Marvin
Cronau
,
Marvin
Szabo
and
Bernhard
Roling
*
Department of Chemistry, Philipps-Universität Marburg, Hans-Meerwein-Strasse 4, 35032, Marburg, Germany. E-mail: roling@staff.uni-marburg.de
First published on 11th November 2021
A single-step high-energy ball milling process was used for the synthesis of a Li5.3PS4.3ClBr0.7 glass ceramic. This material exhibits a Li+ conductivity of 5.2 mS cm−1, the highest value reported so far at room temperature for a Li+ solid electrolyte pellet prepared without any annealing step. We demonstrate the utilization of this electrolyte in an all-solid-state battery exhibiting a composite cathode impedance of only 5 Ω cm2 at room temperature.
The currently most promising group of SEs are sulphide-based materials. Recently, Lee et al. built an ASSB with a capacity retention of over 80% after 1000 cycles using microcrystalline argyrodite-type Li6PS5Cl materials as electrolyte.2 Most sulphide-based SEs are synthesized via a two-step or even three-step synthesis. First the starting materials are mixed and pre-treated in a ball mill. Afterwards, the resulting powder is filled into a quartz ampoule, and the quartz ampoule is heated up to a temperature (powder annealing), at which the material is partially crystallized (glass ceramic) or completely crystallised (microcrystalline material). The powder is then ground and pressed to a pellet (cold-pressed pellet), in some cases followed by pellet annealing. Pellet annealing leads usually to a considerable enhancement of the ionic conductivity due to reduced grain boundary resistances.
Argyrodite-type Li6PS5Cl reaches ionic conductivity values up to 6 mS cm−1 after pellet annealing.3–16 Although this value is sufficient for building working ASSBs, some pellet-annealed, fully microcrystalline SEs with a significantly higher conductivity have been discovered. One way to increase the Li+ conductivity is cation doping of the argyrodites. For instance, Li6.6P0.4Ge0.6S5I exhibits a Li+ conductivity of 5.4 mS cm−1 after powder annealing and cold pressing in comparison to Li6PS5I with only 1.3 μS cm−1. After pellet annealing, the conductivity of Li6.6P0.4Ge0.6S5I increases to 18.4 mS cm−1.17 Li6.6Si0.6Sb0.4S5I outperforms Li6.6P0.4Ge0.6S5I with 14.8 mS cm−1 for the cold-pressed pellet and 24 mS cm−1 for the annealed pellet.18 However, materials containing tetrel atoms are chemically less stable than the parent compound Li6PS5Cl.19–21
Another promising approach to boost the Li+ conductivity is the synthesis of halide-rich Li6−xPS5−xHal1+x (Hal = Cl, Br) solid electrolytes.22,23 For example, the reported Li+ conductivities of Li5.5PS4.5Cl1.5 and Li5.5PS4.5Br1.5 after pellet annealing reach values of about 11 and 12 mS cm−1, respectively, while the conductivity of the recently synthesized mixed-halide microcrystalline argyrodite Li5.3PS4.3ClBr0.7 reaches 24 mS cm−1.24–27
In general, two-step or three-step syntheses are well suited for lab-scale production, but not for commercial large-scale production.28 Furthermore, high-temperature annealing (around 500–600 °C) of SE particles inside the composite electrodes of ASSBs is not possible due to the limited thermal stability of active material particles, like NMC.29 Therefore, the development of highly conductive SE materials, which reach high ionic conductivities without using high-temperature annealing steps during synthesis is of utmost importance.
Herein, we present a single-step ball milling synthesis of a glass ceramic (GC) Li5.3PS4.3ClBr0.7 powder. After cold pressing, the resulting pellet reaches the highest Li+ conductivity reported so far for Li+ solid electrolytes prepared without any annealing step. The simple synthesis in comparison to the materials mentioned above, see Fig. 1, and the already demonstrated long-term stability of glass ceramic solid electrolytes in ASSBs30 makes GC Li5.3PS4.3ClBr0.7 a viable candidate for the use in ASSBs. We demonstrate that the integration of GC Li5.3PS4.3ClBr0.7 into an In/LiCoO2 solid-state battery leads to remarkably low cathode impedance and overall battery impedances.
![]() | ||
Fig. 1 Illustration of different synthesis routes for solid electrolytes and of a home-made cell setup for the characterisation of all solid-state batteries (ASSBs). |
Stoichiometric amounts of the precursors were filled into a zirconia pot with 10 zirconia balls (10 mm diameter) and ball milled for 8.25 hours (850 rpm). In the case of argyrodites, this synthesis route leads in general to nanocrystallites in an amorphous matrix, with low volume fractions of the amorphous phase in the range of 10–15 vol%.27,31–33 The resulting powder (Fig. 2) was used as received without any heat treatment (a detailed description of the synthesis can be found in the ESI†). The X-ray powder diffractogram shows the typical nanocrystalline pattern (average particle size about 18 nm) of many argyrodite-type glass ceramic electrolytes.
![]() | ||
Fig. 2 (a) X-Ray diffraction pattern of the SE. (b) Scanning electron microscopy image (magnification: 5000) of the synthesized powder. |
For the fabrication of pellets for Li+ conductivity measurements, a fabrication pressure of 490 MPa was used inside a CompreDrive laboratory press (rhd instruments, Darmstadt, Germany). The impedance measurements were carried out by means of an Autolab PGSTAT302N (Metrohm Autolab, Utrecht, Netherlands) under an applied stack pressure of 100 MPa. The impedance spectra were taken in a frequency range from 105–101 Hz and in a temperature range from 30–85 °C with an applied rms AC voltage of 10 mV. Since reported Li+ conductivity values depend strongly on the fabrication pressure and stack pressure, we used identical pressure conditions for the fabrication of four solid electrolyte pellets and the subsequent ionic conductivity measurements. These solid electrolytes are: (i) GC Li6PS5Cl prepared by a single-step ball milling synthesis; (ii) microcrystalline (μC) Li6PS5Cl with applied powder annealing at 550 °C; (iii) GC Li5.5PS4.5Cl1.5 prepared by a single-step ball milling synthesis; (iv) GC Li5.3PS4.3ClBr0.7 prepared by a single-step ball milling synthesis.
Fig. 3 shows an Arrhenius plot of the Li+ conductivity of all materials at 30 °C. With 5.2 mS cm−1, the ionic conductivity of Li5.3PS4.3ClBr0.7 clearly exceeds those of the other materials, and even that of powder-annealed Li6PS5Cl.
![]() | ||
Fig. 3 Arrhenius plots of the Li+ conductivity σ of different solid electrolytes (left, temperature range 30 to 85 °C) and table with Li+ ionic conductivity at 30 °C (right). |
To demonstrate the performance of the new glass ceramic Li5.3PS4.3ClBr0.7 electrolyte in an ASSB, the electrolyte was used for the preparation of the separator layer as well as for the preparation of composite cathodes with LiNbO3-coated LiCoO2 active material particles and carbon fibers as conductive additive (70:
30
:
3 wt%). Indium foil was used as anode, which was pressed for one minute with 298 MPa onto the separator layer to form a good anode/separator contact. The lab-scale ASSB was characterized in a home-made cell setup with pressure control, see Fig. 1. For the battery cycling experiments, a pressure of 98 MPa was applied, and the battery was cycled at room temperature with a rate of 0.1 C between the cutoff voltages 2.1 V and 3.7 V. Fig. 4(a) shows the first charge and the first discharge voltage profile of the battery. With a discharge capacity of 147 mA h g−1, the battery exploits the full capacity of LiCoO2. For comparison, an ASSB was built with non-annealed Li6PS5Cl in the composite cathode and in the separating SE layer, which reaches only a capacity of 137 mA h g−1. Impedance spectra of the two ASSBs after the first discharge cycle are shown in Fig. 4(b). The higher ionic conductivity of GC Li5.3PS4.3ClBr0.7 as compared to GC Li6PS5Cl manifest in a lower high-frequency resistance of the GC Li5.3PS4.3ClBr0.7-based battery, which can be identified with the separator resistance, and in a smaller second semicircle, which is usually attributed to the cathode impedance. The cathode impedance is determined by the ion transport resistance of the solid electrolyte within the composite cathode as well as by the charge transfer resistance at the interfaces between the active material particles and the solid electrolyte particles.34 The fast ion transport in the GC Li5.3PS4.3ClBr0.7-based cathode leads to a cathode impedance value of only 5 Ω cm2, which is one of the lowest values reported so far at room temperature.35
In conclusion, we synthesized the glass ceramic Li5.3PS4.3ClBr0.7 electrolyte in a single-step synthesis without any annealing step and integrated this electrolyte in a low-impedance ASSB. To our knowledge, cold-pressed GC Li5.3PS4.3ClBr0.7 exhibits the highest Li+ conductivity (5.2 mS cm−1) reported so far for electrolytes prepared without annealing steps and surpasses that of annealed microcrystalline Li6PS5Cl (3 mS cm−1). The cathode impedance of our lab-scale ASSB was only 5 Ω cm2, one of lowest values achieved so far at room temperature. Using a thinner GC Li5.3PS4.3ClBr0.7-based separator layer would further lower the total impedance to values of 30–40 Ω cm2. Building an ASSB with an area of some 1000 cm2, like in commercial Li+ ion batteries,36 would then result in a battery resistance in the range of 10 mΩ, which is similar to that of such commercial Li+ ion batteries. Together with the already reported favourable pressure-dependent properties,37 and the high long-term stability of glass ceramic SE in ASSBs,30 our results give strong indication that GC Li5.3PS4.3ClBr0.7 prepared by a simple one-step synthesis is a viable candidate for next-generation batteries. In future work, milling protocols for reducing the energy consumption should be developed, e.g. by reducing the rotational speed and by using balls with optimized diameter.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ma00784j |
This journal is © The Royal Society of Chemistry 2021 |