Yan Zeng1 Karim Zaghib2 George Demopoulos1

1, McGill University, Montreal, Quebec, Canada
2, Hydro-Québec, Varennes, Quebec, Canada

With the current proliferation of Li-ion batteries (LIB) beyond mobile devices there is a great need to develop inexpensive and safe cathode materials with high energy and power densities. Li2FeSiO4 (LFS) is a promising LIB cathode candidate because of its high theoretical energy density, nontoxicity, abundance of iron and silicon in the earth’s crust, and structurally safe polyanionic framework. However, challenges with full attainment of 2Li capacity, reversibility and conductivity have not allowed for the full potential of these interesting intercalation compounds to be achieved. In search of deeper understanding of the structural chemistry of LFS towards overcoming some of these challenges in this study the effect of dopants is investigated via theoretical modelling and experiment. In terms of modelling, we apply density functional theory (DFT) calculations to examine the viability of several divalent dopants selected from alkali earth metals and 3d transition metals. Computational results suggest that the formation of Li2Fe1-xMxSiO4 solid solution (M = Mg2+, Ca2+, Sr2+, Co2+, and Ni2+, x = 0 – 6.25%) is energetically favourable. For the lithiation/delithiation of the first Li+, which corresponds to the redox couple Fe2+/Fe3+, less than 3.0% volume variation is calculated indicating that good structural stability can be maintained in all the doped LFS materials. Cell voltage is found to increase slightly after doping and shows a linear relation with the doping concentration. The intercalation of the second Li+, on the other hand, requires a voltage higher than 4.5 V vs. Li/Li+ but more importantly is accompanied by a large volume change (around 21%) even with the help of Fe-site substitution. Experimentally, doped-LFS was synthesized with crystal formation control via hydrothermal method. The use of low temperature (e.g. 200oC) provides us the chance to explore the rarely studied low-temperature orthorhombic phase with S.G. Pmn21 as opposed to previous works that have looked the high-temperature monoclinic phase with S.G. P21/n. Magnesium and cobalt substitution of iron in the LFS framework were achieved from sulfate precursor systems, whereas calcium and strontium were doped out from chloride systems due to the limited solubility of their sulfate salts. Nickel, however, was not miscible with LFS, which agrees with previously reported results. Electrochemical performances of doped-LFS as cathode material were investigated in detail via galvanostatic charge/discharge, cyclic voltammetry (CV), and electrochemical impedance spectroscopy (EIS) measurements.