Anode-free lithium metal battery (AFLMB) is one of new promising energy storage systems. In contrast to lithium-ion battery, the negative electrode of AFLMB does not contain any active materials, resulting in a two-fold increase in energy density. As a result, the charging mechanism on the negative electrode in AFLMB is based on electrodeposition of lithium, rather than intercalation of lithium ions. Unfortunately, that causes a series of problems, such as lithium dendrite formation, volume expansion, broken SEI and dead lithium accumulation, leading to the low cycling stability of AFLMB. Consequently, interfacial characteristics of the negative electrode (usually copper) play a critical role in determining the morphology of lithium deposits which largely affects the issues mentioned above. Understanding the mechanism of lithium electrodeposition across the electrode-electrolyte interface is crucial to design a better copper electrode for long-cyclability AFLMB.The interfacial characteristics of copper are mainly influenced by crystal phase, surface roughness and defects. While the effects of crystal phase and surface roughness on lithium electrodeposition have been extensively studied, the role of micro-scale defects has been seldom discussed, despite their potential to strongly influence the surface properties of copper. Therefore, this work focuses on elucidating the elementary process about how the defects on copper affect the lithium nucleation and subsequent growth. Specifically, we use (111)-nanotwinned and (111)-preferred orientation copper foils with the same surface roughness to conduct a series of electrochemical measurements. The presence of nanotwin structure can reduce the grain boundary and atomic step density, a type of the defects with high surface energy, on the copper. Before lithium electrodeposition, solvents and anions in an electrolyte adsorb on the electrode-electrolyte interface to form the electric double layer. During this process, the atomic steps with high surface energy on copper can establish strong adsorption with the substance in the electrolyte. The solvents adsorbed on the defects may hinder the reduction of anion. Accordingly, the (111)-nanotwinned Cu foil with fewer defects induces the formation of LiF-rich solid electrolyte interphase (SEI), which is beneficial to the transportation of lithium ions. In cyclic voltammetry test, the (111)-nanotwinned Cu foil exhibits higher reduction current density of the precipitation of LiF (Fig. 1).For lithium electrodeposition, during the initial stage, lithium adatoms tend to adsorb on the defects with high surface energy on the copper surface, where other adatoms accumulate to form stable nucleus. Subsequently, the nucleus on each nucleation sites might have different growth direction and crystal phase, resulting in a lot of independent lithium grains. In other words, compared to the (111)-preferred orientation substrate, the (111)-nanotwinned Cu foil with less nucleation sites provides more space for the nucleus to grow in 2-D direction, which leads to the lower nucleation overpotential (Fig. 2 a). That result also facilitates the formation of larger lithium grains with smaller surface area exposed to the electrolyte (Fig. 3). As shown in Figure 2 (b), quasi-steady state over-potential of Li plating on the (111)-nanotwinned Cu is generally larger than that on the (111)-preferred orientation Cu foil, which has been ascribed to the high local current density on larger lithium grains. The quasi-steady state overpotential of Li plating on the (111)-preferred orientation Cu foil gradually increased with the cycle number, which might be attributable to the relatively unstable SEI and the accumulation of dead lithium with the GCD cycling. Furthermore, the (111)-nanotwinned Cu foil always exhibits a lower quasi-steady state overpotential of lithium stripping in comparison with the (111)-preferred orientation substrate at each cycle, probably resulting from the uniform Li deposit and the special LiF-rich SEI formation due to the nanotwinned structure. In the cycling stability test (Figure 2 c), for the Cu//Li cell using the (111)-nanotwinned Cu foil, the coulombic efficiency was kept being around 90% in this 160-cycle test. However, for the Cu//Li cell using the (111)-preferred orientation Cu foil, the coulombic efficiency is generally lower than that of the cell using the (111)-nanotwinned Cu foil at the cycle number > 15. In addition, the coulombic efficiency starts to obviously decay at the 135th cycle and becomes lower than 5% at the 160th cycle.To sum up, we proposed a mechanism about how the defects affect the reaction at electrode-electrolyte interface. The nanotwinned columnar grains reduces the atomic density on the copper surface, which effectively reduces the formation of lithium dendrite, the accumulation of dead lithium and broken SEI. The (111)-nanotwinned Cu foil is proved to be a superior substrate for lithium deposition and stripping, significantly improving the Li+/Li redox reversibility and cycling stability. Figure 1
Read full abstract