The effect of the binder morphology on the cycling stability of Li–alloy composite electrodes

https://doi.org/10.1016/S0022-0728(01)00532-0Get rights and content

Abstract

It is demonstrated that the design of the composite electrode, or more precisely the morphology and distribution of the binder poly(vinylidine fluoride) (PVdF) within the composite electrode, has a significant impact on the cycling performance of Li storage alloy (Sn/SnSb) electrodes. Different binder morphologies and distributions have been obtained by using different solvents for the slurry preparation, such as 1-methyl-2-pyrrolidinone (NMP), in which PVdF is dissolved, yielding electrodes with a homogeneously and finely distributed binder, or decane, in which PVdF is only dispersed, yielding electrodes in which the original particle morphology of the binder powder is preserved. In constant current cycling tests carried out in an excess of electrolyte, the electrodes with the ‘dispersed’ binder show far better cycling capacities and stabilities than those with the ‘dissolved’ binder. This is explained by the different binding strengths, swelling behaviour in the electrolyte, electrode porosities, and possible ‘buffer’ effects of the compact and the finely distributed binders.

Introduction

The key issue for the application of lithium storage metals and alloys in rechargeable Li-ion cells is the control of their volume changes which occur during cycling. Only by a proper choice of the active material and a proper design of its morphology is it possible to obtain a material, which can be cycled several times [1]. However, the volume changes are not only a challenge to the active material itself, but also to the protective film (the solid electrolyte interphase) which is formed at the electrode  electrolyte interface [2], and to the maintenance of the integrity of the composite electrode during repeated cycling. Especially, the binder has to meet strict requirements. On the one hand it must counteract the dispersive forces caused by the expanding metallic host matrix and withstand extensive swelling in the polar organic electrolyte, since both effects would result in a loss of internal contact; on the other hand it must retain a certain flexibility, since too rigid a system would simply crack and crumble. Obviously, the chemistry of the binder and — as will be shown here — also its distribution within the composite electrode play a vital role.

Most of the hitherto published work on binders deals with the testing of new binder materials and their interactions with the active material and the electrolyte (some recent articles are e.g. Refs. [3], [4], [5], [6], [7], [8]). In addition, it was found that the mixing method for the preparation of the electrode slurry [9] and the mixing sequence of the slurry components [10] had some influence on the capacities and on the cycling stabilities.

Yang et al. [6], [7] compared the cycling behaviour of Sn/SnSb-based composite electrodes with poly(vinylidene fluoride) (PVdF) and high-density polyethylene (PE-HD) as the binder and found that the composite electrodes with PE-HD showed a better cycling performance than those with PVdF. They explained this by the stronger swelling of PVdF in the polar electrolyte, which together with the large expansion of the active material results in fast loss of internal contact. What appears interesting is the different preparation of the composite electrodes: PVdF was dissolved in 1-methyl-2-pyrrolidinone (NMP), whereas PE-HD was used dry without any solvents. The question arises as to whether the different cycling stability is solely a matter of the chemistry of the binder or whether it is also influenced by the different processing methods, which surely affect the distribution of the binder within the electrode. To elucidate this point, composite electrodes were investigated, which have exactly the same composition and the same binder, but which were processed in different ways (cf. also Ref. [11]).

Section snippets

Li–alloy composite electrodes

Sub-micron crystalline Sn/SnSb powder (82 wt%) [1] acting as active material, Ni powder (10 wt%) [1] acting as conductive additive, and PVdF (8 wt%; Aldrich, Mw=∼534 000) acting as binder were blended, and NMP (Aldrich, 99%; Fluka, p.a.) or decane (Aldrich, 99+%) were added to give a slurry. After thorough mixing under ultrasonic conditions, the slurry was painted onto both sides of a stainless steel mesh of wire diameter 25 μm and mesh width 67 μm (mesh 280) (Spörl, Germany) acting as the

Results and discussion

Two types of Sn/SnSb composite electrodes were compared which have exactly the same composition but which were processed in two different ways. For the first type, NMP was used for the slurry preparation, as an example of a solvent, which dissolves PVdF. For the second type, decane was used, as an example of a dispersant, which does not substantially dissolve PVdF. The rest of the composite electrode preparation was the same (as outlined in Section 2). The results obtained may be generalised to

Conclusions

Two types of Sn/SnSb composite electrodes with identical compositions were compared, where: (i) NMP (in which PVdF is dissolved); or (ii) decane (in which PVdF is only dispersed) were used for the slurry preparation. In case (i) the binder was homogeneously distributed within the composite electrode, forming thin threads of diameters of 30 nm and less between the active materials. In case (ii) the original particle morphology of the binder with diameters of ∼200–300 nm was maintained, and the

Acknowledgements

Financial support by the Austrian Science Fund and the Oesterreichische National Bank in Project 12768 and in the Special Research Program ‘Electroactive Materials’ as well as support by Mitsubishi Chemical Co. (Japan) are gratefully acknowledged. Furthermore, the authors thank Merck (Germany) for the donation of electrolytes.

References (16)

  • M. Wachtler et al.

    J. Power Sources

    (2001)
  • W.F. Liu et al.

    J. Power Sources

    (1997)
  • M.N. Richard et al.

    J. Power Sources

    (1999)
  • J. Yang et al.

    J. Power Sources

    (1999)
  • Z. Liu et al.

    J. Power Sources

    (1998)
  • K.M. Kim et al.

    J. Power Sources

    (1999)
  • J.-M. Tarascon et al.

    Solid State Ionics

    (1996)
  • J. Yang et al.

    Electrochem. Solid-State Lett.

    (1999)
There are more references available in the full text version of this article.

Cited by (82)

  • Unveiling thermal decomposition kinetics of Single-Crystalline Ni-Rich LiNi<inf>0.88</inf>Co<inf>0.07</inf>Mn<inf>0.05</inf>O<inf>2</inf> cathode for safe Lithium-Ion batteries

    2022, Chemical Engineering Journal
    Citation Excerpt :

    Meanwhile, the decreased intensity of 533.5 eV implies the decomposition of organic carbonates (R2CO3). In the F1s spectra (Fig. 2c), the peak at 687.4 eV for the pristine arise from PVDF [24] and decomposition products from LiPF6. The weak shoulder peak at 685 eV corresponds to LiF [22].

  • Towards water based ultra-thick Li ion battery electrodes – A binder approach

    2019, Journal of Power Sources
    Citation Excerpt :

    However, an increase in dry film thickness goes along with a variety of drawbacks like longer diffusion pathways, reduced electronic conductivity and higher mechanical stress during drying [10]. As state of the art, N-methyl-pyrrolidone (NMP) is used as process solvent in combination with polyvinylidene difluoride (PVdF) as binder [11]. PVdF itself is a costly fluorinated polymer in comparison to non-fluorinated sodium-carboxymethylcellulose (CMC) as it is five times more expensive, (depending on the supplier).

View all citing articles on Scopus
View full text