
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Authors: 
 
 
 
 
 
 
 
 
 
 
 In Hyuk Son Jong Hwan Park Seongyong Park Kwangjin Park Sangil Han Jaeho Shin Seok-Gwang Doo Yunil Hwang Hyuk Chang Jang Wook Choi Abstract Improving one property without sacrificing others is challenging for lithium-ion batteries due to the trade-off nature among key parameters. Here we report a chemical vapor deposition process to grow a graphene–silica assembly, called a graphene ball. Its hierarchical three-dimensional structure with the silicon oxide nanoparticle center allows even 1 wt% graphene ball to be uniformly coated onto a nickel-rich layered cathode via scalable Nobilta milling. The graphene-ball coating improves cycle life and fast charging capability by suppressing detrimental side reactions and providing efficient conductive pathways. The graphene ball itself also serves as an anode material with a high specific capacity of 716.2 mAh g−1. A full-cell incorporating graphene balls increases the volumetric energy density by 27.6% compared to a control cell without graphene balls, showing the possibility of achieving 800 Wh L−1 in a commercial cell setting, along with a high cyclability of 78.6% capacity retention after 500 cycles at 5C and 60 °C. Introduction The ongoing development of lithium-ion batteries (LIBs) has resulted in remarkable improvements in various aspects of their operation, including energy density, power density, cycle life, and safety. At present, mobile IT devices account for a significant portion of LIB applications, where the operating specifications of current state-of-the-art LIBs are mostly satisfactory , , . However, as electric vehicles (EVs) have penetrated LIB markets, key electrochemical properties have imposed more challenging standards; while higher energy densities are desired for increased driving mileage, enhanced reaction kinetics are demanded for fast charging and high rate operations. Safety is also definitely a critical factor for EV applications. A formidable challenge lies in overcoming the trade-off relation among these key properties; it is usually nontrivial to improve one property without sacrificing others. 1 2 3 Such a trade-off relation is particularly obvious between energy density and fast charging (or power capability). While the use of nanomaterials as active components ,  and incorporation of carbon nanomaterials , ,  as conductive agents have demonstrated improved charging rates by reducing ion diffusion distance and internal resistance, most of those approaches still require further improvement before they can be implemented in current LIB technology. The tap-densities of nanomaterials are far lower than those of conventional micro-particles, which is detrimental to volumetric energy density, a crucial factor for many LIB applications than the gravimetric counterpart. In a similar context toward reduction in internal resistance, the integration of carbon nanomaterials such as graphene has proven effective in increasing the electric conductivity of lab-scale electrodes. However, achieving a uniform distribution of a small weight content (i.e.,<3 wt%) of carbon nanomaterials in a large-scale slurry process has yet to be demonstrated. The modification of existing active materials by doping with foreign elements ,  or stoichiometric control , ,  can also increase the diffusivity of Li ions and thus the charging rate of conventional active materials with micrometer dimensions. Most of those approaches are, however, effective at the expense of specific capacity. 4 5 6 7 8 9 10 11 12 13 Furthermore, beside energy density and fast charging, achieving long cycle life, especially at high temperatures (i.e., 60 °C), still remains a challenge for advanced LIBs that incorporate high-capacity electrode materials. While nickel (Ni)-rich and Li-rich layered oxide materials are considered as upcoming cathode materials due to their superior specific capacities compared to those of classical LiCoO2 counterparts, it is well known that the inevitable cation-mixing during charging degrades the layered framework to spinel or rock-salt structures, along with unwanted side reactions at the electrode surface. Considering the efforts undertaken so far, one of the most realistic solutions for immediately achieving fast charging of cathode materials with a negligible loss in energy density and cycle life involves finding conductive protective materials that can be coated uniformly on the active materials with a minimal content. Finding advanced anode materials is also essential because current graphite anodes suffer from Li metal deposition upon high rate charging. Here we report a graphene–silica (SiO   ) assembly, called a graphene ball (GB), as a coating material for high-capacity Ni-rich layered cathode materials as well as an LIB anode material. Each GB is composed of a SiO   nanoparticle center and surrounding graphene layers, constituting a three-dimensional (3D) popcorn-like structure. The SiO   nanoparticles play a crucial role in multiple aspects, such as avoiding the formation of a silicon carbide (SiC) layer at the SiO   –graphene interface during graphene growth, ensuring uniform coating of GB onto the cathode material, and providing a high specific capacity when GB is used as an anode material. The uniform coating of GB on the Ni-rich layered cathode enhances the interfacial stability with the electrolyte and the electronic conductivity over the electrode, improving the cyclability and fast charging capability of the cathode substantially. Taking the unique advantages of GB, the full-cell consisting of the GB-coated cathode and GB anode demonstrates the possibility of high volumetric energy density near 800 Wh L−1 in a commercial cell condition, together with 78.6% capacity retention after 500 cycles at 5C and 60 °C. x x x x Results Synthesis of graphene ball In order to synthesize the popcorn-like GB, we developed a chemical vapor deposition (CVD) process for graphene growth. CH4 gas was fed into a furnace in the presence of SiO2 nanoparticles with diameters of 20–30 nm at 1000 °C. At this temperature, CH4 is decomposed to generate hydrogen atoms, which can subsequently reduce SiO2 to SiO   (  < 2). OH− is also simultaneously produced via the following reaction: x x The key feature in this reaction is that the produced SiO   provides catalytic sites for graphene growth, and OH− serves as a mild oxidant to facilitate the graphitic carbon formation toward graphene , . The role of mild oxidants on graphitic carbon formation was recently rationalized  in a way that, without a mild oxidant, the formation of amorphous carbon is more dominant. However, in the presence of a mild oxidant, it was discovered that the reaction toward CO or CO2 formation is preferred for carbon atoms that would otherwise have evolved into amorphous carbon. In fact, the facile growth of graphene using metal catalysts is known to occur in a similar manner , , , where the oxygen remaining on the metal surface leads to the preferential reaction toward graphitic carbon formation. The presence of a mild oxidant prevents the formation of silicon carbide (SiC) layer at the surface of SiO   because the oxidant prevents the reduction of Si toward SiC formation . x 14 15 16 17 18 19 x 19 In our reaction scheme based on the exclusive use of CH4, at temperatures below 900 °C, graphene barely grew. In contrast, at 1050 °C and above, graphene growth became too aggressive toward graphite formation (Supplementary Fig.  ). Notably, the growth temperature of 1000 °C is lower than those (1100–1400 °C) usually employed for CVD-based graphene growth based on Si and SiO2 substrates without metal catalysts , , . The critical role of the growth environment was clarified by a control experiment (Supplementary Fig.  ); following the conventional CVD processes using metal catalysts, CH4 and H2 were fed in a 3:7 volumetric flow ratio at 1000 °C. Because the decomposition of CH4 was hindered by the large quantity of H2, graphene growth was indeed very slow, revealing the importance of the exclusive use of CH4 on the efficient graphene growth in our synthesis protocol. Overall, the choice of reactants and reaction temperature in the present reaction scheme is distinct from previous CVD-based counterparts employing metal catalysts and plays a key role in producing a graphene–SiO   assembly with the 3D hierarchical structure. 1 20 21 22 2 x A series of scanning electron microscope (SEM) (Supplementary Fig.  ) and transmission electron microscope (TEM) analyses (Fig.  ) conducted during the course of the CVD process reflect the changes in graphene growth and morphology. After 5 min of the CVD process, a 2–3-layer-thick graphene film was vaguely observed along the surface of the SiO2 (Fig.  , e), with the powder color turning black (Supplementary Fig.  ). The graphene growth became more prominent after 30 min of the CVD process, where 3D sheet morphology was clearly visible (Fig.  , f). The high-magnification TEM image (Fig.  ) captures the hexagonal atomic arrangements in a section of graphene, which reflect its basal plane. After 30 min, it was also observed that the sizes of the SiO2 nanoparticles decreased from 28.2 to 15.3 nm on average (Supplementary Fig.  ). This occurred because graphene growth involves the aforementioned reduction of SiO2 and subsequent OH− formation. Hence, the overall graphene growth is fairly analogous to popcorn expanding in the 3D bulky structure, produced at the expense of the original kernels (Fig.  ). 3a–f 1a–g 1b 3g 1c 1g 4 1h   Graphene growth from SiO2 nanoparticles.  –  TEM characterization   before CVD growth,   after 5 min growth, and   after 30 min growth (scale bars, 50 nm).  –  Their respective magnified images (scale bars, 10 nm).   Higher magnification image of graphene after 30 min growth and its atom-level view from the white box (inset) (scale bar, 2 nm).   Graphical illustration of popcorn-like graphene growth from SiO2 nanoparticles a c a b c d f g h The graphene growth was further elucidated by a series of additional analyses. X-ray diffraction (XRD) spectra obtained during the growth indicate that as the CVD process progresses, the peaks at 2  = 26.44° and 44°, corresponding to the (002) and (101) planes of graphite, grew continuously (Fig.  ), supporting once again the formation of graphitic carbon , , . The graphene growth was also reflected in the peaks at 284.5 and 291.5 eV in the C 1s branch of X-ray photoelectron spectroscopy (XPS) analysis (Fig.  ) , . The peak at 284.5 eV is assigned to C–C and C = C bonds, which appeared prominently after 30 min of growth due to graphitic carbon formation. The progressive XPS peaks in the Si 1s branch of XPS analysis (Fig.  ) reconfirm  the reduction of SiO2 during the CVD process. With increasing growth time, the peak near 103 eV shifted to lower binding energies due to the decreased oxidation state of Si. Based on the presence and intensity of the carbon signal in energy dispersive X-ray (EDX) mapping (Fig.  ), it was found that even after 5 min, graphene growth indeed took place along the surface of the SiO   , and after 30 min, the growth spread over the entire image area, while Si and O remained homogeneously distributed. The weight portion of graphene, and the results of Raman analysis, specific surface area, and conductivity of GB at different CVD growth time are summarized in Table  . As for the specific surface area, its decrease with increasing reaction time is attributed to the progressive graphitization during which graphene layers are merged and stacked. Detailed Raman and thermogravimetric (TG) analyses spectra are presented in Supplementary Fig.  . θ 2a 23 24 25 2b 22 23 2c 25 2d–g x 1 5   Analysis of graphene-ball during its growth.  –  Monitoring graphene growth during the course of the CVD process using   XRD,   XPS in C 1s branch, and   XPS Si 2p branch.  –  EDX elemental mapping with respect to C, O, and Si after  ,   5 min and  ,   30 min graphene growth from SiO2 (scale bars, 50 nm) a c a b c d g d e f g Coating of graphene ball onto Ni-rich cathode material Because of its atom-thick-layered morphology, graphene has been considered to be one of the most ideal materials for surface protective coating of high capacity layered cathode materials. However, in the case of well-known solution processes , , ,  based on Hummer’s method, achieving a uniform coating with a limited content of graphene is nontrivial due to agglomeration, which arises from the strong electrostatic interaction between graphene sheets. In an effort to overcome the agglomeration issue, a different approach, such as the crushing of graphite employing a ball-milling process, has been introduced . However, this process is usually slow (>6 h) and, more critically, can harm the crystallinity of the host material. It should also be noted that amorphous carbon coating by CVD processes is not applicable to Ni-rich oxide materials ,  because the CVD processes extract transition metals (TMs), perturbing the original crystal structure of the host material. 26 27 28 29 30 31 32 In contrast, the synthesized GB in this study can be coated on the surface of commercial LiNi0.6Co0.1Mn0.3O2 powder (size = ~10 µm) at a minimal content of 1 wt% via a mild Nobilta milling process. The coating process is also quite fast (~10 min at 3000 rpm). The SiO   seeds play a critical role in enabling this uniform coating at low weight contents. The SiO   seeds account for 57.5 wt% of the GB (TG result in Supplementary Fig.  ) and in the coating process, their presence allows the GB to exert a shear force onto the surface of the LiNi0.6Co0.1Mn0.3O2, producing intimate interaction and even penetration. To confirm this phenomenon, the same coating process was undertaken with a physical mixture of 0.5 wt% SiO2 and 0.5 wt% graphene. This sample suffered from crystalline damage on the surface region of the NCM due to aggressive collisions between LiNi0.6Co0.1Mn0.3O2 particles as well as between LiNi0.6Co0.1Mn0.3O2 particles and the reactor wall (Supplementary Fig.  ). These phenomena were amplified by graphene agglomeration that occurred in the absence of the 3D hierarchical structures containing SiO   particle centers. x x 5b 6 x The uniform coating of GB was confirmed by SEM and TEM analyses. Hereafter, LiNi0.6Co0.1Mn0.3O2 before and after GB coating is denoted as pristine NCM and GB-NCM, respectively. Low magnification SEM images of the NCM before and after GB coating (Fig.  , Supplementary Fig.  ) showed very similar morphologies because of the uniform distribution of the low content GB. However, a higher-magnification SEM image of GB-NCM indicates that GB penetrated into the spaces between the primary particles of NCM (Fig.  ). The penetration of GB was more clearly observed in TEM images (Fig.  , f). In Fig.  , a graphene sheet was observed in the open space above an NCM particle (see red dotted lines). Some graphene sheets were aligned along the surface of the NCM and displayed a consistent interlayer distance of 3.3 or 3.4 Å (Fig.  ). 3a, b 6a, b 3d 3e 3e 3f   Coating of GB onto LiNi0.6Co0.1Mn0.3O2. SEM images of LiNi0.6Co0.1Mn0.3O2  ,   before (scale bars, 2 µm and 200 nm, respectively) and  ,   after GB coating (scale bars, 2 µm and 200 nm, respectively).  ,   TEM images of graphene sheets (scale bars, 5 nm).   Located between primary particles of LiNi0.6Co0.1Mn0.3O2 and   coated along primary particles of LiNi0.6Co0.1Mn0.3O2.  ,   EDX elemental mapping of LiNi0.6Co0.1Mn0.3O2 with respect to C and Si (scale bars, 2 µm).   Before and   after GB coating.   Raman spectra of GB, pristine LiNi0.6Co0.1Mn0.3O2 and GB-coated LiNi0.6Co0.1Mn0.3O2.   XPS spectra of the same three samples in C 1s branch.   Schematic illustration of the GB coating onto LiNi0.6Co0.1Mn0.3O2 a c b d e f e f g h g h i j k EDX elemental mapping before and after the GB coating (Fig.  ) also indicates the homogeneous distribution of GB over the entire area of each NCM particle. In Raman analysis (Fig.  ), pristine NCM did not exhibit any peaks, whereas the coating of GB in the case of GB-NCM was reflected in its D, G, and 2D peaks, which were commensurate with those of bare GB. Similarly, graphitic carbon formation represented by C–C and C = C bonds was detected in the XPS C 1s spectrum of the GB-NCM (Fig.  ). A satellite peak at 291.5 eV also supports graphitic carbon formation . Importantly, even 1 wt% addition of GB improved the conductivity of NCM powder significantly when tested in a pelletized form; pelletized pristine NCM and GB-NCM exhibited conductivity values of 1.13 × 10−3 and 9.72 × 10−2 s cm−1, respectively. Overall, the unique hierarchical structure of GB enables its uniform coating over micrometer-size NCM powder at the minimal content (Fig.  ) and supplements the relatively low conductivity of NCM. 3g, h 3i 3j 23 3k Battery testing To test the effect of the GB coating on battery performance, both pristine NCM and GB-NCM were first examined in a half-cell configuration, using a Li metal disk as a counter and reference electrode. We first discuss the electrode conditions that we employed, to emphasize the commercial viability of the GB-based approach. In current commercial LIB cathodes based on LiCoO2 (for example, those used for smart phones), the active material occupies 96–97 wt% of the electrode to maximize the volumetric energy density . In the case of Ni-rich active materials, because of their lower electronic conductivity, a content of ~92 wt% is presently accepted as an optimal value, with a conductive agent and binder each accounting for ~4 wt%. 3 In our case, in an attempt to compensate the relatively low conductivity of Ni-rich cathodes and thus increase the content of active material to the levels of LiCoO2-based electrodes, an electrode consisting of 97 wt% NCM active material, 0.5 wt% super P, 1.0 wt% GB, and 1.5 wt% polymeric binder (polyvinylidene fluoride, PVDF) was prepared (see details in Supplementary Table  ). Since SiO   seeds constitute 57.5 wt% of the GB (Supplementary Fig.  ), graphene itself occupies only 0.425 wt% 1 x 5b

Part of HackerNoon's growing list of open-source research papers, promoting free access to academic material.

ສຽງນີ້ຖືກຜະລິດເປັນພາສາຕົ້ນສະບັບຂອງເລື່ອງ!

Lempohoho, Bophelo bo Bolelele: Mokhoa oa Graphene o ntlafatsang Libeteri tsa Lithium

About Author

ຄຳເຫັນ

ວາງປ້າຍ

ບົດຄວາມນີ້ໄດ້ຖືກນໍາສະເຫນີໃນ

Related Stories

Starting 2025 with New Features: Settings Dashboard, HackerNoon Decoded, Updated Search UI, and More

Chowa Widgets: Smarter Writing at Your Fingertips

The Branding Writing Contest: Winner Announcement!

Welcome to HackerNoon Decoded: The Best of 2024 Tech Blogging

Starting 2025 with New Features: Settings Dashboard, HackerNoon Decoded, Updated Search UI, and More

Chowa Widgets: Smarter Writing at Your Fingertips

The Branding Writing Contest: Winner Announcement!

Welcome to HackerNoon Decoded: The Best of 2024 Tech Blogging

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps

Lempohoho, Bophelo bo Bolelele: Mokhoa oa Graphene o ntlafatsang Libeteri tsa Lithium

About Author

ຄຳເຫັນ

ວາງປ້າຍ

ບົດ​ຄວາມ​ນີ້​ໄດ້​ຖືກ​ນໍາ​ສະ​ເຫນີ​ໃນ

Related Stories

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps

ບົດຄວາມນີ້ໄດ້ຖືກນໍາສະເຫນີໃນ