Understanding Contamination in Li-ion Battery Raw Materials

Summary

Impurities and contaminants in the material used in the production of Li-ion batteries can have catastrophic impacts on the finished products. As such, monitoring of the quality and cleanliness of materials throughout the production process is essential if contaminants are to be found and their sources controlled. This monitoring must start with the raw materials produced at the mine and continue through to the final battery grade powders.

This application note demonstrates a scanning electron microscopy (SEM) based solution for automatic detection and identification of impurities in battery raw material powders using an Ultim Max EDS detector and AZtecBattery. By taking samples at different steps of the production process, it is possible to identify where contaminants are introduced; thereby allowing sources to be identified and solutions developed.

Introduction

Li-ion batteries have been a key enabling technology over the last decade and are considered integral to EV (Electric Vehicle) development. The materials used in the production of Li-ion batteries must be of high purity in order for the batteries made from them to have the expected performance and lifetime and also to prevent dangerous failures caused by metallic particles penetrating isolating barriers.

Quality control and monitoring of materials throughout the manufacturing process is crucial as even very small amounts of contaminants can have catastrophic consequences. For the metals used in batteries, the process starts at the mine with the verification of the purity of the material which has been produced. This material is typically transported to a materials processing facility which creates the battery precursor material which is in turn supplied to the battery manufacturing plant. In this process there are several potential sources of contamination, meaning that it is vital to verify if the materials are contaminant-free at each stage.

Schematic cross section of a cylinder battery. Inset: NCM811 cathode precursor powder particles

For the final precursor material to be of "battery grade" it is often the case that impurities must be removed from the powder which can be both a time consuming and expensive process. Therefore, it is important to have a simple, reliable and fast process to identify impurities to ensure that the final products are suitable.

Cathodes in Li-ion batteries are typically made from nickel, cobalt and manganese although other chemistries are also being developed. Cobalt is of high demand for battery production and the supply of new mined material is struggling to keep up with the demand, with implications for its price. This makes it even more essential to ensure that the material is not contaminated and therefore wasted.

In this application note we consider contamination of CoO powder. Here the primary concern is contamination by large particles (diameter above 1 µm) of Ag, Cu or Zn as these are the contaminants that have been found to be most likely to lead to failure of the final product.

SEM-based Analysis

Contamination particles can be identified in the SEM using BSE (Backscatter Electron) imaging detectors in combination with EDS analysis. Particles with higher mean atomic Z number appear brighter in BSE images and the EDS system can be used to identify those particles. This approach is effective when performed manually but is very laborious when large numbers of particles are analysed and can be open to (often unconscious) bias.

AZtecBattery provides an automated detection and analysis method, which makes it easy to rapidly locate, identify and classify particles distributed over a large sample area while maintaining high spatial resolution.

Figure 1 - a)Thresholded BSE image used to identify particles. b) Selected Cu rich feature. c) Spectrum from the selected feature

Figure 1 shows an example of the automated process; first a grey level threshold is applied to the BSE image in order to identify the location of the particles and measure their morphology. EDS data is then acquired from the detected particles. On the basis of this compositional information a classification scheme is used to automatically count the number of particles of different types containing unwanted elements.

Analysis

Three samples were analysed using AZtecBattery: one sample from a mine, one sample from a delivery truck and one sample from a processing plant.

The sample from the mine and the sample from the processing plant consisted of powder material dispersed onto a carbon tab, whereas the sample from the truck was collected from a dust filter.

All three samples were analysed with AZtecBattery using an Ultim Max 170 EDS detector in combination with BSE imaging. A threshold for the grey level in the BSE image was defined to ensure that EDS data was only collected from potential contaminant particles to ensure high analysis speeds. The analysis was performed at an accelerating voltage of 20 kV covering a 1 mm² area on all samples.

Figure 2 – Number of contamination particles from samples taken at different steps along the production process.

The raw material from the mine did not contain Zn or Ag and only a small amount of Cu. The dust sample from the delivery truck contained a large amount of Cu rich particles which could explain the Cu in the sample taken at the processing plant. The Ag contaminants appear to have been introduced at a later stage in the process.

There are several processes which can be used to remove the contamination particles from the powder material; however, each is expensive and time consuming. By identifying the stage at which contamination occurs in the way described here, a significant amount of both time and money can be saved by solving the problem at source. In this specific case, it was determined that implementing simple cleanliness controls in the transport stages would have a significant positive effect in reducing contamination. Furthermore, the data indicates that contamination also took place in the processing plant suggesting that further investigation of the sources of the particles found in this stage could indicate additional actions that could be taken to prevent contamination from other process lines or the handling of material.

Conclusion

Ultim Max EDS detectors and AZtecBattery offer an easy and fast solution for identification of contaminants, allowing the quality of the material to be monitored all the way from the mine to the final product. This is important because contamination can be introduced at any step along the process from the initial production of raw material to manufacture of the final product. The cost of failing to deal with this contamination is potentially very high as failed batteries can have severe consequences.