Achievement of the High-Capacity Retention Rate for the Li[Ni0.8Co0.15Al0.05]O2 (NCA) Cathode Containing an Aqueous Binder with CO2 Gas Treatment Using the Cavitation Effect (CTCE)

The Li[Ni0.8Co0.15Al0.05]O2 (NCA) cathode containing an aqueous binder for Li ion batteries was fabricated by the CO2 gas treatment method using the cavitation effect (CTCE) with the continuous mixing device process. The CTCE is a simple process for decreasing the pH value of the cathode slurry containing an aqueous binder from the alkaline region to neutral region. With the CTCE, the retention rate of the discharge capacity at the 50th cycle with respect to that at the 2nd cycle achieved 92%. It was confirmed that almost all of the NCA particle surfaces were uniformly covered with a Li2CO3 layer after the CTCE. The Li2CO3 layer formed on the NCA particle surfaces and prevented increasing the film resistance and charge transfer resistance because the layer suppressed the electrolyte decomposition and stabilized the surfaces. These phenomena lead to improvement of the cyclability. Moreover, a part of the Li2CO3 layer on the NCA particle surfaces contains carbon black (CB) because CB is uniformly dispersed by the mixing equipment using the cavitation effect. This result improved the electrical conductivity of the NCA powder, resulting in increased capacities for all the cycles. © The Author(s) 2019. Published by ECS. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any medium, provided the original work is properly cited. [DOI: 10.1149/2.0471903jes]

The Li[Ni 0.8 Co 0.15 Al 0.05 ]O 2 (NCA) cathode for use in Li ion batteries (LIBs) has been widely expected as the primary material for notebook PCs and electric vehicles because it has a higher energy density than the LiCoO 2 (LCO) cathode that is commonly used in commercial LIBs. 1,2 In the case of producing the slurry, N-methyl-2-pyrrolidone (NMP) has been generally used as the solvent for the dispersant of the binder. 3,4 However, NMP causes serious problems which include being harmful for human body, flammable, expensive, etc. 5,6 Therefore, it is strongly demanded that the NMP solvent is replaced by water which is cheap and noncombustible. Indeed, the graphite electrodes using the aqueous binder have already been employed in the commercial LIBs anode. Moreover, prototype cathode for the commercial LIBs such as LiNi 0. 5 2 8 using the aqueous binder has been fabricated. However, using the NCA powder for the cathode active materials, alkaline species, such as LiOH, are contained in the NCA powder as by-products from the NCA powder production process, and the pH value of the NCA slurry is quite alkaline (>11), leading to occurrence of the corrosion reaction of the Al foil current collector when it is coated by the alkaline NCA slurry. The coating of metal oxide such as TiO x 9 on the NCA particle surfaces can prevent the corrosion of the Al foil current collector and improve the cyclability. However, this method raises the process cost.
In our studies, we have recently developed a simple method to remove the alkaline species in the cathode slurry, i.e., the pressurized CO 2 gas treatment (PCT), in which the cathode aqueous slurry is only mixed under a CO 2 gas atmosphere from 0.2 MPa to 0.6 MPa. 10 The pH value of the cathode aqueous slurry changes from alkaline to neutral in only a few minutes because the LiOH dissolved in the cathode slurry promptly reacts with the CO 2 when the CO 2 gas dissolves into it, and a Li 2 CO 3 layer is formed on the cathode particle surfaces. The PCT is a batch process. Therefore, it is expected that the PCT can be used for treatment of the slurry in a laboratory. However, it is * Electrochemical Society Member. z E-mail: kimura.katsuya@aist.go.jp; m-yanagida@aist.go.jp demanded that the CO 2 gas treatment can be a continuous process for mass production. Thereafter, we have developed a CO 2 gas treatment using the cavitation effect (CTCE) as a continuous process in the present study. In the CTCE, the CO 2 gas is simply flowed from the aperture parts to the vena contracta of the mixing device using the cavitation effect while the cathode slurry is produced by this method. A lot of the bubbles containing the CO 2 gas expand and contract by the cavitation effects because of the dynamic change in the pressure. The intense movement of the fluid makes dissolution of the CO 2 gas in the slurry very rapid, resulting in a decrease in the pH of the slurry in a short time. This method can achieve a uniform dissolution of CO 2 in the slurry because many bubbles containing the CO 2 gas can be generated and dissolved in the slurry. Moreover, a lot of the slurry can be treated in a short time by use of the CTCE.
The capacities for the NCA electrode were significantly improved for all the cycles, and a 92% retention rate of the discharge capacity for 49 cycles was achieved using the CTCE with the continuous process.

Experimental
Production of the NCA slurry by the CTCE.-The NCA powder (92 wt%) was mixed with carbon black (CB) (4 wt%) and an acrylic polymer (4 wt%) in water (acrylic binder) by a mixing device using the cavitation effect (Jet Paster, Nihon Spindle Manufacturing Co., Ltd.) along with adding the CO 2 gas (1.0 L/min). The pH value of the slurry after the CTCE indicated 8.5.
Production of the NCA slurry with the PCT.-CB and the acrylic binder were agitated by a planetary centrifugal mixer. The dispersion was mixed under CO 2 gas at 0.55 MPa for 4 minutes in order to dissolve the CO 2 in the dispersion. The NCA powder was added to the dispersion and stirred by the mixer. The slurry was then again mixed under the CO 2 gas at 0.55 MPa for 4 minutes to further neutralize the alkaline species. The weight ratio of NCA/CB/binder was the same as the CTCE. The pH value of the slurry after the PCT indicated 8.5. Fabrication and electrochemical evaluation of the NCA electrodes.-These slurries were pasted on Al foil current collectors and dried at 80 • C in air and at 160 • C under vacuum for 13 hours. The foil was cut into a circular shape (11-mm diameter). The capacity density of the NCA electrodes was ∼1.3 mAh/cm 2 .
The NCA electrode was used as the working electrode with lithium foil as the counter electrode with a polypropylene porous micro membrane (Celgard #2325, Celgard, LLC.) as the separator in CR2032type coin cells. The electrolyte consisted of 1 M LiPF 6 in a mixture of ethylene carbonate (EC) and diethyl carbonate (DEC) (1:1 by volume).
The charge (delithiation)-discharge (lithiation) cycle tests were done using a battery charge and discharge unit (BLS series, Keisokuki Center Co., Ltd.) between 4.2 and 2.7 V for 50 cycles at 30 • C. The charge and discharge currents were 18 mA/g (∼0.1C) for the 1st cycle and 36 mA/g (∼0.2C) for the additional 49 cycles. Cyclic voltammetry (CV) profiles were obtained by an electrochemical workstation (Solartron) between 4.6 and 2.7 V at 30 • C. The AC impedance spectra were measured using the electrochemical workstation (Solartron) from 100k Hz to 0.1 Hz at 4.0 V with a 10 mV amplitude perturbation.
The transmission electron micrograph (TEM) images were obtained by a Tecnai G20 FEI with the 200 kV transmission electron microscope incident electron energy. Fig. 1 shows TEM images of the NCA particle surfaces with the CTCE using the mixing device and the cavitation effect (Figs. 1a and 1b) and PCT using the planetary centrifugal mixer for production of the slurry (Figs. 1c and 1d). After both treatments, most of the NCA particle surfaces were covered with the coating layer that was mainly composed of Li 2 CO 3 because three different lattice distances of the layer obtained from the TEM images coincided with the three strongest lines of the XRD pattern (PDF# 22-1141) for Li 2 CO 3 . Furthermore, the Li 2 CO 3 phase was observed in addition to the NCA phase from the XRD pattern of the NCA electrodes with the PCT and CTCE. However, with the PCT, only a part of the NCA particle surfaces were covered with the Li 2 CO 3 layer (Fig. 1d). On the other hand, with the CTCE (Figs. 1a and 1b), the surface areas of the NCA particle only partly covered with the Li 2 CO 3 layer were significantly decreased. In addition, it was observed that a part of the Li 2 CO 3 layer contained CB (Fig. 1b). The mechanisms for the formation of a more uniform Li 2 CO 3 layer than that with the PCT and the Li 2 CO 3 layer containing CB by using the CTCE are suggested in Fig. 2. With the CTCE, the slurry was strongly agitated by the mixing equipment using the cavitation effect. When the pressure was lower, the bubbles containing the CO 2 gas inside the aggregated powder grew in size (a-ii). The uniform dispersion of the NCA powder and CB was obtained by expansion and contraction of the bubbles due to increasing and decreasing the pressure in the equipment ((a-ii)∼(a-v)). In addition, some of the CO 2 gas had dissolved in the slurry from the many generated bubbles. Therefore, the concentration of the CO 2 in the slurry was uniform (a-v). These phenomena allowed almost all the NCA particle surfaces to be coated with the Li 2 CO 3 layer (a-vi). Moreover, a part of the Li 2 CO 3 layer contained CB (a-vi) because the CB was a uniform dispersion. On the other hand, with the PCT (Fig. 2b), the slurry was mixed by the planetary centrifugal mixer. However, a part of the NCA powder and CB was not uniformly dispersed with only rotation and revolution, leading to an uneven distribution of part of the NCA powder and CB in the slurry (b-ii). Furthermore, the CO 2 gas in the slurry flowed only from a certain direction (b-iii). This phenomenon caused part of the NCA particle surfaces to be partly coated with the existing Li 2 CO 3 layer (b-iv). Fig. 3 shows the discharge capacities of the NCA electrodes with the CTCE and PCT. The discharge capacities of the 1st, 2nd and 50th cycles, and retention rates of the discharge capacity at the 50th cycle with respect to that at the 1st and 2nd cycles are summarized in Table I. The discharge capacities at the 50th cycle for the NCA electrodes with the CTCE and PCT were 161 and 144 mAh/g, respectively, which were much higher than that for the NCA electrode without these treatments (i.e., 106 mAh/g). In spite of using the aqueous binder, the high retention rate of 92% was achieved during the 49 cycles. The discharge capacity at the 50th cycle reached 161 mAh/g, which was attributable to a decrease in the resistance of the NCA electrode for all the cycles as described later. Fig. 4 shows CV profiles of the NCA electrodes after the 50th cycle with the CTCE (Fig. 4a) and PCT (Fig. 4b) at different scan rates. It was clearly observed that the absolute values of the anodic and cathodic peak intensities increased with an increase in the scan rate. The Li ion diffusion coefficient (D Li ) was evaluated by the Randles-Sevcik equation. 11,12 I p = 0.4463n 3/2 F 3/2 C Li S R −1/2 T −1/2 D Li 1/2 v 1/2 [1] where I p is the absolute values of the anodic and cathodic peak currents, n is the number of electrons, F is Fraday's constant, C Li is the Li ion concentration of the electrolyte, S is the geometric area of the electrode, R is the gas constant, T is the absolute temperature, and v is the scan rate. Fig. 4c shows the values of the cathodic peak intensities (I p ) versus v 1 / 2 plots for the NCA electrodes with the CTCE and PCT after the 50th cycle. The D Li values can be calculated from the slopes of the I p -v 1 / 2 plots. The D Li of the NCA electrode with the CTCE was determined to be 1.3 × 10 −10 cm 2 /s, which is 2.3 times higher than that of the NCA electrode with the PCT (i.e., 5.7 × 10 −11 cm 2 /s). This result means that the Li 2 CO 3 layer formed by the CTCE better stabilized the NCA particle surfaces than that formed by the PCT, as described later, which is responsible for the good cyclability of the NCA electrode with the CTCE. Moreover, Table II shows cell voltage differences between the anodic and cathodic peaks ( V) in the CV curves (Fig. 4). The polarization of the NCA electrode with the CTCE for all the scan rate was smaller than that with the PCT. This result may mean that the physicochemical protection of the Li 2 CO 3 layer on the NCA particle surfaces by the CTCE is more effective than that by the PCT, as described later. Fig. 5 shows Nyquist plots of the NCA electrodes with the CTCE (Fig. 5a) and PCT (Fig. 5b) after the 1st, 30th, and 50th cycles. Fig. 5c shows the equivalent circuit. The Nyquist plots consisted of a small intercept, semicircles in the high and middle frequencies, and a quasistraight line in the low frequency region. The value of the intercept with Z Re corresponds to the series resistance (R s ). 13 The diameter of the small semicircle at the high frequency is assigned to the Li ion transport resistance through the film layer on the NCA electrode surfaces (R f ). 13,14 The diameter of the large semicircle in the middle frequency corresponds to the charge transfer resistance (R ct ). 14,15 The quasi-straight line corresponds to the Warburg impedance (Z w ). Fig. 6 shows the R f (Fig. 6a) and R ct (Fig. 6b) values of the NCA electrodes with the CTCE and PCT estimated from Fig. 5. The R f and R ct after the 1st cycle with the CTCE were 4.4 and 8.6 , respectively, which were lower than those with the PCT. In addition, increasing the R ct value with an increase in the cycle number was significantly suppressed compared to that with the PCT. It is believed that an increase in the R ct value during the cycling was degradation of the cathode particle surfaces probably due to dissolution of the metal ions of the cathode powder. 16 Therefore, the Li 2 CO 3 layer formed on the NCA particle

Results and Discussion
surfaces with the CTCE was more stabilized the surfaces than that with the PCT as described later. Fig. 7 shows the proposed structure for the NCA electrodes with the CTCE (Fig. 7a) and PCT (Fig. 7b) before the cycling. The NCA particle was almost totally covered with the Li 2 CO 3 coating layer before the cycling (Fig. 1). It is known that Li 2 CO 3 is electronically insulating because the bandgap of the Li 2 CO 3 is calculated to be 7.07 eV. 17 Therefore, the electrical conductivity of the surfaces was very low due to the formation of the Li 2 CO 3 layer on the surfaces.  This phenomenon caused the increasing R f , resulting in decreasing the capacity. However, with the CTCE (Fig. 7a), a part of the Li 2 CO 3 layer on the NCA particle surfaces contained CB. Therefore, the electrical conductivity of the NCA powder was significantly improved. Therefore, the R f after the 1st cycle was lower than that with the PCT (Fig. 6). Moreover, the R ct with the CTCE after the 1st cycle was also lower than that with the PCT (Fig. 6). This result was most probably due to the improvement of the dispersion of CB by the mixing device using the cavitation effect (Fig. 2). These results increased the capacities of the initial cycles and improvement of the rate characteristics. Furthermore, the surface areas of the NCA powder partially covered with the Li 2 CO 3 layer before the cycle were lower than those with the PCT (Fig. 1). It is known that the Li 2 CO 3 layer formed on the cathode powder 10 and cathode 18 can prevent the electrolyte decomposition and stabilize the cathode particle surfaces during the cycling. Thus, the NCA electrode with the CTCE has greater suppression effects on the electrolyte decomposition and reaction of the electrolyte with the surfaces during the cycling than that with the PCT. These phenomena suppressed the increasing of the R f and R ct values during the cycling (Fig. 6), leading to improvement of the capacities for all the cycles (Table I and Fig. 3).

Conclusions
We studied the NCA electrode containing the aqueous binder with the CO 2 gas treatment using the cavitation effect (CTCE). The CTCE is a simple method for decreasing the pH values of the aqueous cathode slurry. The retention rate of the discharge capacity during the 49 cycles reached 92%. It was observed that almost all the NCA particle surfaces were covered with the Li 2 CO 3 layer after the CTCE. This result suppressed the increasing of the R f and R ct values with an increase in the cycle number because the Li 2 CO 3 layer prevented the electrolyte decomposition and reaction of the NCA particle surfaces with the electrolyte. These phenomena lead to improvement of the cyclability. Furthermore, a part of the CB was absorbed in the Li 2 CO 3 layer formed on the NCA particle surfaces before the cycling. This result improved the electrical conductivity of the NCA powder, leading to increasing capacities for all the cycles. Therefore, the CTCE will be an effective treatment of the NCA cathodes containing the aqueous binder with the continuous process for mass production.