ES3002748T3 - Recovery of nickel and cobalt from li-ion batteries or their waste - Google Patents

Abstract

The present invention falls within the field of pyrometallurgy and describes a method and slag suitable for the recovery of Ni and Co from lithium-ion batteries or their waste. The composition of the slag is defined as follows: 10% < MnO < 40%; (CaO + 1.5*Li2O)/Al2O3 > 0.3; CaO + 0.8*MnO + 0.8*Li2O < 60%; (CaO + 2*Li2O + 0.4*MnO)/SiO2 >= 2.0; Li2O >= 1%; and, Al2O3 + SiO2 + CaO + Li2O + MnO + FeO + MgO > 85%. This composition is particularly suitable for limiting or preventing the corrosion of furnaces lined with refractory bricks containing magnesia. (Automatic translation with Google Translate, no legal value)

Description

DESCRIPTION

Recovery of nickel and cobalt from lithium-ion batteries or their waste

The present invention is in the field of pyrometallurgy and relates more particularly to the recovery of Ni and Co from lithium-ion batteries or their residues.

Electric vehicles have experienced unprecedented growth in recent years, driven, among other things, by new legislation in Europe and China designed to gradually reduce the CO2 footprint of car fleets and limit air pollution in cities. This growth is expected to continue for the next few decades. The adoption of electric vehicles depends largely on the performance of the batteries used to store electrical energy. To achieve maximum energy density while keeping costs under control, rechargeable lithium-ion batteries are currently preferred. Many of these batteries contain cathodes based on the transition metals Ni, Mn, and Co and are therefore also known as NMC batteries. With the growth of the e-mobility market, demand for these metals is also expected to increase significantly.

Demand for Ni and Co may even exceed global production capacity. Co is particularly critical, as it is currently only produced as a byproduct of the Ni and Cu industry. The nickel market is significantly larger than the cobalt market. Most Ni is used for stainless steel production, where Ni purity is less important. High-purity Ni and Co metals or compounds, however, are already in short supply. In view of the above, recovering Ni and Co from spent lithium-ion batteries or their waste is therefore an attractive proposition.

There are several known processes for recycling lithium-ion batteries, where nickel, cobalt, and copper oxides are reduced to metal and concentrated in an alloy phase at high temperature.

WO2017121663 describes slag compositions produced in an industrial process and exposes an effect of MnO on slag viscosity and cobalt recovery. The main slag components described therein are CaO, SiO<2>, AbO3, U<2>O, and MnO (MnO<2>). The MNO concentrations described in those slags are very low and the lesson from this is to limit the amount of MnO in the slag. Elwert et al. (Phase composition of high lithium slags from the recycling of lithium-ion batteries: World of Metallurgy- ERZMETALL, vol. 65 (3), 2012, pp. 163-171) analyzes the phase composition of three different slags with the aim of developing potential processes to beneficiate slags for lithium recovery. Hu et al. (Recovery of Co, Ni, Mn, and Li from Li-ion batteries by smelting reduction-Part II: A pilot-scale demonstration: Journal of Power Sources, vol. 483, 2021, 229089) proposes a recycling process for Li-ion batteries by smelting them with fluxing agents at temperatures above 1500 °C, resulting in a slag and an alloy containing Ni, Co, and Mn. Both Elwert and Hu describe slag compositions with MNO concentrations well below 10%.

Recent lithium-ion batteries, on the other hand, typically contain increasing amounts of Mn. This makes slag formation with small amounts of Mn difficult, as it would require either diluting the Mn by adding large amounts of fluxing agents or choosing more reducing conditions for such processes to send more Mn to the alloy phase rather than the slag phase. The first option increases fluxing agent consumption and the total amount of slag obtained, while the second option increases the complexity of any subsequent hydrometallurgical treatment of the alloy due to its higher Mn content. Both options would significantly degrade process economics.

Patents WO12140951, WO13080266, and WO20013294 propose processes for recycling lithium-ion battery scrap to recover Ni and Co, while fixing impurities such as Fe and P in the slag phase. Although it is stated that Mn could be a major component of the obtained slag, no preferred range or particular effect of MnO in such slags is specified. Furthermore, the emphasis of patents WO12140951 and WO13080266 is on low-melting, Fe-rich slags.

Vest et al. (Friedrich Slag design for lithium recovery from spent batteries: Int. Work. Met. Interact., vol. 9 (93), 2010, pp. 93-106.) describes the theoretical calculations of MnO-rich slags. Wittkowski et al. (Speciation of Manganese in a Synthetic Recycling Slag Relevant for Lithium Recycling from Lithium-Ion Batteries: Metals, vol. 11 (2), 2021, p.

188.) analyzes the phase composition of different lithium-containing slags from the recycling of lithium-ion batteries.

Patents US 2017/229744 A1, CN103924088 and EP3269832 describe bath smelting processes for discarded batteries, which produce an alloy containing Co and/or Ni and a slag rich in SiO<2> and MnO.

No part of the above-mentioned prior art teaches a slag composition according to the present invention.

Even when slags with a relatively higher MnO content are described, none of the aforementioned patents teach the present process conditions, and they do not address any effect of the resulting MnO-rich slag compositions on the furnace walls. Typically, furnace walls are made of refractory bricks. The main composition of such bricks is magnesia. Examples include magnesia bricks (magnesia content typically above 90%) or magnesiachrome bricks (magnesia content typically 50 to 70%). It has been observed that the magnesia in the walls is dissolved by common slags during furnace operation, resulting in wear or corrosion of the furnace walls over time. This is a recurring problem that results in high maintenance costs, as it requires the furnace to be shut down and the refractory bricks replaced at regular intervals. The problem is even more pronounced at higher operating temperatures, such as above 1550 °C.

Therefore, an objective of the present invention is to provide a process for the recovery of Ni and Co from lithium-ion batteries, which at the same time increases the service life of a furnace, operating with a dedicated system of MnO-rich slag and Li<2>O composite such that magnesia-containing refractory bricks corrode less during operation. The achieved wear reduction contributes positively to the overall economics of the present process.

According to a first embodiment, the process for the recovery of Ni and Co from lithium-ion batteries or their waste comprises the steps:

- provide a furnace lined with magnesia-containing refractory bricks;

- providing a load comprising slag formers and lithium-ion batteries or their residues; and,

- melting the charge under reducing conditions, thus obtaining an alloy containing the majority of Ni and Co, and a slag;

characterized in that the slag has a percentage composition by mass according to: 10% < MnO < 40%;

(CaO 1.5*Li2O) / AbOa > 0.3;

CaO 0.8*MnO 0.8*LbO < 60%;

(CaO 2*LbO 0.4*MnO) / SiO<2>> 2.0;

Li<2>O > 1 %; and,

AbO3 SiO<2>+ CaO LbO MnO FeO MgO > 85%.

“Slag formers” means one or more of, for example, CaO, AbO3, and SiO<2>. Other slag formers, well known to those skilled in the art, may also be present. The slag-forming compounds may be added as is or may be obtained on-site from easily oxidizable metals present in the feedstock, such as aluminum.

"Lithium-ion batteries or their waste" means, for example, new or discarded lithium-ion batteries, spent or end-of-life batteries, production or battery waste, electrode materials or pre-processed battery materials, for example, after crushing or sorting, including so-called "black mass". However, these should still contain appreciable amounts of Co and/or Ni.

"Majority" of an element or compound means: more than 50% by weight of the corresponding quantity present in the filler. It may also include a range with a lower value selected from among 55%, 60%, 65%, 70% and 75%, and an upper value selected from among 80%, 85%, 90%, 95% and 100%.

The MnO content in the slag plays a key role in the present invention. A minimum of about 10% in the slag is required to suppress the dissolution of MgO in the slag of the magnesia-containing refractory bricks lining the furnace. It is preferred to have at least 15% MnO. Furthermore, it is beneficial to have an MnO content of 10 to 40% in a relatively small amount of slag, since a decrease in the slag volume contributes to suppressing MgO dissolution. On the other hand, the addition of more fluxing agents, thereby increasing the total volume of the slag and diluting the percentage of MnO, increases the amount of MgO that dissolves in the refractory bricks and therefore has a negative influence.

The upper limit of MnO in the slag is also important, as it helps keep the melting point of the slag sufficiently low. Slags containing up to 40% MnO according to the invention melt below 1500 °C. Lower temperatures are beneficial in suppressing the dissolution of MgO from magnesia-containing refractory bricks along with the amount of MnO. MnO amounts exceeding 40% increase the melting point of the slag above 1500 °C, especially when the slag also contains relatively higher amounts of AbO, such as above 50%, and are therefore less preferred.

According to another embodiment, the MnO content in the slag is 30% or less. The percentages of Mn in the slag are standardized as “% MnO”. However, the exact oxidation state of Mn is not always well defined in such slags. Therefore, manganese oxide (“MnO”) may also represent a mixture of the monospecies MnO with manganese dioxide, MnO<2>. It is assumed, however, that the proportion of monospecies MnO is well above 95%, particularly under the chosen reduction reaction conditions.

MnO is typically green in color, while MnO<2> is typically black-brown or black-gray in color, giving it the name “manganese black.” Only if the monospecies content is high enough will the green color be visible.

According to another embodiment, the CaO content in the slag is 15% or more, preferably 20% or more.

According to another embodiment, the CaO content in the slag is 50% or less, preferably 30% or less.

A minimum of 15% CaO keeps the slag sufficiently fluid and ensures that it can be handled easily.

Higher amounts of CaO, such as 20% or more, are preferred, as CaO helps suppress Mg dissolution from magnesia-bearing refractory bricks, since Ca and Mg share similar chemical sites in the slag. Operating the process with a preferred upper limit of 30% CaO helps keep the slag melting temperature below 1600 °C. Too high an amount of CaO, such as above 50%, significantly increases the slag melting temperature and should therefore be avoided.

It has also been observed that, in addition to MnO and CaO, Li<2>O also suppresses the dissolution of Mg from magnesia-containing refractory bricks, while increasing amounts of SiO<2> have a negative effect.

This is reflected in the equation (CaO 2*Li2O 0.4*MnO) / SiO<2>> 2.0. When recycling lithium-ion batteries, the amount of Li<2>O in the slag is expected to be significant.

According to another embodiment, the AbO3 content in the slag is <50%, preferably <40%, since excessively high amounts of AbO3 increase the melting point in the slag.

According to another embodiment, the Fe content in the slag is 25% or less, preferably 10% or less.

According to another embodiment, the sum of AbO3, SiO<2>, CaO, LbO, MnO, FeO and MgO is 90% or more, preferably 95% or more.

According to another embodiment, (CaO 2*LbO 0.4*MnO) / (SiO<2>+ 0.2*AbO3) is > 1.5.

Slag composition and operating temperature are critical considerations for the process described herein. The present invention strikes a balance between compounds that protect the furnace walls (e.g., MnO, CaO, LbO) and compounds that adversely affect the furnace walls, either being unavoidable (e.g., AbO from Al in the feed) or otherwise necessary (e.g., SiO<2> needed to lower the slag melting point). Controlling the composition in the slag further allows for adequate slag fluidity and minimal superheat in the slag at the desired operating temperature. A temperature as low as possible, while remaining above the melting point of the alloy, is preferred.

This balance is reflected in the process conditions presented, as well as in the composition of the slag itself.

According to another embodiment, the charge melting step is carried out at a temperature of at least 1400°C, to ensure complete melting of the metallurgical charge, and at most 300°C above the liquidus point of the slag, preferably at most 100°C above the liquidus point in the slag. The lower limit is preferred to prevent even partial solidification of the alloy and slag produced. The upper limit is preferred to prevent overheating of the slag. Higher temperatures promote the dissolution of Mg from the magnesia-containing refractory bricks. Therefore, lower temperatures are generally preferred to reduce wear and to save energy. Overheating of the slag has a negative impact on the partial dissolution of the magnesia-containing refractory bricks in the slag.

The described ratio of CaO and LbO to AbO3 in the slag helps to keep the melting point in the slag sufficiently low, preferably below 1700 °C, more preferably below 1650 °C, even more preferably below 1600 °C and most preferably below 1550 °C.

Despite its beneficial effect on the service life of magnesia-containing refractory bricks, restricting the combined amounts of CaO, MnO and Li<2>O in the slag is equally relevant, since the melting point in the slag would be too high when CaO 0.8*MnO 0.8*Li2O exceeds 60%.

According to another embodiment, the melting step comprises the additional steps:

- sample the slag;

- cool the slag sample and evaluate its color; and,

- if the slag sample is green, terminate the melting stage; or,

- If the slag sample is not green, continue with the melting stage after adjusting the pO<2> level to achieve higher reduction conditions.

“Slag sampling” means taking a small sample of the slag and evaluating its color, while the process continues under the chosen conditions.

Color assessment can be easily performed visually. Compared to a chemical analysis of the slag, monitoring the color change provides a rapid and efficient indication that the slag contains a certain minimum percentage of MnO. As will be described in more detail below, it has also been observed that the green color of the slag also indicates that the majority of the Co contained in the feed is being reduced, incorporating it into the alloy. Without wishing to impose any theory, in fact, the color change to green is believed to result from the reduction of MnO<2> to MnO, but also from the reduction of typically darker oxides of, for example, Ni and Co.

Visual inspection is a quick and easy way to determine whether process conditions need to be adjusted, optimized, and/or terminated, which can save time and operating costs. Color checking is therefore a reliable indication of the progress of the metallurgical operation.

In the present context, “green” is defined as the color with hue, lightness and saturation in the following range according to ASTM D1535-14 (2018):

- Tone from 5GY to 5BG;

- Luminosity: >3; and,

- Saturation: >3.

Examples of green are shown in the “Geological Rock-Color Chart with Genuine Munsell Color Chips” produced by Munsell Color in 2009.

The operating conditions are selected to oxidize a majority of the Mn in the slag and to reduce a majority of the Co and Ni in the alloy. Preferably, >90% of the Co and Ni in the alloy are recovered, more preferably >95%, and most preferably >98% to achieve maximum process economy. The pO<2> level of the present process is easily adjusted to achieve these favorable efficiencies.

According to another embodiment, the level of pO<2> is adjusted to 10'7 > pO<2>> 10'12, preferably to pO<2>< 10'8, more preferred to pO<2>< 10-85 and most preferred to pO<2>< 10'9

The preferred pO<2> levels of 10.8, 10.85 and 10.9, as well as the limit of 10.12, therefore represent conditions of greater reduction compared to a pO<2> level of 10.7.

According to another embodiment, the color of the slag is green. During charge smelting under reducing conditions, the color of the slag typically changes from black-gray or black-brown to green as the process progresses. According to another embodiment, the furnace is an electric furnace. The use of an electric furnace or electric arc furnace (EAF) allows for greater flexibility if higher operating temperatures are desired or needed. Another advantage is that it allows for taking advantage of off-peak electricity prices or electricity produced from environmentally friendly, green sources, such as local wind power plants. According to another embodiment, a Li-containing metallurgical slag comprises a mass percentage composition according to:

10% < MnO < 40%;

(CaO 1.5*LI2O) / AI<2>O<3>> 0.3;

CaO 0.8*MnO 0.8*LbO < 60%;

(CaO2*LbO 0.4*MnO) / YES<2>> 2.0;

LÍO<2>O > 1 %; and,

AI<2>O<3>+ YES<2>+ CaO LÍ<2>O MnO FeO MgO > 85%.

According to another embodiment, the metallurgical slag containing Li has a green color.

According to another embodiment, the MnO content in the Li-containing metallurgical slag is 30% or less.

According to another embodiment, the content of CaO in the Li-containing metallurgical slag is 15% or more, preferably 20% or more.

According to another embodiment, the content of CaO in the Li-containing metallurgical slag is 50% or less, preferably 30% or less.

According to another embodiment, the content of AbO3 in the Li-containing metallurgical slag is 50% or less, preferably 40% or less.

According to another embodiment, the Fe content in the Li-containing metallurgical slag is 25% or less, preferably 10% or less. In an FeO-rich slag containing more than 10% FeO, or even more than 20% FeO, the CoO cannot be reduced to metallic Co without also transferring relatively large amounts of metallic Fe to the alloying phase. This significantly increases the costs of any follow-up hydrometallurgical treatment of the resulting alloy and is therefore less desirable.

According to another embodiment, the Li-containing metallurgical slag is used as a slag former in a pyrometallurgical recycling process. The resulting metallurgical slag contains CaO and SiO<2> and can therefore be used as a slag former in further operations.

According to another embodiment, the Li-containing metallurgical slag is used as a slag former in the process according to the first embodiment, thereby partially or completely replacing the slag formers in the step of providing a charge comprising slag formers.

Reusing the produced metallurgical slag in new operations allows for greater flexibility in choosing operating conditions, such as the pO<2> level of the process. For example, if more oxidizing conditions were used, thereby sending more Co and/or Ni into the slag, these valuable metals would not be lost but would be recovered in a new operating cycle, perhaps using more reducing conditions to recover more Co and/or Ni.

When reusing this metallurgical slag as a slag former in a new process or as a starting slag for the same process, it should be noted that the incoming batteries or their waste may contain additional amounts of compounds such as Al, Mn, or Li, which will end up in the slag after oxidation. Consequently, the respective amounts of AbO, MnO, or Li<2>O in the slag will increase. While the amount of Li<2>O is considered less critical, the amounts of AbO and MnO have a more direct influence on the melting temperature. In another embodiment, the preferred upper limit for MnO is 30% to allow a margin for incoming fresh MnO from a new charge.

Due to the fresh incoming compounds, the metallurgical slag according to the invention can only be reused for a limited number of cycles. To decide whether the slag can be further reused, the slag composition must be analyzed and compared with the composition specifications described herein. Stripping and reusing at least part of the metallurgical slag by diluting it with fresh slag formers is a viable long-term option.

As described above, the slag according to the invention helps significantly suppress the dissolution of MgO from the refractory bricks. However, it cannot be completely avoided. This leads to another positive side effect of slag recycling. The dissolved MgO from the refractory bricks (caused by minor dissolution of MgO during previous smelting processes) accumulates in the slag, and this, along with MnO, LbO, and CaO, tends to suppress further corrosion of the refractory bricks. This makes reusing the obtained slag particularly attractive.

The following examples are provided to further illustrate embodiments of the present invention.

Example 1

The dissolution of MgO from the walls of magnesia-containing crucibles was measured during the use of several different slag compositions. Various compounds contained in lithium-ion batteries or their residues, respectively, their oxides, such as FeO, AbO3, Li<2>O and MnO, were melted together with CaO and SiO<2> as fluxing agents in a 1 L MgO crucible. The total weight of the added oxides was 1000 g. The ratio of FeO, AbO3, Li<2>O and MnO was chosen to represent a typical composition of existing lithium-ion batteries.

The crucibles were heated gradually at a heating rate of 150 °C/h using an induction furnace. Once the slags were fully molten, the crucibles were held at temperatures of 1400, 1450, or 1500 °C. After 2 h of heating, the molten slags were removed from the crucibles and quenched with water. Table 1 lists the composition of the slags obtained in this example.

Table 1: Composition of the slags obtained

The MgO concentrations in the above slags were relatively low (0.9% to 3.9%). This result indicates that the dissolution of MgO from the crucible wall was adequately suppressed under the selected conditions.

The experiments were performed with slag compositions that did not contain Ni, Co, or Cu, since the amount of these metals in the final slags is typically very low and therefore has virtually no influence on the properties of the slag.

Comparative example 2

The dissolution of MgO from the walls of magnesia-containing crucibles was measured using different slag compositions. Various compounds contained in lithium-ion batteries or their residues, respectively, their oxides, such as FeO, AbO3, and MnO, were melted together with CaO and SiO<2> as fluxing agents in a 1-L MgO crucible. The total weight of the added oxides was 1000 g.

The crucibles were heated gradually at a heating rate of 150 °C/h using an induction furnace. Once the slags were fully molten, the crucibles were held at temperatures of 1400 or 1450 °C for 2 h. After 2 h of heating, the molten slags were removed from the crucibles and quenched with water. Table 2 lists the composition of the slags obtained in this example.

Table 2: Composition of the slags obtained

Compared to slags 1-1 to 1-3 used in Example 1, in this case, the SiO<2> content in the slags was adjusted to be higher, while the CaO, LbO, and/or MnO content was adjusted to be lower. The MgO concentrations measured in the above slags were relatively high (8.7% to 13.2%), indicating that relatively large amounts of MgO from the crucibles were dissolved in the respective slags.

As in Example 1, the slags contained no Ni, Co or Cu.

Explanation of examples 1 and 2

The slags obtained in Example 1 contained less MgO than the slags obtained in Comparative Example 2. No visible degradation of the MgO crucible was observed under the conditions of Example 1, while the crucible walls became thinner under the conditions of Example 2. Slags containing relatively low concentrations of SiO<2> and relatively high combined concentrations of Li<2>O, CaO and/or MnO suppressed the dissolution of MgO, as demonstrated in Example 1. More specifically, the dissolution of MgO in the slag was efficiently suppressed when the ratio (CaO 2 Li<2>O 0.4 MnO) /SiO<2> was 2 or higher.

Example 3

500 kg of spent lithium-ion rechargeable batteries were fed into a 1 m diameter furnace, freshly lined with 200 mm chromium-magnesia refractory bricks. 80 kg of limestone and 20 kg of sand were added along with the lithium-ion batteries. A bath temperature of 1450–1500 °C was maintained, which is adequate to keep both the slag and the alloy sufficiently fluid for easy tapping and handling. Heat was supplied by oxidation of Al and C in the batteries, using submerged O<2> injection. The injection rate was chosen to ensure strongly reducing conditions, i.e., a pO<2> of 10-9. Natural gas was added to compensate for heat losses in the furnace. After 1 h of heating, the alloy and slag produced were separated by tapping. Table 3 shows the analyses of the input and output phases of the process.

Table 3: Input and output phases of the process

During the processing of the batteries, no visible degradation of the magnesia-containing refractory bricks was observed. The MgO concentration in the resulting slag was only 1.2%, equivalent to a loss of 2.3 kg of MgO from the refractory bricks, which is considered low. The ratio (CaO 2 Li2O 0.4 MnO) / SiO2 was 4.3. Therefore, this slag effectively suppressed furnace wall wear.

Comparative example 4

500 kg of spent lithium-ion rechargeable batteries were fed into a 1 m diameter furnace, freshly lined with 200 mm chromium-magnesia refractory bricks. 50 kg of limestone and 50 kg of sand were added along with the lithium-ion batteries. A bath temperature of 1450–1500 °C was maintained, which is adequate to keep both the slag and the alloy sufficiently fluid for easy tapping and handling. Heat was supplied by oxidation of Al and C in the batteries, using submerged O2 injection. The injection rate was chosen to ensure strongly reducing conditions, i.e., a pO2 of 10. Natural gas was added to compensate for heat losses in the furnace. After 1 h of heating, the alloy and slag produced were separated by tapping. Table 4 shows the analyses of the input and output phases of the process.

Table 4: Input and output phases of the process

The MgO concentration in the slag obtained was 9.0%, equivalent to a loss of 19.8 kg of MgO from the refractory bricks and, therefore, significant wear on the furnace walls. The ratio (CaO 2 Li2O 0.4 MnO) / SiO2 was 1.7.

Example 5

500 kg of spent lithium-ion rechargeable batteries were fed into a 1 m diameter furnace, freshly lined with 200 mm chromium-magnesia refractory bricks. 50 kg of limestone and 50 kg of sand were added along with the lithium-ion batteries. A bath temperature of 1450–1500 °C was maintained, which is adequate to keep both the slag and the alloy sufficiently fluid for easy tapping and handling. Heat was supplied by oxidation of Al and C in the batteries, using submerged O<2> injection. The injection rate was chosen to ensure strongly reducing conditions, i.e., a pO<2> of 10.9. Natural gas was added to compensate for heat losses in the furnace. After 1 h of heating, the alloy and slag produced were separated by tapping. Table 5 shows the analyses of the input and output phases of the process.

Table 5: Input and output phases of the process

The MgO concentration in the slag obtained was 2.8%, equivalent to a loss of 7.4 kg of MgO from the refractory bricks. The ratio (CaO 2 U 2 O 0.4 MnO) / SO 2 was 2.8.

Explanation of examples 3, 4, 5

In Example 3 and Comparative Example 4, the same quantity and composition of batteries were fed to the furnace, but with different limestone-to-sand ratios. The resulting slag 3 contained a higher concentration of CaO and a lower concentration of SiO2 than slag 4.

The ratio (CaO 2 Li2O 0.4 MnO) / SiO2 was 4.3 in slag 3 and 1.7 in slag 4, respectively. Only 2.3 kg of MgO dissolved in slag 3, while a significantly higher amount of 19.8 kg of MgO dissolved in slag 4.

In Example 5, batteries with a higher concentration of Mn and Li were fed to the furnace, maintaining the same limestone and sand ratio as in Comparative Example 4. The resulting slag 5 contained a higher concentration of MnO and Li2O than slag 4.

The ratio (CaO 2 Li2O 0.4 MnO) / SiO2 was 2.8, while 7.4 kg of MgO was dissolved in slag 5. This example therefore demonstrates the beneficial effect of the combined MnO and Li2O, while all other reaction conditions remain the same.

Slags containing lower concentrations of SiO2 and higher combined concentrations of Li2O, CaO and MnO are better suited to suppress MgO dissolution, as demonstrated in Examples 3 and 5.

Example 6

500 kg of spent lithium-ion rechargeable batteries were fed into a 1 m diameter furnace, freshly lined with 200 mm chromium-magnesia refractory bricks. 189 kg of slag produced in Example 3 was added along with the lithium-ion batteries. A bath temperature of 1450–1500 °C was maintained, which is adequate to keep both the slag and the alloy sufficiently fluid for easy tapping and handling. Heat was supplied by oxidation of Al and C in the batteries, using submerged O2 injection. The injection rate was chosen to ensure strongly reducing conditions, i.e., a pO2 of 10. Natural gas was added to compensate for heat losses in the furnace. After 1 h of heating, the alloy and slag produced were separated by tapping. Table 6 shows the analyses of the input and output phases of the process.

Table 6: Input and output phases of the process

During battery processing, no visible degradation of the magnesia-containing refractory bricks was observed. The MgO concentration in the slag produced was only 1.2%, equivalent to a loss of 1.4 kg of MgO from the refractory bricks, which represents even less degradation than in Example 3. The ratio (CaO 2 Li<2>O 0.4 MnO) / SiO<2> was 6.8. Therefore, this slag efficiently suppressed the wear of the furnace wall made with magnesia-containing refractory bricks.

General conclusion

The metallurgical slags according to the present invention are suitable for recovering valuable metals, such as Ni and Co, from lithium-ion batteries or their residues, while minimizing the degradation of magnesia-containing furnace refractory bricks.

Claims (20)

1. Process for recovering Ni and Co from lithium-ion batteries or their waste, comprising the following steps: -providing a furnace lined with magnesia-containing refractory bricks; -providing a charge comprising slag formers and lithium-ion batteries or their residues; and, -melt the charge under reducing conditions, thus obtaining an alloy containing the majority of Ni and Co, and a slag; characterized in that the slag has a percentage composition by weight according to: 10% < MnO < 40%; (CaO 1.5*Li2O) / AbO3 > 0.3; CaO 0.8*MnO 0.8*LbO < 60 %; (CaO 2*LbO 0.4*MnO) / SiO2 > 2.0; Li2O > 1 %; and, AhO3 SiO<2>+ CaO Li<2>O MnO FeO MgO > 85%. 2. Process according to claim 1, wherein the MnO content in the slag is 30% or less. 3. Process according to claim 1 or 2, wherein the CaO content in the slag is 15% or more, preferably 20% or more, and wherein the CaO content in the slag is 50% or less, preferably 30% or less. 4. Process according to any one of claims 1 to 3, wherein the AbO3 content in the slag is 50% or less, preferably 40% or less. 5. Process according to any one of claims 1 to 4, wherein the Fe content in the slag is 25% or less, preferably 10% or less. 6. Process according to any one of claims 1 to 5, wherein the sum of A bO3, SiO2, CaO, LbO, MnO, FeO and MgO is 90% or more, preferably 95% or more. 7. Process according to any one of claims 1 to 6, wherein (CaO 2*LbO 0.4*MnO)/(SiO2 0.2*AbO3) is > 1.5. 8. Process according to any one of claims 1 to 7, wherein the step of melting the charge is carried out at a temperature of at least 1400 °C and at most 300 °C above the liquidus point in the slag, preferably at most 100 °C above the liquidus point of the slag, thereby avoiding overheating. 9. Process according to any one of claims 1 to 8, the melting step comprising the following steps: -sampling the slag; -cool the slag sample and evaluate its color; and, -If the slag sample is green, terminate the fusion stage; or -If the slag sample is not green, continue the fusion stage after adjusting the pO<2> level to achieve higher reduction conditions. 10. Process according to any one of claims 1 to 9, wherein the level of pO2 is adjusted to 10'7 > pO<2>> 10-12, preferably to pO<2>< 10'8, more preferred to pO<2>< 10'85 and most preferred to pO<2>< 10'9. 11. Process according to any one of claims 9 or 10, wherein the color of the slag is green according to the STM D1535-14 standard. 12. Process according to any one of claims 1 to 11, wherein the oven is an electric oven. 13. Metallurgical slag containing Li having a percentage composition by weight according to: 10% <MnO < 40%; (CaO 1.5*LÍ<2>O) / AI2O3 > 0.3; CaO 0.8*MnO 0.8*Li<2>O < 60%; (CaO 2*LÍ<2>O 0.4*MnO) / YES2 > 2.0; LÍ2O > 1%; and, AI2O3 YESO2 CaO L2O MnO FeO MgO > 85%. 14. Li-containing metallurgical slag according to claim 13, wherein said slag has a green color according to ASt M D1535-14. 15. Li-containing metallurgical slag according to claim 13 or 14, wherein the MnO content in the slag is 30% or less. 16. Li-containing metallurgical slag according to any one of claims 13 to 15, wherein the CaO content in the slag is 15% or more, preferably 20% or more, and wherein the CaO content in the slag is 50% or less, preferably 30% or less. 17. Li-containing metallurgical slag according to any one of claims 13 to 16, wherein the AbO3 content in the slag is 50% or less, preferably 40% or less. 18. Li-containing metallurgical slag according to any one of claims 13 to 17, wherein the Fe content in the slag is 25% or less, preferably 10% or less. 19. Use of the metallurgical slag containing Li according to any one of claims 13 to 18 as a slag former in a pyrometallurgical recycling process. 20. Use of the Li-containing metallurgical slag according to any one of claims 13 to 18 as a slag former in the process according to any one of claims 1 to 12, thereby partially or completely replacing the slag formers in the step of providing a charge comprising slag formers.