Lithium metal has long been considered the ideal anode material for lithium-based batteries due to the high inherent capacity and low standard potential of the Li+/Li redox couple. While the first batteries contained lithium metal electrodes, it was soon realized that these electrodes gave rise to unacceptable safety issues [1], [2], [3], [4], [5]. When such batteries are repeatedly charged and discharged, uncontrolled lithium metal growth typically gives rise to the formation of dendritic lithium nanostructures that eventually can short-circuit the battery [1], [2], [3], [4], [5], [6]. The lithium metal electrodes were, therefore, replaced with the much safer, albeit significantly less energy-dense, graphite electrodes still employed in contemporary lithium-ion batteries.

However, the inherent advantages of lithium-metal electrodes have recently given rise to a significant renewed interest in the development of approaches aimed at solving the lithium dendrite problem. These activities can be explained by the present development of the next-generation of Li-based batteries, mainly Li-S and Li-O2 batteries, which require access to stable lithium-metal anodes [1], [2], [3]. In general, the repeated lithium dissolution and deposition gradually transform the Li metal anode into a porous network of dendritic or mossy lithium with poor mechanical stability. The increasing surface area also promotes reactions between the highly reactive lithium electrode and the organic solvents producing a solid-electrolyte interphase (SEI) layer [7], [8]. These irreversible reactions, in which the electrolyte is reduced, also cause the capacity of the battery to decline. Hence, the uncontrolled 3D lithium metal deposition gives rise to both safety issues and a general loss of battery performance.

During the last four decades, numerous strategies have been developed to gain control over the lithium metal growth and to solve the dendrite problem. Most of these strategies can be summarized into three different categories: (1) improving the properties of the SEI layer via the addition of different additives or the use of solid/polymer electrolytes; (2) decreasing the current density by employing electrodes with high surface areas; and (3) electrostatic shielding of lithium electrodes via the addition of alkali metal ions (e.g. Cs+ and Rb+). The first approach has so far received the most attention since it is widely assumed that 3D lithium growth (i.e. dendritic or mossy) is caused mainly by the SEI layer breaking up, thereby exposing a preferential point of lithium growth. However, despite the substantial research efforts, the dendritic lithium growth problem remains essentially unsolved, largely because fundamental understanding of the phenomenon still is lacking.

Uncontrolled 3D lithium growth has essentially become accepted as a natural feature of lithium metal electrodes. From an electrochemical point of view, planar (i.e. two-dimensional or 2D) growth of lithium metal should, however, still be possible to obtain provided that appropriate experimental conditions favoring 2D growth are employed. In the present study, we propose that the lithium dendrite problem mainly stems from the fact that a lithium electrode in a 1?M Li+ electrolyte operates close to its equilibrium potential since this only would give rise to the formation of a few lithium nuclei. The preferential growth of these few nuclei (or the absence of a large number of equally sized nuclei covering the entire electrode surface) is the starting point for the 3D growth. Our approach is therefore aimed at creating a high lithium nuclei density to obtain 2D deposition conditions. As significant overpotentials are needed to generate high nuclei densities [10], [11], [12], such conditions are, however, not readily realized with conventional electrolytes [9].

In our work, we show that dendrite-free 2D lithium metal growth can be achieved by decreasing the Li+ concentration in the electrolyte and by introducing a supporting salt which essentially eliminates the migration of Li+. The low lithium concentration enables the realization of a large overpotential during a short nucleation pulse, which generates a large number of similarly sized lithium nuclei. In this way, the deposition becomes diffusion-controlled and the deposition can also be further stabilized using a pulsed electrodeposition designed to alleviate the local variations in the Li+ concentration at the electrode surface. It is shown that the decrease in the Li+ concentration in the electrolyte favors the attainment of constant lithium deposition conditions, which facilitates the attainment of 2D lithium deposition. In addition, the deposition conditions vary significantly with time using a conventional electrolyte. The risk of the formation of porous or dendritic lithium is therefore significantly higher when using a conventional electrolyte. Analyses of the voltage profiles obtained with the low Li+ concentration and conventional electrolytes also indicate the formation of pits on the electrode surface when using the conventional electrolyte. These pits can then serve as preferential nucleation points during the subsequent lithium deposition step to yield 3D rather than 2D lithium deposition. In contrast, steady-state deposition indicating a stable and abundant nuclei density leading to two-dimensional Li growth is demonstrated for the low Li+ concentration electrolyte during 100 cycles.

In this issue, the cover of Materials Today shows an electron micrograph image of a lithium electrode featuring stacked planar lithium metal layers grown in a lithium battery during 100 cycles. 2D or planar Li growth was achieved using an electrolyte containing a low Li-ion salt concentration (i.e. 20?mM) and a supporting salt (i.e. 1?M TBAPF6). A short (i.e. 10?ms) nucleation step was initially used to create a large number of similarly sized lithium nuclei on the electrode surface. The subsequent 2D lithium deposition was then performed employing pulsed electrodeposition to obtain the very smooth deposits seen in the cover image. The false-color electron micrograph was captured using a Zeiss Merlin high-resolution scanning electron microscope (HR-SEM), and the HR-SEM study was performed without exposing the lithium metal samples to air, using an inert atmosphere sample transfer device.

Planar lithium Electrochemical strategies circumventing dendritic lithium growth


The authors would like to acknowledge Professor Kristina Edström for funding the SEM transfer device and Gustav Nyholm for valuable discussions. The authors would also like to express gratitude for the beneficial SEM discussion with Dr. Linus von Fieandt and Victoria Sternhagen. Financial support from The Swedish Research Council [VR. 2015-04421], The Ångström Advanced Battery Center (ÅABC), and STandUP for Energy is also gratefully acknowledged.

Recommended articlesCiting articles (0)

Further reading:

[1]Y. Gou, et al.

Adv. Mater., 29 (2017), p. 1400007

View Record in Scopus

[2]D. Lin, et al.

Nat. Nanotechnol., 12 (2017), p. 194

CrossRefView Record in Scopus

[3]W. Xu, et al.

Energy Environ. Sci., 7 (2014), p. 513

ArticleDownload PDFCrossRefView Record in Scopus

[4]K. Zhang, et al.

Adv. Energy Mater., 6 (2016), p. 1600811


[5]X.-B. Cheng, et al.

Adv. Sci., 3 (2016), p. 1500213

CrossRefView Record in Scopus

[6]J.-G. Zhang, W. Xu, W.A. HendersonLithium metal anodes and rechargeable lithium metal batteries

Springer Int. Publ. (2017)

Google Scholar

[7]E. Peled

J. Electrochem. Soc., 126 (1979), p. 2047

CrossRefView Record in Scopus

[8]D. Aurbach, Y. Gofer

J. Electrochem. Soc., 136 (1989), p. 3198

CrossRefView Record in Scopus

[9]C. Heubner, et al.

J. Electroanal. Chem., 759 (2015), p. 91

ArticleDownload PDFView Record in Scopus

[10]A. Pei, et al.

Nano Lett., 17 (2017), p. 1132

CrossRefView Record in Scopus

[11]G. Oltean, et al.

Electrochim. Acta, 56 (2011), p. 3203

ArticleDownload PDFView Record in Scopus

[12]M. Paunovic, M. SchlesingerFundamentals of Electrochemical Deposition

John Wiley & Sons Inc. (1998)

Google Scholar

Read full text on ScienceDirect

DOI: 10.1016/j.mattod.2019.02.017