Lithium nickel manganese spinel (LiNi_0.5Mn_1.5O₄, LNMO) is a promising cathode material for high-energy applications due to its operating potential around 4.7 V vs. Li/Li⁺. Lab-scale LNMO/graphite and also LNMO/silicon cells show very stable cycling behavior as long as the cells contain excess active lithium, which is typically achieved by prelithiating the anode.^1,2 In contrast, if the anode is not pre-lithiated and thus the amount of active lithium is limited, strong capacity fade is observed.³ This effect is more severe in LNMO cells with silicon-based anodes,⁴ where the expansion and contraction of the silicon particles during (de-)lithiation leads to continuous exposure of fresh, unpassivated silicon surface, causing ongoing electrolyte reduction and consuming the limited lithium reservoir in the cell. Different methods to prelithiate silicon anodes on a lab scale have been developed,^5–7 however these often include metallic lithium, which is difficult to handle on an industrial scale.

Alternatively, so-called “sacrificial salts” have been suggested as an additional source of active lithium.⁸ These lithium salts are added to the cathode composite electrode. During the initial charge, the anion of the sacrificial salt is oxidized to a gas, thereby releasing the lithium cation; the gas can then be removed after formation. Lithium oxalate has a high specific capacity of 525 mAh/g, but was disregarded as a sacrificial salt for 4 V cell chemistries due to its high oxidation potential around 4.6 V vs. Li/Li⁺.⁸ However, this potential matches well with the charge/discharge plateaus of LNMO. Further, lithium oxalate releases only CO₂ during oxidation, which is innocuous towards the lithium-ion cell chemistry.⁹

In the present study, we use lithium oxalate as a capacity enhancer in LNMO/graphite and LNMO/silicon-graphite (Si:G) full cells. First, we test the addition of 2.5% or 5% of lithium oxalate to LNMO composite electrodes with 90% active material in half cells, showing an increased charge capacity and lithium inventory of 10% or 20%, respectively. Using online electrochemical mass spectrometry (OEMS), we demonstrate the quantitative oxidation of lithium oxalate during the first charge of a LNMO-lithium oxalate composite electrode by tracking the CO₂ evolution. To investigate the behavior of the increased lithium inventory in full cells, we cycle LNMO composite electrodes containing 0%, 2.5% or 5% lithium oxalate against balanced graphite anodes. Further, we also test LNMO composite electrodes with 5% or 0% lithium oxalate against silicon-graphite electrodes (see Figure 1) in an FEC-containing electrolyte. Interestingly, the cells with lithium oxalate (green symbols) showed less capacity fade and higher coulombic efficiency throughout cycling than their counterparts without lithium oxalate (black symbols). As the removal of CO₂ from lithium oxalate oxidation after formation leads to a similar capacity fading rate as in LNMO/Si:G cells without lithium oxalate, we attribute the improved cycling performance to the presence of CO₂. An assessment of the FEC consumption by post-mortem ¹⁹F-NMR in cells with or without lithium oxalate further indicates that CO₂ is an effective SEI-forming additive for silicon-graphite anodes, and that a combination of FEC and CO₂ potentially improves the lifetime of commercial full cells with silicon-graphite anodes.

References:

1. L. Hu, K. Amine, and Z. Zhang, Electrochem. Commun., 44, 34–37 (2014).
2. K. Fridman, R. Sharabi, R. Elazari, G. Gershinsky, E. Markevich, G. Salitra, D. Aurbach, A. Garsuch, and J. Lampert, Electrochem. Commun., 33, 31–34 (2013).
3. L. Hu, Z. Zhang, and K. Amine, Electrochem. Commun., 35, 76–79 (2013).
4. J. Arrebola, A. Caballero, H. Göhr, L. Hernán, J. Morales, and L. Sánchez, Electrochem. Commun., 11, 1061–1064 (2009).
5. C. Jarvis, M. Lain, M. Yakovleva, and Y. Gao, J. Power Sources, 162, 800–802 (2006).
6. I. Seong, K. Kim, and W. Yoon, J. Power Sources, 189, 511–514 (2009).
7. N. Liu, L. Hu, M. McDowell, A. Jackson, and Y. Cui, ACS Nano, 5, 6487–6493 (2011).
8. D. Shanmukaraj, S. Grugeon, S. Laruelle, G. Douglade, J.-M. Tarascon, and M. Armand, Electrochem. Commun., 12, 1344–1347 (2010).
9. J. Besenhard, M. Wagner, M. Winter, A. Jannakoudakis, P. Jannakoudakis, and E. Theodoridou, J. Power Sources, 44, 413–420 (1993).

 

Acknowledgements:

This work is financially supported by the BASF SE Battery Research Network.

Figure 1: Specific charge/discharge capacity (closed/open symbols, upper panel) and coulombic efficiency (lower panel) of LNMO/Si:G full cells with 5% Li₂C₂O₄ (green circles) or 0% Li₂C₂O₄ (black squares). The LNMO electrodes (90% AM, 2.4 mAh/cm²) and Si:G electrodes (35% nano-Si, 45% graphite, 3 mAh/cm²) were combined with a glassfiber separator soaked with 80 µL of LP57 (EC:EMC 3:7 in weight, 1M LiPF₆) + 5% FEC in 2032-type coin cells. Cells were cycled between 4.0-4.8 V at C/2 and 25 °C. All symbols represent the average of two identical cells.