Cobalt and lanthanide recovery from batteries

Project Information

COLABATS

Grant agreement ID: 603482

Project closed

Start date 1 October 2013

End date 30 September 2016

Funded under

• Specific Programme "Cooperation": Environment (including Climate Change)

Total cost € 4 605 586,96

EU contribution € 3 588 506,00

Coordinated by

C-TECH INNOVATION LIMITED
United Kingdom

Final Report Summary - COLABATS (Cobalt and lanthanide recovery from batteries)

Executive Summary:
The feasibility of hydrometallurgical separations for Li-ion and NiMH battery waste based on use of a deep eutectic solvent (DES) has been demonstrated (the “Colabats process”). Successful application of this technology would allow recovery rates of greater than 50% from Li-ion cells which is not possible with current pyrometallurgical processes. The DES used in the demonstration was a mixture of lactic acid, choline chloride, citric acid and water. It has the property that it will dissolve most metal oxides and metals, is composed of non-toxic ingredients, and is much less costly than other ionic liquids. This is a key advantage compared to other processes based on costly and hazardous ionic liquids. The process requires further development and it may be that other waste streams with combined metal content may be suitable as feedstock in addition to waste batteries.
The project took “black mass” from an existing pilot scale pre-treatment process that involved the discharging, shredding and loss of solvents from waste batteries. This was the input material for the newly developed process. The Colabats process started with leaching the black mass in DES to yield a DES loaded with a mixture of dissolved metals. Sequential metal-specific extraction steps were applied. In these steps the DES is mixed with organic extractants that are selective for particular metals, in non-polar diluents. This step happens in counter-flow mixer-settler or equivalent devices. This is an adaptation of minerals processing technology. The important innovation is the use of the novel DES as the leaching agent because it allows multiple sequential hydrometallurgical extractions to take place, which would not be possible with a conventional mineral acid leach. The DES and organic extractants are used in closed loop. In the demonstration it was shown that the DES and other solvents were chemically stable and did not degrade after repeated use.
The process was demonstrated on 12 kg of Li-ion black mass. A 98% pure cobalt chloride solution was obtained, as well as a cobalt oxalate precipitate, a precursor to cobalt oxide by standard means. In the process developed for Li-ion batteries the selective extractions of metals from the DES were made with reagents LIX860 (for copper), DEHPA (for magnesium, manganese and lanthanides) and CYPHOS101 (for cobalt, zinc). The process as applied to NiMH differed in detail. An extraction of lanthanides was made using DEHPA followed by extraction of nickel using LIX 63.
Further development of the process is needed. For both Li-ion and NiMH processes the details of the specification of the DES and organic phases need further work to achieve steady-state operation.


Project Context and Objectives:
Commercial recycling processes for Li-Ion and NiMH batteries are predominantly pyrometallurgical. Developments in battery technology over recent years have focussed on Li-Ion chemistries, and the amount of metal they contain has fallen in newer battery types. This reduction in the metal content is a limiting factor for furnace-based recycling since non-metallic content is lost. As demands grow for more material from waste to be recovered, furnace-based recycling faces mounting challenges.
The European Batteries Directive (2006/66/EC) sets a Recycling Efficiency (RE) target for the amount of material that must be recovered from waste batteries. For Li-Ion and NiMH batteries this target is 50% by weight (excluding water). The metal content of some Li-Ion batteries has fallen to not much more than 50% by weight, making it important that more than just the metal content is recovered in the recycling process. Currently challenges with the Recycling Efficiency approach include those below.
The Recycling Efficiency calculation is complex, and it is widely agreed that results across different member states are not consistent. There has been an adjustment incorporated in the RE calculation for Electric Vehicle batteries that allows the weight of the outer case to be included in the calculation. In practice this means that in some cases the steel casing and copper busbars can be recycled, the battery cells discarded, and the battery recycling efficiency target still be met. There are calls that the 50% RE target should be applicable at a cell level as well as at the battery level.
Some Li-Ion and NiMH batteries contain materials on the critical raw materials list. The current RE calculation method does not specify what materials should be recovered, and some recycling processes are losing critical raw materials whilst still achieving the RE target.
Li-Ion batteries comprises more than six different specific chemistries (Lithium Cobalt Oxide, Lithium Iron Phosphate, Lithium Titanate etc.) containing different materials (e.g. cobalt, manganese, iron, nickel, aluminium, titanium, rare earths, graphite, etc). Current labelling requires only that the battery be identified as “Li-Ion” however, and pyrometallurgical processes are not adapted to recover multiple metallic and non-metallic components. More variations in battery chemistry will appear on the market as technology develops.
European Commission proposals for the circular economy note that a likely future requirement to recover more critical raw materials. Also with a recast of the Batteries Directive due in 2018 it is possible that the Recycling Efficiency calculation may be adjusted to encourage more targeted recycling of certain materials. It is also possible that, in time, the Recycling Efficiency target could be raised as technology allows more materials to be recovered.
The CoLaBATS process is a metal extraction process and preserves materials that are destroyed in pyro processing (electrolyte, graphite, plastic), making recycling efficiency results of greater than 50% possible. The objectives of the work follow from the context above and are:
To recover more that 50% of material from waste batteries. This means recovering more components which are lost in pyro processes. For a lithium-ion battery that means recovering cobalt, aluminium, copper, plastic, and graphite.
To be able to recover specific critical materials such as cobalt whilst having the ability to adapt to adapt to changing battery chemistry.
To have a process which is environmentally viable. This means that it operates in closed loop mode so far as its principal components are concerned.
To have a process which can be operated efficiently at smaller scale than the current pyro process, thus encouraging more players and operation at more locations. To develop technology which can be developed and implemented by small to medium sized companies.


Project Results:
This document is a summary of the work carried out in project Colabats and highlights the main results. The project ran from October 2013 to September 2016. The project website is https://web.archive.org/web/20190829092457/http://colabats.eu/

Novel solvents for use in recycling batteries
Task-specific ionic liquids and deep eutectic solvents were selected as the technology to develop in this project. Task-specific ionic liquids (TSIL) are ionic liquids that have physical properties adjusted and selected for specific tasks – for example by varying hydrophobicity or binding character. It was known for example that the ionic liquid betaine bis(trifluoromethylsulfonyl)imide will dissolve the water insoluble complex nickel dimethylglyoxime, and that when this solution is mixed with hydrochloric acid at 50oC and then cooled to 5oC, the complex is split, generating an acidic nickel solution and a solution of dimethylglyoxime in ionic liquid. Deep eutectic solvents (DES) are related to ionic liquids and are formed when two or more materials with higher melting points are brought together in a specific ratio to form a material with a lower melting. Like TSILs, DES are non-aqueous but polar solvents and are known in applications where metals are dissolved, such as electroplating. The idea in the Colabats project was to see whether TSILs and DES could be designed and used to recover metals from battery waste materials.
Waste batteries
Before getting into the chemistry the first task was to understand in what form the waste battery material should be used. Waste batteries are hazardous and need to be handled carefully. They present fire and explosion risk from rapid and uncontrolled electrical discharge, and they contain harmful substances such as volatile and flammable solvents, perfluorinated electrolytes, carcinogenic and toxic metal salts. The team analysed the physical construction and chemical composition of a range of lithium-ion (Li-ion) and nickel metal hydride (NiMH) batteries. Large batches were selected and analysed for use in the project. It was important to do this since collected waste batteries are variable in chemical composition. After the analysis the team selected a cobalt-bearing batch of Li-ion batteries and a nickel and lanthanide bearing batch of NiMH batteries to work with. The Colabats process development work did not include measures to deal with the discharging hazards, nor the recovery of solvents or electrolytes from Li-ion batteries because there are existing processes to deal with these issues. The team did have to consider the physical form however, and the material used was from a high-speed shredding process, with magnetic removal of some ferrous content in the Li-ion case, and removal of most paper and plastic on shaker tables. Further laboratory preparation by drying and size fractioning was also carried out, to reduce the amount of casings (iron) present in the NiMH case.
Designing a chemical recovery system
Many novel TSILs and DES were synthesised by the team. These were evaluated for their ability to dissolve various complexes of the metals of interest. Various different classes of ionic liquid and deep eutectic solvent were considered. At the same time consideration was given to the overall process design – how the chemical process would operate. This was influenced by some constraining factors, including the need to deal with several metals, the need to have affordable materials, the need to safeguard human health for the operators, the need for materials to be environmentally acceptable, and the need for the overall process to be cost-effective. From all of these considerations and the evaluation of the novel solvents the team decided to focus effort on a particular deep eutectic solvent.
The Colabats Process
The process chosen involves a DES composed of choline chloride, lactic acid, citric acid and water. This material has the property that it will dissolve metal oxides and metals of interest, including cobalt, nickel and lanthanides. The idea is that waste batteries are leached with the DES and then then metals dissolved in the DES are selectively extracted. This is achieved using other types of ionic liquid or “extractants” used commercially in the minerals processing industry. The DES is mixed with these extractants in a non-aqueous diluent (kerosene in this case) and then the two phases allowed to separate. When the right materials are selected there is a selective transfer of metals from the DES to the organic phase, with a corresponding transfer of ions (usually protons) going the other way. In this way the process involves both DES and TSIL (extractants).
The key advantage of this process over a conventional mineral acid leach is that it allows the use of multiple and sequential extractions of metals from the DES, and therefore the separation of metals. A process was therefore developed using a non-selective DES leach followed by sequential liquid-liquid extractions using commercially available extractants LIX860, DEHPA and then CYPHOS101 in kerosene diluents. This is the core of the Colabats process. Different extractants can be used for different metals depending on the waste material specification.
Nickel extraction using LIX63 was evaluated as an alternative batch process but was slower and not implemented in the demonstration. Methods for the selective removal of zinc and cadmium from DES were developed and lithium extraction was investigated but these were not implemented in the demonstration.
Novel ionic liquids and synthetic methods
A number of hydrophobic ionic liquids were synthesized and characterized, based on cheap and readily available long-chain phosphonium cations and hydrophobic anions and including phosphonate and phosphate based systems.
A number of protic ionic liquids with low toxicity was prepared and evaluated. Methylimidazolium acetate was selected as the lead ionic liquid candidate in the project but was not selected for the demonstration.
A continuous-flow method for the synthesis of ionic liquids was created. This may be of use in reducing the cost of ionic liquids in future.
Equipment
The leach apparatus consisted of a 100 L round-bottomed glass reaction vessel with heating mantle and stirrer. It had a working fluid volume of 60 L. Material was pumped in and out using a pneumatic diaphragm pump. The filter press was of standard design but with no metal in contact with process fluid. The leach process was optimised, to two hours in the case of Li-ion and 30 minutes for NiMH.
The application of ultrasound was evaluated in order to improve the leaching effectiveness, whether rate or equilibrium. At the bench scale it did indeed give an improved leaching performance, but not enough to warrant inclusion in the demonstration.
A mixer-settler design was developed that coped with the different viscosities of the process fluids, especially the high viscosity of the DES. The design implemented in the demonstration featured modular units of three counter-current stages with flow directly from one to another and with flow controlled in and out of the three-stage unit by peristaltic pumps. The units were fabricated from transparent PVC which was necessary to allow visual monitoring of the process. The mixer-settler units were housed in a unit enclosed with curtains and fitted with air extraction in order to reduce solvent vapour concentration.
Li-ion and NiMH demonstrations
12 kg of Li-ion battery black mass was treated. All steps were operated and the DES was found to be chemically stable over 5 reuses. There was no detectable contamination of the DES by extractants. All extractants were stable over multiple re-use cycles with no degradation of performance. Cobalt chloride strip solution was obtained at 98% purity. The final output was a cobalt oxalate precipitate, obtained at 93% purity, based on the combination of strip solutions obtained (some at lesser purity that 98% because of non-optimised conditions).
The NiMH demonstration featured the extraction of mixed lanthanides by means of DEHPA extraction and strip. This was followed by nickel extraction using LIX 63 reagent. The formulation of the DES needs modifying since the inclusion of citric acid gave problematic nickel citrate precipitates. Further work showed that this could be avoided by replacing citric acid with an oxidising agent such as iodine or sodium bromide but more work is needed on this. In addition the relatively slower kinetics of the nickel extraction and stripping mean further optimisation of that process is needed.
1200 kg of DES were synthesised in total. Some was used in process development and some in the demonstrations. It was characterization by FT-IR, NMR1H, pH, viscometry and DSC.
There is further work required to optimise the process. In particular it needs to be run in steady state condition with regard to the hydrometallurgical extraction. This should be the basis of the follow-on project.
Chemical analysis of process chemicals
The chemical stability of the DES was evaluated using NMR and was good, with no chemical changes detectable. It was also shown that the chemical stability of the DES could be monitored using chromatographic techniques. An HPLC strategy was developed that allowed to monitor the individual components choline chloride, lactic acid and citric acid. This result is important for the commercial development of the process because it means that there is a an affordable analytical technique to monitor the process chemicals. The NMR techniques used in the project are too expensive and time-consuming for commercial process monitoring application.
Recycling of DES using ion exchange resins and solvents was shown to be possible. Different solvents and ion exchange resins were evaluated and it was shown that the DES could be purified by these means. This is a possible option to use when the DES becomes degraded and needs to be reconditioned.
Next steps: improving the efficiency of the process
The ratio of DES to black mass used in the leaching step was 50:1 and for each batch the 60 kg of DES was heated from room temperature to 85C. This energy input was the most significant energy requirement of the process and adversely affected the costing and environmental impact of the process. If the temperature of the DES can be kept close to the temperature at which leaching takes place then the heating energy input is significantly reduced and the cost and environmental profile are improved. On the prototype plant this was not possible because the facilities did not allow the mixer-settlers to be used with hot solvents. A suitably designed unit could be arranged to do this, with closed vessels and explosion proof motors.
Regarding nickel extraction the kinetics of stripping of the nickel from the LIX63 need further improvement. A second route developed in the project featured the extraction of nickel from DES using the novel pyridyl pyrazole compound PyPzC10H21 and DEHPA in decanol-modified solvent 70. This gave rapid extraction of nickel which is attributed to the PyPz compound having some surface activity. This result featured a high decanol content diluent and raises the possibility of using biodiesel as organic solvent rather than kerosene. This route is promising and warrants further attention.
Regarding cobalt extraction the use of CYPHOS101 may be replaced by LIX 63 if the kinetics of extraction can be improved. This may involve different mixer-settler apparatus giving more intensive mixing and longer settling times.
Modification of the DES specification is required for NiMH, and for both NiMH and Li-ion the specification needs to be refined for more extended operation.
Lithium extraction needs to be resolved, including the use of pre-leach in water followed by precipitation or electrodialysis. Methods for the selective removal of zinc and cadmium were developed, using commercial reagents Cyanex 301 for zinc and Cyanex 302 for cadmium. These methods were shown to be feasible but were not included in the demonstration.
A follow-on project should target a specific composition of waste stream and optimise the process based on the specification of that material. That will allow a simplification of process and optimisation of a more limited set of reagents.
Other opportunities
Electro-winning of nickel directly from DES is a very attractive potential process and was proven in principal but the efficiency was too low to justify inclusion in the demonstration. Different additives gave improved leaching and electrochemical recovery but further work was needed to define an extraction processes with sufficient efficiency.
The principle of using supercritical carbon dioxide to extract lanthanides from DES was established. This would have the considerable advantage of eliminating all additional solvents from the process, and would be especially valuable for the NiMH system. The efficiency was low and it was not pursued further in the project.
Conclusion
A process for the selective recovery of metals from waste batteries was developed and trialled. The feasibility of using a non-selective leach in a benign eutectic solvent followed by hydrometallurgical separation based on liquid-liquid extraction has been established. Further development is required on a specific waste material to establish steady-state running of the process.


Potential Impact:
The potential impact of this project is that hydrometallurgical recycling processes will be developed that allow complete recycling and recovery of all material in waste batteries and related waste. The hydrometallurgical process is an alternative to pyrometallurgical processes which do not allow the recovery of plastic and graphite for example. It is therefore an enabler to attaining full recovery of materials. If such processes can be developed then regulation will be able to advance in due course and the overall environmental impact of making and selling batteries will be reduced.
The project has shown that a hydrometallurgical process is feasible. This is a first and necessary step to prompt further development of the technology, with more specific economic targets. The project took a chemical concept at TRL 2 (technology concept formulated) and advanced it to TRL 5 (technology validated in relevant environment) by means of a demonstration of the process. The process is now at a stage where end-users with specific requirements can assess it and develop it further. It will be necessary to specify particular waste compositions in order to take the next step of improving the cost-effectiveness of the process. That be a particular source of waste batteries for example.
A likely impact is the development of a business model for the technology including the operation of the process and the necessary supporting activities. The process has very different capital and operational requirements that make it suitable for operation by SMEs. Whether the process is owned and operated by the end-user or contracted for from a service company will depend on the economic and operational details. It might be for example that the process itself is mobile – containerised - and travels to the waste locations. There will be a necessity for analytical support services – of the waste itself, of the products, and of the process fluids.
Development work is required to advance the process to the next level, probably to TRL 7 in a follow-on project. This will require integration with front-end battery sorting and preparation (discharging, shredding, physical separation) as well as downstream processing of product (as precipitate for example). Follow-on laboratory work at TRL 3-4 to further develop and refine the chemistry, to develop the nickel removal process further, and to replace kerosene with biodiesel, will be necessary also. The process may find use in the treatment of other mixed wastes where a low-temperature leaching in DES to remove metals but leave plastics or other organic materials behind is desirable.
The intended overall impact of the project is that it prompts further development of hydrometallurgical metal recovery technology that will find a place in waste treatment and reduce the burden that primary extractive metal industries place on the world, against a back-drop of rising consumption of batteries and related items.

List of Websites:
https://web.archive.org/web/20190829092457/http://colabats.eu/
Dr R J Crawford. [e-mail]