Kinetics of Li-ion transfer reaction at LiMn2O4, LiCoO2, and LiFePO4 cathodes

Kinetics of LiFePO4, LiMn2O4, and LiCoO2 cathodes operating in 1 M LIPF6 solution in a mixture of ethylene carbonate and dimethyl carbonate was deduced from impedance spectra taken at different temperatures. The most striking difference of electrochemical impedance spectroscopy (EIS) curves is the impedance magnitude: tens of ohms in the case of LiFePO4, hundreds of ohms for LiMn2O4, and thousands of ohms for LiCoO2. Charge transfer resistances (Rct) for lithiation/delitiation processes estimated from the deconvolution procedure were 6.0 Ω (LiFePO4), 55.4 Ω (LiCoO2), and 88.5 Ω (LiMn2O4), respectively. Exchange current density for all the three tested cathodes was found to be comparable (0.55–1·10−2 mAcm−2, T = 298 K). Corresponding activation energies for the charge transfer process, Ect#\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {E}_{ct}^{\#} $$\end{document}, differed considerably: 66.3, 48.9, and 17.0 kJmol−1 for LiMn2O4, LiCoO2, and LiFePO4, respectively. Consequently, temperature variation may have a substantial influence on exchange current densities (jo) in the case of LiMn2O4 and LiCoO2 cathodes.


Introduction
Rechargeable Li-ion batteries have become major power sources in a growing number of applications, including the automotive industry. This creates new requirements such as high rate performance during both charging and discharging. To describe the rate limiting process, all internal system resistances must be known. Among several processes taking place in Li-ion batteries, the Li + charge transfer reaction at both the cathode and anode seems to be essential to gain insight into the internal cell impedance. The most studied anode materials are different carbons and LiMn 2 O 4 , while LiCoO 2 and LiFePO 4 are basic cathode materials. Kinetics of anodic [1][2][3][4][5][6][7][8] and cathodic processes [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] have been studied. Results of reported studies are not comparable. However, one of the most important factors is the resistance (impedance) associated with the charge transfer process, which has not been fully clarified for cathodes. The general aim of this study was to compare the charge transfer process taking place at LiMn 2 O 4 , LiCoO 2 , and LiFePO 4 cathodes using the same methodology and electrolytes.
Tested cathodes were prepared on golden current collectors by a casting technique from a slurry of electrode material, carbon black, and PVdF suspension in NMP. After solvent (NMP) evaporation at 120°C in a vacuum, a layer of the electrode, containing the active material, the electronic conductor (CB), and a binder (PVdF) was formed.

Procedures and measurements
Electrochemical properties of the cells were characterized using electrochemical impedance spectroscopy (EIS) and galvanostatic charging/discharging tests. The cycling measurements were taken with the use of the ATLAS 0461 MBI multichannel electrochemical system (Atlas-Sollich, Poland). Impedance spectroscopy measurements were performed using the Gamry 1000 multichannel electrochemical system (USA) at different temperatures. Tested anodes were separated from metal-lithium counter and reference electrode by the glass microfiber GF/A separator (Whatmann, 0.4-0.6-mm thick), all placed in an adopted Swagelok® connecting tube. Typically, the mass of the lithium was ca. 31 mg (0.785 cm 2 ), while that of cathodes was 3.0-4.0 mg. The cells were assembled in a glove box in the dry argon atmosphere. After electrochemical measurements, the cells were disassembled and the cathodes were washed with DMC and dried in vacuum at room temperature. The morphology of electrodes was observed with a scanning electron microscope (SEM, Tescan Vega 5153). The BET surface of pristine electrode materials was determined with an Autosorb iQ apparatus (Quantochrome Instruments, UK) and particle size distribution with a Zetasizer Nano ZS (Malvern Instruments Ltd., UK).

Results and discussion
BET surface, size distribution, and SEM images Specific surface area of cathode materials from BET analysis was between 15.4 and 2.05 m 2 g −1 (Table 1). Particle size distribution is shown in Fig. 1. LiFePO 4 and LiCoO 2  Impedance and kinetic parameters Figure 5 shows impedance spectra of tested cathode materials at room temperature. All curves consist of two semicircles at the high frequency region and a straight line at low frequencies (in the case of LiFePO 4 , only a part of the first semicircle can be seen). The most striking difference is the impedance (resistance) magnitude: tens of ohm in the case of LiFePO 4 , hundreds of ohm for LiMn 2 O 4 , and thousands of ohm for LiCoO 2 . Impedance spectra were deconvoluted according to the equivalent circuit shown in Fig. 6. It was selected from a library of circuits based on two time constants due to best correlation of fits with experimental data. Passivation film (R f ) and charge transfer (R ct ) resistances obtained from the deconvolution procedure are shown in Table 1. Both R f and R ct given in ohms are expressed versus the geometrical surface area of electrodes (1.27 cm 2 ). It can be seen from Table 1 that resistance of the passivation film differs by two orders of magnitude depending on the electrode material: it was 4.9 Ω for LiFePO 4 , ca. 20 times more for LiMn 2 O 4 (102 Ω), and as high as 645 Ω for LiCoO 2 . Charge transfer resistances (R ct ) estimated from the deconvolution procedure were between 6.0 and 88.5 Ω (Table 1). Those resistances can be expressed against the anode real surface area A (estimated from BET measurements S = m·S BET ) as R ct ·S (expressed in Ωcm 2 ). Charge transfer resistances may be converted into surface area independent of exchange current densities: Both R ct ·S and j o values are given in Table 1. It can be seen that while charge transfer resistances differ for one order of Kinetic parameters of cathodes may be compared to the corresponding data for the metallic lithium [9,[25][26][27][28][29][30] anode. A typical reported value is of the order of 10-10 −1 mA cm −2 , depending on the solvent, electrolyte, and its concentration [30]. Exchange current density values, expressed versus the active material specific surface, suggest that the kinetics of the charge transfer taking place at the cathode is slower in comparison to metallic lithium, while the surface area is larger. To our knowledge, available literature does not present exchange current densities for LiC 6 (graphite) anodes working together with classical LiPF 6 electrolytes. Generally, R ct values determined from impedance spectroscopy for a lithiated graphite anode can be found in the literature; however, resistance depends on the electrode size (real surface area), which usually is not mentioned. Therefore, the j o value for LiC 6 was measured in the present study in the same way as for cathodes (the impedance spectrum for the LiC 6 anode is shown in Fig. 7) and data are listed in Table 1. It can be seen that the exchange current density is comparable to that characteristic of anodes (0.77·10 −2 mA cm −2 ).
Values of R f and R ct measured at different temperatures and plotted as −lnR = f(T −1 ) provide electrode size-independent charge transfer activation energy E # . Figure 8 shows the Arrhenius plot for a LiFePO 4 cathode as an example. Both E # f and E # ct values for all cathodes are shown in Table 1. It can be seen that activation energies for the Li + ion conduction in passivation film and charge transfer reaction are similar in the case of LiMn 2 O 4 and LiCoO 2 while in the case of the LiFePO 4 cathode, both E # f and E # ct are considerably lower. Activation energies for the charge transfer process taking places at cathodes can be found in the literature [11][12][13][14][15]. Again, a comparison of data is difficult due to different solvents, electrolytes, and their concentrations. Some data are reported for polymer electrolytes. In addition, the state of cathodes intercalation is different, which is equivalent to a  [14]). Both values were measured at a potential of ca. 3.9 V versus metallic lithium. The corresponding activation energy measured at a higher potential (4.2 V), equivalent to a lower degree of intercalation, was reported to be much lower (ca. 25 kJmol −1 [15]).

Compatibility of cathodes with the LiC 6 anode
Li-ion batteries typically contain a carbon anode, which capacity (for graphite ca. 370 mAh g −1 ) is usually higher in comparison to that characteristic of cathodes (Table 2). However, the capacity of the CF x material is very high (ca. 900 mAh g −1 ), and hence, literature kinetic data for this cathode [31] are also shown in Table 2. The ratio of the cathode and graphite mass should be equal to the corresponding ratio of specific capacity (n c /n G = q c / q G ) to maximize active material utilization. The ratio m c /m G indicates also the mass of the cathode compatible with 1 g of carbon material. It can be seen from Table 2 that the mass ratio n c / n G for LiMn 2 O 4 , LiCoO 2 , and LiFePO 4 cathodes is between 3.08 and 2.18 in contrast to the CF x material (n CFx /n G = 0.41). On the other hand, from the point of view of power, the ratio of anodic and cathodic charge transfer resistance should be close to 1 (a more resistive electrode determines the operation rate). The ratio of cathodic and anodic resistances R c /R G can be calculated from the ratio of electrode surface S G /S c and exchange current densities j G o /j c o : The R c /R G ratio values shown in Table 2 fall within a broad range between 0.13 and 10.04. While LiMn 2 O 4 and LiCoO 2 cathodes (R c /R G ≈ 1) are kinetically compatible with the carbon anode, the situation in the case of other cathodes is different. The LiFePO 4 material shows resistance by one order of magnitude lower than the equivalent amount of carbon. In contrast, the CF x cathode is characterized by resistance by one order of magnitude higher than the equivalent amount of the anode.
Corresponding activation energies for the charge transfer process, E # ct , differed considerably: 66.3, 48.9, and 17.0 kJmol −1 for LiMn 2 O 4 , LiCoO 2 , and LiFePO 4 ,  Table 2 Calculated compatibility parameters of graphite anode (G) and cathodes (c): theoretical capacity q, mass n of cathode material compatible with 1 g of graphite (n c = 1 g·(q G /q c )), real surface area of electrode material S (calculated as n·S BET ), ratio of graphite and cathode real surface (S G /S c ), exchange current density (j G o /j c o ), and charge transfer resistance (R c /R G )