Structures and basic properties of Co-COTTTP
Single crystal X-ray diffraction (SCXRD) analysis for Co-COTTTP reveals that it is a 3D MOF crystallizing in the C2/c space group of the monoclinic crystal system (Fig. 1a, Tables S1-S2). The asymmetric unit includes a Co2+, two half-ligand COTTTP, one coordinated water, one coordinated NO3−, one free NO3−, and one CH2Cl2 molecule, forming a charge-balanced stabilizing framework (Figure S2). Each COTTTP ligand coordinated with four Co atoms via the N atoms on the pyridines. (Figure S3). The coordination environment around the Co atom is hexa-coordinated, with four nitrogen atoms from four COTTTP ligands and two oxygen atoms from one water molecule and one nitrate ion (Fig. 1b). The one-dimensional structure of Co-COTTTP is formed by the cis-trans bonding of adjacent Co atoms by COTTTP (Figure S4). Co-N bond lengths were 2.138–2.152 Å and Co-O bond lengths were 2.150–2.181 Å, comparable to those reported Co-MOFs.[18–22] The one-dimensional chains are staggered and ordered along the three crystal axes to form a three-dimensional lattice structure (Figures S5-S7). Simultaneously, rhombic apertures of the same size were formed in the b-axis direction, each containing two Co atoms and four S atoms with a size of 14.8 × 14.7 Å2 (Fig. 1c). The flexible COTTTP ligand and the presence of a large cavity allow the assembly of neighboring individual 3D networks to form a 2-fold interpenetrated framework (Figs. 1d and S8). This sulfur-rich pore contributes to its excellent lithium storage capacity. Topologically, the structure of Co-COTTTP can be reduced to a 2-fold interpenetrated pts topology by simplifying the ligand COTTTP to a 4-connected tetrahedron and the Co atoms to 4-connected nodes (Fig. 1a).
The experimental powder X-ray diffraction (PXRD) pattern of Co-COTTTP shown in Fig. 2a is in better agreement with the simulated one which suggests the phase purity of Co-COTTTP. Furthermore, the PXRD experimental results of Co-COTTTP after immersion in different organic solvents for 12 h verified that Co-COTTTP shows good chemical stability (Figure S9). Co-COTTTP was also able to maintain the framework structure when immersed in pH 3–11 solutions for 3 h, respectively, which proves that Co-COTTTP possesses good acid-base stability (Fig. 2b). The IR spectra of Co-COTTTP and the COTTTP ligand are shown in Fig. 2c. Co-COTTTP shows a sharp absorption peak at 1603 cm− 1, which is a characteristic peak of the C = N group and is caused by a deformation vibration within the pyridine ring. The C = N peak of the complex Co-COTTTP is red-shifted by 13 cm− 1 compared to the COTTTP ligand, which proves the coordination of the pyridine group in the complex. In addition, Co-COTTTP shows a weak peak at 1387 cm− 1 which is presumed to be the characteristic vibration of NO3−.
The solid-state UV-vis-NIR absorption spectrum in Figure S10 showed that Co-COTTTP had a large main absorption band in the range of 200–600 nm, which was mainly derived from the n-π* and π-π* electronic leaps of the COTTTP ligand.[17] The appearance of a broad absorption band in the 900–1100 nm range may be attributed to the d-d transition or metal-to-ligand charge transfer (MLCT).[23–25] The bandgap width of Co-COTTTP is calculated from the Tauc plot to be 2.20 eV, which shows the nature of the semiconductor material (Figure S11). The thermogravimetric analysis (TGA) data of Co-COTTTP (Fig. 2d) revealed that there was only 2.1% solvent loss up to 280°C, after which it started to decompose gradually. Based on this, we selected 400, 500, and 600°C as the carbonization temperatures, respectively.
Microstructure studies of Co-COTTTP-400/500/600
The MOF-derived materials, denoted as Co-COTTTP-400/500/600, were precisely prepared by controlling the pyrolysis temperature in N2 atmosphere using Co-COTTTP as a precursor, following the procedure illustrated in Fig. 3a.
To gain a comprehensive understanding of the framework structures of Co-COTTTP-400/500/600, a series of measurements were carried out. First, the absence of PXRD patterns for Co-COTTTP-400/500/600 in comparison to crystalline Co-COTTTP suggests that Co-COTTTP-400/500/600 exists in an amorphous state (Fig. 3b). These findings are consistent with TGA results of Co-COTTTP which indicate the decomposition of the organic skeleton above 300°C (Figure S12). Second, the nearly unchanged morphologies of Co-COTTTP-400/500/600 relative to Co-COTTTP in scanning electron microscopy (SEM) images suggest a similar rigid framework for COTTTP-400/500/600 (Figure S13). Additionally, SEM elemental mapping of COTTTP-400/500/600 reveals uniform dispersion of C, N, O, S, and Co atoms within the skeleton without significant agglomeration (Fig. 3d, Figures S14-15). At last, N2 adsorption was conducted to examine the microstructure and specific surface areas of COTTTP-400/500/600. The N2 isotherms of Co-COTTTP-400/500/600 exhibited type II curves (Fig. 3c), indicating the presence of a microporous structure in these materials. Compared to Co-COTTTP (Figure S16) Co-COTTTP-400, Co-COTTTP-500, and Co-COTTTP-600 displayed increased surface areas of 0.64, 29.76, and 77.44 m2·g− 1, respectively. These findings suggest that pyrolysis of Co-COTTTP at 400/500/600°C contributes to larger specific surface areas, which not only enhances the Li+ embedding/de-embedding rate but also facilitates electrolyte solution penetration.
To obtain more detailed information about the composition of Co-COTTTP-400/500/600, several characterization techniques including FT-IR, Raman spectroscopy, HR-TEM, and XPS were performed. First, FT-IR analysis revealed the gradual reduction in characteristic vibration of C = N in the COTTTP ligand, indicating the decomposition of the ligand (Figure S17). The absence of peaks corresponding to organic vibrations suggests the formation of N, O, and S-doped carbon materials. Second, Raman spectra (Fig. 3f) were used to quantitatively study graphitization, by analyzing the intensity ratio of the D-band (located at 1334 cm− 1, reflecting the carbon defects in the lattice) to the G-band (located at 1573 cm− 1, reflecting the ordered lattice structure of the carbon material). The values of ID/IG of Co-COTTTP-400/500/600 were 0.47, 0.79, and 0.97, respectively, indicating that higher pyrolysis temperatures result in a greater abundance of defect sites. Third, TEM images along with EDX mapping demonstrated the presence of inorganic compounds consisting of Co and O within the N, O, S-doped graphene (Figures S18-21). Additionally, high-resolution TEM analysis of Co-COTTTP-500 (Figs. 3g-h) revealed the existence of Co3O4 nanoparticles, as evidenced by lattice fringes observed at spacings of 0.28 Å, 0.24 Å, and 0.21 Å, corresponding to the (220), (311), and (200) crystallographic planes of Co3O4 and CoO, respectively.[26–31] Notably, the largest Co3O4 nanoparticle with a size of 24 nm was found in Co-COTTTP-500, suggesting that a pyrolysis temperature of 500°C is optimal for the aggregation of large Co3O4 nanoparticles, potentially enhancing the Li+ storage capacity and electrode performance. In order to obtain electronic configuration and local coordination information of Co-COTTTP-500, X-ray absorption spectroscopy (XAS) data of Co-K edge were collected. According to the X-ray absorption near edge structure (XANES) curve of Co-COTTTP-500 sample (Fig. 3i), weak front edge peaks can be observed, which indicates the existence of tetrahedral configuration in the coordination environment of Co atoms. Fourier transformed R-space in the extended X-ray absorption structure (EXAFS) region of Co-COTTTP-500 shows a significant peak at 1.55 Å, which is related to the coordination characteristics of Co-O, while the weaker peak at 2.35 Å represents the coordination characteristics of Co-Co (Fig. 3j). In order to further clarify the coordination atoms, the Co-O K-edge wavelet transform was used to indicate that the K position has coordination characteristics close to Co-O and Co-Co (Fig. 3k). Using Artemis to fit the EXAFS of the sample for quantitative analysis of the coordination environment, the coordination number of Co-O and Co-Co scatterings fitted by Co-COTTTP-500 was 4.5 and 1.2, respectively, indicating that Co-COTTTP-500 is a structure similar to Co3O4 with a small amount of CoO (Figures S22-23, Table S3). At last, X-ray photoelectron spectroscopy (XPS) was employed to investigate the elemental composition and valence states of Co-COTTTP-400/500/600. The presence of C, O, N, S, and Co elements in the XPS spectra is consistent with the elemental mapping results obtained from SEM/TEM. In the Co 2p spectrum (Fig. 4c), additional peaks located at 785.5 eV and 796.6 eV, presumed to be characteristic peaks of Co(III) in Co3O4, were observed.[32–35] In the S 2p spectrum (Fig. 4b), the new characteristic peaks appearing at 168.5 eV and 169.7 eV were assigned to C-SOX-C, which demonstrates the formation of sulfur oxides.[36–38] It can be clearly seen that new characteristic peaks appeared at 400.4 eV and 404.5 eV for graphite N and pyridine N oxides (Fig. 4d).[35, 39–41] The results of XPS testing confirmed the presence of sulfur oxides and graphitic nitrogen, thereby validating the successful synthesis of N, O, S-doped graphene materials from cobalt-based MOFs (Fig. 4a, Figures S14-15).
Furthermore, we also conducted solid-state UV-vis-NIR tests on Co-COTTTP-400/500/600 (Fig. 3e). After high-temperature pyrolysis, the Co-COTTTP derivatives all exhibit the typical black morphology of carbon materials, and their absorption ranges appear to be significantly broader than those of Co-COTTTP, which also makes them more absorbent in a wider range of wavelengths. Among them, the absorption range of Co-COTTTP-400 is slightly narrower than that of Co-COTTTP-500/600, which may be attributed to the fact that high-temperature d pyrolysis can change the energy band structure and the layout state of electrons of the materials, which in turn extends their absorption range. Meanwhile, the band gap width of the Co-COTTTP derivatives gradually becomes narrower with increasing pyrolysis temperature (Figure S25), which also allows the electrons to easily jump from the valence band to the conduction band to form an electric current and improve the conductivity of the material.
Transformation of MOF precursors to Co-COTTTP-400/500/600 anodes for lithium-ion batteries
Benefiting from high porosity, large surface area, and modifiable functional groups, composite materials derived from MOFs consisting of porous carbon and metallic oxide can serve as high-performance anode materials for lithium-ion batteries. The cyclic voltammetry (CV) curves of Co-COTTTP-500 anodes present reduction peaks at 1.52 and 0.41 V in the first cathodic scan, and these peaks might originate from the insertion of Li+ into the framework of Co-COTTTP-500 and the formation of solid electrolyte interface (SEI) 42,43 (Figure S26). Meanwhile, the initial discharge curve at a current density of 0.1 A/g is consistent with the result (Figure S27). The peak at 0.99 V in the anodic scan is ascribed to the extraction of Li+ from the framework of Co-COTTTP-500. One couple of redox peaks (0.88/0.99 V) remain almost unchanged in the subsequent three cycles, corresponding to the highly reversible insertion/extraction of Li+ into/from Co-COTTTP-500. In addition, the redox peaks of the Co-COTTTP-400/500/600 anodes occur at almost the same potential, indicating that they undergo an identical lithium storage process even with different pyrolysis temperatures (Figure S28). Nevertheless, different carbonization temperatures can cause differences in the distribution of carbon materials and metal cobalt compounds in derivatives, resulting in differences in electrochemical performance. The capacitive behavior of the Co-COTTTP-500 anode was thoroughly investigated, with kinetics analyzed using CV measurements. Figure 5a clearly illustrates CV curves at varying scan rates ranging from 0.1 to 2.0 mV s− 1. These profiles maintain a remarkably similar shape across different scan rates. Typically, the scan rate (v) and the measured current (i) conform to a specific relationship as outlined in previous studies.[44, 45]
$$\:logi=b\bullet\:logv+loga$$
The empirical parameters are denoted by a and b. As indicated in prior studies, signifies diffusion-controlled behavior, whereas a value of 1 represents ideal capacitive behavior.[44] Utilizing the aforementioned formula, and by fitting a linear regression between logi and logv, we can calculate the b-values. Figure 5b illustrates that the slope of the fitted line is approximately 0.773 (b-values) at 1.93 V, indicative of rapid kinetics arising from the pseudocapacitive effect.
In addition, Fig. 5c exhibits the detailed cycling performance of Co-COTTTP-500 anode, the initial specific capacity of 2430.6, 712.3, 501, 387.4, 264.4 mAh g− 1 at 0.1, 0.5, 1.0, 2.0, 5.0 A g− 1 is achieved, respectively. The corresponding charge-discharge profiles of Co-COTTTP-400/500/600 at different current densities are displayed in Figure S29. The samples that have not undergone carbonization and those subjected to carbonization temperatures of 400 and 600°C demonstrate inadequate rate cycling performance and limited cycle stability. The galvanostatic cycling stability of the Co-COTTTP-500 anode reveals a high reversible specific capacity of 1005.7 mAh g− 1 after 100 cycles at 0.1 A g− 1 (Fig. 5d). Even increasing the current rate to 1 A g− 1 and prolonging the cycling time to 800 cycles, they can still deliver a high capacity of 549.1 mAh g− 1 (Fig. 5e). The above results indicate that Co-COTTTP-500 has excellent lithium storage capacity and electrochemical stability, and can be used as high performance for lithium-ion batteries (Table S5).
To comprehensively elucidate the enduring stability of Co-COTTTP-500 anodes through a rigorous examination, the FTIR and XPS of the electrode sample cycling 800 cycles were conducted. The results of the FT-IR curve (Figure S30) show that there is no obvious change in functional groups, indicating the integrity structural of anodes even after extensive cycling. Additionally, the XPS testing (Figure S31) corroborated the preservation of the elemental valence states, which were found to be in alignment with the initial measurements. Collectively, these findings underscore the exceptional cycling stability of Co-COTTTP-500, affirming its stability and reliability in practical applications.
Notably, Co-COTTTP-500 exhibits satisfactory specific capacity and cycle stability compared to Co-COTTTP-400/600. In general, low-temperature carbonization of higher nitrogen precursors results in highly nitrogen-doped carbon materials. However, their poor graphitization and lower nitrogen-to-oxygen ratio can make them less conductive. Increasing the carbonization temperature will help to enhance the electrical conductivity, which will facilitate the rapid transfer of electrons and Li+, but the high temperature will reduce the total nitrogen content and decrease the reactivity at the same time. Co-COTTTP-400 has more nitrogen content but lower graphitization and relatively poorer electrical conductivity, so it is less active as an anode material for lithium-ion batteries. Simultaneously, Co-COTTTP-400 has a low specific surface area due to incomplete pyrolysis at a low temperature, which is not conducive to Li+ transport and electrolyte penetration, and therefore exhibits a low capacity. Both Co-COTTTP-500 and Co-COTTTP-600 exhibit high specific capacity, although Co-COTTTP-500 has a higher specific capacity as well as extended calendar life, attributed to the appropriate pyrolysis temperature which endows it with larger cobalt oxide active sites, giving it a stronger Li+ affinity and hence a higher specific capacity. However, the higher temperatures can remove part of the organic components of the ligand, but they also reduce the total nitrogen content of the material, which partially reduces the reactivity and is unfavorable for Li+ migration. Thus, the appropriate carbonization temperature is crucial for improving the electrochemical performance of MOF-derived anodes.