Sandipan
Maiti
a,
Tanumoy
Dhawa
a,
Awadesh Kumar
Mallik
b and
Sourindra
Mahanty
*a
aFuel Cell & Battery Division, CSIR-Central Glass & Ceramic Research Institute, Kolkata 700032, India. E-mail: mahanty@cgcri.res.in
bBioceramics & Coating Division, CSIR-Central Glass & Ceramic Research Institute, Kolkata 700032, India
First published on 12th January 2017
We report herein a facile metal–organic framework (MOF) derived route for the synthesis of carbon embedded CeO2 (CeO2@C) with a pre-designed shape-specific morphology by varying the organic linker and by using PVP as the structure directing agent. It is found that the general morphological features of the parent MOF are mimicked by the derived oxide. Four different linkers have been used to prepare CeO2@C particles with three different shapes – spherical, bar-shaped and thin plate-like. A probable formation mechanism is discussed based on metal–ligand coordination. Influence of the morphology on the electrochemical properties as a lithium-ion battery (LIB) anode has been studied in coin cells vs. Li/Li+. The spherically shaped CeO2@C-14 shows a superior performance with a maximum specific capacity of 715 mA h g−1 at 0.05 mA cm−2, good rate performance (413 mA h g−1 at 0.5 mA cm−2) and cycling stability (∼94% capacity retention after 100 cycles). The present results demonstrate that the major limitations of metal oxide anodes – volume expansion during lithiation/delithiation, rate performance and capacity fading upon cycling – could be overcome to a great extent by adopting the two-way approach of morphology design through the MOF route and in situ embedded carbon matrix.
A schematic representation of the overall synthesis process is shown in Scheme 1 (also, see Scheme S1, ESI†).
Fig. 1 X-ray diffractograms of the synthesized MOFs (a) Ce-1,4 BDC MOF, (b) Ce-1,2,4 BTC MOF, (c) Ce-1,3,5 BTC MOF, and (d) Ce-1,2,4,5 BTEC. |
The observed diffraction peaks of Ce-1,4-BDC MOF, shown in Fig. 1a, are in good agreement with the reported XRD data of the Ce2(bdc)6·(H2O)4 MOF.63 Similarly, the diffractograms observed for the Ce-1,3,5-BTC MOF (Fig. 1b) and Ce-1,2,4,5-BTEC MOF (Fig. 1d) correspond to the reported XRD data of the Ce(1,3,5-BTC)(H2O)6 MOF64 and [Ce(H2btec)1/2(btec)1/2(H2O)]n MOF,65 respectively. Though there is no previous report of the Ce-1,2,4-BTC MOF, the observed X-ray diffractogram (Fig. 1c) closely resembles that of the Gd(1,2,4-BTC)(H2O)3·H2O MOF of the lanthanide MOF family.66 Further evidence for the coordination of carboxylate ligands with Ce can be obtained from FTIR. In the FTIR spectra, it is found that the characteristic bands (1686–1720 cm−1) of the non-ionised carboxylic groups of the linkers have disappeared while some new bands appear in the wavenumber region of 1612–1531 cm−1 (Fig. S1, ESI†). This observation proves that the cerium ions have been coordinated with the carboxylate ligands successfully to form the MOFs.62,67
Fig. 2 shows the X-ray diffractograms of the resulting CeO2@C, derived from different MOFs. The observed diffraction peaks could be indexed to (111), (200), (220), (311), (222), (400), (331) and (420) planes conforming to a cubic fluorite structure of the ceria phase with space group Fm3m. A somewhat broadened nature of the diffraction peaks originates from a low crystallite size and the presence of lattice strain. The obtained lattice parameter for each of the CeO2@C is also in good agreement with JCPDS data (file number 01-075-0390). The full width at half maximum (FWHM) values, average crystallite size and lattice strain for each of the CeO2@C are listed in Table S1, ESI.† The average crystallite size has been found to be in the range of 3.3–6.2 nm. These results suggest that difference in the carboxylate linker in the MOF does not significantly affect the crystal structure of the resulting CeO2. As the parent MOFs were heat-treated in an argon atmosphere (constant argon flow), the derived oxides contain carbon, generated in situ from the organic moiety. Thermogravimetric analyses (TGA) indicate the presence of 26.8, 20.9, 7.9 and 28.3 wt% of carbon in CeO2@C-14, CeO2@C-124, CeO2@C-135 and CeO2@C-1245 respectively (Fig. S2, ESI†).
Fig. 2 X-ray diffractograms of CeO2@C (CeO2@C-14, CeO2@C-124, CeO2@C-135 and CeO2@C-1245) derived from the corresponding MOF. |
Raman spectroscopy is an excellent tool for characterization of carbon containing materials. Therefore, Raman spectra of the synthesized CeO2@C have been recorded in the range 200–2000 cm−1 and are shown in Fig. 3 and also in Fig. S3 and S4, ESI.† CeO2 is characterised by observation of a typical band at ∼460 cm−1 corresponding to the F2g symmetry.61,68 On the other hand, the characteristic D-band (disordered) and G-band (graphitic) of carbon appear at ∼1350 and ∼1580 cm−1 respectively.69 Interestingly, CeO2@C-124 and CeO2@C-135 show sharp peak at ∼458 cm−1 and relatively weaker peaks at ∼1350 and 1580 cm−1. On the other hand, CeO2@C-14 and CeO2@C-1245 show very weak peak at ∼458 cm−1 and relatively stronger peaks at ∼1350 and 1580 cm−1 which may be due to the deep embedment of CeO2 crystallites in the carbon matrix. From the peak intensity ratio of D and G bands, the nanographite size has been calculated70 and tabulated in Table S2, ESI† along with other Raman spectra information. The nanographite crystal size is found to increase in the order: CeO2@C-135 < CeO2@C-1245 < CeO2@C-124 < CeO2@C-14. Further, it can be seen that all the Raman peaks are shifted from their theoretical positions (Table S2, ESI†). The CeO2 peaks are downshifted indicating the presence of tensile stress in the lattice, which revalidate the XRD results. On the other hand, the nanographite peaks are upshifted due to compressive stress present in the lattice.
Fig. 3 Raman spectra of CeO2@C (CeO2@C-14, CeO2@C-124, CeO2@C-135 and CeO2@C-1245) in the range 200–2000 cm−1. |
Morphology and microstructural features of the synthesized MOFs and the derived oxides have been examined by FESEM and TEM. Fig. 4a–h show the FESEM images of Ce-MOFs and corresponding CeO2@C. It is observed that the general morphological features of the parent MOF are mimicked by the derived oxide. However, depending on the ligand used, the morphology of the MOF varies. Ce-1,4-BDC MOF and Ce-1,2,4-BTC MOF are formed as spherical particles. The corresponding oxides, CeO2@C-14 and CeO2@C-124, are also spherical with diameters of 0.7–1.5 μm and 8–15 μm respectively (Fig. 4a–d).
On the other hand, the Ce-1,3,5-BTC MOF shows a randomly orientated bar-shaped morphology with sharp edges. The dimensions of the bars in the derived CeO2@C-135 are 2–10 μm in length and 0.2–0.6 μm in diameter (Fig. 4e and f). In the case of Ce-1,2,4,5-BTEC MOF and CeO2@C-1245, the morphology can be described as twisted bundles of thick plate shaped particles with uneven edges (Fig. 4g and h). The variation in the morphology of different MOFs results from the varying metal–ligand coordination: the molecular structure of the carboxylic acid ligands and possible metal–ligand coordination for each of the carboxylate precursor are shown in Fig. S5 and S6, ESI.† Based on this, a probable formation mechanism of cerium based MOFs is presented in Scheme 2. Initially, heterogeneous nucleation and stepwise growth of metal–ligand networks occur under solvothermal conditions.71 BTC, BTEC or BDC ligands bridge with the Ce3+ ions through the oxygen atoms of the -COO- in the aromatic ring to form 1D ribbon-like molecular motifs due to the steric alignment of the carboxylate groups having flexible coordination. Then, the next ligand molecule coordinates with this Ce containing as-formed structure along the same direction. This coordination-induced assembly process went on until the weakening of reactants in the solution resulting in the formation of well-dispersed nanorods. Here, the organic ligands not only act as the reactant but also may act as the possible crystal face inhibitor in the formation process of nanorods. This phenomenon supports the formation of oriented nucleation leading to anisotropic growth of the nanorods. The anisotropic growth process could also be guided by the polar solvent: the final products, being insoluble in the polar solvents, would have minimum involvement with the solvent, which actually helps in 1D self-assembly through molecular stacking interactions.72 Interpenetration of these metal–ligand networks leads to the formation of a 3D framework which in turn, reorganise themselves into bar shaped (Ce-1,4-BDC & Ce-1,3,5-BTC) or plate shaped (Ce-1,2,4-BTC & Ce-1,3,4,5-BTEC) morphologies with different dimensions depending upon the organic linkers used. Similar to the present observations, Rieter et al. have reported that for Gd MOFs (Gd-1,4-BDC and Gd-1,2,4-BTC), the usual tendency of the 1,4-BDC ligand is to form nanorods/nanobars whereas 1,2,4-BTC ligands lead to nanoplates as the basic morphology.66 In the present study, PVP plays a dual role as a stabilising agent during the metal–ligand co-ordination as well as a structure-directing reagent by regulating the subsequent growth and assembly of the unit structures (bars and plates) through the bundle formation.73 In the case of Ce-1,4-BDC, due to the shorter length of the nanobars, the assembled bundles take a spherical shape. Similarly, due to the smaller dimensions, the nanoplates of Ce-1,2,4-BTC are also assembled in a pseudo-spherical morphology (Scheme 2 and Fig. S7a, ESI†). Jin et al. showed that in the absence of PVP, 1,4-H2BDC bridged Gd(1,4-BDC)1.5(H2O)2 forms a spindle-like shape instantly before miming the well-dispersed straw-like or bar-shaped morphology.73 On the other hand, the structural units of Ce-1,3,5-BTC and Ce-1,2,4,5-BTEC assemble themselves by their edges due to their relatively bigger dimension of the macrobars and macroplates (Scheme 2, Fig. S7b and c, ESI†).
Scheme 2 Formation mechanism of Ce-1,4-BDC, Ce-1,2,4-BTC, Ce-1,3,5-BTC and Ce-1,2,4,5-BTEC MOF precursors. |
TEM, HRTEM and SAED patterns of the MOF derived CeO2@C are shown in Fig. 5. The TEM image of CeO2@C-14 reveals a double shell spherical architecture (Fig. 5a) whereas, CeO2@C-124 shows a pseudo-spherical morphology with relatively bigger geometrical dimensions (Fig. 5b). On the other hand, bar shaped and plate-like morphologies have been observed for CeO2@C-135 and CeO2@C-1245 respectively (Fig. 5c and d), corroborating the observations in FESEM. The appearance of clear lattice fringes in HRTEM (Fig. 5e–h) indicates good crystallinity of the MOF-derived CeO2@C; the observed spacing of 0.27 and 0.32 nm correspond to the (200) and (111) planes of cubic CeO2. Further, the observation of spots arranged in circular rings in the selected area diffraction (SAED) patterns confirms the polycrystalline nature of the synthesized CeO2@C (Fig. 5i–l). N2 adsorption–desorption isotherms and corresponding pore size distributions of the MOF-derived oxides are shown in Fig. S8, ESI.† For all the cases, the isotherms are convex in nature up to P/P0 = 1.0, which is typical for type III isotherms. A discrete hysteresis loop starting from P/P0 = 0.45 reveals a porous nature and indicates the presence of interparticle as well as structural pores. The BET surface area values are calculated to be 159, 63, 73 and 39 m2 g−1 respectively for CeO2@C-14, CeO2@C-124, CeO2@C-135 and CeO2@C-1245.
Fig. 5 TEM, HRTEM and SAED patterns of (a–c) CeO2@C-14, (d–f) CeO2@C-124, (g–i) CeO2@C-135 and (j–l) CeO2@C-1245. |
Fig. 6 Cyclic voltammetry (CV) plots of (a) CeO2@C-14, (b) CeO2@C-124, (c) CeO2@C-135 and (d) CeO2@C-1245. |
The CVs were recorded for three consecutive cycles at a scan rate of 0.1 mV s−1 in the potential window of 0.01–3.0 V vs. Li/Li+. The nature of CVs in all cases is similar: a broad cathodic peak at ∼0.7 V and two broad anodic peaks centered around 0.15–0.21 V and around 1.23–1.26 V. The peak at ∼0.7 V followed by a sloppy nature up to 0.01 V can be attributed to the reductive transformation of CeO2 to Ce2O3, lithiation in the carbon matrix and also to the formation of an SEI layer in the first scan.42 It is observed that the extent of the sloppy region to 0.01 V varies with the amount of embedded carbon present in the material: being a minimum for CeO2@C-135 (7.9 wt% of C, Fig. S2, ESI†) and maximum for CeO2@C-1245 (28 wt% of C, Fig. S2, ESI†). Therefore, the first anodic peak at ∼0.15–0.21 can be assigned to the de-lithiation from the carbon matrix. On the other hand, the prominent second anodic peak at ∼1.23–1.26 V occurs due to the oxidation reaction, Ce2O3 → CeO2. To confirm the conversion reaction, ex situ XPS studies were carried out on the CeO2@C-135 electrode after discharging (0.01 V). Fig. 7a shows the survey scan indicating the presence of C and O along with Ce. As shown in Fig. 7b, a number of peaks could be observed in the high resolution Ce 3d spectrum. Ce 3d core-level photoemission is very sensitive to the local valence revealing the oxidation state. The Ce 3d region has well separated spin–orbit components and each spin–orbit component is further split by multiplet splitting. The six characteristic components of CeO2 are clearly visible: three Ce 3d3/2 peaks at 917.1, 907.8 & 901.4 and three Ce 3d5/2 peaks at 898.8, 889.0 & 882.9 eV. On the other hand, the characteristic components of Ce2O3 could also be clearly discerned: two Ce 3d3/2 peaks at 904.2 & 900.0 and two Ce 3d5/2 peaks at 885.9 & 881.5 eV respectively. The presence of a peak at 917.1 eV indicates the existence of Ce4+. On the other hand, appearance of a shoulder at 881.5 eV is a signature of Ce3+. These results indicate the co-existence of Ce3+ and Ce4+ suggesting transformation of CeO2 to Ce2O3 during discharge. The presence of Ce4+ could be due to incomplete reduction during the first discharge. Further, the core-level O 1s spectrum clearly suggests the transformation of CeO2 into Ce2O3 by showing a shift of the O 1s level from 529.8 eV (CeO2) to a higher binding energy of 532.1 eV (Ce2O3) (Fig. 7b).74–78 These results are also in accordance with the observations of Su et al.42 where in situ TEM studies reveal a reversible transformation between fluorite CeO2 and cubic Ce2O3 during the electrochemical process. Thus, the electrochemical reactions can be represented as
2CeO2 + 2Li+ + 2e ↔ Ce2O3 + Li2O | (1) |
Fig. 7 XPS spectra of CeO2@C-135 after discharging up to 0.01 V: (a) survey scan (b) Ce 3d and (c) O 1s. |
Fig. 8 represents the galvanostatic discharge–charge (GCD) profiles for the initial formation cycle for CeO2@C in the potential window of 0.01–3.0 V at a current density of 0.05 mA cm−2. Initial discharge/charge capacities of 1344/596 mA h g−1, 1032/546 mA h g−1, 517/301 mA h g−1 and 850/470 mA h g−1 have been found for CeO2@C-14, CeO2@C-124, CeO2@C-135 and CeO2@C-1245, respectively with first cycle coulombic efficiency (%) [(charge capacity/discharge capacity) × 100] of 44.3, 52.9, 58.2 and 55.3%. The somewhat low coulombic efficiency is common for transition metal oxide anodes and can be accounted for side reactions with the non-aqueous electrolyte as well as the formation of a solid electrolyte interface (SEI) layer on the electroactive materials.79,80 Interestingly, the bar-shaped or plate-like morphology of the particles (CeO2@C-135 and CeO2@C-1245) leads to a marginally better coulombic efficiency than the spherical morphology (CeO2@C-14, CeO2@C-124). On the other hand, the obtained charge capacities are higher for the spherically shaped particles than for bar shaped or plate-like particles. In order to check the cycling performance at different current densities, the CeO2@C cells have been tested for 100 continuous cycles at current densities between 0.05 and 0.5 mA cm−2 and the results are shown in Fig. 9a–d.
Fig. 9 Electrochemical cycling performance of (a) CeO2@C-14, (b) CeO2@C-124, (c) CeO2@C-135 and (d) CeO2@C-1245 at current densities of 0.05, 0.1 and 0.5 (CeO2@C-14) mA cm−1. |
In all cases, there is an increasing tendency for the first 5–10 cycles. This behavior can be related to the wetting and gradual activation of the electrode till equilibration is reached and a steady SEI is formed.79,80 Except for CeO2@C-135, all samples show a reversible capacity exceeding 600 mA h g−1 which is retained for 100 cycles (Fig. 9). The low specific capacity of CeO2@C-135 might have resulted from a low embedded carbon content (7.9 wt%) impeding the charge transfer process. At the 100th cycle, CeO2@C-14 showed specific charge capacity values of 674 mA h g−1, 590 mA h g−1 and 453 mA h g−1 at current densities of 0.05, 0.1 and 0.5 mA cm−2 respectively, with nearly 99% coulombic efficiency (Fig. 9a). These values are higher than the previously reported values for CeO2 spheres,43,44,47,48 brick-like CeO2,49 rhombus shaped CeO2,50 or CeO2–graphene composites,45,46 demonstrating the superiority of the MOF derived route.43–50 A detailed performance comparison is given in Table S3, ESI.† In order to investigate the contribution of Super P, if any, to the overall capacity electrochemical characterization has been conducted with a Super P carbon electrode. It is found that Super P carbon is not a good electroactive material and contributes only 4.0% and 2.1% to the overall capacity at current densities of 0.05 mA cm−2 and 0.5 mA cm−2 (Fig. S9, ESI†).
The rate performance and capacity restoration capability of CeO2@C have been studied at different current densities ranging from 0.05 to 0.5 mA cm−2 and the results are shown in Fig. 10a–d. All the CeO2@C electrodes show a similar trend but with different achievable capacity values. At a slow current density of 0.05 mA cm−2, after the initial increasing trend, a somewhat steady capacity of ∼715 mA h g−1 is observed at the end of the 10th cycle in the case of CeO2@C-14 (Fig. 10a). With stepwise increase in current density, the capacity decreases. But, even at a relatively high current density of 0.5 mA cm−2, a reversible capacity of ∼413 mA h g−1 is obtained at the end of the 40th cycle. Further, the cells show good capacity restoration ability. When the current density is reverted to 0.05 mA cm−2 (cycle 60–100), a capacity value of ∼670 mA h g−1 is observed; which means that ∼94% of the capacity is restored with respect to the steady capacity obtained at the 10th cycle (1–10 cycles). In the case of CeO2@C-124, ∼109% capacity restoration has been observed after the 100th cycle with respect to the value observed at the 10th cycle (Fig. 10b). CeO2@C-135 (Fig. 10c) and CeO2@C-1245 (Fig. 10d) also showed a similar rate performance and capacity restoration ability though with lower capacity values. These results further support the superior performance of MOF-derived spherically shaped particles compared to bar shaped or plate-like particles or spherically shaped CeO2 particles prepared by other routes.43,44,47–50 Inert calcination of cerium MOFs produces carbon embedded CeO2 where the crystallite size is found to be in the range of 3.3–6.2 nm. The carbon matrix, apart from providing charge transportation pathways, would also buffer the stress/strain due to volume expansion. Thus, the reported volumetric change of ∼17% or a radial expansion of ∼5.4% found in ceria nanoparticle having the diameter of 10–40 nm would be mitigated to a great extent.42Fig. 11a–d show the electrochemical impedance spectra (EIS) for CeO2@C at three different cycling stages. The nature of the impedance plots is typical of conventional LIBs, a depressed semicircle followed by a sharp straight line and can be fitted to Randles like equivalent circuits (Fig. S10, ESI†). It should be noted that the analysis of the electrochemical properties of active materials based on two-electrode impedance measurements could lead to incorrect conclusions due the influence of the lithium metal counter electrode on the full cell impedance.81 In the present case, the EIS data can be regarded as close to the actual impedance behaviour on the assumption that the electrochemical activity of lithium metal is extremely high compared to that of the active materials and, hence, the effect of the impedance of lithium metal on the full cell impedance may be considered as negligible.82–85
Fig. 10 Electrochemical rate performances of (a) CeO2@C-14, (b) CeO2@C-124, (c) CeO2@C-135 and (d) CeO2@C-1245 at different current densities ranging 0.05–0.5 mA cm−1. |
Fig. 11 Electrochemical impedance spectra (EIS) of (a) CeO2@C-14, (b) CeO2@C-124, (c) CeO2@C-135 and (d) CeO2@C-1245 at different cycling stages. |
The fitted impedance data are given in Table 1. In the high frequency region, intercept on the real axis (Z′) denotes the solution resistance (Rs) consisting of intrinsic resistance of the electrode material, ionic resistance and contact resistance between the electrode and current collector. All the CeO2@C cells show a low Rs value of ∼3 Ω except for CeO2@C-135 (16 Ω).
Cycling intervals | R s (ohm) | R SEI (ohm) | R ct (ohm) | C SEI (μF) | Q (CPE) (μF) | Diffusion admittance, Y0 (mMho) [W] | C (mF) |
---|---|---|---|---|---|---|---|
CeO2@C-14 | |||||||
As assembled | 2.6 | — | 111 | — | 4.7 | 5.9 | 19.4 |
After formation cycle | 4.4 | 2.6 | 107 | 5.2 | 6.4 | 7.1 | — |
After 100th cycle | 7.7 | 4.8 | 137 | 3.6 | 5.1 | 9.2 | — |
CeO 2 @C-124 | |||||||
As assembled | 2.8 | — | 137 | — | 4.9 | 4.4 | 11.8 |
After formation cycle | 5.3 | 3.8 | 142 | 3.9 | 5.7 | 6.5 | — |
After 100th cycle | 13.4 | 5.2 | 91.6 | 2.2 | 3.3 | 7.5 | — |
CeO 2 @C-135 | |||||||
As assembled | 16 | — | 151 | — | 4.0 | 7.4 | 6.9 |
After formation cycle | 22.3 | 4.2 | 101 | 2.2 | 5.5 | 7.8 | — |
After 100th cycle | 23 | 6.6 | 108 | 1.7 | 4.7 | 9.6 | — |
CeO 2 @C-1245 | |||||||
As assembled | 3.1 | — | 132 | — | 4.9 | 6.0 | 12.8 |
After formation cycle | 3.4 | 3.6 | 120 | 3.7 | 5.2 | 7.4 | — |
After 100th cycle | 11.5 | 4.8 | 65 | 3.3 | 4.3 | 8.4 | — |
Also, Rs does not increase significantly after 100 cycles indicating good stability of the electrode. The diameter of the depressed semicircle, formed up to the mid-frequency region, represents the charge transfer resistance (Rct). The Rct value is found to be the lowest (111 Ω) for CeO2@C-14 (containing 26.8 wt% embedded carbon), while it is the highest (151 Ω) for CeO2@C-135 (containing 7.9 wt% embedded carbon). No drastic increase in Rct is observed with cycling explaining the steady cycling performance of the CeO2@C cells. The sharp straight line from the mid-frequency to the low frequency region can be linked to the Warburg resistance (W) which is related to the ionic diffusion in the electrolyte/electrode interface. In the as-assembled state, the Warburg line shows a deviation from the usual intercept angle of 45° and tends to become parallel to the Y-axis. This may have resulted from the capacitive nature of these MOF derived materials forming Helmholtz double layers in the as-assembled state. However, after the formation cycle, the Warburg line approaches 45°. Also, from the diffusion admittance (Y0) values, which is inversely related to W, it is found that the diffusion process becomes faster with cycling. In order to examine the change in the morphology upon cycling, an ex situ FESEM study was undertaken.
It is observed that the particles are not agglomerated and retains the overall sizes and shape, but the inner structure underwent a considerable change forming a cage-like structure with interconnected thin plates (Fig. 12). This may have resulted through interpenetration and consequent assembly of the constituent nanobars. During the formation of Ce-1,4-BDC MOF, the spherical morphology was formed due to the self-assembly of nanobars in the presence of PVP, a morphology directing agent. This basic spherical morphology of Ce-1,4-BDC MOF is inherited by CeO2@C-14. However, during repeated lithiation/delithiation processes under a potential gradient; the nanobars may reorient themselves in a plate-like assembly to allow the smooth passage of ionic species. This could lead to the gradual roughening of the surface ultimately leading creation of open tunnels within the spherical shape after 100 consecutive cycles. Such kind of open tunnel structure would also explain the increased diffusion admittance with cycling as observed in the EIS study.
Footnote |
† Electronic supplementary information (ESI) available: Synthesis scheme, FTIR spectra of MOFs, XRD analysis data, TGA, Raman spectra, Raman spectra analysis data, molecular structures of ligands, possible metal–ligand co-ordination in MOFs, additional FESEM micrographs of MOFs, N2 adsorption–desorption isotherm and pore size distribution, electrochemical study of Super P carbon, equivalent circuit models, Comparison of electrochemical results with the literature. See DOI: 10.1039/c6se00026f |
This journal is © The Royal Society of Chemistry 2017 |