XRF Analysis of Lithium Iron Phosphate Cathodes

Introduction

LiFePO₄, normally referred to as LFP, is a major cathode material used by lithium-ion battery industry. LFP has its advantage in superior safety and lower material cost compared to other popular chemistries like NMC (LiNixMnyCo1-x-yO2). Though LFP has lower energy density than NMC, this gap is diminishing fast with new battery manufacturing and assembling techniques, like Cell to Pack (CTP) and blade battery designs.

There are several approaches for the manufacturing of LFP, such as high-temperature solid-state fusion of FePO₄ precursor with Li₂CO₃, co-precipitation of LiFePO₄ precursor followed with high-temperature sintering, carbothermal reduction of Li, P and Fe precursors, and hydrothermal synthesis. Each approach has its own pros and cons in the process and the final product quality control. Whatever is the manufacturing method, a strict control on the molar ratios of the raw materials and the elemental composition of the final product is critical to the performance of the assembled battery.

X-Ray Fluorescence (XRF) spectroscopy, with its advantage of high stability and simple sample preparation, is an ideal technique to analyze major elements in LFP cathode manufacturing, from raw material to the final product. A study is shown in this report as an example.

Instrumentation

Malvern Panalytical Zetium floor standing WDXRF spectrometer equipped with an Rh anode SST-mAX tube was used in these measurements. Designed to meet the most demanding process control and R&D applications, the Zetium XRF spectrometer leads the market in high-quality design and innovative features for sub-ppm to percentage level analysis of elements ranging from Be to Am. The Zetium is equipped with the latest version of our SuperQ software, including our Virtual Analyst that enables even non-expert users to easily set up applications. 

Sample preparation

Samples were prepared in two ways: Borate fusion and pressed pellets. Due to the limitation in matrix match, calibration samples were prepared only for the fusion application. The method precision of the pressed pellet application was evaluated with a single-point calibration.

Fusion preparation

To create a calibration for the fusion method, synthetic reference materials were prepared using Fe₂O₃ and Li₃PO₄ mixtures in varying stoichiometric ratios. 

To achieve an optimal precision for Fe in the sample, Co was used as an internal standard. A flux doped with 10% Co₂O₃ is used during the fusion. A dilution ratio of 1:10 was used for the calibration standards, while 0.8:10 was used for the samples in order to match the concentration. The samples were then fused with TheOX automatic fusion machine at 1120℃.

Pellet preparation

5g of LFP powder was dried and pressed into pellet with boric acid as backing.

Measurement procedure

Below measurement conditions were used for fuse bead samples. Total measurement time was 160 seconds per sample. Li₂O is used as a balanced compound.

Item		Tube		Optical path				Time (s)
Compound	Line	kV	mA	Tube filter	Collimator	X-tal	Detector
(Co)	KA	60	40	None	150 µm	LiF 220	Scint.	20
Fe2O3	KA	60	40	None	150 µm	LiF 220	Scint.	60
P2O5	KA	24	100	None	550 µm	Ge 111-C	Flow	80

Table 1. Measurement conditions

The same measurement conditions were used for press pellet samples, except that Co was not present in these. 

Calibration and results

Figures 1 and 2 show the calibrations using fuse bead synthetic reference samples. A single-point calibration (not shown here) was made for the pressed pellet samples.

Figure 1. Fusion calibration for Fe₂O₃

Figure 2. Fusion calibration for P₂O₅

Since data consistency is quite important for process control, sample preparation and measurement repeatability were tested to validate the XRF application on LiFePO₄ samples. The results are given in tables 2 and 3 below.

Pressed pellet				Fused bead
Sample	P2O5	Fe2O3	(Li2O)	Sample	P2O5	Fe2O3	(Li2O)
1	47.640	48.200	4.160	1	47.677	48.108	4.215
2	47.679	48.129	4.193	2	47.675	48.219	4.105
3	47.683	48.095	4.223	3	47.677	48.106	4.217
4	47.465	48.084	4.588	4	47.618	48.047	4.335
5	47.789	48.164	4.047	5	47.683	48.139	4.179
Average (%)	47.624	48.134	4.242	Average (%)	47.666	48.124	4.210
Minimum result (%)	47.465	48.084	4.047	Minimum result (%)	47.618	48.047	4.105
Maximum result (%)	47.789	48.200	4.588	Maximum result (%)	47.683	48.219	4.335
RMS (%)	0.174	0.048	0.204	RMS (%)	0.027	0.063	0.083
Relative RMS (%)	0.37	0.10	4.82	Relative RMS (%)	0.06	0.13	1.97

Table 2. Sample preparation repeatability for pressed pellet and fusion sample preparations

Pressed pellet				Fused bead
Measurement	P2O5	Fe2O3	(Li2O)	Measurement	P2O5	Fe2O3	(Li2O)
1	47.789	48.164	4.047	1	47.736	48.106	4.158
2	47.780	48.152	4.068	2	47.737	48.163	4.100
3	47.759	48.177	4.064	3	47.731	48.201	4.068
4	47.757	48.173	4.070	4	47.740	48.155	4.105
5	47.770	48.152	4.078	5	47.736	48.243	4.020
6	47.779	48.152	4.069	6	47.724	48.203	4.073
7	47.785	48.122	4.093	7	47.721	48.243	4.037
8	47.796	48.125	4.079	8	47.709	48.283	4.008
9	47.762	48.158	4.080	9	47.716	48.229	4.055
10	47.672	48.181	4.147	10	47.686	48.247	4.067
Average (%)	47.765	48.156	4.080	Average (%)	47.723	48.207	4.069
Minimum result (%)	47.672	48.122	4.047	Minimum result (%)	47.686	48.106	4.008
Maximum result (%)	47.796	48.181	4.147	Maximum result (%)	47.740	48.283	4.158
RMS (%)	0.035	0.020	0.027	RMS (%)	0.017	0.053	0.044
Relative RMS (%)	0.07	0.04	0.65	Relative RMS (%)	0.04	0.11	1.09

Table 3. Sample measurement repeatability for pressed pellet and fusion sample preparations

Summary and Conclusion

Above data demonstrate the capability of XRF in LFP material analysis. It shows high precision in sample analysis loop with fully automatic sample preparation and measurement. With further refinement, it is possible to reach precision similar to titration methods.