| dc.description.abstract |
This study investigates the layered oxide cathode with NMC-type LiNixMnyCozO2 as the alternative cathode material for lithium ion batteries. This material has attracted the researcher’s interest as alternative cathode material due to its low cost and less toxicity as compared to the most widely commercialised lithium cobalt oxide (LCO). Lithium nickel manganese cobalt oxide (often abbreviated as NMC) is a type of cathode material used in lithium-ion batteries. It's a popular choice because it offers a balance of high energy density, good cycling stability and relatively low cost compared to other cathode materials. In this study we investigate the stability properties of Ni0.3Mn0.5Co0.2CO3 and Ni0.3Mn0.5Co0.2O2, respectively. In particular, we focus on the manganese rich compositions and minor amounts of nickel and cobalt. We further doped both system (Ni0.3Mn0.5Co0.2CO3 and Ni0.3Mn0.5Co0.2O2) with fluorine, titanium, niobium and chromium to check if their contributions could improve or disprove the behaviour of Ni0.3Mn0.5Co0.2CO3 and Ni0.3Mn0.5Co0.2O2 materials. Firstly, the structural, electronic, mechanical and vibrational properties of Ni0.3Mn0.5Co0.2CO3, Ni0.3Mn0.5Co0.2O2 and their doped systems have been calculated using the density functional theory employing the pseudo-potential plane-wave approach within the local gradient approximation with the Hubbard parameter U for strongly correlated transition metals. The structural property calculations included the equilibrium lattice parameters, density and energy of formations while electronic properties included the partial density of states (PDOS), total density of states (TDOS) and band structures for all the systems. Furthermore, mechanical properties investigated the elastic constants, Pugh ratio and anisotropy while vibrational properties investigates the phonon dispersion curves for Ni0.3Mn0.5Co0.2CO3, Ni0.3Mn0.5Co0.2O2 and their doped systems. The calculated lattice parameters and energy of formation could be used for benchmarking in the future since no similar work was found in literature for comparison. Moreover, the calculated energy of formations revealed the relatively low and negative values for all the systems, suggesting thermodynamic stability. With the band structures, we found that Ni0.3Mn0.5Co0.2CO3 and Ni0.3Mn0.5Co0.2O2 structures were semiconductors with a direct gap of 0.004 eV and 0.036 eV with their doped systems also indicating metallic characteristics. Moreover, the partial density of states for our materials and their doped systems were also found to be metallic as there was no energy band gap observed at the Fermi line. Furthermore, the elastic constants revealed that all our systems recorded 21 independent elastic constants which falls within the triclinic lattice systems. For a material to be considered mechanically stable within the triclinic system, there are conditions to be satisfied, hence Ni0.3Mn0.5Co0.2CO3 satisfied all the conditions suggesting mechanical stability while Ni0.3Mn0.5Co0.2O2 did not satisfy all the conditions implying mechanical instability. The phonon dispersion curves revealed that Ni0.3Mn0.5Co0.2CO3 was vibrationally stable while Ni0.3Mn0.5Co0.2O2 was vibrationally unstable due to the presence of negative vibrations along the Brillouin zone. Furthermore, the phonon dispersion curves for doped systems revealed that some are vibrationally stable while some are vibrationally unstable. Secondly, since the study focuses on manganese rich systems, cluster expansion technique was used to generate phases in the manganese rich side. From the results, various phases with varied concentrations and symmetries were produced by the ground-state phase diagram. The accuracy of new structures during cluster expansion fitting is indicated by the cross validation score (CVs) for all of the generated new structures being less than 5meV per active atom position. Since all of the developed structures have CVs below 5meV, this indicate that our calculations were valid and the newly generated structures will work realistically. From the phase diagram, we noticed that all the predicted phases are in the negative energy of formations side (miscible constituent) which indicate thermodynamic stability. Moreover, of all the phases generated within the diagram, only phases in the manganese rich side were explored by using first principles calculations to further confirm their stability properties by determining their structural, electronic, mechanical and vibrational properties. The energy of formation results revealed that all the phases are thermodynamically stable while electronic properties revealed metallic characteristic for all the phases in the Mn-rich side. For mechanical properties, we found that few phases did not satisfy the triclinic conditions which implies mechanical instability while other phases were found to satisfy the conditions, indicating mechanical stability. Lastly, the carbonate co-precipitation method was used in this study to synthesize the transition metal carbonate precursors using a 4L stirred tank reactor (CSTR) under steady state circumstances. We produced Ni0.28Mn0.53Co0.19CO3 and Ni0.17Mn0.67Co0.17CO3 which was later lithiated to form LiNi0.33Mn0.53Co0.14O2 and LiNi0.17Mn0.67Co0.17O2 as our layered cathodes. Both the lithiated samples were further characterized for thermogravimetric analysis, x-ray diffractions, morphologies, EDX and XRF. Thermogravimetric analysis revealed thermodynamic stability for both samples while XRD’s also managed to produce the most crystalline peak at 003 indexing. The scanning electron microscopy was also tested to determine the particle size and distribution for both samples and the results revealed a homogeneous particle distribution in each sample. We further collaborated with University of Kent for the usage of the synchrotron beam of the Diamond light source to determine the effect of fluorination on our NMC samples. In particular, we wanted to check if fluorination reduces or increases the oxidation states of metals within our samples and results revealed that fluorination does not change the oxidation state of our samples. |
en_US |