Lithiation Mechanism and Improved Electrochemical Performance of TiSnSb-Based Negative Electrodes for Lithium-Ion Batteries

Abstract

Lithium alloying materials are promising candidates to replace the current intercalation-type graphite negative electrode materials in lithium-ion batteries (LIBs) due to their high specific capacity and relatively low cost. Here, we investigate the electrochemical performance of TiSnSb regarding its charge/discharge cycling stability as a negative electrode material in LIB cells. To assess a more practical performance with respect to a limited active lithium content in LIB full-cells, we evaluate the impact of pre-lithiation for TiSnSb with respect to the cycling stability in a NCM111||TiSnSb cell setup. The observation of the individual electrode potentials reveals comprehensive insights into the ongoing cell chemistry, showing that clear performance improvements can be achieved via pre-lithiation. Furthermore, the lithiation mechanism of TiSnSb is systematically studied via ex situ 7Li magic-angle spinning (MAS) nuclear magnetic resonance (NMR), ex situ X-ray diffraction, and static ex situ 119Sn wideband uniform rate smooth truncation Carr–Purcell Meiboom–Gill (CPMG) WCPMG NMR experiments. For comprehensive references regarding the isotropic 7Li shift of the Li–Sb intermetallic phases, all thermodynamically stable Li–Sb intermetallics of the binary Li–Sb systems have been synthesized and subsequently characterized by 7Li MAS NMR. Combined, our measurements for lithiated TiSnSb do not give any evidence for the formation of Li–Sn and Li–Sb intermetallics related to crystalline bulk phases (Li7Sn3, Li7Sn2, Li3Sb, and Li2Sb) as has been previously reported. In contrast, unique insights obtained from static ex situ 119Sn WCPMG NMR and ex situ XRD measurements reveal the formation of ternary Li–Sb–Sn species during lithiation, which can be assigned to the intermetallic phase Li2.8SbSn0.2. Additionally, the 7Li MAS NMR measurements combined with the observed discharge capacity reveal a second Li species, which we assign to an amorphous Li–Sn phase.

Pushing the Detection Limit of Static Wideline NMR Spectroscopy Using Ultrafast Frequency-Swept Pulses

“…In particular for static UW ssNMR, the large nuclear spin interactions (e.g., the chemical shift anisotropy (CSA) or the quadrupolar interaction) dominate the inhomogeneous contribution, often causing the inhomogeneous contribution to dominate the entire transverse relaxation behavior, i.e., T 2 * < τ d ≪ T 2 . In such cases, CPMG acquisition has been extremely successful to boost the sensitivity despite using relatively long WURST pulses, e.g., as demonstrated for 14 N, 35 Cl, 127 I, 195 Pt, or 207 Pb, where a large number of echoes (often more than 100) can be recorded, ,,,, and WURST pulses can be tuned to minimize line shape distortions. , However, a significant homogeneous contribution may hamper signal acquisition under conventional WCPMG conditions to an extent that barely a single echo can be recorded. , This requires appropriate WURST-pulse-design strategies to be applied to minimize τ d , in turn extending the applicability of static UW ssNMR to new fields of materials.…”

“…Using UF WURST pulses with short ring-down delays and narrow acquisition windows, τ can be reduced significantly, τ̃ < 20 μs. A compelling example for a more extreme challenge and first-hand application of UF WURST pulses has been included in our recent study of the lithiation mechanisms for titanium tin antimony (TiSnSb)-based electrode materials in lithium-ion batteries . Here, ex situ 119 Sn WCPMG NMR was employed to unravel the different phases TiSnSb forms after lithiation.…”

“…The 119 Sn WCPMG NMR spectrum shown in Figure was ultimately acquired employing 10 μs UF WURST-5 pulses sweeping over Δ = 1 MHz ( R = 100 kHz/μs), where the additional amplitude truncation (N = 5) was used to be able to further reduce ringdown delays. This ultimately allowed us to not only acquire a meaningful 119 Sn NMR spectrum and extract the corresponding shift-anisotropy parameters (blue line), but also to obtain unique insights into the lithium mechanism by comparison with other static 119 Sn fingerprint NMR signals …”

“… 119 Sn WCPMG NMR of Li 2.8 Sn 0.2 Sb (suggested composition for the TiSnSb electrode after lithiation) obtained at 7.05 T using a 10 μs UF WURST-5 pulse (Δ = 1 MHz, R = 100 kHz/μs). The numerical model with δ iso = 22.5 kHz, Δδ = 247.5 kHz, and η = 0.08 is shown in blue lines.…”

“…14,27 However, a significant homogeneous contribution may hamper signal acquisition under conventional WCPMG conditions to an extent that barely a single echo can be recorded. 29,30 This requires appropriate WURST-pulse-design strategies to be applied to minimize τ d , in turn extending the applicability of static UW ssNMR to new fields of materials. As commonly employed in modern frequency-sweeping schemes, 32−34 the rf-amplitude of WURST pulses is attenuated in the beginning and at the end of irradiation.…”