In-situ XRD study of a Chromium doped LiNi0.5Mn1.5O4 cathode for Li-ion battery

In-situ XRD study of a Chromium doped LiNi0.5Mn1.5O4 cathode for Li-ion battery CHLADIL, L.; KUNICKÝ, D.; KAZDA, T.; VANÝSEK, P.; ČECH, O.; BAČA, P. Journal of Energy Storage Volume 41, September 2021, 102907, Pages 1-15 ISSN: 2352-152X DOI: https://doi.org/10.1016/j.est.2021.102907 Accepted manuscript © 2021. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ dspace.vutbr.cz In-situ XRD Study of a Cr Doped LiNi0.5Mn1.5O4 Cathode for Li-ion Battery 1 2 L. Chladila,b, D. Kunickýa, T. Kazdaa,b, P. Vanýseka,b, O. Čecha,b and P. Bača a,b 3 4 a Department of Electrical and Electronic Technology, Brno University of Technology, 5 Technicka 10, 616 00 Brno, Czech Republic 6 b Centre for Research and Utilization of Renewable Energy, Faculty of Electrical 7 Engineering and Communication, BUT, Technicka 10, 616 00 Brno, Czech Republic 8 9 Abstract 10 This paper deals with structural (in-situ XRD) and electrochemical characterization 11 of high-voltage lithium-ion cathode materials LiMn2O4 (LMO), LiNi0.5Mn1.5O4 (LNMO), 12 and LiCr0.1Ni0.4Mn1.5O4 (LCNMO) prepared by solid-state synthesis. Structural in-situ X13 ray diffraction spectra were measured by an affordable Rigaku diffractometer. Our 14 synthesis route produced the samples with similar morphologies where the average particle 15 sizes were 1.11 µm and 1.46 µm for LNMO and LCNMO respectively. Results of the 16 Rietveld analysis brought detailed insight into two-phase structure transitions for LMO and 17 three-phase transitions for LNMO and LCNMO. XRD study revealed differences in the 18 structural behavior of LMO and LNMO prepared by solid-state synthesis compared to the 19 results of other authors using the sol-gel synthesis route. In the case of chromium-doped 20 LNMO, our results indicate ability of the chromium metal to effectively reduce Mn3+ 21 content while the ordering of the structure increases. Chromium doping also promotes a 22 larger lattice parameter in a fully delithiated state than in the case of undoped LNMO. 23 Therefore, lowering of the volume changes was observed and faster phase II/III transition 24 taking place, when Ni3+/Ni4+ redox pair was oxidized, was also identified. Cr doping of 25 LNMO also promotes the reaching the lattice parameters of phases in both phase transitions 26 and thus could reduce the internal stress of active material under high C-rate cycling. 27 Results thus suggested that chromium doping can improve the stability of the inner 28 structure and performance at higher charging C-rates even though the structure goes 29 through a three-phase region during charging as undoped LNMO. The evaluation of 30 diffusion coefficients of Cr-doped LNMO revealed increased diffusivity in a full discharge 31 state and as the cathode underwent the cycling the differences in diffusivity seemed to be 32 more pronounced. 33 34 Introduction 35 Lithium-ion batteries are progressively being deployed in electric vehicles (EVs) 36 and in a grid energy storage where a long lifetime is a significant requirement. Majority of 37 car drivers travel daily only between 0-80 km while the range of the EVs is about 400 km. 38 This would seem already sufficient range. However, there is an increased application of 39 autonomous vehicles and electric-powered trucks where the 400 km range might be close 40 to the daily demand of 100 % DoD [1]. From a long-term stability point of view, deep 41 cycling significantly accelerates irreversible electrode changes especially in high-voltage 42 Li-ion systems, and thus promotes fast capacity decay. Many researchers are therefore 43 targeting ways to stabilize the structures by doping by various elements to increase the 44 battery life-span while utilizing fully its capacity. 45 Very well-known electrode material LiFePO4 was widely used in EVs, hybrid EVs, 46 etc. This phospho-olivine material has been first introduced by Padhi et al. in 1997 [2]. It 47 is considered to be better than LiCoO2 in some aspects such as safety at high temperatures, 48 overcharges, and in its non-toxicity [3]. 49 Other materials that are promising better performance are high voltage materials 50 based on LiMn2O4, mainly doped with 0.5 mole percent of nickel – LiNi0.5Mn1.5O4 51 (LNMO) [4] [5] [6]. This material has higher redox potential (around 4.7 V) because of the 52 transition of Ni2+ to Ni4+ [6] and higher gravimetric energy density (around 700 Wh/kg). It 53 is also more environmentally friendly than LiCoO2 as it does not contain potentially toxic 54 and rare cobalt. LNMO generally could be synthesized as an ordered structure and a 55 disordered one which exhibit P4332 space group and Fd-3m respectively. In the first case, 56 Ni atoms occupy 4a and 12d sites, whereas in the case of Fd-3m nickel is randomly 57 distributed in the octahedral 16d sites [7]. Each phase exhibits different parameters such as 58 conductivity, Li+ diffusivity, capacity retention at high C-rate and also different long-term 59 performance [8] [7]. Effect of different Ni content in LiNixMn2-xO4 was published in a 60 complex study developed by Duncan et al. [9]. They found that the transition between the 61 ordered and disordered structure proceeds in a narrow Mn3+ content window that was 62 almost linearly dependent on the amount of Ni substitution for Mn. Because the presence 63 of Mn3+ is associated with dissolution of active mass and therefore with a gradual fade of 64 the electrode capacity [8], therefore there are general attempts to reduce the amount of 65 Mn3+. However, its presence is also accompanied by beneficial oxygen deficiency in the 66 structure that improves the electronic conductivity in the bulk. The other important aspect 67 of high-voltage LNMO spinels is so-called disordering which is generated by randomized 68 occupancy of nickel on the manganese 16d site positions. The main increase in disordering 69 was found for 0.35>x>0.4 in LiNixMn2-xO4 and typically developed LiNi0.5Mn1.5O4 with 70 synthesis temperature around 700 °C and slow cooling exhibited ordered structure with a 71 low amount of Mn3+ [9]. Such ordered structure shows good cyclability but worse capacity 72 retention in comparison to disordered one prepared by fast cooling in the final stage of 73 synthesis. Disordering of the structure by post-synthesis temperature treatment is one way 74 how to influence the disordering without significant increase of Mn3+ which brings the 75 long-term material instabilites. 76 LNMO generally suffers from the dissolution of manganese into the electrolyte 77 during cycling at higher temperatures. This leads to a capacity decrease and worse 78 structural stability [10]. This could be caused by the Jahn-Teller distortion which occurs in 79 octahedral and tetrahedral compounds. The observation was published in 1937 by H. A. 80 Jahn and E. Teller [11] and states that in an electronically degenerate state, a nonlinear 81 molecule undergoes distortion to remove the degeneracy by lowering the symmetry and 82 thus by lowering the energy. Although known for eight decades, the details of the Jahn83 Teller distortion remained elusive. Another problem of Mn-based oxides is the oxygen 84 release during higher temperatures. This phenomenon can be expressed for example as 85 LiMn2O4 → LiMnO2 + 1/3 Mn3O4 + 1/3 O2. There is a phase transition during charging 86 and discharging which leads to significant structural changes, thus leading to worse 87 structural stability during cycling [12]. 88 There were used many different doping elements to try to improve LMO or LNMO 89 performance. LMOs doping by phosphorus drastically improved cyclability and rate 90 capability even at elevated temperatures [13], doping by zinc [14] also improved 91 cyclability, and the increasing concentration of zinc provided long-term cycle stability and 92 superior reproducibility. Materials moderately doped by molybdenum [10] had increased 93 structural stability even at higher temperatures. The double doping by indium and sulfur 94 [15] optimized the cycle stability, discharge capacity, and high rate discharge ability. 95 Cycling performance and specific capacity improvement were done by doping with 96 chromium [16]. Promising doping materials are from the lanthanides series. For example, 97 gadolinium [17] has significantly improved structural stability, stability at high discharge 98 rates, and reversible capacities, which was attributed to suppression of the Jahn-Teller 99 distortion. Exceptional structural stabilization was achieved by scandium doping [18]. 100 There are also studies of new materials ranging from vanadium oxides [19] [20] to organic 101 compounds [21] [22]. 102 As charging and discharging continue in lithium-ion batteries, the crystallographic 103 structure of the electrodes is changing. This has a significant effect on the overall 104 performance of the batteries. The charging of cathodes is tied to the decrease of lithium 105 ions as the lithium ions leave the structure. From a structural viewpoint, a small decrease 106 of lithium concentration leads to a decrease of ion occupancy that could change the lattice 107 constant only by a small amount. A gradual loss of lithium planes during delithiation leads 108 to a complete transformation of the cathode materials structure [8] [9] [23]. Phase transition 109 is different for ordered and disordered LNMO and also varies with different Mn3+ content. 110 Whereas the well-ordered structure exhibits three-phase transition during 111 lithiation/delithiation [9], disordered structure with a high amount of Mn3+ is manifested 112 by one two-phase transition. Our analysis of available literature shows that two-phase 113 transition was generally measured especially for structures with Mn3+ content around 20 % 114 or more [23] [24]. 115 Delithiation is in any case accompanied by significant changes in the lattice 116 dimensions regardless of the presence of the transitional phase and this results in volume 117 changes of materials during charging and discharging. These changes can be seen from X118 ray diffractograms and can be calculated by the Rietveld analysis. It is also known that 119 volume changes directly influence the long-term stability of the materials, and therefore a 120 precise measurement of these structural parameters in a different state of charge (SoC) is 121 essential for assessing the expectations for long-term usability [25]. 122 Yoon et al. found out that the structural stability of the cathode material is affected 123 by the post-calcination cooling rate [26]. Gao et al. improved stability by annealing the 124 base material, prepared by the solid-state method, for two hours after fine grinding [24]. 125 Also, the structure itself on a micro-scale affects stability [27] [28]. 126 A recent study by Kondracki et al. [23] used copper to verify the stabilization of 127 the structure. This was successfully verified as the volume changes during charging were 128 only 3.3 %. It is attributed to the reduction of the free volume within the unit cell and 129 improved kinetics of lithium diffusion. 130 Our structural and electrochemical study is focused on the usage of chromium in 131 LiNi0.5Mn1.5O4 and how it can help stabilize the structural properties and how it could 132 influence the diffusion properties of the cathode. The main benefit of this work lies mainly 133 in the comparison of pure LNMO and Cr-doped LNMO prepared by the same synthesis 134 route which allows clear identification of the effect of chromium on all properties of 135 LNMO masses such as Mn3+, the structure arrangement, the position of phase transitions 136 and transition speed, diffusivity of Li+, etc. 137 The results in this paper were obtained using the basic diffractometer Rigaku 138 Miniflex 600HR, which is widely used in industrial quality inspection, equipped with an 139 in-situ electrochemical cell attached to the diffractometer positioning system. Using X-ray 140 diffractometry in connection with electrochemical impedance spectroscopy (EIS) can bring 141 deeper insight into the electrochemical processes and structural changes taking place inside 142 the materials and thus can provide accurate feedback on new ways of material optimization. 143 144 Experimental 145 LiMn2O4 (LMO), LiNi0.5Mn1.5O4 (LNMO), and LiCr0.1Ni0.4Mn1.5O4 (LCNMO) 146 were prepared by a solid-state synthesis. LMO was produced by NEI Corporation 147 distributed by Sigma-Aldrich and LNMO and LCNMO were prepared by own solid-state 148 synthesis from carbonates (Li2CO3 and MnCO3) and oxides (NiO and Cr2O3) precursors 149 obtained also from Sigma-Aldrich. Preparation is described in detail in [29] and [30]. 150 Materials were prepared by a two-step solid-state synthesis process contain annealing at 151 600 °C for 10 hours and annealing at 900 °C for 15 hours. Cooling was realized by opening 152 the furnace and therefore the cooling had a higher speed in the first phases of cooling and 153 within 1.5 hours the samples reached approx. 350 °C. 154 Electrodes were prepared from the slurry containing active material which was 155 mixed in the mass ratio 80:10:10 with polyvinylidene difluoride (PVDF) binder and Super156 P carbon. N-methyl-2-pyrrolidone (NMP) has been added to create a dense pasty mass. 157 Electrodes for galvanostatic and EIS testing were prepared from electrode slurry by 158 doctor-blade casting on aluminum foil. Then the electrodes were dried and pressed by the 159 pressure of 31.4 kN/cm2. A disc electrode with a diameter of 18 mm (2.54 cm2) was cut 160 out of the coated aluminum foil and investigated in El-Cell© ECC-Std electrochemical test 161 cell. The cell assembly was carried out in an M-Braun argon-filled glove box. Metallic 162 lithium was used as the counter electrode and a 1.5 mol·dm-3 LiPF6 EC:DMC 1:2 w/w 163 electrolyte was soaked in a glass fiber separator. Galvanostatic charge and discharge 164 cycling were utilized as a method for the investigation of capacity retention with the 165 potential window from 3.0 V to 5.1 V versus Li/Li+. Polarization and impedance studies 166 were performed on Biologic VSP potentiostat. 167 For the in-situ XRD study, the slurry was stirred with a magnetic stirrer for 24 hours 168 and then coated onto an expanded thin metal aluminum foil stamped in the shape of a coin 169 and dried at 50 °C for 1 hour. After drying the second layer of the active substance was 170 applied, it was followed by drying at 50 °C for 1 hour and then followed by drying at 171 105 °C for at least 24 hours. After thorough drying, the sample was pressed by pressure 172 24.5 kN/cm2. The assembly of the cell was done in an argon-filled glove box Jacomex. The 173 electrode was then put into an ECC-Opto-Std cell by El-Cell® that has been adapted for 174 direct insertion into the chamber of the powder diffractometer Rigaku Miniflex 600 HR. A 175 separator made from glass fibers was used and the cell was filled by electrolyte EC:DMC 176 (1:1) with 1 mol·dm-3 LiPF6. A higher EC:DMC ratio was preferred for short-term in-situ 177 measurement due to the better conductivity of the mixture [31]. Metal lithium was used as 178 both the reference and counter electrodes (0.75 mm thick foil). 179 Charging proceeds inside of the diffractometer chamber, as shown in Figure 1, in 180 the range of 10-110 degrees. As an X-ray transparent window was used a Kapton foil. The 181 charging rate was set to be 0.05 C with the overall duration around 20 hours. X-ray 182 diffractograms were obtained operando. 183 184 185 Figure 1 Modified Opto-Std cell by El-Cell mounted in the chamber of the powder diffractometer 186 187 Structure and morphology of LNMO and LCNMO 188 Figure 2 shows the SEM images of LNMO and LCNMO particles. Both powders 189 contain clusters of octahedral particles whose diameters vary approximately from 0.5 µm 190 to 3 µm. For a better characterization of the powders, SEM images using the ImageJ 191 (National Institutes of Health) program were analyzed and diameters of the selected group 192 of particles were measured for both samples. Measured particle sizes were divided into 193 bins and particle profile distribution in Figure 3 was compiled. 194 195 196 Figure 2: SEM image of a) LiNi0.5Mn1.5O4 particles and b) LiCr0.1Ni0.4Mn1.5O4 particles. 197 The particle size profile revealed that LNMO has basically smaller particles than 198 the Cr doped one and most of the particles are in the range from 0.7 µm to 1.1 µm, and 199 1.1 µm to 1.7 µm respectively. The average particle diameter is 1.11 µm for LNMO and 200 1.46 µm for LCNMO. The surface area of the powders can be calculated using Eq. 1: 201 𝐴 = ∑𝑆𝑝 ∑𝑚𝑝 (1) 202 a) b) where A is the surface of the powder, Sp is the surface of each measured particle selected 203 for particle profile analysis which is calculated from their diameters (assuming a spherical 204 particle shape) and mp is the mass of each analyzed particle. Calculated values of powder 205 surface and parameters of the electrode for EIS investigation (weight of active material mel 206 and calculated area of active material/electrolyte interface Ael for each electrode) are 207 summarized in Table 1. Although the surface area was not determined by the much more 208 precise BET technique, our calculated values are in good agreement with the BET surface 209 parameter of samples of other authors [32] who synthesized the powders with similar 210 particle diameter and published the BET surfaces slightly above 1.0 m2g-1. 211 212 Figure 3 Distribution of particle sizes of LiNi0.5Mn1.5O4 (LNMO) and LiCr0.1Ni0.4Mn1.5O4 (LCNMO) 213 Table 1 Calculated parameters of LNMO and LCNMO powders 214 215 Material davg [µm] A [m2g-1] mel [g] Ael [cm2] LNMO 1.11 0.865 4.61·10-3 39.9 LCNMO 1.46 0.765 5.76·10-3 44.1 216 Crystallographic study 217 The structural characterization was done first on pristine materials. All samples 218 LMO, LNMO, and LCNMO indicated a symmetric Fd-3m space group which is, unlike 219 primitive cubic structure (P4332), typical for the synthesis where the temperature exceeds 220 700 °C. The results of the lattice cell parameter a and the lattice volumes from the initial 221 Rietveld analysis for each analyzed structure are shown in Table 2 and the values agree 222 with literature [23] [33]. It can be seen that addition of chromium ions does not change the 223 structure parameters significantly. According to [34] chromium doping of LMO leads to a 224 gradual reduction of lattice parameter a as the concentration of chromium increases 225 because of the replacement of the larger Mn3+ by smaller Cr3+ ions. This resulted in a 226 0.38 % change of the lattice constant when the Cr3+ replaced 25 % of Mn3+ positions. When 227 we replace the same amount of Mn3+ ions with Ni3+, the lattice parameter is reduced by 228 0.61 %. Replacement of Ni3+ by Cr3+ at a higher level could lead to an increase of the 229 parameters a to some extent. In this state, the part of larger Ni3+ ions are replaced by smaller 230 Cr3+ but without noticeable changes in the lattice constant which in turn could lead to better 231 permeability for Li+ ion during lithiation/delithiation. 232 XRD patterns of pristine materials are shown in Figure 4. There were detected 233 Mn2O3 impurities in the LMO pattern. No impurities were detected in other patterns. It can 234 0 2 4 6 8 10 12 14 16 18 20 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 Occurrence Particle size [µm] LiNi₀․₅Mn₁․₅O₄ LiCr₀․₁Ni₀․₄Mn₁․₅O₄ be seen that all patterns display diffraction peaks with their attributed lines at 18° (111), 235 36° (311), 38° (222), 44° (400), 48° (331), and 64° (440). The intensity ratios of (311) and 236 (400) peaks vary from sample to sample which indicates the different occupancy of cations 237 between tetrahedral and octahedral 16d sites of the spinel structure. The ratio is lower for 238 LiCr0.1Ni0.4Mn1.5O4 indicating that the chromium ions occupy mainly the octahedral 239 positions. This is consistent with the study [35] and the same tendency of occupation was 240 reported also in the case of Cu-doping of LNMO [23]. Other authors [33] assume that in 241 fact, two atoms of Cr3+ replace a pair Ni2+ + Mn4+ and thus help to maintain the Mn4+ 242 oxidation state unchanged which in turn leads to a longer life-span of the structure. 243 Moreover, the substitution of Cr can gain the structure disorder in such a way, that Ni2+ 244 and Mn4+ are ordered on octahedral sites at a shorter range only [36] and this is expected 245 to increase the diffusion properties of the bulk. 246 247 Table 2. Structural characterization of pristine materials LMO, LNMO, and LCNMO. 248 Material Space group Lattice constant [Å] Lattice volume [Å3] LiMn2O4 Fd-3m a= 8.212 557.03 LiNi0.5Mn1.5O4 Fd-3m a= 8.171 545.73 LiCr0.1Ni0.4Mn1.5O4 Fd-3m a= 8.169 545.79 249 250 Figure 4 XRD pattern of pristine a) LMO with marked Mn2O3 impurity phase b) LNMO and c) Cr-doped LNMO 251 (LCNMO) 252 Figure 5, Figure 7, and Figure 10 display the part of the XRD patterns for LMO, LNMO, 253 and LCNMO, where the changes in structure can be easily read. Measured diffraction 254 spectra were analyzed by the Rietveld refinement and the results of the lattice constant in 255 dependence on the state of lithiation xLi are shown in Figure 6, Error! Reference source 256 not found., and Figure 12 for each analyzed sample. The summary of the lattice parameters 257 in lithiated (0 % SoC) and delithiated (100 % SoC) states for all measured materials is 258 shown in Table 3. At the start of the measurement, the cathode materials are from the 259 synthesis fully lithiated and thus the bottom lines in Figure 5 represent the state, where 260 xLi = 1. At the fully charged state – top line – the lattice parameters reach a value a = 8.05 261 Å which can be attributed, according to [37], to a 10 % of residual lithium concentration. 262 Therefore, we marked this state as xLi = 0.1 at 100 % SoC of the electrode. Because each 263 material was charged a slightly different length of time, the number of XRD scans was also 264 different and varied from 20 scans for LMO to 22 scans for LNMO. Therefore, between 265 0 1000 2000 3000 4000 5000 6000 7000 10 20 30 40 50 60 70 80 90 100 Intensity (cps) 2-theta (°) (111) (311) (222) (400) (331) (333) (440) (531) 0 1000 2000 3000 4000 5000 6000 7000 10 20 30 40 50 60 70 80 90 100 Intesity (cps) (111) (311) (222) (400) (331) (511) (440) (531) 0 500 1000 1500 2000 2500 10 20 30 40 50 60 70 80 90 100 Intensity (cps) (111) (311) (222) (400) (331) (511) (440) (531) * * * *Mn2O3 a) b) c) 410 Figure 11 Selected spectra of XRD patterns of Cr-doped LNMO obtained during the delithiation process that confirms 411 the three-phase transition. Phase I: LiCr0.1Ni0.4Mn1.5O4, phase II: Li0.5Cr0.1Ni0.4Mn1.5O4, and phase III: Cr0.1Ni0.5Mn1.5O4. 412 413 414 Figure 12 Results from the Rietveld analysis of XRD spectra for LCNMO (Cr-doped LNMO) in a different 415 state of lithiation xLi: a) Lattice parameter changes of LiCr0.1Ni0.5Mn1.5O4 (phase I) corresponding to 416 discharged state, Li0.5Cr0.1Ni0.5Mn1.5O4 (phase II) and Cr0.1Ni0.5Mn1.5O4 (phase III) for partially and fully 417 charged state b) Content ratio changes of the identified phases. 418 58 58.5 59 59.5 60 60.5 61 Intensity [a.u.] 2-theta [°] Phase I Phase II Phase III 0% 20% 40% 60% 80% 100% 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mass fraction [wt. %] Lithium concentration xLi phase I phase II phase III 8.00 8.05 8.10 8.15 8.20 8.25 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Lattice parameters [Å] phase II/III phase I/II Table 3 Lattice parameters in lithiated and delithiated state of disordered LMO, LNMO, and LCNMO 419 prepared by solid-state reaction. 420 Material Lithiated Delithiated Lattice par. [Å] Lattice vol. [Å3] Lattice par. [Å] Lattice vol. [Å3] LiMn2O4 a= 8.2117 553.73 a= 8.0493 521.52 LiNi0.5Mn1.5O4 a= 8.1708 545.50 a= 8.0176 515.39 LiCr0.1Ni0.4Mn1.5O4 a= 8.1705 545.44 a= 8.0304 517.86 421 Figure 13 represents the summary of volume changes of characterized materials in 422 comparison with similar materials LMO, LNMO, and Cu-doped LMNO prepared by the 423 sol-gel technique [23] and also in comparison with our already published study focused on 424 the well commercialized LFP cathode material [25]. In the case of LMO volume changes, 425 the comparison indicates that these changes are more pronounced for materials prepared 426 by the sol-gel synthesis method than those prepared by the solid-state synthesis. This has 427 an origin in the lower lattice parameters of a fully lithiated structure for cathode material 428 prepared by solid-phase synthesis. On the other hand, our LNMO exhibits lowering of 429 volume changes which took place on the same scale as in the case of LMNO prepared by 430 sol-gel synthesis [23]. A substitution of nickel by chromium brings further stabilization of 431 the structure from the volume changes point of view and this improvement could therefore 432 help to the long-term stability of the electrode system. 433 434 Figure 13 Summary of volume changes of different disordered LMO-based cathode and commercial LFP [2], [25], [23]. 435 Published LMO, LNMO, and Cu doped LNMO were prepared by sol-gel synthesis and our measured samples were 436 prepared by solid-state reaction. 437 Lithium Diffusivity Study Using Electrochemical Impedance Spectroscopy 438 To evaluate the effect of chromium doping on the diffusivity of lithium in the bulk of 439 the active mass, electrochemical impedance spectroscopy (EIS) study of both materials was 440 performed and the Nyquist diagrams, shown in Figure 14 were obtained. The EIS spectra 441 show the impedance behavior of the electrodes in discharged (fully lithiated) state at 442 potential E=2.86 V. The electrodes were subjected to two voltammetric cycles before EIS, 443 therefore we assume that a significant part of the SEI layer was already formed. Spectra 444 5.82% 5.52% 5.06% 6.92% 7.13% 5.75% 3.26% 6.92% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0% LMO LNMO LNMO+Cu LNMO+Cr LFP Volume changes measured published comprised of a semicircle measured in the frequency range 100 kHz to 76 Hz, where the 445 minimum of -Z” was measured, and of an inclined line at low-frequency range down to 446 100 mHz. The semicircle represents the charge transfer process and the inclined line 447 represents the diffusion of lithium ions into the bulk of the electrode material. Diffusion is 448 in the impedance response described by the Warburg coefficient σw and in an equivalent 449 circuit, this process represents the Warburg impedance component W2. Coefficient σw can 450 be obtained from the fit of the spectra by an equivalent circuit (typical equivalent circuit 451 for fully discharge state is in Figure 15) or from the Eq. 2 452 𝑍´ = 𝑅𝑒𝑙 + 𝑅𝑐𝑡 + 𝜎𝑤𝜔−1 2 (2) 453 454 where Rel and Rct are the resistance of the electrolyte and the charge transfer resistance 455 respectively and ω is the angular frequency in the low-frequency region. Coefficient 𝜎𝑤 456 can be derived from the slope of the real part of impedance Z´ vs. the reciprocal square root 457 of the angular frequencies (ω−1/2). The interpolated dependencies, including the equation 458 of the linear interpolation, are shown in Figure 16. Per the findings of other authors who 459 studied the diffusion properties of cathode materials such as LiFePO4 [39], LiNiMnO [40], 460 LiNiCoMnO [41] or Al, Mg, and Co-doped LiMnO [42], we use a similar approach to 461 determine the diffusion coefficient DLi+ that can be calculated from Eq. 3 462 𝐷𝐿𝑖+=2𝑅2𝑇2 𝑐𝐿𝑖 2𝑛4𝐹4𝐴2𝜎𝑤 2 (3) 463 464 where DLi+ represents the lithium diffusion coefficient, R is the universal gas constant, T is 465 the absolute temperature in kelvin, cLi is the lithium concentration in the cathode material 466 (calculated by as 23845 mol·m-3), n is the number of electrons transferred in charge transfer 467 reaction, F is the Faraday’s constant, A is electrode surface area and σw is above mentioned 468 the Warburg coefficient. 469 The exchange current density can be calculated then from Eq. 4: 470 𝑗0=𝑅𝑇 𝑛𝐹𝑅𝑐𝑡𝐴 (4) 471 Where Rct is the charge-transfer resistance and in this study were obtained by fitting the 472 spectra using EC-lab analysis tools. An equivalent circuit in Figure 15 was used for the 473 simulation of the spectra. 474 475 476 0 100 200 300 400 500 0 100 200 300 400 500 600 -Z" [Ω] Z' [Ω] LiNiMnO LiNiCrMnO 0 10 20 30 40 010 20 30 40 50 -Z" [Ω] Z' [Ω] 477 Figure 14 EIS Nyquist plots of the electrode response in the frequency range 0.1 Hz – 100 kHz for LNMO and LCNMO 478 479 Figure 15 Equivalent circuit for simulation of impedance behavior of the electrodes – response in Figure 14 480 481 482 Figure 16 The relationship between Z´ and ω−1/2 at low frequencies for LNMO and LCNMO 483 484 Table 4 contains the values of Rct, double-layer capacitance Cdl, Warburg 485 coefficient σw, and also the values of calculated diffusion coefficient DLi+ and exchange 486 current density j0. Diffusion coefficients were calculated both from the values of the 487 electrode area denoted as DLi+ el and from the value of the active surface area of the active 488 substance denoted as DLi+ as according to Table 1. The diffusion coefficient in our 489 calculation reaches value 1.45×10-14 cm2s-1 for LNMO in a full discharge state which is in 490 good agreement with the result for LMO in the study [42] and is a much higher value than 491 in the case of Co-doped LNMO in the study [40] whose authors also use the electrode area. 492 Typical results of the GITT technique bring the values in the range of order 10-10 – 10-12 493 cm2s-1 for all states of charge of LMNO [40]. It has to be noted that the results of different 494 authors vary depending on what surface area was used for the calculation as well as on 495 what description of concentration of lithium inside of the electrode was selected 496 (concentration of Li in bulk or electrolyte soaked into electrode mass). Although there are 497 different approaches in different studies, it has to be concluded that this method can be still 498 used for the comparison of materials within the scope of each individual publication. From 499 our calculation of the diffusion coefficient and the exchange current density, we can 500 conclude that Cr-doping of LMNO increases diffusion values inside the bulk of material 501 as well as the activity of such improved material expressed by the higher value of the 502 exchange current density. This fully corresponds with the improved performance of the Cr503 doped LNMO during a high rate of 5 C cycling [30]. 504 505 Table 4 Nyquist analysis of LNMO and Cr-doped LNMO cathode in a discharged state in the second cycle 506 y = 26.43x + 18.23 y = 20.86x + 13.62 0 10 20 30 40 0 0.2 0.4 0.6 -Z´[Ω] ω-1/2 [Ω-1/2] LiNi₀․₅Mn₁․₅O₄ LiCr₀․₁Ni₀․₄Mn₁․₅O₄ Re [Ω] Cdl [F] Rct [Ω] σw [Ω Hz-1/2] CL [F] DLi+ el [cm2s-1] DLi+ as [cm2s-1] j0 [A cm-2] LiNi0.5Mn1.5O4 4.795 1.29·10-5 13.11 28.81 3.60·10-3 1.45·10-14 5.69·10-17 7.89·10-4 LiCr0.1Ni0.4 Mn1.5O4 4.177 9.72·10-6 10.05 18.95 4.44·10-3 2.32·10-14 7.47·10-17 1.03·10-3 507 The graph of discharge capacities in Figure 18 b) shows the performance of the electrodes 508 in the first 65 cycles and under different C rates. Cr-doped LNMO exhibits a much higher 509 discharging capacity especially at the highest 5 C rate despite larger particle diameters and 510 smaller active surface area. This can be explained mainly by different diffusivity inside the 511 bulk of the active mass. During the 5 C rate, the LMNO exhibits a gradual increase of 512 discharge capacities that could be attributed to some formation effect of the high rate 513 cycling. To assess the changes in electrode diffusion properties, diffusion coefficients DLi+ 514 were evaluated also in a different state of galvanostatic cycling – after 30, 55, and 65 cycles 515 when the electrodes were in a full discharged state. The slope of real impedance on the 516 reciprocal square root of the angular frequencies (ω−1/2) for both materials is displayed in 517 Figure 17. Corresponding diffusion coefficients DLi+ calculated on the geometric electrode 518 area are plotted above the galvanostatic dependencies in Figure 17 b). A significant 519 decrease of DLi+ after 30 cycles from the initial one can be easily explained by the different 520 states of electrodes. Initial diffusion coefficients were measured in full discharge state 521 when x = 1 for LixNi0.5Mn1.5O4 or LixNi0.4Cr0.1Mn1.5O4 whereas after 30 cycles of 0.5 C 522 rate some residual charge remain in the material and therefore x<1. According to [23], 523 a decrease of almost two orders of DLi+ can be expected when the state of charge increases 524 by about 20 % at the start of charging. LNMO exhibits a much higher drop of DLi+ which 525 could be due to a lower degree of active mass lithiation in the discharged state. Another 526 evaluation of the diffusion coefficient was after high rate cycling - 55 cycles and after 527 65 cycles. A partial increase of the diffusion coefficient for both electrodes was detected 528 already at the end of the high-rate test where the discharge rate was 1 C and mainly after 529 the end of cycling, where the same state of lithiation is expected as before the high-rate 530 cycling. It can be concluded that the diffusion coefficient was higher for Cr-doped LNMO 531 in each state of galvanostatic cycling and also that high-rate cycling helps to increase the 532 diffusion coefficient inside the electrode mass. 533 534 535 Figure 17 The relationship between Z´ and ω−1/2 at low frequencies for LNMO (left) and Cr doped LNMO (right) 536 537 y = 164.9x + 22.7 y = 102x + 25.4 y = 69.3x + 35.2 0 50 100 150 200 250 0 0.5 1 1.5 -Z´[Ω] ω-1/2 [Ω-1/2] y = 62.87x + 20.23 y = 54.23x + 26.16 y = 21.24x + 31.33 0 20 40 60 80 100 120 0 0.5 1 1.5 ω-1/2 [Ω-1/2] 3 cycle 55 cycle 65 cycle 538 Figure 18 Values of a) calculated diffusion coefficients for LNMO and LCNMO in the corresponding state of 539 galvanostatic cycling b) discharge capacities during cycling with high-rate galvanostatic test 540 541 542 Conclusions 543 LMO (LiMn2O4), LNMO (LiNi0.5Mn1.5O4), and LCNMO (LiCr0.1Ni0.4Mn1.5O4) 544 prepared by solid-state synthesis were characterized by in-situ X-ray spectroscopy, and the 545 effect of chromium doping was evaluated by electrochemical impedance spectroscopy. The 546 in-situ XRD technique gives a detailed insight into the lattice parameter changes as well as 547 the results elucidate the phase changes during the delithiation of the electrode. All 548 synthesized structures exhibit disordered structure with Fd-3m space group symmetry. 549 Among the main detected changes for LMO, need to be mentioned the smaller 550 lattice parameters in a fully lithiated state in the case of solid-state synthesis compared to 551 sol-gel, resulting in lesser volume changes [23]. LNMO material prepared by solid-state 552 synthesis at 700 °C exhibits approx. 14 % Mn3+ content and the potential difference of 553 nickel redox pairs reached 58 mV confirming the strongly disordered nature of the material. 554 Although LNMO contains such high Mn3+ content and highly disordered structure, the in555 situ X-ray Rietveld analysis has been able to recognize three-phase transition during 556 delithiation. Comparison of the cathode behavior and phase evolution of our disordered 557 LNMO material prepared by solid-state synthesis with materials prepared by the sol-gel 558 technique with higher Mn3+ content revealed differences mainly in presence of second 559 transition phases Li0.5Ni0.5Mn1.5O4. 560 Chromium doping of LNMO influences both Mn3+ content and the disordering of 561 the structure. Whereas the Mn3+ content was reduced by 37 %, nickel redox plateaus 562 60 70 80 90 100 110 120 130 140 010 20 30 40 50 60 70 Capacity [mAh g-1] Cycle LiNi₀ ․ ₅Mn₁ ․ ₅O₄ 0.5 C 1 C 2 C5 C 2 C 1 C0.5 C 1E-16 1E-15 1E-14 1E-13 010 20 30 40 50 60 70 DLi+ [cm2s-1] a) b) showed greater potential separation and thus point to a further increase in the disordering 563 of the structure. LCNMO also stabilized the structure changes in several aspects. First of 564 all, it increased the lattice parameters of the fully delithiated phase and thus effectively 565 reduced the volume changes during cycling. Second, the position and width of both two566 phase regions change in comparison with LNMO. For undoped LNMO, the two-phase I/II 567 region was narrow and sharp whereas the two-phase II/III region was gradual and much 568 broader. Cr-doped LNMO showed a completely opposite characteristic and thus whereas 569 phase I/II transition is consuming broader delithiation range, transformation within phase 570 II/III transition took place much faster. However, chromium doping, unlike other doping 571 elements, is not able to reduce this three-phase change to twoor one-phase. 572 The impedance study supplemented by evaluation of the diffusion coefficient 573 revealed the improved diffusivity of Li in Cr-doped LNMO in the discharged (fully 574 lithiated) state, and during cycling, the chromium further intensifies the positive effect on 575 diffusion properties, which results in much higher discharge capacities mainly at high-rate 576 cycling. Effect of some kind of “formation” of an electrode from a diffusivity point of view 577 under high rate cycling was observed for the LNMO cathode material. 578 Doping of LiNi0.5Mn1.5O4 thus represents a very promising way in terms of 579 improving structural stability, diffusivity, and also durability of high-voltage batteries. 580 Further research on other materials or different approaches while synthesizing materials 581 are needed to improve the performance of batteries based on LNMO. 582 583 Acknowledgment 584 585 This research work has been carried out in the Centre for Research and Utilization of 586 Renewable Energy (CVVOZE). This work was supported by the specific graduate research 587 of the Brno University of Technology No. FEKT-S-20-6206. 588 589 References 590 591 [1] J. Harlow, X. Ma, J. Li, E. Logan, Y. Liu, N. Zhang, L. Ma, S. Glazier, M. Cormier, M. Genovese, S. Buteau, A. Cameron, J. Stark and J. Dahn, "A Wide Range of Testing Results on an Excellent Lithium-Ion Cell Chemistry to be used as Benchmarks for New Battery Technologies", Journal of The Electrochemical Society, vol. 166, no. 13, pp. A3031-A3044, 2019. [2] A. Padhi, K. Nanjundaswamy and J. Goodenough, "Phospho-olivines as PositiveElectrode Materials for Rechargeable Lithium Batteries", Journal of The Electrochemical Society, vol. 144, no. 4, pp. 1188-1194, 1997. [3] A. Yamada, S. Chung and K. Hinokuma, "Optimized LiFePO4 for Lithium Battery Cathodes", Journal of The Electrochemical Society, vol. 148, no. 3, pp. A224-A229, 2001. [4] Y. Gao, K. Myrtle, M. Zhang, J. Reimers and J. Dahn, "Valence band of LiNi x Mn 2 − x O 4 and its effects on the voltage profiles of LiNi x Mn 2 − x O 4 /Li electrochemical cells", Physical Review B, vol. 54, no. 23, pp. 16670-16675, 1996. [5] J. Xu, S. Dou, H. Liu and L. Dai, "Cathode materials for next generation lithium ion batteries", Nano Energy, vol. 2, no. 4, pp. 439-442, 2013. [6] H. Yoo, E. Markevich, G. Salitra, D. Sharon and D. Aurbach, "On the challenge of developing advanced technologies for electrochemical energy storage and conversion", Materials Today, vol. 17, no. 3, pp. 110-121, 2014. [7] J. Kim, S. Myung, C. Yoon, S. Kang and Y. Sun, "Comparative Study of LiNi 0.5 Mn 1.5 O 4 - δ and LiNi 0.5 Mn 1.5 O 4 Cathodes Having Two Crystallographic Structures", Chemistry of Materials, vol. 16, no. 5, pp. 906-914, 2004. [8] J. Zheng, J. Xiao, X. Yu, L. Kovarik, M. Gu, F. Omenya, X. Chen, X. Yang, J. Liu, G. Graff, M. Whittingham and J. Zhang, "Enhanced Li ion transport in LiNi0.5Mn1.5O4 through control of site disorder", Physical Chemistry Chemical Physics, vol. 14, no. 39, 2012. [9] H. Duncan, B. Hai, M. Leskes, C. Grey and G. Chen, "Relationships between Mn 3 Content, Structural Ordering, Phase Transformation, and Kinetic Properties in LiNi x Mn 2– x O 4 Cathode Materials", Chemistry of Materials, vol. 26, no. 18, pp. 53745382, 2014. [10] T. Kazda, J. Vondrák, A. Visintin, M. Sedlaříková, J. Tichý and P. Čudek, "Electrochemical performance of Mo doped high voltage spinel cathode material for lithium-ion battery", Journal of Energy Storage, vol. 15, pp. 329-335, 2018. [11] H. Jahn and E. Teller, "Stability of polyatomic molecules in degenerate electronic states - I—Orbital degeneracy", Proceedings of the Royal Society of London. Series A - Mathematical and Physical Sciences, vol. 161, no. 905, pp. 220-235, 1937. [12] Y. Xie, Y. Jin and L. Xiang, "Understanding Mn-Based Intercalation Cathodes from Thermodynamics and Kinetics", Crystals, vol. 7, no. 7, 2017. [13] Y. Hou, K. Chang, H. Tang, B. Li, Y. Hou and Z. Chang, "Drastic enhancement in the rate and cyclic behavior of LiMn2O4 electrodes at elevated temperatures by phosphorus doping", Electrochimica Acta, vol. 319, pp. 587-595, 2019. [14] P. Chand, V. Bansal, . Sukriti and V. Singh, "Investigations of spinel LiZnxMn2−xO4 (x ≤ 0.03) cathode materials for a lithium ion battery application", Materials Science and Engineering: B, vol. 238-239, pp. 93-99, 2018. [15] Z. Su, Z. Lu, X. Gao, P. Shen, X. Liu and J. Wang, "Preparation and electrochemical properties of indiumand sulfur-doped LiMnO2 with orthorhombic structure as cathode materials", Journal of Power Sources, vol. 189, no. 1, pp. 411-415, 2009. [16] W. Pang, J. Lee, Y. Wei and S. Wu, "Preparation and characterization of Cr-doped LiMnO2 cathode materials by Pechini's method for lithium ion batteries", Materials Chemistry and Physics, vol. 139, no. 1, pp. 241-246, 2013. [17] K. Vediappan, K. Prasanna, S. Shanmugan, R. Gnanamuthu and C. Lee, "Structural stability and electrochemical properties of gadolinium-substituted LiGd x Mn 2-x O 4 spinel as cathode materials for Li-ion rechargeable batteries", Applied Surface Science, vol. 449, pp. 412-420, 2018. [18] S. Bhuvaneswari, U. Varadaraju, R. Gopalan and R. Prakash, "Structural stability and superior electrochemical performance of Sc-doped LiMn2O4 spinel as cathode for lithium ion batteries", Electrochimica Acta, vol. 301, pp. 342-351, 2019. [19] L. Chen, H. Wu, H. Wang, L. Chen, X. Pu and Z. Chen, "Tailoring NaVO3 as a novel stable cathode for lithium rechargeable batteries", Electrochimica Acta, vol. 307, pp. 224-231, 2019. [20] X. Jian, J. Tu, Y. Qiao, Y. Lu, X. Wang and C. Gu, "Synthesis and electrochemical performance of LiVO3 cathode materials for lithium ion batteries", Journal of Power Sources, vol. 236, pp. 33-38, 2013. [21] A. Slesarenko, I. Yakuschenko, V. Ramezankhani, V. Sivasankaran, O. Romanyuk, A. Mumyatov, I. Zhidkov, S. Tsarev, E. Kurmaev, A. Shestakov, O. Yarmolenko, K. Stevenson and P. Troshin, "New tetraazapentacene-based redox-active material as a promising high-capacity organic cathode for lithium and potassium batteries", Journal of Power Sources, vol. 435, 2019. [22] G. Sun, Y. Hu, Y. Sha, C. Shi, G. Yin, H. Zhang, H. Liu and Q. Liu, "An insoluble naphthalenediimide derivative as a highly stable cathode material for lithium-ion batteries", Materials Chemistry and Physics, vol. 236, 2019. [23] Ł. Kondracki, A. Kulka, A. Milewska and J. Molenda, "In-situ structural studies of manganese spinel-based cathode materials", Electrochimica Acta, vol. 227, pp. 294302, 2017. [24] C. Gao, H. Liu, S. Bi, H. Li and C. Ma, "Investigation the improvement of high voltage spinel LiNi0.5Mn1.5O4 cathode material by anneal process for lithium ion batteries", Green Energy & Environment, 2020. [25] L. Chladil, D. Kunický, P. Vanýsek and O. Čech, "In-Situ X-Ray Study of Carbon Coated LiFePO 4 for Li-Ion Battery in Different State of Charge", ECS Transactions, vol. 87, no. 1, pp. 107-114, 2018. [26] J. Yoon, D. Kim, J. Um, M. Jeong, W. Oh and W. Yoon, "Effect of local structural changes on rate capability of LiNi 0.5 Mn 1.5 O 4−δ cathode material for lithium ion batteries", Journal of Alloys and Compounds, vol. 686, pp. 593-600, 2016. [27] Y. Shen, X. Ju, J. Zhang, T. Xie, F. Zong, D. Xue, X. Lin, J. Zhang and Q. Li, "A convenient co-precipitation method to prepare high performance LiNi0.5Mn1.5O4 cathode for lithium ion batteries", Materials Chemistry and Physics, vol. 240, 2020. [28] X. Ren, Y. Wang, Q. Xiao, G. Lei and Z. Li, "Excellent electrochemical performances of high-voltage LiNi0.5Mn1.5O4 hollow microspheres synthesized by a static co-precipitation method", Materials Letters, vol. 248, pp. 97-100, 2019. [29] T. Kazda, J. Tichý, O. Cech, M. Sedlarikova, P. Čudek, T. Binar, J. Švarc and P. Vyroubal, "Influence of Conditions on the Process of Spinel Based Cathode Material Synthesis", ECS Transactions, vol. 81, no. 1, pp. 127-133, 2017. [30] T. Kazda, O. Cech, J. Vondrak, A. Strakova Fedorkova, A. Visintin, P. Cudek and V. Kasparek, "Comparative Study of the Properties of High Voltage Spinel Cathode Material Depending on the Method of Synthesis", ECS Transactions, vol. 63, no. 1, pp. 15-21, 2014. [31] E. Logan, E. Tonita, K. Gering, L. Ma, M. Bauer, J. Li, L. Beaulieu and J. Dahn, "A Study of the Transport Properties of Ethylene Carbonate-Free Li Electrolytes", Journal of The Electrochemical Society, vol. 165, no. 3, pp. A705-A716, 2018. [32] M. Börner, P. Niehoff, B. Vortmann, S. Nowak, M. Winter and F. Schappacher, "Comparison of Different Synthesis Methods for LiNi 0.5 Mn 1.5 O 4 -Influence on Battery Cycling Performance, Degradation, and Aging", Energy Technology, vol. 4, no. 12, pp. 1631-1640, 2016. [33] W. Zhu, D. Liu, J. Trottier, P. Hovington, C. Gagnon, A. Guerfi, K. Zaghib, A. Mauger, H. Groult and C. Julien, "In Situ XRD Study of the Phase Evolution in LixMn1.5Ni0.5O4 as 4.7-Volt Positive Electrode Materials for Li-Ion Batteries", ECS Transactions, vol. 58, no. 14, pp. 35-40, 2014. [34] S. GURPREET, P. AMRISH, S. ANJAN and G. SUDIPTO, "Synthesis and characterization of citric acid assisted Cr doped lithium manganese oxide spinel", Ceramics – Silikáty, vol. 53, no. 4, pp. 260-267, 2009. [35] D. Liu, J. Hamel-Paquet, J. Trottier, F. Barray, V. Gariépy, P. Hovington, A. Guerfi, A. Mauger, C. Julien, J. Goodenough and K. Zaghib, "Synthesis of pure phase disordered LiMn1.45Cr0.1Ni0.45O4 by a post-annealing method", Journal of Power Sources, vol. 217, pp. 400-406, 2012. [36] W. Zhu, D. Liu, J. Trottier, C. Gagnon, A. Mauger, C. Julien and K. Zaghib, "In-situ X-ray diffraction study of the phase evolution in undoped and Cr-doped LixMn1.5Ni0.5O4 (0.1 ≤ x ≤ 1.0) 5-V cathode materials", Journal of Power Sources, vol. 242, pp. 236-243, 2013. [37] J. Hunter, "Preparation of a new crystal form of manganese dioxide: λ-MnO2", Journal of Solid State Chemistry, vol. 39, no. 2, pp. 142-147, 1981. [38] C. Gao, H. Liu, S. Bi, H. Li and C. Ma, "Investigation the improvement of high voltage spinel LiNi0.5Mn1.5O4 cathode material by anneal process for lithium ion batteries", vol. 6, no. 1, pp. 114-123, 2021. [39] Y. Cui, X. Zhao and R. Guo, "Improved electrochemical performance of La0.7Sr0.3MnO3 and carbon co-coated LiFePO4 synthesized by freeze-drying process", Electrochimica Acta, vol. 55, no. 3, pp. 922-926, 2010. [40] M. Mohamedi, "Electrochemical investigation of LiNi0.5Mn1.5O4 thin film intercalation electrodes", Electrochimica Acta, vol. 48, no. 1, pp. 79-84, 2002. [41] L. Wang, J. Zhao, X. He, J. Gao, J. Li, C. Wan and C. Jiang, "Electrochemical Impedance Spectroscopy (EIS) Study of LiNi1/3Co1/3Mn1/3O2 for Li-ion Batteries", International Journal of Electrochemcal Science, vol. 7, no. 1, pp. 345 - 353, 2012. [42] C. Snyders and E. Ferg, "Electrochemical Impedance Spectroscopy (EIS) study of doped spinel manganese cathode oxide materials synthesized for Li-ion batteries", Materials Today: Proceedings, vol. 5, no. 4, pp. 10450-10459, 2018. 592 593