Surface Coating of NCM-811 Cathode Materials with g-C3N4 for Enhanced Electrochemical Performance.

Li(Ni0.8, Co0.1, Mn0.1)O2 (NCM-811) cathode materials have been commercialized recently, aiming to increase the specific capacity and specific energy of the lithium-ion battery for practical applications in electric vehicles. The surface coating has been proved to be an effective approach for the stabilization of NCM-based cathodes, which could reduce the structural instability and prevent surface reactions between the cathode materials and electrolytes. Herein, we demonstrate the facile synthesis of graphitic carbon nitride (g-C3N4)-coated NCM cathodes with both the sonication-assisted liquid exfoliation method (g-C3N4NS@NCM-811) and chemical vapor-assisted coating method (g-C3N4@NCM-811). It is discovered that coating with a thin g-C3N4 layer could reduce the impedance of the NCM-811 cathode material, as well as increase the cycle stability of the cathode material, leading to increased capacity retention from 130 mA h/g (for the pristine sample) to 140 mA h/g after 225 cycles. While the coating of thick g-C3N4 nanosheets could hinder the lithium intercalation, resulting in significant loss of specific capacity. This study paves the way toward practical applications of the g-C3N4-coated NCM-811 cathode materials.

Introduction

The recent rapid development of electric vehicles requires tremendous efforts to improve the specific energy of lithium battery systems to solve the “mileage anxiety”. Among all the commercially available lithium-ion battery cathode materials, Ni-rich-layered oxides (LiNiCoMnO2, denoted as NCM, where x + y + z = 1) exhibit the highest specific energy density,[1−5] which has been considered as the potential next-generation cathode materials for the complete electrification of transportation. In particular, Li(Ni0.8, Co0.1, Mn0.1)O2 (denoted as NCM-811) has been commercialized recently, which could significantly minimize the use of expensive Co resources while also ensuring considerable stability of the battery system.[6−8] Meanwhile, it is well known that during the de-lithiation process, the transition-metal ions could dissolve into the electrolyte due to the surface side reactions (e.g., with the residual water or HF in the electrolyte), leading to structural instability and capacity fade.[9−11]

The surface coating has been proved to be an effective avenue for the stabilization of NCM-based cathode materials.[12−14] The thin surface coating layer acts as an artificial cathode electrolyte interface (CEI), which can prevent direct contact between an NCM cathode and electrolyte, mitigating the solid–liquid reaction while also enhancing the mechanical stability. Various coating materials have been investigated so far, including reduced graphene oxide,[15] metal oxides, such as MgO and Al2O3,[16−18] florides (i.e., AlF3),[19] lithiated oxides (e.g., Li2TiO3),[20] polymers (i.e., polyimide),[21] and so forth. They have been shown to significantly increase the structural stability of the cathode materials and hence, the electrochemical performances can be enhanced. However, these coating techniques typically require expensive equipment or raw materials, which hinder the practical commercial applications.

Graphitic carbon nitride (g-C3N4) has been shown to exhibit excellent chemical resistance and a moderate band gap (∼2.7 eV[22]), which has been widely explored for applications in photocatalysis,[23−25] sodium-ion battery anode,[26] lithium-ion battery anode,[27] separator for lithium–sulfur battery,[28] and so forth. In particular, the excellent chemical resistance and moderate band gap of C3N4 make it a good potential coating material for the NCM cathode, which is stable in the electrolyte during cycling, while also ensuring proper electronic and ionic transport across the C3N4/NCM interface. Moreover, g-C3N4 has a similar layered structure to graphite, where the layers are bound by relatively weak van der Waals forces. For g-C3N4, the lattice spacing between layers is about 3.26 Å,[29−31] which can act as a channel for Li+ diffusion.

Herein, in this study, we designed a surface-coating routine of the NCM-811 cathode materials with graphitic carbon nitride (g-C3N4) via both sonication-assisted liquid exfoliation (g-C3N4NS@NCM-811) and chemical vapor-assisted coating methods (g-C3N4@NCM-811). It is discovered that coating with a thin g-C3N4 layer (thin g-C3N4@NCM-811, ∼20 nm) could reduce the impedance of the NCM-811 cathode material, as well as increase the cycle stability of the cathode material, leading to the increase of specific capacity from 130 mA h/g (for the pristine sample) to 140 mA h/g after 225 cycles. While the thick coating of g-C3N4 (thick g-C3N4@NCM-811, ∼70 nm) could lead to a large decrease of specific capacity (decreases to ∼115 mA h/g after 225 cycles). Furthermore, the g-C3N4 nanosheet coating (g-C3N4NS@NCM-811) by simple mechanical ball milling forms a much thicker (∼100 nm) and inhomogeneous coating, which could hinder the intercalation of Li, resulting in almost no energy storage capacity.

Experimental Section

Materials Synthesis

The schematic illustration of the synthesis procedures for the g-C3N4-coated NCM-811 cathode materials is shown in Figure . Two types of cathode materials are designed, namely, the g-C3N4 nanosheet-coated NCM-811 (denoted as g-C3N4NS@NCM-811) and g-C3N4 surface-coated NCM-811 (denoted as g-C3N4@NCM811). To synthesize the g-C3N4NSs@NCM-811 (Figure a), melamine (99%, Shanghai Macklin Biochemical Co., Ltd.) was first sintered and decomposed at 550 °C to obtain g-C3N4. Then, 400 mg yellowish g-C3N4 particle was dispersed in 200 ml deionized water. The sonication-assisted liquid exfoliation method was further employed, where the water/g-C3N4 mixture was treated with sonication for 75 min to form a g-C3N4 nanosheet suspension. Then, the suspension was purified via centrifugation at a speed of 5000 rpm for 5 min. Consequently, 400 mg NCM-811 (purchased from Ningbo Rongbay New Energy Technology, Co. Ltd., China) was added, stirred vigorously with a magnetic stirrer, filtered, and dried at 80 °C for 12 h. Meanwhile, to obtain the g-C3N4@NCM-811, a chemical vaporization-assisted synthesis method is used (Figure b). Different amounts of melamine were added to control the thickness of coating layers, where 1 g/2 g melamine powder and 5 g NCM-811 powder were put separately in a closed crucible and heated to 350 °C for 2 h to ensure the full vaporization of melamine. Then, they were further heated to 550 °C for 4 h to decompose the surface melamine to form g-C3N4. The thin g-C3N4@NCM-811/thick g-C3N4@NCM-811 can be obtained after furnace cooling for the samples with 1 g/2 g melamine powder.

Figure 1

Schematic illustrations of the synthesis process for (a) g-C3N4NSs@NCM-811 and (b) g-C3N4@NCM-811.

Structural Characterization

Structural characterizations were performed to investigate the phase, morphology, and elemental distributions of the cathode materials. X-ray diffraction (XRD) measurements were employed with Cu-Kα (MiniFlex 600 Diffractometer, Rigaku, Japan, λ = 1.54059 Å) to unveil the crystal structure of the coated and uncoated NCM-811, with a scan rate of 10°/min in the range of 5–90°. The surface morphology was further characterized by field-emission scanning electron microscopy (SEM, Hitachi S-4800, 5 kV) and high-resolution transmission electron microscopy (HRTEM, Hitachi HT-7700, 120 kV).

Battery Assembly

The cathode was obtained with 80 wt % of the as-prepared NCM-811 materials, 10 wt % Ketjen Black (Lion Co., Japan), and 10 wt % polyvinylidene fluoride (PVDF, Arkema, France). They were ground for 10 min in an agate mortar to ensure full mixing before dissolving in a NMP (99.5%, Shanghai Aladdin Bio-Chem Technology Co., Ltd., China) solvent and consequently stirred vigorously for 5 h to form a homogenous slurry. It was further cast on an Al foil and dried at 80 °C for 12 h in vacuum before pouching into round cathode disks. They were then calendered under 20 MPa and dried in vacuum at 80 °C for another 4 h to further remove the residual NMP solvent and water. The CR2025 standard coin cells were further assembled, with a Li metal anode (China Energy Lithium Co., Ltd., Tianjin, China) as the counter electrode and 1.0 M LiPF6 in EC/DMC = 3:7 vol % with 5.0% FEC additives (Suzhou Duoduo Chemical Tech. Co., Ltd., China) as the electrolyte and separator (Celgard Co. Ltd., USA).

Electrochemical Testing

The electrochemical performances were tested using a Land battery testing station (Wuhan LAND Electronic Co. Ltd., Wuhan, China) under a voltage range of 2.6–4.2 V. The first cycle was tested at 0.1 C, followed by 1 C for the consequential cycles (1 C = 205 mA/g).

Results and Discussion

XRD measurements are performed to unveil the phase of the material. As shown in Figure a, the XRD pattern for the pristine NCM-811 matches well with a pure α-NaFeO2-type crystal structure (space group R3̅m, JCPDS #09-0063) with no obvious secondary phase peaks.[32] The corresponding crystal surfaces for the major peaks have been indexed in the figure. The splitting of the two peaks, namely, (006)/(012) and (018)/(110), can be attributed to the hexagonal ordering of the layered structure for NCM-811.[33,34] Meanwhile, it is also widely accepted that the intensity ratio of the (003)/(104) peaks is determined by the degree of cation disorder, where the lower intensity ratio leads to a higher degree of Li+/Ni2+ cation disorder.[35−37] The measured intensity ratio of the (003)/(104) peaks for the pristine sample is 1.4, showing a cation-ordered phase. Both the g-C3N4NSs@NCM-811 and g-C3N4@NCM-811 exhibit a similar pattern to the pristine sample, with a slight change in the peak intensities, indicating that the majority of the phase of the material does not change during the coating procedures. The intensity ratio of the (003)/(104) peaks for the g-C3N4NSs@NCM-811 and g-C3N4@NCM-811 decreases to 1.3 and 1.2, respectively. This reveals that the coating procedure, in particular, the heat treatment, could increase the Li+/Ni2+ cation disorder. This can be attributed to the oxygen release during the heat treatment. While it is reported that increasing the cation disorder leads to a poorer rate performance and cycle performance due to its hinderance to Li+ transport with disordered Ni2+.[38,39] Because the intensity ratio only changes slightly, the influence of the cation disorder on the cathode performance should be minimal. Notably, no obvious C3N4 peaks can be discovered in both samples. It can be understood that because the g-C3N4 coating layer is very thin it cannot be detected by the XRD technique.

Figure 2

Structural characterizations for the coated/uncoated NCM-811 cathode materials. (a) XRD patterns of the cathode materials. SEM image of (b) pristine NCM-811 particle, (c) g-C3N4NSs@NCM-811, (d) thin g-C3N4@NCM-811, and (e) thick g-C3N4@NCM-811. (f–i) Magnified view of (b–d), respectively. EDS mapping of (j) g-C3N4NSs@NCM-811, (k) thin g-C3N4@NCM-811, and (l) thick g-C3N4@NCM-811 particle, the Ni, Co, Mn, and N elements are shown.

The SEM images for the pristine NCM-811, g-C3N4NSs@NCM-811, thin g-C3N4@NCM-811, and thick g-C3N4@NCM-811 are displayed in Figure b–i. It can be observed that the pristine samples are spherical in shape, with a diameter ranging from 10 to 20 μm (Figure b). The large spherical particle is formed by the assembly of small nanosized primary particles. The particle structure is preserved for the coated NCM-811 samples, as shown in Figure c–e. Meanwhile, the magnified view of the g-C3N4NS@NCM-811 shows the formation of a thick coating layer on the NCM-811 surface (Figure g). Whereas the surface coating layer is not as obvious in the g-C3N4@NCM-811 (Figure h,i). The EDS mappings for the g-C3N4NS@NCM-811 (Figure j), thin g-C3N4@NCM-811 (Figure k), and thick g-C3N4@NCM-811 (Figure l) show that the distribution of the transition metals (Ni, Co, and Mn) and N elements are homogenous in the particle region, while the N signal is brighter near the particle edge for the g-C3N4NS@NCM-811 sample. This indicates a thicker and inhomogeneous coating of the g-C3N4 nanosheets on the NCM-811 particle surface with pure mechanical mixing.

TEM characterizations are performed for the pristine NCM-811, g-C3N4 nanosheet, g-C3N4NSs@NCM-811, and g-C3N4@NCM-811, as depicted in Figure . As shown in Figure a, the pristine NCM-811 particle (shown in black contrast) has a well-defined shape interface without an obvious surface coating layer. The surface rupture can be discovered, which can be attributed to the nanosized primary particles. The g-C3N4 nanosheet prepared by the sonication-assisted liquid exfoliation method is presented in Figure b, which forms a flake-like irregular shape with a size of ∼100 nm. For the g-C3N4NSs@NCM-811, while the overall shape is similar to the pristine sample, a thick coating layer (∼100 nm) can be discovered on the black NCM-811 particle (Figure c). It can also be observed that flakes similar to the g-C3N4 nanosheet are further attached to the NCM-811 surface beside the direct surface coating. Meanwhile, a thin homogenous coating layer can be discovered for the thin g-C3N4@NCM-811 (Figure d), showing that g-C3N4 has been homogeneously coated on the NCM-811 particle. With the increase in the weight ratio of melamine from 20% to 40%, the g-C3N4 coating layer increases from ∼20 to ∼70 nm. Overall, the SEM and TEM characterizations confirm the successful surface coating of the g-C3N4 on the NCM-811 particle for both g-C3N4NSs@NCM-811 and g-C3N4@NCM-811, whereas the coating layer is much thicker for g-C3N4NSs@NCM-811. This demonstrates that the thickness of the g-C3N4 coating layer can be adjusted by changing the melamine/NCM-811 ratio.

Figure 3

TEM image of (a) pristine NCM-811 particle, (b) g-C3N4NS, (c) g-C3N4NSs@NCM-811, (d) thin g-C3N4@NCM-811, and (e) thick g-C3N4@NCM-811.

The electrochemical performance is further tested and given in Figure . The differential capacity analysis is performed for the three samples (Figure a). It can be found that the pristine NCM-811 exhibits multiple peaks between 3.5 to 4.2 V for both charge and discharge cycles, consistent with previous reports.[40,41] A slight right shift can be discovered for the g-C3N4@NCM-811, which can be attributed to the slightly increased overpotential with surface coating. Whereas no peaks can be discovered for the g-C3N4NSs@NCM-811, indicating that the coating of thick g-C3N4 nanosheets could hinder the intercalation process for the NCM-811 cathodes. The electrochemical impedance spectroscopy (EIS) measurements are further performed, which demonstrates increased charge-transfer resistance (Rct) from 412.7 Ω to 498.1 Ω for the g-C3N4NSs@NCM-811 as compared to the pristine sample. While the series resistance Rs also increases from around 2.0 to 14.1 Ω for the g-C3N4NSs@NCM-811 (Figure b). This indicates that thick g-C3N4 nanosheet coating would increase both series resistance and charge-transfer resistance, which can be understood as because of bulk g-C3N4 being a semiconductor with a band gap of 2.7 eV. Interestingly, the charge-transfer resistances of the g-C3N4@NCM-811 samples reduce from 412.7 Ω to around 320 Ω. The reduced charge-transfer resistance indicates the prohibition of the surface side reactions with electrolytes, which could improve the surface stability and the electrochemical performance. The charge–discharge profile is presented in Figure c. The charge profile for the pristine sample and g-C3N4@NCM-811 almost overlap with each other, with a plateau of ∼3.75 V and a linear slope between 3.8 to 4.2 V, corresponding to the multiple peak configurations in the dQ/dV curve. While for the discharge curve, the thin g-C3N4-coated cathode shows a slightly lower capacity under the same voltage above 3.7 V, while a higher capacity below 3.7 V, indicating that the coating could enhance the overall structural stability during the discharge process. As the coating layer of g-C3N4 becomes thicker, the specific capacity of cathode materials decreases. Meanwhile, the g-C3N4NSs@NCM-811 shows almost no plateau and capacity during the charge and discharge. The specific capacity and coulombic efficiency during cycling are further given, under 1C, both pristine NCM-811 and thin g-C3N4-coated NCM-811 exhibit an initial specific discharge capacity of ∼170 mA h/g, consistent with a previous report (Figure d).[10] While the specific capacity almost coincides for the first 50 cycles, the thin g-C3N4@NCM-811 demonstrates higher capacity retention after 50 cycles. It is shown that after 225 charge/discharge cycles, the discharge capacity decreases to ∼130 mA h/g for the pristine NCM-811, whereas the g-C3N4@NCM-811 shows a higher capacity of ∼140 mA h/g, proving that the coating could increase the overall cycle stability of the NCM-811 cathodes. The improved stability during cycling can be attributed to two factors: (1) g-C3N4 can act as an artificial CEI, which can prevent the dissolution of transition-metal ions on the electrolyte. (2) g-C3N4 can reduce the charge-transfer resistance because it prohibits the surface reaction between the NCM cathode and electrolyte that could form insulating CEI components such as LiF. Whereas for the g-C3N4NSs@NCM-811, the specific capacity is <5 mA h/g, indicating that the intercalation process is hindered after the surface coating of thick g-C3N4 nanosheets. It is worthwhile to note that coulombic efficiency during the cycling for both thin and thick g-C3N4 coating increases, which demonstrates that the g-C3N4 coating layer can prevent the side reactions between the NCM-811 cathode and electrolyte.

Figure 4

Electrochemical performance for the coated/uncoated NCM-811 cathode materials. (a) dQ/dV with respect to electric potential (V). (b) EIS measurements. (c) Charge/discharge curve. (d) Discharge capacity (hollow dots) and coulombic efficiency (filled dots) during cycling at 1 C.

Conclusions

In conclusion, we have synthesized the g-C3N4-coated NCM-811 cathode materials via the sonication-assisted liquid exfoliation method (g-C3N4NS@NCM-811) and chemical vapor-assisted coating method (g-C3N4@NCM-811). It is discovered that the surface coating of a thin g-C3N4 layer (thin g-C3N4@NCM-811) could reduce the impedance of the NCM-811 cathode materials, as well as increase the cycle stability of the cathodes, leading to increased capacity retention from 130 mA h/g for the pristine sample to 140 mA h/g after 225 cycles. Meanwhile, for the g-C3N4NS@NCM-811, the impedance increases while the specific capacity is <5 mA h/g, showing that the lithium intercalation is severely hindered by the thick coating layer. We hope this study could spur further interest in the research and practical applications of the g-C3N4 surface-coated NCM-811 cathodes.