269
Beyond Electrochemical Lithium Ion Battery Characterization. Correlation of Chemical and Microstructure Analysis, a Unique Pathway for Studying Performance Characteristics and Failure Mechanisms

Tuesday, 31 May 2016: 12:25
Sapphire Ballroom A (Hilton San Diego Bayfront)
S. Freitag (Carl Zeiss Microscopy GmbH)
Introduction: Lithium ion batteries (LIB) are considered to be a promising solution for various energy storage applications owing to their high energy density. These batteries have grown in popularity in recent years and are often found integrated into portable electronic devices and electric vehicles. In order to support adoption of LiBs for future applications, the formation, operation, and failure properties must be understood. By designing appropriate battery microstructures, reports have indicated that the charge-discharge characteristics of the battery cell may be controlled.

Methods: This requires a coupled approach of chemical and structural fabrication with image-based and analytical characterization, in order to both design and validate a particular cell geometry model. Image-based characterizations have recently demonstrated some unique abilities to support LIB research, owing to advancements in single imaging instruments as well as correlative microscopy workflows. While scanning electron microscopy (SEM) reveals information about aging effects of for example grains on a micro and nanometer scale, Raman spectroscopy is the tool of choice when it comes to organic and inorganic material identification. ZEISS has worked closely with approved Raman suppliers to integrate the Raman system into the SEM.

Experiment: Two cylindrical batteries type 18650 have been used for the examinations. One cell is shown in its initial condition (new) the other cell was cycled at 25 °C with 8 °C over 480 cycles resulting in a SOH (state of health) of approximately 64 %. The cells were discharged and cut open in a glove box under protective Argon atmosphere. Subsequently the electrode foils were separated, washed in DMC and dried. Small samples were cut from the electrodes and the separator. The samples were mounted in resin. A mechanical preparation by grinding and polishing was used to create the cross sections of the electrodes.

Results: In this study, EDX images showed that the cathode contains of two different active materials: one Mn-rich phase which can be identified as LiMn2O2 and a Co,Ni-rich phase with the approximate stoichiometry of LiNi0.33Co0.33Mn0.33O2 (NCM111). Another observation is the formation of cracks within the active material particles of the cycled electrode. Those cracks are presumably due to volume changes occurring during intercalation/deintercalation of lithium and the subsequent evolution of stress and strain inside the cathode material grains. The measured anode thickness difference was approximately 18 % and the measured cathode thickness changed by 7 %. The crumbling of particles normally leads to contact loss between particles and between particles and current collector, thus increasing impedance. While EDX measurements revealed information about the existing elements and metals, Raman spectroscopy is not able to detect metals but adds information about elements below Z=4 (H, He, Li, Be) and organic as well as inorganic material. In this study, Raman spectroscopy was used to distinguish between different polymers, carbon modifications as well as cathode materials like LiMn2O or LiCoO2. In addition the homogeneity of binder material was investigated. Raman micrographs were overlaid on the SEM data , showing the multi-phase composition of the anode as containing graphite (bright blue) and amorphous carbon (dark blue). Furthermore, by extending this characterization to the other layers, two different polymers were identified within the separator and the cathode was identified as  Li(Mn,Ni,Co)O2. The most impressive results were changes in the trilayer separator of polypropylene(PP)/polyethylene(PE / polypropylene(PP). The comparison of the separator in a new and cycled battery element revealed a change of the orientation of the PP chains molecule orientation in the polypropylene. While the new battery consisted of uniaxial polypropylene, the old battery is of bi-axial polypropylene. The molecule orientation changes the properties of polypropylene in various ways. Polymers, by definition, are long chain molecules in which the atoms are bound to one another by means of strong covalent bonds. High strength and stiffness values in the chain direction are expected since the applied load would then be opposed by the covalent bond themselves. Most of the commercial polymers exhibit strength and stiffness values far below their theoretical limits. To improve these mechanical properties for example highlychain extended/oriented polymers are manufactured. Biaxial polypropylene can be stretched at lower temperatures compared to uniaxial polypropylene. And because of the biaxial orientation, good mechanical properties are achieved, for example strong chemical resistance, abrasion resistance, and good wettability with organic solvents. Also the pore and void size, which normally lies between 35 % and 50 % changes. Hence the observed new molecule orientation can influence significantly the performance of the Li-ion battery.