Large scale simulation studies of lithiated Li-Mn-O nanoarchitectures

Abstract

The vast majority of typical cathode materials for lithium ion batteries suffers from severe specific capacity due to structural changes that occur during the process of cycling, which may lead to fractures within the battery material. Nanostructured materials can act as lattice hosts to accommodate lithium during the charging and discharging process of lithium ion batteries. Nanoporous materials specifically have been studied and found to have large surface areas and volume density, which allows them to flex within their pores during cycling. However, not much has been done to apprehend their electrochemical and mechanical properties at nanoscale especially at intermediate lithium concentrations where the structures undergo phase transitions and how these changes or transitions affect cycling. For that reason, this study seeks to understand the implications that come with structural changes and how they affect the mechanical performance of the battery material; and eventually, come with a better structure that can withstand harsh battery conditions as severe capacity fade. Computational simulation methods, using the amorphisation and recrystallisation technique employing the DL_POLY code, were used to generate and simulate the spinel LiMn2O4 nanoarchitectures. The conductive ion host capabilities of the Li-Mn-O composite nanoarchitectures (porous and bulk) with co-existing layered-spinel (Li2MnO3, LiMnO2, LiMn2O4 and Mn3O4) materials are investigated to predict their structural, electrochemical and mechanical properties during the discharge process. These properties are very important since they provide information on microstructural features and possible volume variations that may occur during cycling. Furthermore, the amorphisation and recrystallisation technique can assist in investigating the effects of diffusion induced stress during the cycling process. The discharge process was simulated by chemically lithiating the Li1+xMn2O4 nanoarchitectures and bulk at different lithium concentrations, where 0 ≤ x ≤ 1. To be precise, the structures were lithiated to five different concentrations of Li1.00Mn2O4, Li1.25Mn2O4, Li1.50Mn2O4, Li1.75Mn2O4 and Li2.00Mn2O4. The simulated structures are nanoporous 75 Å, nanoporous 69 Å, nanoporous 67 Å and the bulk (63 Å). The structures were amorphised and recrystallised at 1700 K, then characterised by interrogating their total radial distribution functions (RDFs) and their X-ray diffraction patterns (XRDs). Furthermore, the structural integrity of the materials was investigated through microstructural and structural changes in terms of volume variation, defects and grain evolution with increasing lithium concentration. The total RDFs and structural snapshots for the Li-Mn-O composites show efficiently and spontaneously recrystallised structures that evolved into single and multiple grains. The multiple grained structures are observed mainly at the Li1.75Mn2O4 concentration. However, few to no grain boundaries for nanoporous 69 Å recrystallised structures at NST and NVT are observed, respectively. Meanwhile, for the bulk recrystallised under the NST ensemble, the structure remains a single grain throughout the discharge process. The Li1.75Mn2O4 is the concentration where the structures change symmetry from cubic to tetragonal phase. The harvested microstructures show a wealth of crystallographic defects that reduce with increasing lithium content. The increase in lithium concentration also influences the abundance of the spinel Mn3O4, which reduces with lithiation. The microstructures also show the layered-spinel components co-existing within structures. The XRD validates the co-existence of layered-spinel structural composites by characterising the signature and fingerprint peaks of these materials when compared to the experimental. The XRDs for the simulated structures depict significant shifts, splits and broadening of peaks with lithiation. The nanoporous, unlike the bulk structures, have channels (pores) that allow for better lithium atom transportation and diffusion. The lattice dimension of these nanoporous structures is relative to the pore size. This implies that a material with a larger lattice will have a bigger pore that can possibly accommodate more lithium atoms on its structure and cavity. Further investigations on the nanoporous structures reveal that pore sizes reduce with increasing lithium concentration under NVT ensemble recrystallisation, except for nanoporous 69 Å at Li1.75Mn2O4, where the pore is observed to increase in size. Recrystallising the structures under the NST ensemble result in the pores increasing in size from Li1.00Mn2O4 to Li1.25Mn2O4, and thereafter reduce with increasing concentration where some pores completely close up at Li1.75Mn2O4. However, full lithiating the structures to Li2.00Mn2O4 under both ensembles results in the materials almost regaining their original pore sizes. The structures were also exposed to temperature to investigate their lithium diffusivity and stability with lithiation. The diffusion coefficients of lithium in the structures with increasing concentration and temperature also show that pore size influences diffusivity. This is because nanoporous 75 Å shows the highest lithium diffusion compared to its counterparts followed by nanoporous 69 Å, then nanoporous 67 Å and lastly, the bulk, where the stability of the materials is maintained. The structures also undergo volume expansion during lithiation under the NST ensemble and they all maintain their structural integrity throughout. However, nanoporous 69 Å at Li1.75Mn2O4 is observed to be resilient to expansion. Meanwhile, the bulk volume drops drastically from Li1.75Mn2O4 to Li2.00Mn2O4 concentration. The reason for the nanoporous structures to withstand lithiation without fail is attributed to their ability to flex within their pores and larger surface areas; meanwhile, for the bulk, the Mn3O4 walls could be collapsing. The mechanical properties of the nanoporous materials investigated through stress-strain calculations, with the imposed uniaxial stress of 0.1 GPa, show RDFs which have broad and flat peaks with increasing radial distance r, implying the disorder in atom arrangement. Meanwhile, the XRDs show peaks merging to form one broad peak; while other peaks shift and split. The stress-strain plots reveal that nanoporous 69 Å especially at Li1.75Mn2O4 has the highest yield strength when compared to its counterparts. This indicates that nanoporous 69 Å is more robust since it can evolve into one or few grains, resist expansion, fracture and also have efficient lithium diffusion at varied temperatures. This implies that nanoporous 69 Å material has the potential to curb the current battery setbacks such as volume expansion that causes cracks within the battery material during cycling causing battery degradation.