Explaining
battery voltage fade

Lithium-rich layered oxides (LLROs) are promising cathode materials for better rechargeable batteries for electric vehicles. They are plagued by a phenomenon called voltage fade, however: When these batteries go through series of charge–discharge cycles, their voltage fades, and with it the amount of energy it can hold and later release for use. An international team led by the University of California, San Diego, has uncovered how this process occurs at the nanoscale. Combined measurements at the Advanced Photon Source (APS) of Argonne National Laboratory (ANL) in the USA and the PETRA III synchrotron source at the DESY research centre in Germany revealed that a mobile network of nanoscale defects forms in the LLRO material during charging, with dislocation density increasing upon repeated charging. These dislocations dramatically alter the local lithium environment and contribute to the voltage fade. Based on their findings, the team devised a method to recover the original high voltage functionality: Heat-treating the cathode materials eliminated most of the defects and restored the original voltage.

The team uncovered the formation of the dislocations using in situ 3D Bragg coherent X-ray diffractive imaging (BCDI) on nanoparticles made of LLRO and of a classical layered oxide for comparison. The BCDI technique enabled them to directly image the interior of the nanoparticles during battery charging – under operating conditions and at nanoscale resolution. The measurements at PETRA III were conducted using a LAMBDA detector, whose small pixel size proved crucial to the success of the experiment: “In a coherent experiment, we measure interference patterns called speckles,” explains co-author Michael Sprung, scientist in charge of the P10 Coherence Applications beamline at PETRA III. “To successfully conduct such a BCDI study, the speckles need to have a certain minimum size with respect to the pixel size of the detector. In this case, they needed to cover at least five pixels.” The pixel area of 55 µm by 55 µm of the LAMBDA detector is nearly a factor of 2 smaller than that of conventional detectors. “This small pixel size was crucial to actually carrying out the experiment at all, due to restrictions of the minimum reachable beam size at the diffractometer setup”, says Sprung.