Morphology parameters of C/Li2MnSiO4 composites calcined at different temperatures.
Technological development of portable devices, e.g, mobile phones, laptops, etc., as well as progress in electrical vehicles (EV) and hybrid electrical vehicles (HEV) technologies require batteries efficient in volumetric and gravimetric energy storage, exhibiting large number of charge/discharge cycles and being cheap and safe for users. Moreover, materials used in energy storage and conversion systems should be environmentally friendly and recyclable. Currently, rechargeable lithium-ion batteries (LIBs) are the most popular portable energy storage system, mostly due to their highest energy density among all others rechargeable battery technologies, like Ni-Cd or Ni-MH cells which reached the theoretical limit of performance. Commercially available LIBs are based on layered lithium cobalt oxide (LiCoO2) or related systems, which are expensive and toxic. These materials are unstable in an overcharged state, thus the battery safety is affected, especially in high power (20-100 kWh) applications for EV, HEV and renewable energy systems. The bigger battery capacity results in more energy accumulated, thus operational safety is a key issue. On the other hand, lifetime and capacity retention of LIBs in changeable operation conditions (from -30°C to +60°C, average lifetime 2-4 years) are a challenges to develop new materials and cell assembly technologies.
2. Li-ion battery technology
First rechargeable lithium cells taking advantages of intercalation process were developed in year 1972 . The cells Li/Li+/LixTiS2 revealed 2V potential and relatively low gravimetric capacity. Applications of metallic lithium as anode material resulted in common cell breakdown due to formation of dendritic structures on anode during cell cycling. The problems forced research and development of new intercalation materials for lithium batteries. In the eighties a new conception of lithium cell was proposed (so called Li-ion batteries or “rocking-chair batteries”), which consisted of application of two different lithium intercalation compounds as anode and cathode materials [2-4]. As anode material a graphite intercalated with lithium was used while cathode materials were based on layered 3d transition metal oxides.
2.1. Layered LiCoO2 oxide and related systems
The first LixC6/Li+/Li1-xCoO2 Li-ion battery system was commercialized in 1993 by Sony Co. In Fig. 1 a working mechanism during discharge cycle of Li-ion cell is presented.
The electrochemical reaction at the graphite anode side can be written as (1):
and suitably at LiCoO2 cathode side (2):
Commercially available LIBs based on layered lithium cobalt oxide (LiCoO2) or related systems (LiNiyCo1-yO2, LiMnyCo1-yO2, LiMn1/3Ni1/3Co1/3O2) reveal reversible capacity 130-150 mAh/g and working potential 3.6-3.7 V. The oxide materials are expensive and toxic due to cobalt content and are unstable in an overcharged state, thus the safety is strongly affected. This is related to strong oxidizing behavior of the charged layered oxide cathode in contact with organic electrolyte what may lead to combustion or even explosion [4, 5]. Unfortunately, this effect may be also increased by application of nanosized materials with high surface area.
2.2. Spinel LiMn2O4 and related systems
An alternative material for cathode based on layered oxides LiMyCo1-yO2 (M=Ni, Mn, Fe) is LiMn2O4 spinel. LiMn2O4 reveals a little lower reversible capacity 120 mAh/g at working potential 4 V, but the material is distinctly cheaper and nontoxic. However, application of spinel as cathode material in commercial Li-ion batteries is retarded by phase transition observed near room temperature, i.e. battery operation temperature, and related to Jahn-Teller distortion of Mn3+ ions, what resulted in capacity fading. Stabilization of cubic spinel structure is possible by controlled formation of cationic defects , by lithium  or 3d metal substitution  into spinel lattice as well as by isoelectronic sulfur substitution [9, 10]. Instability of high oxidation state of transition metal in spinel structure observed in charged state of cathode leads to oxygen evolution, similarly to the layered oxide cathodes, and reaction with electrolyte. Application of spinel based materials is limited to cheap battery packs for EV .
2.3. Olivine LiFePO4 cathode material
An interesting and very promising group of insertion materials are LiMXO4 (M= metal 3d, X=S, P, As, Mo, W) type compounds , with LiFePO4 among them. LiFePO4 lithium iron phosphate of olivine structure is chemically and thermally stable material with relatively high gravimetric capacity 170 mAh/g at working potential 3.5 V. The material is cheap, nontoxic and environmental friendly. Its high chemical stability towards electrolyte, related to strong P-O bonds, significantly improves safety of LIBs. However, very low electrical conductivity of LiFePO4 system (10-9 S/cm @RT) requires application of carbon coatings and composite formation [13, 14].
3. Orthosilicates Li2MSiO4 – New high capacity cathode materials
Application of orthosilicates Li2MSiO4 (M=Fe, Mn, Co, Ni) compounds as insertion materials for LIBs was firstly proposed by Prof. Goodenough [12, 15]. The materials mainly crystallizes in orthorhombic system of Pmn21 space group in olivine structure [16-18]. The Li2MSiO4 olivine structure may be described as wavy layers of [SiMO4]∞ on ac axis plane and connected along b axis by LiO4 tetrahedra . Within the layers every tetrahedron SiO4 shares its corner with four next MO4 tetrahedra. Lithium ions occupy the tetrahedral sites (LiO4) between two layers and share 3 and 1 oxygen atoms with the layers. In fact, diffusion of lithium ions in this structure is possible only through the canals formed by LiO4 tetrahedra. Li2MSiO4 silicates reveal a possibility of reversible insertion of two lithium ions per molecule, so this leads to exchange of two electrons by transition metal. As a results the silicates reveal very high theoretical gravimetric capacity, up to 330 mA/g. However, structural limitation due to sharing of LiO4 between two wavy layers, should affect a little the capacity. Calculations of electrochemical potential of lithium insertion/deinsertion show a two stage process:
Depending on transition metal potential of the reaction (3) vary from 3.2 V (for Fe) to 4.1-4.4 V (for Mn, Co, Ni), while the potential of reaction (4) is in range 4.5-5.0 V. Deinsertion of the second lithium ion requires applying high potential above 4.5 V, and this is a challenge for electrolyte. Orthosilicates Li2MSiO4 (M=Fe, Mn, Co, Ni) materials reveal very low electrical conductivity (10-12 - 10-15 S/cm @RT) and carbon coating of the materials is required. On the other hand, downsizing of material grains should improve electrochemical performance . Very strong covalent bonding of Si-O results in high chemical stability towards electrolyte, strongly increasing the safety of LIBs based on silicates. The materials are nontoxic, environmental friendly and cheap. The properties mentioned above provide a challenge for developing a composite cathode materials based on Li2MSiO4 orthosilicates.
4. Li2MnSiO4 nanostructured cathode material
Li2MnSiO4 is a member of dilithiumorthosilicates Li2MSiO4 (M = Fe, Mn, Co) family, thanks to strong covalent Si-O bond, it shows high thermal and chemical stability. High theoretical capacity 333 mAh/g, low production costs, safety and optimal working potential make them an attractive cathode material. Very low electrical conductivity (10-12 - 10-15 S/cm @RT) can be improved by coating with conductive carbon layers (CCL) and by grains size downsizing [18-20]. The properties of Li2MnSiO4 material comes to a conclusion that cathode material for LIBs based on this compound should be prepared as C/Li2MnSiO4 nanocomposite. Thus, the special preparation techniques have to be applied in terms to obtain nanocrystaline grains of Li2MnSiO4 cathode material coated by conductive carbon layers (CCL). During last few years several different technics of Li2MnSiO4 synthesis were proposed. Hydrothermal synthesis [21-24] as well as solid-state reactions [25-28] can lead to one phase product but the control of the grain size is significantly limited. Sol-gel method is one of the soft chemistry techniques which can be used in synthesis of nanosized lithium orthosilicates [29-33]. Sol-gel processes, especially Pechini’s method is a simple technic, characterized by low cost and low temperature of treatment, resulting in homogenous, high purity materials.
4.1. Preparation of nanostructured Li2MnSiO4
Li2MnSiO4 was produced using sol-gel synthesis – Pechini type. Starting reagents were: lithium acetate dihydrate (Aldrich), manganese acetate tetrahydrate (Aldrich), tetraethoxysilane (TEOS) (98%, Aldrich) as a source of silicon, ethylene glycol (POCh), citric acid (POCh) and ethanol (POCh). Thanks to chelating metal ions in solution by citric acid the cations can be mixed at the molecular level and the stoichiometric composition can be achived. The reactants were mixed in a molar ratio 1:1:18:6:4:16 - Mn:Si:C2H6O2:C6H8O7:C2H5OH:H2O. Based on previous studies it was affirmed, that using 20% excess of lithium acetate leads to one-phase product. All reagents were dissolved in glass reactor under constant argon flow (Ar 5.7). As a solvent, only stoichiometric amount of distilled water was used. Heating water to 35 ˚C assure fast and complete dissolution of metal acetates. Prepared mixture was heated to 60 ˚C and few drops of concentrated hydrochloric acid was added to initiate polymerization of metal citrates using ethylene glycol and TEOS. Reaction was conducted for 24h in close reactor. Obtained gel was aged for 3 days at 60 ˚C in close reactor (Ar atmosphere) and for 3 days at 60 ˚C in an air-drier (Air atmosphere).
Thermogravimetric analysis (TGA) coupled with simultaneous differential thermal analysis (SDTA) and mass spectrometry evolved gas analysis (MS-EGA) of precursor were performed in Mettler-Toledo 851e thermo-analyzer using 150 μl corundum crucibles under flow of air/argon (80 ml/min), within temperature range 20–800 °C with heating rate of 10 °C/min. The simultaneous MS-EGA was performed in on-line joined quadruple mass spectrometer (QMS) (Thermostar-Balzers). The 17, 18 and 44 m/z mass lines, ascribed to OH, H2O and CO2 species respectively, were collected during the TGA experiments. TGA of the Li2MnSiO4 precursor are shown in the Fig. 2 and Fig. 3.
According to TGA curve (Fig. 2a) a complete decomposition of gel organic matrix occurs above 500 °C. Evolved gas analysis confirmed the disintegration of organic components. Amount of active material can be estimated at about 12 wt.% of the precursor. Basing on TGA results calcination conditions were chosen. Li2MnSiO4 precursor was calcined under Ar flow at 600, 700, 800 and 900 °C.
4.2. Properties of nanostructured Li2MnSiO4
X-ray powder diffraction patterns of the samples were collected on BRUKER D2 PHASER using Cu Kα radiation = 1.5418 Å. In Fig. 4 XRD patterns of Li2MnSiO4 obtained at different temperatures (600, 700, 800 and 900 °C) are collected.
In the sample calcined at 600 °C all diffraction lines can be attributed to pure Li2MnSiO4 phase (Pmn21). In other specimens obtained at higher temperatures, MnO and Li2SiO3 impurities starts to appear. Formation of MnO and Li2SiO3 in samples calcined at temperatures above 700 °C can be connected, with Li2MnSiO4 decomposition. Average crystallites sizes calculated from diffraction lines broadening based on Scherrer equation are in the range of 17-19 nm depending on calcination temperature of the sample (Tab. 1). The calcination under carefully selected conditions leads to a transformation of a part of organic compounds from the gel matrix into thick carbon layer covering the active material. The temperature programmed oxidation (TPO) performed using TGA equipment confirmed presence of carbon in samples. Fig. 5b presents exemplary TG/DTG/SDTA curves from TPO conducted under air flow in temperature range of 20-1000 °C. The amounts of carbon in samples calcined in different temperatures were calculated from TPO measurements and are collected in Table 1.
Measurements of the specific surface area of the samples were performed in Micrometrics ASAP 2010 using BET isotherm method. About 500 mg of each sample was preliminary degassed at 250–300 °C for 3 h under pressure 0.26–0.4 Pa. Then, N2 sorption was performed at pressure of 8 104 Pa. One of pore size distribution plot with the adsorption-desorption curves inset is presented in Fig. 5a. Exact values of specific surface area and average pores diameters are presented in table 1.
Low temperature N2-adsorption (BET) measurements of C/Li2MnSiO4 composites show that specific surface area of samples decrease with increasing calcination temperature. Decrease of specific surface area is connected with graphitization of carbon during calcination at higher temperatures.
|Sample name||Precursor calcination temperature||Carbon content||Specific surface area (BET)||Average pore dimension (BJH)||Crystallite size (from XRD)|
|C/Li2MnSiO4@600||600 °C||34%||169 m2/g||50 Å||17 nm|
|C/Li2MnSiO4@700||700 °C||32%||152 m2/g||72 Å||18 nm|
|C/Li2MnSiO4@800||800 °C||32%||144 m2/g||73 Å||19 nm|
|C/Li2MnSiO4@900||900 °C||35%||57 m2/g||73 Å||19 nm|
Electrical conductivity was measured using the AC (33Hz) 4-probe method within temperature range of -40÷55 °C. The carbon coated composite powders were so elastic that the standard preparation of pellets was impossible. The fine powder samples were placed into a glass tube and pressed by a screw-press between parallel gold disc electrodes (∅=5 mm) till the measured resistance remained constant. The results of electrical conductivity measurements are gather in Fig. 6. All composites exhibit good electrical conductivity (up to 0,36 S/cm for C/Li2MnSiO4@900) in comparison with Li2MnSiO4 itself ( ̴ 10-15 S/cm at room temperature - RT). It can be observed that temperature dependence of conductivity for composites calcined at higher temperatures (700, 800 and 900 °C) seem unaffected, by temperature (metallic-like behavior). Conductivity value and its invariability against temperature indicate that carbon layers consist of graphite-like domains. In case of C/Li2MnSiO4@600°C the carbon layers is more disordered and probably consist of an activated-like carbon. Those results are consistent with BET measurements in case of graphitization process occurring at higher temperatures.
Selected samples were investigated using transmission electron microscope (TEM) (Fig. 7 and Fig. 8). Micrographs were collected on TECNAI G2 F20 (200 kV) coupled with an energy dispersive X-ray spectrometer (EDAX).
TEM micrographs reveal well dispersed lithium manganese orthosilicate grains (dark grey and black dots) in the carbon matrix. Figures 7a; 7c; 7d present C/Li2MnSiO4@600 at different magnifications. Fig. 7b shows STEM-HAADF (Scanning Transmission Electron Microscopy – High Angular Annular Dark Field) image of the C/Li2MnSiO4@600 composite. Energy dispersive X-ray spectroscopy (EDS) (Fig. 7b1; 7b2) confirmed presence of lithium manganese silicate in the spots marked in the micrograph 7b (Cu peaks in all EDS spectra originate from TEM copper grid). High resolution electron microscopy image (HREM) shows single grain of Li2MnSiO4 (Fig. 7e). Corresponding to that IFFT (Inverse Fast Furrier Transform) image (Fig. 7f) clearly shows a crystalline structure of obtained material which can be identified as Li2MnSiO4. Grains in the composite have sizes in a range of 5-10 nm. Microstructure of a C/Li2MnSiO4 composite calcined at 700 °C is shown in Fig. 8. C/Li2MnSiO4@700 °C composite has a bimodal distribution of a crystallite size. TEM micrographs show presence of small (5-10 nm) Li2MnSiO4 crystallites, analogues of C/Li2MnSiO4@600 °C and a bigger, well crystallized silicates grains, in a range of 50-150 nm (Fig. 8a, c, d). Both types of crystallites in the sample have the same composition which is confirmed by EDS analysis presented in fig. 8b1, b2, b3.
STEM-HAADF image (Fig. 8b) from the same region as a bright field Fig. 8a shows points from where EDS analysis was conducted. All of lithium manganese orthosilicate crystallites in C/Li2MnSiO4@700 °C are well dispersed and fully covered in amorphous carbon. HREM image in Fig. 8d1 with IFFT in Fig. 8d2 shows crystalline structure of Li2MnSiO4 grain.
5. C/Li2MnSiO4 composite cathode material
Nevertheless obtained composites show homogeneous distribution of particles in carbon matrix and exhibit good electrical conductivity they work poorly in a battery cell (see paragraph 5.2). Thickness of primary carbon coating on active material grains strongly limits the electrochemical performance of composite.
5.1. CCL/Li2MnSiO4 composite preparation and properties
Preparation of composites with well define morphology of carbon layers and optimal carbon content was achieved by burning out the primary carbon and recoating Li2MnSiO4 nanosized grains with conductive carbon layers (CCL/Li2MnSiO4 composites). Burning out of primary carbon formed in the sample during synthesis process was carried through calcination of C/Li2MnSiO4@600 and C/Li2MnSiO4@700 under air flow at 300 °C for 3 h. CCL/Li2MnSiO4 composites were produced by wet polymer precursor deposition on active material grains and subsequent controlled pyrolysis [14, 20, 34]. Poly-N-vinylformamide (PNVF) obtained by radical-free polymerization of N-vinylformamide (Aldrich) with pyromellitic acid (PMA) additive (5-10 wt%) was used as a carbon polymer precursor . To achieve an impregnation, Li2MnSiO4 grains were suspended in water polymer solution (8-15 wt%). Finally, samples were dried up in an air drier at 90 °C for 24 h. Prepared samples were pyrolyzed at 600 °C for 6 h under inert atmosphere (Ar).
Burning out primary carbon from the surface of lithium manganese silicate in air atmosphere leads to partial decomposition of active material. Fig. 9 shows XRD patterns of C/Li2MnSiO4@600, Li2MnSiO4@600 without carbon and CCL/Li2MnSiO4@600 composite after coating with CCL from polymer precursor.
In diffraction patterns of Li2MnSiO4@600 it is clearly visible that after burning out primary carbon layer new phase appears in the sample (LiMn2O4). Proposed carbon coating process can reverse decomposition of Li2MnSiO4. After coating Li2MnSiO4 (Pmn21) is a dominant phase in CCL/Li2MnSiO4@600 and all the lines, except for very low intensity line around 19°, can be associated with Li2MnSiO4 phase. CCL preparation procedure influence active material grain size. Average crystallite size calculated using Scherer equation in CCL/Li2MnSiO4@600 is ca. 25 nm. TEM analysis confirms increase of crystallites diameters. Li2MnSiO4 grains observed under TEM are in the range 25-75 nm. TEM micrographs of CCL/Li2MnSiO4@600 are presented in Fig. 10.
Bright field micrographs (Fig. 10a, c, d) shows crystallites of Li2MnSiO4 covered with conductive carbon layers (CCL). Formed carbon coatings adhere well to the surface of active material grains, no voids are visible at the CCL/Li2MnSiO4 interface (Fig. 10e). Silicate grains are uniformly covered with approximately 4-5 nm thick CCL. EDS analysis once again confirm presence of lithium manganese silicate in the composite (Fig. 10b, b1). Carbon content in samples was calculated from TPO measurements and they are displayed in Table 2.
|Sample name||Precursor calcination temperature||Carbon content|
5.2. Electrochemical properties of C/Li2MnSiO4 composite cathode
The charge-discharge cycling studies of Li/Li+/(C/Li2MnSiO4) cells were conducted in a four electrode configuration using CR2032 assembly between 2.7 and 4.7 V at C/200 rate at room temperature. LiPF6 solution 1M in EC/DEC (1:1) was used in the cells as an electrolyte. The galvanostatic measurements were carried out using ATLAS 0961 MBI test system. Charge-discharge tests were conducted for all CCL composites listed in table 2, composites with primary carbon (C/Li2MnSiO4@600, C/Li2MnSiO4@700 and C/Li2MnSiO4@800) and standard composites (Li2MnSiO4@600_CB and Li2MnSiO4@700_CB). Standard composites were prepared by mixing Li2MnSiO4 powder with commercial carbon additive – carbon black (CB). 15 wt.% of carbon was used. Results collected from charge/discharge tests are presented in fig. 11-16.
C/Li2MnSiO4@600 sample did not show any reversible capacity (Fig. 11a). Lack of electrochemical activity of C/Li2MnSiO4@600 is connected with too high carbon loading. High carbon content (34%) is responsible for limiting of ionic conductivity in the composite and for surface polarization. Galvanostatic cycling studies of C/Li2MnSiO4@700 and C/Li2MnSiO4@800 revealed reversible capacity in a range of 30 mAh g-1. Fig. 11b show charge and discharge capacity for 7 cycles of C/Li2MnSiO4@700 composite. After first four cycles reversible capacity stabilizes at about 30 mAh g-1. Very poor performance of this sample is also connected with high amount of carbon in prepared composite (32%).
Standard composites obtained by mixing active silicate materials with carbon black (15%) show extremely low reversible capacity as well (Fig. 12a and 12b). In this case, carbon additive does not provide sufficient electrical contact between active material grains. Due to this fact, even under low C/200 rate samples performed very poorly in charge/discharge tests.
Fig. 13-16 show cycling behavior of CCL composites with different amounts of carbon. Delithation of CCL composites after initial charging is in the range of 48-68% (charged CCL/Li2MnSiO4@600_12%C – correspond to composition Li1.04MnSiO4 and charged CCL/Li2MnSiO4@700_10%C – correspond to composition Li0.64MnSiO4). The observed capacity loss in the first cycles is related to SEI formation in CCL. During lithium extraction, Mn3+ and Mn4+ appear, and at the same time Li2MnSiO4 undergoes decomposition caused by Jahn-Teller distortion associated with changes in lattice parameters during Mn3+ → Mn4+ transition . Despite the fact that crystalline structure collapses, discharge capacity is around 90-110 mAh/g. CCL composites produced from Li2MnSiO4 calcined at 600 °C exhibit better coulombic efficiency than Li2MnSiO4 calcined at 700 °C, the reversible capacity after 10 cycle is close to 100 mAh/g. The results indicate that coulombic efficiency of cathode material depends on grain size and homogeneity in grain size distribution, while cell capacity is limited by carbon coating performance.
Dilithium manganese orthosilicate – a high energy density cathode material – was successfully synthesized by sol-gel Pechini method. Encapsulation of nanosized grains of Li2MnSiO4 in carbon matrix, resulted from organic precursor, avoided further sintering. Different crystallite size was obtained, in particular, nanoparticles within range of 5-10 nm. High chemical stability of this material under a highly reductive environment was observed. Application of carbon coating improved electrical conductivity of cathode material to the satisfactory level of ~10-0.5 S/cm. Thickness of carbon coating on active material grains strongly limits the electrochemical performance of composite. Formation of CCL/Li2MnSiO4 composites significantly improved electrochemical performance of cathode materials, showing a reversible capacity of 90-110 mAh/g after 10 cycles. Electrochemical tests indicated that composite preparation should be optimized in terms of carbon content, CCL performance and homogeneity of a crystallite size. It was found that coulombic efficiency of cathode material depends on grain size and homogeneity in grain size distribution, while cell capacity is limited by carbon coating performance.
This work has been financially supported by the Polish National Science Centre under research grant no. N N209 088638, and by the European Institute of Innovation and Technology under the KIC InnoEnergy NewMat project. The part of the measurements was carried out with the equipment purchased thanks to the financial support of the European Regional Development Fund in the framework of the Polish Innovation Economy Operational Program (POIG.02.01.00-12-023/08). One of the authors (M.Ś.) acknowledge a financial support from the International PhD-studies programme at the Faculty of Chemistry Jagiellonian University within the Foundation for Polish Science MPD. TEM analysis were carried out in Laboratory of Transmission Analytical Electron Microscopy at the Institute of Metallurgy and Material Science, Polish Academy of Sciences.