The challenge of recycling lithium-ion batteries  

Current battery recycling is confined to recovering raw materials from the scrap produced by gigafactories. But with the adoption of lithium-ion batteries (LIBs) increasing, a much richer vein will soon emerge, as the first wave of electric vehicles (EVs) reach the end of their lives.  And with it, an additional challenge emerges managing the growing waste generated by these energy storage applications. Traditional recycling methods, such as pyrometallurgy and hydrometallurgy, have faced certain hurdles in efficiently recovering valuable materials, mainly due to the presence of the organic components present in the battery binders and separators. Some methods are labour-intensive, whereas others need lots of energy or are environmentally harmful. But each hurdle is another opportunity for innovation.  

Optimisation of shredding pre-treatment: TES develops and validates their mechanical process for battery recycling 

Leveraging over 20 years’ experience in battery recycling, TES partners oversee the development of a shredding process for batteries using its patented in-house technology. Over the first two years of the RHINOCEROS project, they have been conducting various activities using commercial battery packs, from evaluation and deactivation to pack-to-cell dismantling.  

TES have been prioritising safety measures and efficiency, ensuring the system extracts the residual energy stored in batteries, preventing it from going to waste. Following the discharge process, TES applies its proprietary mechano-physical pre-treatment, which involves pre-shredding under an inert atmosphere. This is followed by physical separation techniques to recover polymers, base metals (Al, Cu, Fe), and black mass containing valuable active materials such as Co, Ni, Mn, Li and graphite. 

Despite the effectiveness of the current mechano-physical treatment, TES has identified areas for improvement. Moving beyond the state-of-the-art [SoA], TES has optimised its process with a pre-drying step to minimise solvent residues and enhance delamination and separation efficiency. These advancements have led to significant improvements in material recovery and product quality, making the recovered black mass more suitable for downstream hydrometallurgical processing. 

• Electrolyte recovery rate: exceeding 95 % 
• Polymer recovery rate: exceeding 80 % 
• Lithium recovery rate: surpassing 95 %, with the majority retained in the black mass 
• Impurity levels of Cu and Al: reduced below 1 % 

Optimisation of thermal pre-treatment process  

However, reintegration of cathodic and anodic materials into the production chain of new batteries remains a challenge. Thermal pre-treatment of the batteries remains paramount in deactivating, releasing and separating the battery materials. But before reaching the thermal pre-treatment, batteries undergo a sorting process according to the cell type. Discharging is not mandatory prior to the pre-treatment. 

ACCUREC [ACC] partners have been modifying their thermal pre-treatment process, including a preceding stage for cell solvent recovery, which places their process beyond the SoA. To explore this technological path, they experimented various parameters altering the working temperature and the pressure of solvent distillation, to address safety concerns and stabilise the background electrolyte. A selective recovery of solvent was possible under specific temperature and pressure conditions achieving high purities >99.6 %. However, additional refining steps are necessary prior to further use in battery manufacturing.  ACC partners managed to deactivate batteries completely and produce high-purity black mass [BM] with recovery yields of >95 %, using high temperature pyrolysis and a two-stages screening process.  

Their process demonstrates that the distillation process prior to the thermal treatment allows the extraction of solvents used in the cell and streamlines high recovery rates of active material. However, their research concludes that the processing of end-of-life [EoL] batteries remains a trade-off between recycling efficiency and purity of the obtained fractions due to the large number of compounds and lack of control over the starting material.  

Electrolyte recycling

The safe and environmentally responsible recovery of electrolytes is crucial in the LIBs recycling process. The SoA non-aqueous liquid electrolytes consist of a conductive salt (lithium hexafluorophosphate, LiPF₆) dissolved in a mixture of organic solvents and additives. Many of these components are highly volatile, flammable, and classified as toxic. The current state-of-the-art LiB recycling process begins with the production of black mass, obtained through a series of pre-treatment steps. But uncontrolled evaporation of volatile electrolyte solvents during these stages presents major safety hazards. The research group at Technical University of Chalmers [CHA] have been exploring various processing pressure and temperature parameter settings for their supercritical carbon dioxide [scCO₂], with a particular focus on the extraction of ethylene carbonate [EC].

CHA researchers successfully developed and validated the scCO₂ process for electrolyte recycling from LIB black mass. The results demonstrated a >95 % overall recovery, with individual yields of >99 % for DMC, EMC, and DEC and 95 % for EC, using specific process conditions. The recovered EC retains its original structural integrity, confirming its chemical stability and purity.

But beyond the high purity and selectivity of carbonates, the developed scCO₂ does not require any chemical reagents. Moreover, the used CO₂ can be indefinitely recycled within the system, turning it into a clean separation process that eliminates toxic gas during processing.

Reactive milling for black mass production

The research group at KIT received black mass samples from partners ACC and TES. Only the samples from TES were fit to study the mechanochemical transformation of BM, as the ACC supply was already in a reduced state. The BM supplied by TES consists mostly of NMC (lithium nickel manganese cobalt oxides) cathode material and graphite, which was found to slow down the reaction kinetics.

KIT researchers applied their reactive milling on the BM provided by TES to reduce the cathode material to its metallic form and produce water-soluble lithium salts. They optimised the process using three types of mills [two planetary mills of different sizes and a SPEX shaker mill], engaging three types of reagents: Al, Mg and Ca. They also investigated the influence of milling parameters on the process duration and identified the most optimal parameters for faster kinetics. The planetary mill showed better results than the shaker mill for Mg and Al, while Ca performed well in the SPEX mill.

In depth studies performed by KIT concluded that the process is a mechanically induced self-propagating reaction, which is initiated by mechanical energy, such as the impact energy from the milling balls. Once started, it propagates through the material on its own, similar to a chain reaction. This type of reaction is highly exothermic, releasing a significant amount of heat, which further drives the reaction forward. Small amounts of volatile compounds, such as residual solvents or moisture, were found to have a significant effect on the reaction time. The basis of the volatiles interference is not clear yet, but it seems to be important to remove these compounds before the milling process to ensure a more efficient and faster reaction.

Further experiments are needed to scale up the process and ensure that it can be applied to larger batches.

Vacuum pyrolysis and Supercritical CO₂ [scCO₂] processing for binders recovery

Binders such as polyvinylidene fluoride [PVDF] and polytetrafluoroethylene [PTFE] for cathodes, and carboxymethyl cellulose [CMC] and styrene-butadiene rubber [SBR] for anodes, are particularly difficult to separate during mechanical processing. This reduces the efficiency of metal recovery and can lead to the formation of environmentally harmful substances, such as per- and polyfluoroalkyl substances [PFAS].

Thermal treatment methods, while effective in removing these organics, generate hazardous emissions like hydrogen fluoride [HF] and phosphoryl fluoride [POF₃], necessitating stringent gas treatment measures. Additionally, battery separators, often composed of complex polymers like polyethylene [PE] and polypropylene (PP), pose further challenges due to their low economic value and the presence of residual hazardous materials.

ACC has optimised vacuum pyrolysis at 550°C, which enabled the complete decomposition of PVDF, and thus, the effective removal of binders and separators from LIB waste, while recovering fluorinated compounds from decomposition products. Generated HF was captured using two different methods: through potassium hydroxide [KOH] washing (producing potassium fluoride [KF]) or a calcite filter (producing fluorite).

Supercritical carbon dioxide [scCO₂] process technology has emerged as a promising alternative for LIB recycling, offering properties that are adjustable to organic compounds. However, PVDF dissolution in scCO₂ alone requires extreme conditions, making industrial implementation challenging. Chalmers researchers have explored possible co-solvent scCO₂ systems to enable the recycling of binder directly from mechanically treated BM. They developed a specialised extraction process using scCO₂ with a co-solvent (dimethyl sulfoxide, DMSO) to recover the PVDF binder from industrial battery waste. The process achieved a 55.6 % recovery rate of PVDF under mild conditions.

The scCO₂ technology was also applied to electrolyte and separator recycling, achieving over 90 % separator recovery with high purity.

© visual: ImageFlow via Adobe Stock

Researchers at University of Agder (UiA) are working on the automated sorting and dismantling of lithium-ion batteries (LIBs) that facilitates their reuse for second life applications.

During the first reporting period, UiA designed a simulator within a virtual environment, which allowed researchers to collect necessary data and parameters, and additionally identify potential bottlenecks that may occur in the actual disassembly process. Beyond collecting data without any physical experiment, the simulation environment brings the benefit of being cost- and time-efficient, allowing for safe and flexible robotic programming without disrupting the production.

According to the simulation environment that covers the entire disassembly process, from automated discharging to sorting, the entire process can run with a total duration spanning between 12 and 14 minutes. The detailed results of this activity will soon be publicly available in a new scientific paper titled addressing  the evaluation of deep reinforcement learning for job shop problems.

During the past six months, UiA researchers have constructed a virtual simulator to train the Machine Learning (ML) algorithms. Deep learning methods have already been applied for the Job Shop problem for finding the optimal disassembly sequence when the dependence matrix is known. Next development steps will entail training the algorithms to enable automatic disassembly of Electric Vehicle (EV) batteries without prior knowledge, while optimising procedures and enhancing safety.

© AdobeStock Photos

Author: KIT

During the third semester, researchers from KIT further studied and improved the conditions for the mechanochemical transformation of black mass (BM) into metallic black mass (MBM). Since BM supplied by ACC is already in a reduced state, they focused on reducing BM supplied by TES. This BM consists mostly of NMC (lithium nickel manganese cobalt oxides) cathode material and graphite, which was found to slow down the reaction kinetics. The reduction of the cathode active material by the metallic reducing agent result in the formation of the transition metals along with lithium oxide (Li₂O) and the oxide of the respective reducing agent, which can be monitored by X-ray diffraction.

In contrast to the previous two semesters, researchers switched from shaker mills to planetary mills, which enable control of the rotation speed and larger volumes that can be processed. Various parameters such as ball-to-sample ratio (BSR), ball size, total load and rotation speed were investigated to optimise for a short reaction time.

Main take-aways

In general, the higher the BSR, the more mechanical energy can be transferred per gram of powder which results in a more intense milling and a faster reaction; however, this limits the throughput. Larger balls, on the one hand, lead to higher kinetic energies. On the other hand, fewer balls are used to keep the BSR constant resulting in a lower collision frequency. The maximum rotation speed is lower to prevent damage to the grinding media.

With Calcium as the reducing agent, no reaction was achieved at all. An unfavorable combination of ductility and size of the calcium pieces seems to resist further size reduction, which is required for the reaction.

Aluminium has the advantage of being used as a current collector and is already present in the black mass. However, during the reaction, LiAlO₂ is formed, which is limiting the subsequent Li extraction efficiency in WP5. This problem can be avoided when magnesium is used as the reducing agent, which proved to be more reactive than aluminium but doesn’t form other lithium compounds than Li₂O.

Compared to the shaker mill, a higher reaction rate was observed in the planetary mill. Researcher from KIT achieved a complete conversion of the lithium transition metal oxide in the planetary mill within 3 h using Mg as the reducing agent. In a larger version of the mill, the required milling time increases to 8 hours. Here, further investigations are planned for the next months.

Read previous article on the pre-treatment operations: Pre-treatment operations: Reactive milling for the production of metallic black mass

© Photo: Adobe

RHINOCEROS attending Shifting Economy Week

From 21 to 25 November 2023, the city of Brussels hosted the Shifting Economy Week, an annual event dedicated to showcasing transformative projects that aim to pave the way to an economy that is low-carbon, regenerative, and equally circular. The 2023 exhibitors’ line-up included, among other regional stakeholders, our partner Watt4Ever (W4E), industrial partner specialised in the development of innovative solutions for energy storage and management. W4E leveraged its presence at Shifting Economy Week to to raise awareness about the importance of circular economy principles in the context of the battery industry.

During the same event, W4E’s CEO, Aimilios Orfanos, was invited to speak at the BeCircular conference, an event dedicated to presenting concrete examples of circular economy approaches put in place by Brussels-based companies. He shared insights from W4E’s experience in developing second-life battery systems for electric vehicles, emphasising their potential benefits in terms of environmental impact and cost savings. Simultaneously, the CEO also highlighted the challenges faced by the industry in implementing circular business models, including regulatory barriers and market incentives.

RHINOCEROS at its second participation at Circular Wallonia Days

A few days after attending Shifting Economy Week, W4E represented the RHINOCEROS project at the Circular Wallonia Days, held on 13 and 14 December 2023. Centred around advancing the circularity of the batteries value chain, the event brought together stakeholders from academia, industry, and government to discuss strategies for improving the sustainability of battery production, use, and disposal. The focus topics covered recycling technologies, supply chain transparency, and policy measures to support the transition to a circular battery economy.

© Photo credits: Watt4Ever

Part of the Cluster Hub “Production of raw materials for batteries from European resources”, the RHINOCEROS consortium received the online visit of FREE4LIB representatives during the second day of the Consortium meeting held in Gothenburg. This initiative including stakeholders involved in different European R&D initiatives goes beyond building a knowledge exchange ecosystem to address common topics related to EU-funded projects; it paves new collaboration routes and synergies aiming at driving innovations for the recycling of batteries and the production of raw materials for battery applications from primary and secondary resources available in Europe.

Represented by Julius Ott (industrial engineer with expertise in circular economy at Karl-Franzens-Universität Graz) and Pau Sanchis (senior policy officer Eurobat), the FREE4LIB presentation focused mainly on the Digital Battery passport and the relevant legislative situation at European level.

Pau Sanchis referred to the Digital Battery Passport in the context of the new regulation on batteries and waste batteries which entered into force on 17 August 2023.  According to this update, thoroughly explained in Art. 77, the battery passport should contain information “relating to the battery model and information specific to individual battery, including resulting from the use of that battery”.

“Batteries should be labelled in order to provide end-users with transparent, reliable and clear information about batteries and waste batteries. That information would enable end-users to make informed decisions when buying and discarding batteries and waste operators to appropriately treat waste batteries. Batteries should be labelled with all the necessary information concerning their main characteristics, including their capacity and the amount of certain hazardous substances present. To ensure the availability of information over time, that information should also be made available by means of QR codes which are printed or engraved on batteries or are affixed to the packaging and to the documents accompanying the battery and should respect the guidelines of ISO/IEC Standard 18004:2015. The QR code should give access to a battery’s product passport. Labels and QR codes should be accessible to persons with disabilities, in accordance with Directive (EU) 2019/882 of the European Parliament and of the Council (17).”

The policy officer emphasised the role of the standardisation process on the Battery Passport, which requires the Commission to adopt implementing decision requesting European Standardisation Organisation to develop standards in support of Ecodesign by December 2023. Standards regarding the technical design and operation of the Battery Passport are expected to complement provisions under Art. 78. According to the timeline presented in the regulation, the first application of the battery passport is expected in 2027. From 18 February 2027 onwards, “all batteries shall be marked with a QR code as described in Part C of Annex VI.

The new regulation aiming at strengthening sustainability rules for batteries and waste batteries will be supported by various secondary legislation pieces which will ensure all the requirements will be developed and implemented effectively. The QR code shall provide access to the following:

• for light means of transport (LMT) batteries, industrial batteries with a capacity greater than 2kWh and electric vehicles batteries, the battery passport in accordance with Article 77.
• for other batteries, the applicable information referred to in paragraphs 1 to 5 of this Article, the declaration of conformity referred to in Article 18, the report referred to in Article 52(3) and the information regarding the prevention and management of waste batteries laid down in Article 74(1), points (a) to (f).
• for starting, light, and ignition (SLI) batteries, the amount of cobalt, lead, lithium or nickel recovered from waste and present in active materials in the battery, calculated in accordance with Article 8.

The policy overview presentation was complemented by the technical presentation undertaken by Karl-Franzens-Universität Graz, that will develop a data model of the digital battery passport platform aiming to close the information gap between beginning-of-life (BoL) and end-of-life (EoL) battery lifetime. Relying on knowledge generated previously by Univ. of Graz, the researchers set an objective to define clear user roles and establish access to certain information. Up to this moment, the work carried out has been focusing on data collection and data handling (data points sorting) on the other side. This specific work encountered various challenges, notably the users willingness to share information, process standardisation, the variety of products, the recycling cost/revenue ratio, the dynamic development of the legislative framework, to name a few.

Learn more about the progress on the battery passport on the FREE4LIB website.

Within Work package 2 – Selection, characterisation and supply, partners Watt4Ever [W4E] and Accurec [ACC] assembled the database and the parameters for module selection, which will further facilitate streamline the development of electric vehicles (EV) 2nd life batteries.  

The criteria were selected based on the 2nd life partner input and were adapted at module level. Partners generated a database using a sample of 200 commercial and passenger vehicles grouped in the following categories: Battery Hybrid Electric Vehicle (BEV), Mild Hybrid Electric Vehicle (MHEV) and Plug in Hybrid Vehicle (PHEV). The input received contributed to the identification of the required acceptance criteria that will help select the best modules for 2nd life Battery Energy Storage System (BESS).  

Selection criteria

The database of 200 BEV/PHEV/MHEV batteries and their characteristics, including a summary of technical information for each model, has been generated. Due to the mechanical, technical and software challenges that need to be overcome for efficient module integration in deployable LV and HV BESS, the database includes the following parameters as criteria for 2nd life applications: size, capacity, Cell Management Unit (CMU), casing and cell configuration. The chosen criteria should improve the security of the dismantling process and facilitate access to the module level for each battery pack. Simultaneously, these parameters will simplify the integration of 2nd life modules in battery energy storage or other systems.

Mechanical design criteria	Electrical design criteria
Physical Properties	Cell configuration
Capacity	Casing
	BMS/CMU

Mechanical design criteria 

For 2nd life applications, module level parameters are the ones that give insight on the desirability. First set of parameters are the Physical Properties. Second life integrators already have knowledge of module integration and have specific designs to accommodate the battery modules, if the module sizing is the commonly used roughly the size of a shoe box 350*150*120 mm, it already makes the integration a lot easier, faster and cheaper. By adding Capacity, power density can be calculated, and thus the technical design can be improved to maximise the capacity of the system. 

Electrical design criteria 

Cell configuration give an insight in to the voltage of the module, but also allows to wonder about the possible Battery Management System (BMS) use in the battery. Casing protects the cells from puncture, grinding, shorting out, but most importantly, expansion. With regards to BMS,/CMU, on various cases, partners noticed that OEMs integrate internal CMU on the module level, which can facilitate their use for second life applications. On the contrary, in cases where no CMU is available, one needs to be provided by the OEM or 3rd party.  

Module acceptance criteria for low voltage (LV) 48V systems

This design would allow to accommodate shoebox size modules from different OEM, while also keeping the same design of the box with small adjustments.

Mechanical design criteria

• Casing– Open top or Alu Jacket 
• Size: 350*150*120 mm ± 100 mm on all axis 
• CMU: External multimodule, external single module or internal and all can be reused if OEM CMU unit communications gateway is possible. 
• Cell amount: 3s-12s 
• Voltage: 10V-30V 
• Chemistry: NMC with risk mitigation or LFP

Module acceptance criteria for high voltage (HV) system 

For HV battery energy storage system, modules with higher voltage are prioritised, due to their capacity to save time and budget otherwise spent on wiring and building expenses.  External CMU from 3rd party is the preferred solution to internal CMU, as it minimises any interference and bug risks with the Energy Management System.  to be used to minimize any interference and bug risks with the Energy Management System.  

• Size: >350*150*120 mm ± 100 mm on all axis 
• Cell amount: 12s-30s 
• Voltage: 40V-100V