Short-Range Order and Unusual Modes of Nickel Redox in a Fluorine-Substituted Disordered Rocksalt Oxide Lithium-Ion Cathode

Abstract

Fluorine substitution for oxygen in cation-disordered lithium-excess transition metal oxides (Li1+xTM1-xO2) used as lithium-ion cathodes was recently demonstrated to improve the reversibility of the processes taking place on charge and discharge by reducing the amount of oxygen loss on charge and preventing major structural rearrangements at high voltage. Yet, little is understood about how fluorine incorporates the oxide structure and impacts its electrochemical properties. Here, we use a combination of experimental (solid-state Nuclear Magnetic Resonance (NMR) spectroscopy) and theoretical techniques (density functional theory (DFT) calculations and Monte Carlo simulations) to investigate the evolution of the local structure around fluorine and lithium, and the oxidation state of redox-active nickel during charge and discharge of the Li1.15Ni0.45Ti0.3Mo0.1O1.85F0.15 (LNF15) cathode. We show that fluorine doping introduces short-range order in as-synthesized LNF15, by incorporating in lithium-rich sites with five or six lithium nearest-neighbors. We observe the emergence of new signals in the ex situ 19 F NMR spectra taken at high states of charge, which we tentatively assign to undercoordinated, diamagnetic fluorine environments seen in our computed models. Our theoretical results also suggest that octahedral nickel ions directly bonded to fluorine follow a different oxidation mechanism than those surrounded by six oxygens, forming Ni 3+ intermediates instead of oxidizing from Ni 2+ directly to Ni 4+. While the oxidation of Ni 2+ towards Ni 4+ is incomplete in oxides, due to overlap between the oxygen and nickel valence states, this result suggests that fluorination may be an efficient strategy to utilize the Ni 2+ /Ni 4+ redox reservoir to a greater extent.

Effect of Fluorination on Lithium Transport and Short‐Range Order in Disordered‐Rocksalt‐Type Lithium‐Ion Battery Cathodes

“…Previous reports have established that most DRX systems can achieve at least 5% of fluorination by solid state synthesis, a value that is within the solubility limits predicted by our calculations in Table . Therefore, we first evaluate the impact of 5% F‐incorporation at 1273 K by comparing the 0‐TM percolation in Li 1.2 M x 1 M' x 2 O 1.9 F 0.1 and Li 1.2 M x 1 M' x 2 O 2 for the seven (M, M') pairs.…”

“…The phase diagram using the qualitative pair‐interaction model of Section 2.1 is shown in Figures S3 and S4 (Supporting Information) and defines the temperature and F‐contents achievable. We chose 7.5% F substitution which was previously estimated to be the thermal F solubility limit at 1273 K in several DRX materials . The percolation map in Figure shows the predicted 0‐TM capacity as function of temperature and Li‐excess.…”

“…In addition to the general trends that apply to all systems, one system‐specific trend can be observed in Figure 5e: The two Ni 2+ ‐based systems need less Li excess to achieve good 0‐TM percolation than the Mn 2+ ‐based systems, especially at low Li‐excess ( x < 1.1). This makes Ni 2+ system of interest to achieve high Li capacity, but needs to be weighed against the issue that in Ni systems electron capacity limitations may dominate because the Ni 4+ valence state is typically difficult to access in DRX materials due to its overlap with the oxygen states . Furthermore, the choice of high‐valent TM species also appears to have an impact on percolation, although no definitive trends emerge: when paired with Ni 2+ , Ti 4+ requires less fluorination to reach a percolating amount of 0.6 Li/f.u at moderate Li excess content ( x < 1.2) than the Nb 5+ analogue, while the opposite is true for Mn 2+ .…”

“…Cation SRO is known to exist in DRX materials and varies significantly depending on the TM species present . With the introduction of fluorine, the strong bonding preference between Li and F atoms induces additional SRO . As suggested in previous work, this interaction may give rise to a significant modification of the Li‐ion transport upon F incorporation, but an in‐depth understanding of the effect of fluorination on bulk Li transport is lacking.…”

“…Experimental and computational studies have demonstrated that LiF bonds are strongly preferred over MF bonds (M = other metal cation species) in F‐DRX materials. Therefore, the Li–F interactions are considered net attractive (negative energy contributions), whereas the M–F interactions are considered net repulsive (positive) as compared to the average interaction (see schematic in Figure S1, Supporting Information).…”