Elsevier

Journal of Power Sources

Volume 224, 15 February 2013, Pages 260-268
Journal of Power Sources

Development and costs calculation of lithium–sulfur cells with high sulfur load and binder free electrodes

https://doi.org/10.1016/j.jpowsour.2012.10.004Get rights and content

Abstract

A binder free thick film sulfur cathode based on a carbon structure with carbon nano tubes (CNT) is introduced. The sulfur mass can be varied between 3 and 20 mg cm−2 electrode leading to sulfur loads that are several times as high as in slurry electrodes. The electrode structure and thickness is examined by SEM, the surface area measured by BET and the in-plane conductivity is determined by a modified 4-point measurement. The achieved capacities for these extremely high sulfur loads are around 900 mAh g−1 sulfur at a current of 0.64 mA cm−2 electrode.

Additionally the price of future Li–S cells (18650), both conventional sulfur slurry cathode and here introduced binder free CNT sulfur cathode were calculated and compared with a lithium-ion system (NCA-graphite).

Highlights

► Binder free sulfur cathodes based on CNT for Li–S batteries. ► High sulfur loads. ► Influence of electrode surface and structure on electrochemical performance. ► Costs calculation for 18650 cell.

Introduction

Lithium–sulfur cells are regarded as one of the most promising systems for next generation batteries due to their high theoretical capacity, the abundant and cheap sulfur resources and lithium-ion comparable cathode production techniques. One great benefit vs. Li–O2 is the encapsulated system which greatly reduces the complexity of a future Li–S cell and prevents parasitic influence from outside.

Drawbacks of Li–S are the low operating voltage of around 2.15 V during discharge, the isolating property of both sulfur and Li2S, the volume change of approximately 79% based on full transformation of sulfur to Li2S, safety problems due to the lithium metal anode and a polysulfide shuttle mechanism first described by Mikhaylik et al. [1].

During the last ten years the electrochemical properties of the Li–S system have been greatly improved. Having only 10–20 cycles with low and unstable capacities at the beginning, the barrier of 100 relatively stable cycles with high capacities around 800–1000 mAh g−1 sulfur have been reached in a recent publication [2]. An example for high cycle stabilities at high currents (1.3 mA cm−2 = 1C; sulfur load: <1 mg cm−2 electrode) with good capacities can be found in Schuster et al. [3].

Nevertheless most sulfur cathodes in the previous investigations had sulfur loads below 2 mg cm−2 electrode [4]. Additionally, the sulfur fraction in sulfur electrodes is greatly below the common 80% active material level in lithium-ion cathodes. Fig. 1 gives the fraction of the active material in cathodes for recent Li–S publications (total 87) during the last 10 years. Most of the evaluated work (84%) is about electrodes using elemental sulfur as active cathode material (indicated by

in Fig. 1); 8% used sulfur composite cathodes (indicated by
in Fig. 1) in which the sulfur was bonded to a conductive carbon chain (e.g. polyacrylonitrile + sulfur). The other 8% of the previous work are based on Li2S cathodes that can work without a lithium metal anode (indicated by
in Fig. 1). So it can be concluded that most of the cathodes and especially the sulfur composites have an active material fraction below 60%. Keeping in mind that the porosity of sulfur cathodes (that has to be filled with electrolyte) is regularly above 30% to sustain active material volume change (S8 ↔ Li2S: 78%), the gravimetric energy density of such a Li–S cell won't be above current lithium–ion level.

In order to obtain high energy densities it is important to have high sulfur utilization, high loads and high fractions of active material in the electrode. We therefore introduce a binder free thick film electrode based on CNT, coated on a conductive carbon structure that can contain very high loads of active material (3–20 mg cm−2 electrode).

Section snippets

Preparation of CNT-electrodes

A 3D carbon nonwoven-based gas diffusion layer (SGL Carbon, Sigracet GDL 10AA, weight: 8.5 mg cm−2 +/− 1.4 mg cm−2) with a dimension of approximately 6 × 20 cm was used as substrate for the growth of CNTs by a CVD process. The detailed procedure is described in Refs. [5], [6]. Different series of CNT coated carbon structure electrodes were prepared by varying the CNT mass per cm² electrode in order to evaluate its influence on the electrochemical and physical electrode properties.

The electrodes

CNT growth on carbon structure substrates

Fig. 2 shows scanning electron micrographs (SEM) of the received carbon structure (column a) and the CNT coated structures (column b with 4.7 mg cm−2 electrode and column c with 11.0 mg cm−2 electrode). The carbon fibres have a thickness of around 10 μm and build up a network with a distance of around 30–100 μm to the neighbouring fibres. It is clearly visible that the CNT coating increases the thickness of the electrode by pushing the carbon fibres apart. Additionally the high macro porosity

Conclusion

A binder free CNT coated carbon structure electrode was introduced. The sulfur loads can be several times higher than in all so far published sulfur cathodes. The electrochemical results can be stable for 100 cycles and more if shortcuts by lithium dendrites can be avoided. The achieved capacities depend on the sulfur load and the applied current and are between 800 and 900 mAh g−1 sulfur at 0.64 mA cm−2. The capacities related to the sulfur mass, the CNT mass and the carbon structure mass can

Acknowledgements

This research was financed by the German Bundesministerium für Bildung und Forschung (BMBF) through the project STROM: “AlkaSuSi-Alkalimetal sulfur and silicon”. We are grateful for the support. We would further like to thank Jan Tomforde (BASF) and Holger Schneider (BASF) for discussing the integrity of the price calculation.

References (11)

  • M. Rao et al.

    Electrochemistry Communications

    (2012)
  • S.S. Zhang et al.

    Journal of Power Sources

    (2012)
  • S. Dörfler et al.

    Chemical Physics Letters

    (2011)
  • M. Hagen et al.

    Journal of Power Sources

    (2012)
  • W.F. Howard et al.

    Journal of Power Sources

    (2007)
There are more references available in the full text version of this article.

Cited by (0)

View full text