Elsevier

Solid State Ionics

Volume 272, April 2015, Pages 1-8
Solid State Ionics

Microstructure and ionic conductivity of LLTO thin films: Influence of different substrates and excess lithium in the target

https://doi.org/10.1016/j.ssi.2014.12.005Get rights and content

Highlights

  • Epitaxial growth of Li3xLa2/3  xTiO3 thin films on oriented SrTiO3, LaAlO3 and MgO substrates.

  • Unavoidable growth of La2Ti2O7 secondary phase at the oxygen pressure required to maintain low electronic conductivity.

  • Secondary phases act as blocking layer to lithium ion diffusion.

  • Up to 10% excess of lithium in the target is necessary to compensate lithium loss.

Abstract

The deposition of single phase Li3xLa2/3  xTiO3 (LLTO) thin films remains very challenging. The growth of the perovskite phase is in competition with the insulating La2Ti2O7 phase when prepared at high oxygen pressure by PLD. Nevertheless, we have achieved epitaxial growth of LLTO on different (001) oriented substrates such as LaAlO3, SrTiO3 and MgO despite a large lattice mismatch of up to + 8.8%. We also determined the percentage of lithium excess in the target necessary to reach a maximum ionic conductivity. However, the presence of the blocking La2Ti2O7 phase strongly hinders the lithium ion migration and reduces the total conductivity compared to bulk properties.

Introduction

Ceramic electrolytes have attracted much attention in the field of Li-ion battery development as they represent a safer alternative to the flammable organic electrolytes. Their development, and in particular as thin films, is fundamental to the development of new power storage units in microelectronic components [1], [2], [3]. To achieve the requirements of battery technology, these electrolytes must present properties equivalent to conventional organic electrolytes, i.e., high Li-ion conductivity to facilitate fast lithium transport between the anode and the cathode, and low electronic conductivity to avoid losses and maintain efficiency [4]. Among existing ceramic materials, the perovskite lithium lanthanum titanate (Li3xLa2/3  x1/3−2xTiO3, LLTO) has attracted much attention since the discovery of its high Li-ion conductivity [5], [6]. The presence of vacancies on the A-site, and the ordering of La-rich and La-poor layers, enhance the mobility of lithium ions that migrate from one dodecahedral A-site cage to the next one through square planar windows formed by four oxygen atoms [7], [8]. The highest measured conductivities are obtained when an adequate balance between the lithium content and the number of vacancies in the A-site position of the perovskite structure is achieved. These results have been achieved in bulk ceramic samples for an LLTO composition with x = 0.11, which leads to a lattice conductivity as high as 1 mS⋅cm −1 at room temperature [9]. However, the grain boundaries of these ceramics present a large series resistance that dramatically decreases the total conductivity to values below 0.01 mS⋅cm −1. The reasons for the large grain boundary resistance to Li-ion migration are still not fully understood and, despite recent improvements in the processing of these ceramics [10], a higher total conductivity is necessary for efficient practical applications.

The development of thin film technology over the last few decades has opened a new area of research with the introduction of nanoscale materials and hetero-structures. The properties of a material can be tuned via the influence of the substrate as thin films often develop large stress during their growth, particularly in epitaxial growth conditions. The stress-induced lattice distortions can significantly affect the physicochemical properties of materials. Reports on this interfacial effect are well documented in the literature, for SOFC electrolyte materials for instance, where variation of the oxygen ion conductivity of thin films has been reported to depend on the epitaxial strain developed at the interface between mismatched structures [11]. Important improvements in thin film deposition technology enable an epitaxial growth of thin films where layer-by-layer growth can be achieved. This type of growth displays a slow and well-controlled deposition of the material on the substrate and can present crystalline properties close to single crystal and, more particularly, the absence of grain boundaries. An example can be found in ref [12] where highly ordered thin films, free of high angle grain boundaries, were used to directly access the bulk proton conductivity of Y-doped barium zirconate, higher by two orders of magnitude than the total conductivity of ceramic samples of this material. This finding opens new research direction in ionic conductors, i.e., to dispose of blocking grain boundaries, which would have tremendous consequences for Li-ion conductors and, more particularly, LLTO ceramic electrolytes.

Physical vapour deposition techniques have been shown to give rise to a high loss of lithium during matter transport from the target to the substrate; however, this particular issue can be addressed by adding extra lithium to the targets [13]. Here we review the growth conditions of LLTO thin films on different substrates necessary to achieve epitaxial layers when prepared by pulsed laser deposition and we report on the influence of the lithium content of the targets on the ionic conductivity and the film microstructure.

Section snippets

Sample preparation

Targets used for the PLD deposition were prepared by citrate route as this route yields crystalline particles at low temperatures. The method consists of dissolving LiNO3, La(NO3)3 and titanium(IV) oxide bis(2,4-pentanedionate) in a solution of citric acid with a low volume percentage of nitric acid. A gel was formed after evaporation of the organic solvents and then calcined at 600 °C to obtain a powder. The crystalline phase was formed at 800 °C, and the targets were sintered at 1350 °C in air

Characterisation of the targets

X-ray diffraction analyses of the targets indicate that they possess a tetragonal crystallographic structure corresponding to the P4/mmm space group (a = 3.8741 Å and c = 7.7459 Å [15]), as it can be observed in Fig. 1 for the target prepared with 10 mol% excess of lithium. This corresponds to the thermodynamically stable phase of LLTO at room temperature which consists of the layered structure of La-rich and La-poor layers alternating along the c-axis, necessary for Li-ion conduction. Despite high

Conclusions

In this work, we grew epitaxial Li3xLa2/3  xTiO3 thin films by PLD on STO, LAO and MgO substrates using targets presenting different lithium enrichments. We demonstrate that it is necessary to use a lithium-rich target with an excess of at least 10 mol% of lithium to achieve the maximal ionic conductivity. However, microscopy studies reveal that these films present a secondary phase restricting the lithium pathway. This phase identified as La2Ti2O7 increases the resistance of the films as lithium

Acknowledgements

We would like to acknowledge Miss Nuria Goméz for ICP experimental work. F.A. would like to thank the support of the Ministry of Industry of the Basque Country for their financial support through the Etortek program untitled Energigune´13.

References (24)

  • S.D. Jones et al.

    Solid State Ionics

    (1992)
  • S.D. Jones et al.

    J. Power Sources

    (1995)
  • Y. Inaguma et al.

    Solid State Commun.

    (1993)
  • Y. Harada et al.

    Solid State Ionics

    (1999)
  • Y. Inaguma et al.

    Solid State Ionics

    (1994)
  • D.H. Kim et al.

    J. Cryst. Growth

    (2013)
  • J.L. Fourquet et al.

    J. Solid State Chem.

    (1996)
  • G. Van Tendeloo et al.

    J. Magn. Magn. Mater.

    (2000)
  • S. Havelia et al.

    J. Cryst. Growth

    (2008)
  • T. Ohnishi et al.

    Solid State Ionics

    (2012)
  • O. Bohnke et al.

    Solid State Ionics

    (2011)
  • S. Furusawa et al.

    Solid State Ionics

    (2005)
  • Cited by (40)

    • Ion dynamics and positron annihilation studies on polymer ceramic composite electrolyte system (PVA/NaClO<inf>4</inf>/Y<inf>2</inf>O<inf>3</inf>): Application in electrochemical devices

      2022, Ceramics International
      Citation Excerpt :

      As the frequency of the applied field increases, the dipoles will not rotate sufficiently, so their oscillation will lag behind that of the field and hence decrease of dielectric constant nearly up to 1000 Hz. The dipole polarizability is negligible above the frequencies around 1000 Hz, and dipoles do not have enough time to switch the field direction and cease/delay to follow the applied alternating field, so the dielectric constant stays constant with frequency [75]. It is clear from Fig. 11(a) that the dielectric constant is high for all doped PVA composites than pure PVA+Salt.

    • Lithium solid-state batteries: State-of-the-art and challenges for materials, interfaces and processing

      2021, Journal of Power Sources
      Citation Excerpt :

      Still, special care must be taken when depositing Li-containing films, since recent reports have shown significant off-stoichiometries due to substantial Li loss during the material transfer from the pellet to the substrate. In order to compensate this, several strategies have been proven to be effective, such as the overlithiation of the ablated target [375] or the addition of thin film Li reservoirs in the substrate by means of multilayer deposition [374]. Important efforts are being done nowadays on the scaling up of PLD process to industrial scale, mainly associated to the development of large area deposition systems.

    View all citing articles on Scopus
    View full text