fluids, particularly as temperatures and pressures drop below the critical point of water, which decompresses to form an aggressive solvent that harbours the incompatible elements.

Other elements, such as fluorine, generate such a depression of the solidus that they become a self-fulfilling component – one that remains in a molten state (morphing into the late-stage aqueous fluids) because they act as fluxing agents and their ever-increasing concentration creates an ever-reducing solidus temperature. As fractional crystallization in the magma chamber advances, the fluid phase composition trends towards a complex eutectic point, while the remaining magma has a relatively low viscosity. High-level intrusives solidifying within a few kilometres of the earth’s surface may transgress the critical point of water, explosively propelling the remaining low-viscosity fluids into brittle fractures of the surrounding country rock. This is how pegmatites are formed.

So, it is the behaviour of water and the fluxing agents that provides the key to reversing the process in a synthetic, controlled environment, resulting in the dissolution of silicate minerals at low temperatures and pressures.

VULNERABILITY OF SILICATES TO HYDROTHERMAL PROCESSES

Examination of naturally occurring, low-temperature magmatic processes may provide empirical evidence as to which minerals are best suited to attack by hydrometallurgical processes, as well as the conditions in which that attack may be effectively controlled.

TAKING NATURE ONE STEP FURTHER

The dissolution of ore minerals in sulphuric acid systems is commonplace. Sulphides are oxidised at atmospheric or low-pressure conditions, to a large extent mimicking weathering processes. Taking these natural processes and exaggerating the corrosion conditions in a synthetic system may amplify the kinetics – and this amplification may be great enough to render the process of metal extraction a commercial reality within an acceptable residence time.

High-pressure acid leach processes employed in nickel extraction reduce the dissolution and deportment of the chemical components from millions of years to hours or even minutes. The natural secrets of these chemical processes remain preserved in the lateritic profile for the astute student to observe.

The dissolution of silicates may be considered a reversal of the crystallization processes preserved in the geological record as pegmatites and greisens. To artificially reverse such processes, water must be introduced into the system and other volatiles/fluxes added to the mineral slurry. The halogens, principally fluorine (‘F’) and chlorine (‘Cl’), play an important role in the natural systems. Indeed, at elevated temperatures and pressures, Cl may well be a significant factor in transporting other elemental species. At lower temperatures and pressures, the dominant reactive halogen is F, as recorded in the composition of mineral species – mica in particular – as crystallization proceeds.

Increased concentrations of F in late-stage pegmatite fluids often result in fluorite (‘CaF2’) overgrowths on earlier F-bearing phases such as topaz or tourmaline, which both crystallize earlier than CaF2 and contain significantly less F than CaF2. This attests to the rising concentration of this important halogen in the final stages of pegmatite crystallization. In many cases, the miarolitic cavities within which the CaF2 crystalizes are silent witnesses to the relatively low-pressure conditions under which the pregnant F liquors may carry soluble components. Jacobson (2013) summarised many of these occurrences as follows.

The anorogenicgranites formed from the melting of lower crustal rocks during crustal thinning are due to tension, not compression related to continental collisions or subduction. The miarolitic, vuggy character of these granites is due to the upper part of the magma rising to a shallow emplacement in the crust. This shallower, lower pressure environment allowed water to exsolve from the magma as the pressure decreased and to form bubble-like pegmatite cavities. Almost any NYF [niobium yttrium fluorine] granite with miarolitic pegmatites may contain attractive fluorite crystals.

Although the miarolitic pegmatite of the NYF association is a very specific example, it does serve to demonstrate that aqueous F solutions are the medium for mass transfer in natural systems with temperatures as low as 200 degrees Centigrade and pressures of 200-250 megapascals (London et al, 2012).

Without resorting to the pressures found in geological systems, the activity of F can be harnessed to break the silicon-oxygen (‘Si=O’) bond (Kuang et al, 2012), the very building block of silicate minerals. Moderate concentrations of F are required in solution, as is elevated temperature. Lithium Australia NL (‘Lithium Australia’) has been developing its 100 per cent owned SiLeach™ process to achieve that goal.

CONVENTIONAL PROCESSING

Spodumene has long been the only commercially processed hard-rock lithium mineral. Conventional processing relies on a phase conversion from low-temperature α-spodumene to the higher-temperature polymorph β-spodumene to enhance leach performance in sulphuric acid. While this process – which effectively rejects the gangue elements, leaving lithium as the principal revenue source – is energy intensive, its selectivity is an advantage, since it results in simple processing steps following the initial phase conversion. That said, lack of significant by-product credits is a commercial disadvantage with conventional processing of spodumene.

The cost of producing lithium carbonate(‘Li2CO3’) from a spodumene concentrate of 7 per cent lithium oxide a (‘Li2O’) is about US$4,500 per tonne (‘t’) (Roskill, 2016). It follows that the cost of producing Li2O using similar processes but with lithium mica at half the grade should be twice as much. Deutsche Bank (2016) has confirmed the veracity of this assumption, indicating that the operating costs of Chinese producers using a lithium mica feed approach US$8000/t (see Figure 1 – far right-hand bar).

Figure 1. Li2O cost curve, after Deutsche Bank (2016). If conventional processing technology is used, the lowest-cost production is from brines, followed by spodumene (and petalite) and, lastly, lithium micas.

DEVELOPMENT OF THE SILEACH™ PROCESS

Having studied the paragenesis of pegmatites and greisens, Lithium Australia made the following observations.

  • Many hard-rock lithium occurrences demonstrate crystallization of lithium minerals from magmatic phases with high levels of water and fluorine.
  • Many of the late-stage, relatively low temperature and pressure minerals contain F as a major constituent of the crystal lattice.
  • Li and F increase in the magma towards the eutectic, which arises from partial crystallization of fertile magmas.

A critical component in the genesis of lithium ore minerals is F chemistry and, if high temperatures and high pressures are to be avoided during the extraction of lithium from lithium minerals, that chemistry remains critical in the re-dissolution processes.

SiLeach™ is a broad-spectrum, halogen-based digestion process developed to recover metals from silicates without the requirement to roast. In fact, SiLeach™ is a reversal of the last stages of fertile magma crystallization. Lithium recovery is the principal aim of SiLeach™, although it may be commercially applied to other alkali metals. Aluminium (‘Al’) and silicon (‘Si’) are also available for the production of valuable by-products.

LITHIUM MINERAL ACTIVITY SERIES

Lithium Australia studied a number of both lithium and non-lithium minerals to determine digestion characteristics under comparable conditions using sulphuric acid at elevated temperature (see Figure 2).

Figure 2. Digestion curves for a number of lithium silicates in sulphuric acid at 90 degrees Centigrade. The micas show a range of reaction kinetics, whereas spodumene, a lithium pyroxene, is unreactive.

The fundamental difference between the reactivity of these mineral species is the F activity generated by dissolution of the minerals themselves in sulphuric acid. Zinnwaldite typically contains around 7 per cent F, trilithionite (lepidolite) 4 to 5 per cent F and muscovite around 1 per cent F. Spodumene contains no F in the mineral lattice. Alterations in the contained F within the mica have a profound influence on the reaction rates of micas in sulphuric acid, and changing the amount of F available to react with the silicates can modify the reaction curves to such an extent that it is possible to create conditions for the kinetically slower species to approach the reaction rate of zinnwaldite.

MICA REACTIONS

A mica’s reactivity depends on its structure and the way in which that structure presents active lattice sites, or exposed chemical bonds, to the lixiviant. Reactions that liberate metals from phyllosilicates include:

  • ion exchange;
  • protonation of anions, or
  • destruction/replacement of Si=O or other tetrahedrally coordinated oxygen bonds.

The mica group can be divided into dioctahedral and trioctahedral structures. The latter have more reactive sites per lattice unit that the dioctahedral micas and hence greater ion exchange capacity and greater reactivity in acids. How hydroxyl dipoles (OH) are oriented also influences mineral reactivity.

In strong sulphuric acid solutions, ion exchange reaction mechanisms are less important than the protonation of anions, OH and F in particular. These are only exposed on surfaces perpendicular to the minerals’ perfect cleavage. Indeed, when viewed perpendicular to the C-axis, micas expose only outward-facing oxygen atoms, being the base of silica tetrahedra. The exposed atoms, and the connecting Si=O bonds, are inert to sulphuric acid and consequently the reaction kinetics are controlled largely by the surfaces exposed perpendicular to the cleavage planes.

As the aspect ratio of exposed surfaces in micas (C:A or C:B crystallographic axes) is very large, and as the large surface area is dominated by oxygen at the termination of Si=O bonds, most of the exposed surfaces are unreactive. The abundance of the reactive OH- bonds is low, due to their exposure only on the relatively small surface areas exposed perpendicular to the cleavage. This results in long reaction times in sulphuric acid, as few reactive species, OH and F, are exposed to the attack of hydrogen ions (‘H+’). The reaction of the acid on the mica is generally slow but, as it progresses, it releases F from the lattice, which is then available to attach to the Si=O bonds that form the framework of the mica (see Figure 3).

Figure 3. Hydrogen ions attack micas only on the very limited exposed area perpendicular to cleavage, whereas fluoride ions can attack the vast surface area exposed on cleavage planes,
significantly improving reaction kinetics.

Thus, the rate of attack of F on Si=O bonds, and hence the rate of digestion of a mica in sulphuric acid, is determined by the amount of F the mica can release into solution.

If the amount of F in solution is increased by the addition from a source not derived from the mica, then the reaction kinetics can be changed. For example, the stoichiometric addition of F to lepidolite, resulting in the same concentrations that would be expected from zinnwaldite, will create a digestion rate for lepidolite that is similar to that of zinnwaldite in sulphuric acid.

Taken to the extreme, spodumene, a Li pyroxene, can also be made to react at a similar rate, and it is the control of these reaction rates that forms the basis of the SiLeach™ process (see Figure 4).

Figure 4. Addition of F to sulphuric acid can result in most silicate materials,
including spodumene, having similar dissolution curves.

  BENEFITS OF SILEACH™

SiLeach™ is a non-selective, halogen-based digestion system. By facilitating the destruction of Si=O bonds, it allows access to cations in the silicate lattice that release all metals into solution. Thus, there is the opportunity to recover metals other than lithium from minerals digested using the SiLeach™ process. In the case of the lithium micas, the framework of the lattice provides abundant Si and Al, whereas the octahedral sites harbour abundant Group 1 metals, among them potassium, Li, rubidium and caesium. To a lesser extent, sodium, manganese and iron may also be present in the lattice. The potential for abundant by-product credits, together with the elimination of the requirement to roast, are key elements of the SiLeach™ processing strategy.

While process flowsheet specifics vary with mineral species, the flowsheet for micas is a good example of the steps involved in the process (see Figure 5).

Figure 5.  SiLeach™ flowsheet for lithium mica plant feed

ALTERNATIVES TO SILEACH™

Clearly, conventional processing is an alternative to SiLeach™ but the energy cost is high and revenue is restricted largely to Li. In the case of micas, the release of fluorine, in the form of HF vapour, is a safety issue. Moreover, materials of construction and corrosion of the equipment are of practical concern.

Hydrometallurgical alternatives to SiLeach™ are few, and leaching in sulphuric acid alone is kinetically restrictive on all but the most reactive Li minerals.

Elevating temperature and pressure – the domain of platinum producers and lateritic nickel operations – is an alternative. The application of pressurised reaction vessels certainly presents another dimension to the processing of Li silicates, allowing the options of both acid and caustic systems. Lithium Australia, in association with ANSTO Minerals (a division of the Australian Nuclear Science and Technology Organisation) has been developing the LieNA™ process – a moderate temperature caustic leach that requires pressure containment. LieNA™, which is capable of digesting all the common Li minerals (including spodumene), allows for project-specific process optimisation, facilitating capitalization on unique factors such as local logistics, reagent costs and energy casts.

CONCLUSIONS

While the reactivity of most Li minerals in sulphuric acid is kinetically limited, the kinetics can be improved by adding fluorine to the processing system in a way that minimises the production of volatile F species – HF in particular – that would otherwise present an occupational health and safety risk. The operational perfection of F-assisted leaching is the essence of Lithium Australia’s 100 per cent owned SiLeach™ process.

Raising temperatures, and elevating pressures beyond atmospheric, provides the flexibility to digest Li silicates in either acidic or caustic environments. Caustic digestion is the foundation of the LieNA™ process, currently under development by Lithium Australia.

  ACKNOWLEDGEMENTS

The author wishes to acknowledge the contributions of Lithium Australia, its staff and consultants; in particular, work undertaken at ANSTO Minerals and Murdoch University and partially funded by two Innovation Connections Grants under the Entrepreneurs’ Programme of the Department of Industry, Innovation and Science, plus a grant provided by the Minerals Research Institute of Western Australia.

Pioneers of hydrometallurgical work such as Arthur J Moxham (1925) laid the foundations of much modern sulphuric acid process development and have been an inspiration to all. Enej Catovic led Lithium Australia team’s research, while Andrew Skalski, also at lithium Australia, has begun the task of transforming that research into a commercial reality.

Also instrumental in development of these processes have been Lithium Australia’s shareholders, not only as owners of the Company but as its financiers, and their support is also gratefully acknowledged.

  REFERENCES

  1. Ausenco: Technical Report on the Prefeasibility Study for the Sonora Lithium Project, Mexico (prepared for Baconora Minerals Ltd). April 2016. 186 p.
  2. Deutsche Bank Market Research: Lithium 101. May 2016. 177 p.
  3. Jacobson, M.I.: Fluorite in granitic pegmatites. Rocks and Minerals, 2013, Vol. 88 (2).
  4. Kuang, G., Chen Z., Guo, H. & Li, M.: Lithium extraction method for α-spodumene by fluorine chemical method. Advanced Material Research. 2012, Vol. 524-527, pp 2011-16.
  5. London, D., Morgan, G.B., Paul, K.A. & Guttery, B.M.: Internal evolution of miarolitic granite pegmatites at the Little Three Mine, California, USA. Can Mineral, Aug 2012, Vol 50 (4), pp 1025-54.
  6. Moxham, A.J.: US Patent 1,538,837 Manufacture of aluminium sulphate, May 19, 1922.
  7. Roskill: Lithium. Global Industry, Markets and Outlook to 2025. 2016, 13 Ed., 386 p.
  8. S. Geological Survey (USGS): Mineral commodity summaries 2017. U.S. Geological Survey 2017, 202 p.