Rate-Dependent, Li-Ion Insertion/Deinsertion Behaviour of LiFePO4 Cathodes in Commercial 18650 LiFePO4 Cells

Wednesday, 8 October 2014
Expo Center, 1st Floor, Center and Right Foyers (Moon Palace Resort)
Q. Liu (Indiana University Purdue University Indianapolis), Y. Ren (X-ray Science Division, Argonne National Laboratory), and J. Xie (Purdue School of Engineering and Technology)
Although the performance of LiFePO4 has been improved significantly, the underlying mechanism of lithium ion insertion and de-insertion in FePO4/LiFePO4 (FP/LFP) is still not clearly understood1-4. There are two basic mechanisms occurring during charge/discharge process: solid-solution reaction mechanism and two-phase reaction mechanism. In the first case – the solid solution reaction – only a single phase is involved, and as a result the lattice parameters and unit cell volumes would be expected to change continuously during the charge/discharge cycle.  In the second case – associated with a first order phase transition – the unit cell volumes of both phases remain nearly constant, with the unit cell volume difference of two phases varying by only DV=6.5%.  The distinct nature of these two mechanisms is thus apparent using X-ray diffraction, and as a result, in operando synchrotron HEXRD technique is particular amenable to revealing the underlying mechanisms.

In this study, we have investigated the structural changes that occur in LiFePO4 electrodes in commercial 18650 cells during the charge/discharge process using in operando synchrotron HEXRD. The 18650 cell (APR18650M, 1.1Ah) was provided by A123 Systems with a graphite anode, a LiFePO4 cathode, and a 1.20 M LiPF6 in EC: EMC electrolyte. No special modification was needed for the 18650 cell before characterization. Four equivalent cells have been cycled and in situ characterized under different cycling rates from 0.1 C, 1C, 3C to 5 C. 

The results show that (1) the phase fractions of both LiFePO4 and FePO4 (Fig. 1 a and b) change with SOC at both low rates (i.e. 0.1 C) and high rates (i.e. 1 C), suggesting that the LiFePO4 electrode is undergoing a two-phase reaction mechanism in the entire flat voltage plateau; (2) the unit cell volume of both LiFePO4 and FePO4 (Fig. 1 c and d) changes with the SOC at both low rates (i.e. 0.1 C) and high rates (i.e. 1 C), indicating that the electrode is experiencing the dual-phase solid-solution reaction mechanism, and (3) the difference (DV ) in the unit cell volume between the LiFePO4 phase and the FePO4 phase reduces as the rate increases (not shown here due to the page limit), implying that the region over which the dual-phase solid-solution exists will be compressed with increased rates and, eventually, may be diminished at extremely high rates5. On the other hand, our results suggest that the insertion/deinsertion process is rate dependent and the two different mechanisms, two-phase and solid-solution, co-exist during the charge/discharge process. At very low over-potential (i.e. very low rate charge/discharge), the process is dominated by the two-phase mechanism while it might be dominated by the solid solution mechanism at very high over-potential (i.e. very high rate charge/discharge). Between the two extremes, the process is controlled by both mechanisms with different ratios depending on the rate.  The proposed dual-phase solid-solution mechanism may explain the remarkable rate capability of LiFePO4 in commercial cells.