第一作者：Mark Aarts
通讯作者：Philippe M. Vereecken，Mark Aarts
Correspondence unit: Belgian Microelectronics Research Center
【Background】
The control of lithium metal morphology during electrodeposition has become a major topic in battery research, and many works have focused on the stability of lithium metal anodes during repeated deposition and stripping cycles. However, it is often difficult to visualize the full picture of the post-cycling electrode using scanning electron microscopy (SEM), as illustrated by recent mechanistic insights gained from studying lithium metal growth using various additional techniques in the literature. In addition, there is a wide variety of electrolyte systems, making it difficult to compare results or unravel the effects of physics and chemistry that lead to uncontrolled lithium growth. For example, the growth of smooth lithium due to electrostatic shielding, which was later found to consist of a special self-aligning and dense nanorod morphology.
This columnar topography is very similar to the results obtained by Kanamura et al. first with HF additives and then with trace amounts of water to form HF in situ by hydrolysis of the LiPF6 electrolyte. This growth pattern is often attributed to the LiF solid electrolyte interphase (SEI) layer. So far, the deposition of columnar lithium layers has attracted interest due to their smooth or "dendrite-free" topography and has also shown better results and coulombic efficiency in terms of lifetime. In particular, a unique type of deposition and stripping along the length of the column was observed in cycling, which meant a reduction in local current density and limited volume change in the lithium metal anode, mitigating numerous side reactions in the lithium metal battery.
Recently, however, questions have been raised about the role of LiF, and similar structures have been observed in (non-hydrolyzed) LiTFSI and LiNO3 electrolytes. Since most of the above literature advocates the formation of LiF-rich SEI by electrocatalytic reduction of HF as the cause of columnar growth, a more comprehensive understanding is needed for further optimization.
【Brief introduction of achievements】
Here, Prof. Philippe M. Vereecken and Prof. Mark Aarts from the Belgian Microelectronics Research Center demonstrate the electrodeposition process of columnar lithium topography obtained directly with water as an additive. The authors used non-hydrolyzed LiTFSI and fluorine-free LiClO4 to achieve these morphologies, unequivocally demonstrating that LiF-rich SEI cannot be considered as the sole determinant of this growth pattern. At the same time, the role of the formation step (i.e., the electrochemical reaction that occurs at a potential higher than the lithium deposition potential) was investigated using an ex situ glove box atomic force microscopy (AFM) and electrochemical experiments. No template layer was observed to interpret our experimental results, and the water showed a continuous effect during growth. Finally, the authors studied the growth process using a rotating ring disk electrode (RRDE) and observed the dynamic passivation of this layer during growth, which is due to the continuous formation/dissolution of the LiOH and Li2O surface layers, accompanied by the precipitation of H2 and possibly the formation of LiH. Notably, hydrogen has also been reported as a product of the electrocatalytic formation of LiF in HF-containing electrolytes, suggesting that it is a component of directed growth, an interpretation that aligns the results of this paper with the existing literature and provides mechanistic insights into the design and control of lithium metal topography and surface chemistry.
相关研究成果以“Water as Additive Directing Lithium Electrodeposition”为题发表在ACS Energy Lett.上。
【Core Contents】
First, lithium deposition is tested by a three-electrode battery, using a copper substrate as the working electrode and lithium metal as the counter electrode and a quasi-reference electrode (QRE). Dry tetraethylene glycol dimethyl ether (4G) was used as the solvent, with 1 M LiTFSI or 1 M LiClO4 with different amounts of water as the electrolyte. Figure 1 shows scanning electron microscopy (SEM) images of two lithium depositions after galvanostatic electrodeposition at −1 mA/cm-2. The graph shows that the morphology of the sediment is controlled by the water concentration and that the sediment follows the same trend in quality regardless of the salt used, evolving from moss to columnar as the water concentration increases. The results in Figure 1 contrast sharply with the general explanation that HF-mediated LiF formation is responsible for columnar growth, where LiTFSI is commonly used as a reference salt for hydrolysis resistance and does not produce a columnar structure. It is worth pointing out that these results can be obtained even when the electrolyte is used directly after preparation, since in the case of a LiPF6-based electrolyte, it is usually necessary to wait 48 hours for water to be added.
Figure 1. SEM images of lithium content were deposited using 1 M LiTFSI and LiClO4 salts and different concentrations of water as additives.
As shown in Figure 2, 30 μm can be deposited using a LiTFSI electrolyte containing 25 mM H2O at a current density of 1 mA/cm2. It was found that when the additive concentration was proportional to the deposition current density, a columnar topography could be obtained under a range of experimental conditions. When using LiTFSI, growth is performed at high coulombic efficiency and lower efficiency in LiClO4.
Figure 2. SEM image after deposition of 30 μm lithium at 1 mA/cm2 at 1 M LiTFSI+25 mM H2O (4G).
After demonstrating that the use of water as an additive produces a similar morphology to the HF-based strategy, the authors investigated the underlying mechanisms of lithium columnar growth, and the authors independently observed the formation step (i.e., potential without lithium deposition, U > 0 V) and growth step (U < 0 V). The hypothesis of Kasse et al. was first considered that the physical nanostructure of the surface layer formed during the formation step, rather than the chemical composition, directly guided the growth. To study this surface topography, the electrode was kept at potential U > 0 V, the surface was then characterized using an ex situ glove box AFM.
Figure 3a shows a linear sweep voltammetry (LSV) over 0.1 V, observing two peaks that systematically vary with water concentration, the peak at 1.3 V, which increases with increasing water concentration, and the peak at 1.7 V, which increases and moves to 1.95 V. Figure 3b shows the AFM topography image, with little to no features observed for the sample kept at 2.4 V and in a dry (reference) electrolyte. At 1.7 V, particles tens of nanometers high and 50-100 nanometers wide (larger color axes) appear on the surface. At 1.2 V and 0.1 V, islands are characterized by a few microns wide but only 3 nm high. In particular, it was observed that the density of the islands was an order of magnitude higher for samples maintained at 1.2 V in 25 mM H2O electrolyte than for samples maintained at 0.1 V.
Figure 3. LSV curves of electrolytes with different concentrations of water additives.
Since no clear nanostructure of columnar lithium was observed with atomic force microscopy, the authors replicated the experiments of Kasse et al. to investigate in which part of the electrodeposition process the additive played a role. The authors started with an electrolyte of 25 mM water additive (wet electrolyte), which was replaced by a dry electrolyte at different stages of the process. Figures 4a-c show the resulting topography, with the corresponding voltage−charge traces in Figure 4d, and arrows indicating where the dry electrolyte is introduced. First, in Figure 4A, only the dried electrolyte is used (green traces in Figure 4d). It is important to note that no adhesion was obtained between the lithium and the matrix, and very little material remained after rinsing.
Next, formation is carried out in the wet electrolyte in Figure 4b, and the dry electrolyte is introduced after the potential reaches 2 V in the constant current transient. This allows the sediment to adhere to the substrate without causing columnar growth. Longer formation at 0 V produces a similar moss morphology. Therefore, the wet formation step is essential for the adhesion of lithium to the substrate, which may be due to the formation of the surface layer, but does not result in directional growth. Finally, in Figure 4C, the first 3.7 C/cm2 is deposited in the wet electrolyte, followed by 3.7 C/cm2 in the dry electrolyte (red traces). This experiment clearly shows that water additives are required for continuous columnar growth, as the columnar morphology switches to non-directed growth upon solvent exchange. The results of this paper appear to be contrary to the results of Kasse et al., in which the HF-rich LP30 electrolyte (1M LiPF6 EC/DMC+100 ppm HF) was replaced with the electrolyte of this paper after a constant current cycle to 0 V, and a columnar morphology was obtained. Considering that the formation of HF by in-situ hydrolysis of LiPF6 has resulted in a drop in trace water to 25−50 ppm, a rationalization for this difference is found.
Figure 4. SEM image of the constant current deposition topography in a 4G electrolyte.
To further investigate the formation and growth mechanism of lithium, the authors used a rotating ring disk electrode (RRDE) with a copper disc and a Pt electrode with a collection efficiency of N0=0.256. The ring potential is fixed at 3.5V, where the molecular hydrogen that reaches the ring is oxidized. Figure 5a shows the response of the toroidal current when the disc potential is swept from the open-circuit potential (OCP) to 0.1 V at different concentrations of water, noting that the dried residual water can be brought to the electrode by rotation, and the peak of the disc current is similar to that of the fixed electrode in Figure 3a, indicating that these reactions are limited by passivation or solid-state diffusion, with both the ring current peaking at 0.5 and 1.3 V, and a smaller peak around 2 V, and the LiClO4 electrolyte exhibiting similar behavior. At 1.3 V, the disk and ring currents increased with increasing water concentration, while at 0.5 V, the disk and ring peaks of the drying electrolyte were the highest. To interpret the voltammetry, Figure 5b shows the proportion of the disk current that leads to the formation of hydrogen.
According to the results of this study, the mechanism of growth of columnar lithium forms should be (i) dependent on water additives and not on anions, (ii) formation of an adhesion layer and continuous action during growth, and (iii) reactions involving passivation and hydrogen formation. Possible reactions that meet the criteria are:
After Figures 3 and 5, the surface reaction begins at around 2.2 V, then inhibition occurs at 2 V, inhibition occurs at 1.6 V, hydrogen peaks occur at 1.3 V (hydrogen fraction >30%), and hydrogen is expected to be generated at 2.1 and 1.7 V, respectively, for solid lithium hydroxide and lithium oxide (reactions 1 and 2). In fact, reactions 1 and 2 are very similar to the electrochemical reduction of HF:
Due to the apparent similarity of the interfaces, it is further hypothesized that (iv) the columnar growth process and the columnar growth mechanism of HF-rich electrolytes are the same. In summary, in addition to the dynamic formation/dissolution of the passivation layer, H2 is also an important component in guiding the growth of columnar lithium. For HF, the authors consider this to be a LiH/LiF SEI coating, while for water, it is a LiH/Li2O/LiOH phase.
Figure 5. In different concentrations of water additives, 1 M LiTFSI electrolyte rotated the annular voltammetry from OCP to 0.1 V copper disc Pt ring.
【Conclusion and outlook】
In summary, it is demonstrated that by using water as an additive, the growth thickness of the columnar lithium layer can reach up to 30 μm. The simultaneous use of non-hydrolyzed (LiTFSI) and fluorine-free (LiClO4) electrolytes demonstrated that LiF SEI derived from HF-rich electrolytes could not be considered as the only structure-oriented component of this growth mode. In addition, alternative hypotheses that rely on the formation step (potential above 0 V) were tested, which was also found to be inconsistent with the results of this paper. Rather, it is shown that water additives play a continuing role in guiding electrodeposition. Finally, the process was studied in situ using a rotating ring disk electrode. Based on the continuous precipitation of hydrogen, the authors propose the dynamic formation/dissolution of LiOH, the Li2O passivation layer, accompanied by H2 generation and LiH formation. Since hydrogen has also been reported during the electrochemical reduction of HF, it is believed that hydrogen is a component that guides columnar growth, which makes the results of this paper consistent with the literature on HF-based electrolytes. Considering that LiF is not the only structure-guiding factor, this allows for further design and control of the morphology and surface chemistry of this columnar lithium layer.
【Literature Information】
Mark Aarts,* Sai Gourang Patnaik, Toon Van Roy, Stefanie Sergeant, Maarten Debucquoy, and Philippe M. Vereecken*, Water as Additive Directing Lithium Electrodeposition, 2024, ACS Energy Lett.
https://doi.org/10.1021/acsenergylett.4c00051
Note: Most of the articles reproduced on this site are collected on the Internet, and the copyright of the article belongs to the original author and the original source. The views in the article are only for sharing and exchange, if it involves copyright and other issues, please let us know, and I will deal with it in a timely manner.