Resource demand for the production of different cathode materials for lithium ion batteries

Highlights

• Quantification of resource consumption of five different cathode active materials.

• Process inventories based on data directly provided by industry.

• Most important contributors: metal supply or electricity consumption.

• Cathode powder properties depend on battery design and strongly influence results.

Abstract

With the expansion of the lithium ion (Li-ion) battery market, new materials for lithium ion cathodes are constantly being developed. Especially automotive applications require a decrease in production costs, which often means to increase the content of less expensive metals. However, these composition changes also affect the cathode properties and may also require considerable changes in the processing conditions. There are still considerable gaps regarding life cycle inventories for cathode materials as the availability of commercial process data is limited. The presented study had two main objectives: (1) quantifying the natural resource use for the production and recycling of five cathode materials for Li-ion batteries in a closed loop scenario based on data directly provided by industry and (2) assessing the impact of differences in composition, cathode material properties and production technology. An exergy-based method was employed to assess the cumulative resource use. To get a better view on the impact of the property differences of the cathode materials, resource consumption was expressed per kg, per kWh one cycle and per kWh over the cycle life. The latter was the actual functional unit. The results per kg of cathode were comparable for all (290–346 MJ_ex/kg) but one (622 MJ_ex/kg) of the cathode materials. Metal supply and energy use during cathode material production were the main drivers of natural resource use. Due to the diverging characteristics of the cathode materials the relative results in terms of the functional unit (0.39–0.70 MJ_ex/kWh) differed considerably from the results per kg of cathode material. For example, while the resource use for one of the cathodes was relatively high per kg of material, it was similar to the resource use of two other cathode materials per kWh (cycle life). This implies that it is not sufficient to have good process data to compare the resource consumption of different cathode materials. The properties of the cathode materials, respectively the battery as a whole, have to be carefully determined in function of the application. Two of the cathode materials were developed with the target to reduce cost of feedstock metals while maintaining performance. Indeed, those two cathodes showed low resource use per kg (290–343 MJ_ex) and per kWh (one cycle) (377–463 MJ_ex).

Introduction

For quite a while now Li-ion batteries have been employed in mobile devices like laptops and mobile phones. On top of that they are powering (hybrid) electric vehicles, which might be a “promising technology” (Grünig et al., 2011) for reducing greenhouse gas emissions and which are supported by policies around the world (Lindquist and Wendt, 2011). By now the lithium ion (Li-ion) batteries and lithium polymer batteries make up the large majority of the rechargeable battery market (Goonan, 2012). As the processing of metals, which are constituents of many battery components, is typically quite energy intensive, the question of the resource use in their production is quite important. Indeed, a number of life cycle studies have been performed already with Li-ion batteries as main target (Dunn et al., 2012, Gaines et al., 2011, Majeau-Bettez et al., 2011, Notter et al., 2010, Zackrisson et al., 2010) or as component in an application (Rydh and Sandén, 2005, Samaras and Meisterling, 2008). According to some of those studies the cathode, which in addition to the active material also contains some other materials, is an important component of a Li-ion battery also with respect to environmental impacts. According to Majeau-Bettez et al. (2011) the cathode paste production causes more than 35% of the total global warming impact of a Li-ion battery, while in the study of Notter et al. (2010) the cathode active material accounts for 12.5% of the Cumulative Energy Demand (CED) and 13.8% of the Global Warming Potential (GWP) of the Li-ion battery. Unfortunately, industry data for the modeling of the life cycle inventories (LCI) of anode and cathode materials is usually lacking so that the inventories are approximated based on patents and process descriptions. Regarding battery recycling some more industry data are available, which have been used to model LCIs (Dunn et al., 2012, Fisher et al., 2006). These data are not very complete though or important treatment steps to obtain a material which can be used again for Li-ion cathode production are missing.

The first and foremost objective of the study was to establish gate to gate inventories of production processes for Li-ion battery cathode active materials based on primary industry data. In the following the term cathode material will be used to refer to the cathode active material. As new cathode materials are constantly developed, five different cathode materials were selected representing different stages of development and having different properties. In addition, industrial Li-ion battery recycling processes, recovering usable nickel and cobalt products, were to be inventoried. Subsequently, the cumulative natural resource use to manufacture the Li-ion cathode materials from these recovered nickel and cobalt streams was to be quantified. No specific battery application was targeted due to the diverse nature of the cathode materials. Therefore, no application specific functional unit has been chosen. Nevertheless, a functional unit was to be selected which included at least the more generic characteristics of the cathode materials in order to capture aspects with respect to the service delivered by the various chemistries. The different cathode materials were to be compared, while keeping in mind their different stages of development and slight differences in application. In parts this study is an update and extension of a previous paper by Dewulf et al. (2010), which dealt with only one cathode material and only quantified resource use per kg of material.