Surface Chemistry Evolution in Blended Lithium-Ion Battery Cathode Materials

Abstract

Listen

Translate article

For energy storage applications in which power and safety are paramount, lithium ion batteries (LIB) containing a lithium-iron phosphate (LFP) positive electrode are a common choice. However, due to the relatively low operating potential of LFP and the high proportion of inactive materials in high-power LIB, the energy density of these devices can be well below 100 Wh/kg. Meanwhile, advanced cathode chemistries such as nickel-rich layered compounds and high-voltage spinels could offer improved energy densities, but it’s unclear whether these will ever offer the same power density and safety as LFP cells. Part of the limitation of alternative cathode chemistries is the same avenue by which energy density is increased: significantly higher cathode potentials relative to LFP. One example are spinel-structured manganese oxide compounds with the general formula LiMn2-xMxO4-yNy (LMO) where M are various transition metal oxide cationic dopants and N are typically halogen anionic dopants. Heavily doped spinels have acceptable power, safety, and capacity retention relative to LFP but relatively low specific capacities. Meanwhile, layered nickel-manganese-cobalt compounds with the general formula LiNixCoyMnzO2(NCM) can demonstrate greater than 200 mAh/g when charged to high upper cut-off voltages (4.5 V vs. Li/Li+ and above), but suffer rapid capacity fade as a result. Blends of spinel and layered cathode compounds have been reported as one avenue to achieve a composite electrode with improved power capability, safety, and capacity retention (1) (2) (3). For layered NCM compounds differing explanations for capacity loss when charged above 4.5 V vs. Li/Li+ have been presented. Kang et. Al. and others suggested that capacity loss in NCM materials cycled to high upper cut-off voltages is due to a structural rearrangement at the near-surface of the layered oxide. This structural degradation leads to a dramatic increase in electrochemical polarization, and is in contrast to the formation of a resistive surface film formed by electrolyte decomposition (4) (5) (6). Still, electrolyte additives have been found to suppress structural re-arrangement of NCM compounds, indicating that the electrode-electrolyte interface, rather than a purely structural phenomenon, is involved in NCM capacity loss (5) (6). Thus, from these studies it’s not clear how a simple mechanical mixture of spinel and layered compounds could influence the interfacial degradation of NCM. Our work has found that there are indeed significant changes to LiNi0.5Co0.2Mn0.3O2 surface chemistry as observed by x-ray photoelectron spectroscopy (XPS) after a few electrochemical cycles indicating newly formed C-O and C-F bonds, shown in Figure 1A. Meanwhile, when comparing the potentiodynamic behavior of the layered LiNi0.5Co0.2Mn0.3O2 to high voltage spinel-structured LiNi0.5Mn1.5O4, an apparent irreversible oxidation peak is observed in the layered compound near 4.8 V vs. Li/Li+, which is coincident with the split reversible oxidation peaks observed for the spinel compound. This overlap in apparent reversible and irreversible behaviors may play a key role in determining the surface chemistry of blended cathode materials and will be discussed further during our presentation.