Charakterisierung des Festelektrolytsystems (1−x) Li3PS4 + x LiI sowie dessen Anwendung in Lithiumionen-Festkörperbatterien

In this dissertation, the ion transport properties of the ball-milled solid electrolyte system (1−x) Li3PS4 + x LiI as well as the application of the electrolyte with x = 0.33 in all-solid-state batteries (ASSB) were characterized. The obtained results were published in five articles. Three publications deal with the characterization of the solid electrolyte system (1−x) Li3PS4 + x LiI. Here, the lithium ion conductivity, the lithium ion transport and the structural changes in this system were investigated after a single annealing step of the solid electrolyte (SE). In the other two publications the SE 0.67 Li3PS4 + 0.33 LiI was used in ASSBs. On the one hand the solid electrolyte was applied in the composite cathode with two different active materials. For this purpose, thickness-dependent composite cathodes were examined and the exchange current density was determined for both active materials. On the other hand, the pressure-dependent behavior of the SE 0.67 Li3PS4 + 0.33 LiI in contact to lithium metal was analyzed. These results were used to assess the lifetime of a lithium ion battery. In the first publication, the SE 0.67 Li3PS4 + 0.33 LiI was characterized after a single annealing step. The lithium ion conductivity after this single annealing step at 180 °C increases from 0.8 mS∙cm−1 to 6.5 mS∙cm−1. This huge conductivity enhancement was examined by X-ray diffraction (XRD) and 7Li nuclear magnetic resonance (NMR), especially with respect to the structural changes and the lithium ion dynamic after the single annealing step. It was found that the amorphous structure of the SE is still detectable after the heat treatment. Also the lithium ion dynamic shows only a single broad distribution of lithium ion jump rates, but no bimodal distribution, so that a fast ion transport in the amorphous phase also is established after the single annealing step. Positron-annihilation-lifetime-spectroscopy was carried out to test this hypothesis. With this method, an increase of the free volume in the material after the annealing step was found. Hence, a novel conductivity enhancement mechanism was suggested, which is based on a faster lithium ion transport in the amorphous phase. In a second publication, this mechanism was further investigated by different 7Li NMR measurements. With 7Li spin-lattice relaxation and line-shape analysis the local hopping motion of the lithium ions were ascertained. Furthermore, 7Li field cycling relaxometry and the static field gradient technique were used as well to measure self-diffusion coefficients of the lithium ions. Altogether the hypothesis of a faster lithium ion transport in the amorphous phase after a single annealing step in the SE 0.67 Li3PS4 + 0.33 LiI was confirmed. For further investigations of this conductivity enhancement mechanism, the solid electrolyte system (1−x) Li3PS4 + x LiI was investigated for different compositions in the range x = 0 – 0.5. In the third publication the crystallization temperature of (1−x) Li3PS4 + x LiI was measured by differential scanning calorimetry and all samples were annealed at 10 K below the first crystallization peak. To obtain more information on the crystallization behavior of this electrolyte system, XRD, 7Li NMR and 31P NMR measurements were performed. The systematic characterization of (1−x) Li3PS4 + x LiI showed a different crystallization behavior depending on the LiI amount x. It is mentioned that at all stoichiometries, nanocrystallites were found. The stoichiometries could be subclassified into three groups by XRD and Raman spectroscopy, according to the observed crystallization behavior. In the first group with x = 0 – 0.15, the formation of mostly β-Li3PS4 is found, in the group for x = 0.45 and 0.5, predominantly the formation of Li4PS4I. In both cases a small conductivity enhancement after annealing is observed. On the contrary, a huge conductivity enhancement exists in case of the third group (x = 0.2 – 0.4). The thio-LISICON II phase was identified to be responsible for a drastic conductivity enhancement for a composition of x = 0.2. Also in previous research a superionic thio-LISICON II phase was reported. The conductivity was increased with a higher LiI content up to x = 0.4, however to a lesser extent. The 7Li NMR measurements affirmed this increase in conductivity. Interestingly, in this group the ratio of the thio-LISICON II phase decreased with ascending LiI amount, which did not agree with an increasing conductivity. Hence, it is expected that other nanoscale phases exist with high lithium ion conductivities. Furthermore, 31P NMR measurements confirmed this results and carried out these structural changes in more detail. The observation of nanoscale Li4PS4I for x = 0.33 suggests a highly conductive disordered Li4PS4I phase which increases with increasing LiI amount, but is not detectable by XRD. This disordered phase must be clearly distinguished from the Li4PS4I phase which is formed for x = 0.45 and 0.5, detected by Bragg peaks in the XRD. The same trend can be observed in the Raman spectra. Consequently, it is expected that different nanoscale crystallites influence the conductivity enhancement in the ball-milled solid electrolyte system (1−x) Li3PS4 + x LiI. Especially, there are likely other superionic crystalline phases which have even higher ionic conductivities than the thio-LISICON II phase. Knowing this, a faster ion transport in the amorphous phase after a single annealing step can be excluded, in contrast to the interpretation put forward in the first and second publication. In the fourth publication, the solid electrolyte 0.67 Li3PS4 + 0.33 LiI was used in all-solid-state batteries. Thereby, two different cathode active materials (CAM) were compared. The solid electrolyte 0.67 Li3PS4 + 0.33 LiI was used as separator and as electrolyte in the composite cathode. Indium was used as anode material. Thickness-dependent composite cathode measurements in ASSBs were carried out. A comparative case study of the single crystalline active materials LiCoO2 (LCO) and LiNi0.83Mn0.06Co0.11O2 (NMC) with a LiNbO3 coating was performed to determine exchange current densities. Impedance spectra of ASSBs with various cathode thicknesses were measured in a state of charge of 50 %. For the interpretation of the impedance behavior, the transmission-line model (TLM) was used with the aim to determine the exchange current density of both cathode materials in contact to 0.67 Li3PS4 + 0.33 LiI. For this purpose, several parameters were fixed at experimentally determined values. Other parameters such as the tortuosity ???+, the lithium diffusion coefficient in CAM ????, the non-ideal double layer capacitance parameters ??? and ? as well as the dependence of equilibrium potential on lithium concentration in CAM d?d???⁄ could be assessed by literature, further electrochemical experiments or fitting parameters of the impedance spectra. With this starting values, impedance spectra could be simulated that reproduced the experimental behavior sufficiently well. This enabled us to evaluate values of the exchange current density for both active materials. The determined values are quite different. LCO has an exchange current density of 25.8 A∙m−2 while NMC has a distinctly lower value of 0.11 A∙m−2. They were compared to exchange current densities liquid lithium ion batteries described in literature. For LCO-based electrodes with a liquid electrolyte a value of 0.4 A∙m−2 was reported, which is more than a magnitude smaller than the exchange current density found for ASSBs in this study. However, a value in a range of 2 – 4 A∙m−2 was reported for NMC-based electrodes with liquid electrolytes. This value is more than a magnitude higher than the exchange current density which we estimated for ASSBs. Finally, pressure-dependent measurements of the electrolyte system 0.67 Li3PS4 + 0.33 LiI in contact with lithium metal were executed. The results were published in a fifth article where the lifetime of battery cells with a sulfide solid electrolyte and lithium metal electrodes was investigated. For this purpose, electrochemical measurements as well as X-ray photoelectron spectroscopy and time-of-flight secondary ion mass spectrometry were carried out to examine the interface between lithium metal and the solid electrolyte after lithium stripping and plating experiments. Here, the influence of a solid electrolyte interphase had to be taken into account. A further influencing factor for the lifetime of a lithium metal battery is the interface between the solid electrolyte and lithium metal. On the one hand, lithium vacancies in the lithium metal due to stripping needed to be considered. After plating on the other hand, a seed-like growth behavior of lithium was observed. At this contact spots, higher local current densities might be existent, which lead to faster dendrite growth and thus to a shorter battery lifetime. This effects also were considered in the processing results.