Elsevier

Solid State Ionics

Volume 177, Issues 26–32, 31 October 2006, Pages 2811-2819
Solid State Ionics

Progress in three-dimensional (3D) Li-ion microbatteries

https://doi.org/10.1016/j.ssi.2006.02.048Get rights and content

Abstract

Progressing from 2D to 3D thin-film microbattery structures represents an important paradigm shift in the development of electrochemical power sources. The development process of three-dimensional microbatteries (3D-MB) on a perforated substrate has to solve complicated technical barriers, such as insertion and deposition of a sandwich of high-capacity thin cathode and ion-conducting membrane in long (500 μm) and narrow (< 50 μm) channels. In this study, we investigate composite thin-film cathodes that allow 2.0–3.5 times increase of MB capacity for a given footprint over our previous 3D-MB with conventional thin-film battery cathodes. The composite cathodes are obtained by electrodeposition from the bath modified by PEGDME and PEO additives of different molecular weights and concentrations. SEM, XPS, TOFSIMS and electrochemical characterizations were performed to elucidate the source of the improved MB performance. Experimental results obtained in this work show that the addition of polymeric compounds to the electrolytes significantly improves the structure and electrochemical behavior of electrodeposited molybdenum sulfide cathode materials. A semi-3D-on-MCP cell with a composite cathode exhibits a stable cycle life with about 3.5 mA h/cm2 reversible capacity, 20 to 30 times that of a planar 2D thin-film cell with the same footprint and about twice that of semi-3D cells with pristine cathodes.

Introduction

Microscale technologies have recently allowed the fabrication of biomedical in vivo micromachines, integrated micro-optoelectronic circuits and microelectromechanical systems (MEMS). The power requirements of these microsystems stand behind the increasing number of studies on lithium and lithium-ion solid-state microbatteries (MBs). Thin-film MBs have higher current densities and cell efficiencies than their bulk counterparts, due to the fast ion transport over shorter distances. However, the capacity of such MBs is restricted by battery dimensions and, to date, the best commercial planar (2D) thin-film batteries have a reversible capacity of 0.133 mA h/cm2 [1], which is quite small. Proposed 3D microbattery (3D-MB) architectures, which may overcome the size and energy-density deficiency of the conventional 2D designs, need to contain sufficient active material to power most on-chip MEMS devices and microelectronic circuits for extended periods of time. Some proposed 3D architectures suggest using vertical “posts” connected to a substrate, wherein the layered battery structure is formed around the posts [2], [3], [4]. Other 3D architectures [3], [4] are based on the conformal deposition of electrodes and electrolyte layers on a graphite mesh as anode and cathode current collector. The 3D microbattery recently developed by our group [5], [6] is much simpler and more efficient, as it is based on readily available silicon, glass and even plastic substrates. Its basic attribute is more than an order of magnitude increase in geometrical area gain (and therefore cathode volume gain) per given substrate footprint. This is achieved by having a perforated substrate replace a full one (Fig. 1a). The depends on the number of through microchannels obtained by perforation and on their aspect ratio.

The 3D-MB development process had to solve complicated technical barriers, such as insertion and deposition of a sandwich of thin-film Ni current collector, thin cathode and solid polymer electrolyte (SPE) or hybrid polymer electrolyte (HPE) layers in long (500 μm) and narrow (< 50 μm) channels, as well as on the two remaining flat top and bottom surfaces of the perforated substrate. Each layer is continuous and interconnected between all channels through both top and bottom surfaces (Fig. 1b,c). Full-cell 3D-MBs were fabricated for the first time on silicon and glass microchannel-plate (MCP) substrates [6]. The best MCP-MB reported in [6] exhibited a capacity of about 2 mA h/cm2, in good agreement with the roughly 23-fold area gain over an identical-footprint 2D cell [6]. In these cells, one of the most complicated problems was the manufacturing of the cathode.

The preparation and characterization of thin-film cathodes based on oxides such as LiMn2O4, LiCoO2, Lix(MnyNi1−y)2−xO2, V2O5 or sulfides such as TiS2 and MoS2 has been at the focus of a number of studies [7], [8], [9], [10], [11], [12]. The principal methods used in thin-film cathode preparation are sputtering, evaporation and laser ablation. An important advantage of these methods is that they enable preparation of new cathode materials. However, these processes are essentially useful only for 2D geometries and some need very expensive equipment. In addition, thin cathode films are often annealed after preparation, while for some applications, such as direct deposition of the microbattery on a microelectronic chip, high-temperature treatment is forbidden. The optical, electrochemical and mechanical properties of molybdenum dichalcogenides make them promising semiconductor materials for various applications, such as solar cells, rechargeable batteries and solid lubricants. Despite this, the difficulty in preparing thin films of transition metal dichalcogenide compounds with desired properties held back their wide use in microelectrochemical devices.

We have recently developed an inexpensive and relatively simple electrodeposition method for preparing low-cost and low-toxicity molybdenum and iron sulfide thin cathode layers [13]. A highly adherent, homogeneous, compact molybdenum oxysulfide film about 300–500 nm in thickness can be deposited on nickel-coated silicon and glass substrates. However, an increase in the deposition time, current density and temperature is followed by the formation of cracks, which may be responsible for film peeling and the inability to deposit thicker cathodes. It seems likely that these cracks are caused by high internal stresses that develop during deposition. As a result, the capacity of the Li or Li-ion/MoS2 planar battery is on the order of 40–60 μA h/cm2. Due to the significant area gain provided by the three-dimensional design, the 3D-MB capacity increases by a factor of 20 to 30. This is still insufficient for the powering of some microdevices, for example in smart drug delivery [14] and other autonomous MEMS that may require a 1 mW h battery on a footprint of less than 3 × 3 mm2.

This study is focused on the development of a new type of composite thin-film cathodes, with the goal of producing cheap, simple and electrochemically superior microbatteries.

Section snippets

Experimental

Electrodeposition was carried out in an electrolytic bath containing tetrathiomolybdate (MoS42−) anions as the electroactive species. The bath was prepared by mixing aqueous solutions of Na2S and Na2MoO4 and by adjusting the pH to 7.5–8.0 through addition of KH2PO4. In a sulfidic solution, molybdate anion undergoes sulfidation in four steps that conserve Mo(VI) and lead to tetrathiomolybdate [15]. The monothio- to trithio-intermediates have low stability, but their presence in the final

Results and discussion

The morphology of non-modified pristine and modified electrodeposited cathodes was studied by SEM, and shown in Fig. 2. The figure captions list the various deposition conditions. Fig. 2a shows a non-modified cathode deposited for 10 min at 10 mA/cm2. Cracks that are one-half to one micron thick are clearly seen. The large crystallites are composed of small, closely packed grains. Increase of current density from 1 to 5 mA/cm2 and deposition time for more than 10 min cause strong internal stresses

Conclusions

The addition of PEG and PEO of different molecular weights and concentrations enables fabrication of high-adhesive-to-substrate thin-film cathodes for microbattery application. SEM micrographs show that molybdenum oxysulfide cathode films obtained from the modified electrolyte are predominantly constituted of plate-like crystallites densely covered by small polymer-like single spheres or assemblages of spheres. XPS and TOF-SIMS tests show formation of sulfur-rich molybdenum oxysulfide compounds

Acknowledgements

We express our gratitude to RAMOT - Tel Aviv University Authority for Applied Research and Industrial Development Ltd. for financial support of this project.

References (29)

  • G. Nagasubramanian et al.

    J. Power Sources

    (2004)
  • Y.J. Park et al.

    Solid State Ionics

    (2000)
  • J.B. Bates et al.

    Solid State Ionics

    (2000)
  • G. Meunier et al.

    Thin Solid Films

    (1991 (1 Dec))
  • D. Golodnitsky et al.

    J. Power Sources

    (2006)
  • N.J. Clarke et al.

    Inorg. Chim. Acta

    (1987)
  • E. Bahena et al.

    Electrochim. Acta

    (2004)
  • Albu-Yaron et al.

    Thin Solid Films

    (2000)
  • M.A. Baker et al.

    Appl. Surf. Sci.

    (1999)
  • L. Benoist et al.

    Thin Solid Films

    (1995)
  • E. Schmidt et al.

    Thin Solid Films

    (1995)
  • G. Gonzalez et al.

    Solid State Ionics

    (1996)
  • S.R. Das et al.

    IEEE Spectrum

    (2004 (Oct.))
  • A. Singh et al.

    J. Electrochem. Soc.

    (2002)
  • Cited by (0)

    View full text