Flexible long-chain-linker constructed Ni-based metal-organic frameworks with 1D helical channel and their pseudo-capacitor behavior studies

Journal of Power Sources 377 (2018) 44–51

Contents lists available at ScienceDirect

Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour

Flexible long-chain-linker constructed Ni-based metal-organic frameworks with 1D helical channel and their pseudo-capacitor behavior studies

T

Kuaibing Wanga,c,1, Zikai Wangb,1, Xin Wanga, Xueqin Zhoua, Yuehong Taoa, Hua Wua,c,∗ a

Jiangsu Key Laboratory of Pesticide Sciences, Department of Chemistry, College of Sciences, Nanjing Agricultural University, Nanjing 210095, PR China College of Resources and Environmental Sciences, Nanjing Agricultural University, Nanjing 210095, PR China c State Key Laboratory of Coordination Chemistry, Coordination Chemistry Institute, Nanjing University, Nanjing 210093, PR China b

H I G H L I G H T S 3D Ni-MOFs with 1D helical channels exhibit pseudo-capacitor behaviors. • Two 3D Ni-MOFs directly in electronic energy storage field is scarce to date. • Employing • Excellent rate capability, energy deliverable ability and cycling stability.

A R T I C L E I N F O

A B S T R A C T

Keywords: Ni-MOFs Flexible long-chain-linker Helical channel Pseudo-capacitor

Two novel and isostructural Ni-based MOFs with topological symbol of 422·54·62, namely [Ni2(TATB)2(L)2(H2O)], have successfully synthesized, where L is the flexibly N-donor bid (1,10-bisimidazoledecane) or btd (1,10-bistriazoledecane) linker and TATB is the deprotonation mode from 4,4′,4″-s-triazine-2,4,6triyl-tribenzoic acid (H3TATB). Two types of left- and right-handed helical channels with mean diameter of 11 Å results in large void space in 3D network. When directly use as electrode materials, the as-synthesized Ni-MOFs single-crystal electrodes behave as pseudo-capacitor and deliver high gravimetric capacitance with superior energy deliverable ability and cycling stability. For example, the maximum gravimetric capacitance is 705 F g−1 with the energy density of 29.6 Wh kg−1 at a current density of 1 A g−1. Even after 5000 continuous cycles, the capacitance retention maintains at 92.1%. The good electrochemical performance should be ascribed to the 1D helical channels facilitating the diffusion of OH−. Furthermore, the low bulk solution (0.46 and 0.50 Ω) and charge-transfer resistances accelerate the contact between OH− and active species in the electrode, and consequently result in efficiency Faradaic reaction. This work opens a new way for the directly application of 3D topological MOFs single-crystal with novel interior structures especially porous and channel-like architectures in electronic energy storage field.

1. Introduction Tremendous research effort has been performed aiming at increasing the energy density of supercapacitors (SCs) without sacrificing their high-power capability to be close to or even beyond that of lithium batteries as well as reducing fabrication costs in the past few years [1–6]. For achieving this purpose, numerous electrode materials have been prepared, such as carbon-based materials including graphene, mesoporous carbon and activated carbon, metal oxides, conducting polymers and their composites [7–11]. These SCs electrode materials deliver high surface area, large specific capacitance and excellent stability [10,11]. However, carbon-based materials exhibit low

∗

1

capacitance, metal oxides and conducting polymers have poor conductivity and lack of stability. Therefore, exploring the new electrodes with porous motifs, high surface area and structural stability remains challenging and is still an optimal choice at present. Metal–organic frameworks (MOFs) are a class of porous materials first defined by Yaghi and co-workers [12], which have recently attracted extensive research interests as potential electrode candidates for SCs either served as soft-template for synthesis of nanostructured carbon, metal oxides and their hierarchical composites or in a direct application as a new type of electrode material [13–20]. Especially the later one has attracted enormous attentions due to their high surface areas, tunable architectures, permanent porosity and the growing

Corresponding author. Jiangsu Key Laboratory of Pesticide Sciences, Department of Chemistry, College of Sciences, Nanjing Agricultural University, Nanjing 210095, PR China. E-mail address: [email protected] (H. Wu). Kuaibing Wang and Zikai Wang, are equal to this work.

https://doi.org/10.1016/j.jpowsour.2017.11.087 Received 13 September 2017; Received in revised form 20 November 2017; Accepted 26 November 2017 0378-7753/ © 2017 Elsevier B.V. All rights reserved.

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

0.15 mmol) was used instead of bid in SC-1. Green crystals of SC-2 were collected in a yield of 82%. C76H68N18Ni2O13: FT-IR (KBr pellet, cm−1): 3601w, 3134 m, 2924s, 2852 m, 1693 m, 1596s, 1545s,1428 m,1377s, 1358s, 1292 m, 1235 m, 1107 m, 1016 m, 944w, 864w, 810 m, 774s, 699 m, 631w, 504 m. Electrode preparation. The as-synthesized MOFs crystals were directly used as active materials. The working electrodes were prepared as follows. The mixture containing 75 wt% active materials (SC-1 or SC2 single crystals), 15 wt% acetylene black and 10 wt% polytetrafluoroethylene (PTFE, 1 μm particle size purchased from SigmaAldrich) was well mixed and strongly stirred in isopropanol system, and then was deposited 1 cm2 on nickel foam bar (after treated by diluted hydrochloric acid, size: 1 cm × 5 cm), and the typical mass load of electrode materials ranged in 3.5–5.5 mg after pressed by Manual Rolling Press (MR-100A, MTI Corp.).

number (more than 20000 MOFs have been created [8]). For instance, Yang and co-workers prepared 2D Ni-MOFs delivering good rate capability, large capacitance (1127 F g−1 at 0.5 A g−1) and cycling stability (90% retention over 3000 continuous cycles) [13]. Besides, this group also fabricated various Zn-doped Ni-MOFs and investigated the SCs performance [14]. Based on the same strategy, to solve the poor conductivity of MOFs, Banerjee used rGO as conductive matrixes supporting Ni-doped MOFs that exhibited an energy density of 37.8 Wh kg−1 at 226.7 W kg−1 [15]. Gao and co-workers fabricated a asymmetric SCs including Ni-MOFs/CNTs positive electrode and rGO/C3N4 negative electrode that displayed a maximum energy density of 26.8 Wh kg−1 at 2000 W kg−1 [16]. Except of the poor conductivity, the instability for most MOFs electrode materials during charge-discharge process is also extensively considered a limitation for their application in SCs. The strategy for solving this problem is to fabricate kinetically stable MOFs itself through ligand functionalization [17,18]. For example, Qu and coworkers successfully synthesized novel, nickel-based, pillared MOFs through choosing functionalized bridged ligands with maximum specific capacitance of 552 F g−1 at 1 A g−1 and superior cycle stability (98% after 16000 cycles) [17]. Recently, Sheberla et al. constructed an Nickel-based MOFs (Ni3(HITP)2) with high electrical conductivity that behaved as electric double-layer capacitors and showed a working potential window of ≈1.0 V [19]. In these recent reports, many testing electrodes materials are nano/micro-structured MOFs, which is considered as an effective way to scale-down the size and thus resulted in decreasing the diffusion distance of electrolyte and enhancing the electrochemical performance [20]. However, these nano/micro-scaled MOFs powders are polycrystalline, although the structure is isostructural with the single-crystals. Moreover, the morphology and size are tunable by adjusting the concentration of reactant, solvent, reacting temperature and so forth [21–25]. Thus, nano/micro-sized MOFs itself may turn into an erratic factor in long-cycle test. In this regard, herein, two novel and isostructural 3D Ni-based MOFs with interior 1D helical channels have been successfully fabricated through choosing rigidly tripodal ligand H3TATB and flexibly bridged linker, and directly applied the as-synthesized single-crystals as electrode materials for SCs. The single-crystals are air-stable and insoluble in water and common organic solvents, and the report verifying them as the electrode materials in SCs is scarce to date. They both deliver high specific capacitance and energy density with good rate capability, excellent energy deliverable ability and cycling stability. After 5000 continuous cycles, the capacitance retention for the assynthesized MOFs is 92.1% and 88.7% separately. The maximum specific capacitance for Ni-MOFs crystals is 705 F g−1 at current density of 1 A g−1 and exhibit an energy density of 29.6 Wh kg−1 at 366 W kg−1.

2.2. Characterization Diffraction intensities for SC-1 and SC-2 were recorded on a Bruker SMART Apex CCD diffractometer with graphite-monochromated Mo Kα radiation (λ = 0.71073 A) at 293 K. The structures were solved with the direct method of SHELXS-97 [27] and refined with full-matrix leastsquares techniques using the SHELXL-97 program within WINGX [28]. The Squeeze option of Platon [29] was used to eliminate the contribution of disordered guest molecules for SC-1 to the reflection intensities. The formulas of SC-1 and SC-2 were calculated by considering the results of thermogravimetric analysis, element analysis and singlecrystal structures, respectively. The high R factor for SC-1 may be related to the disordered guests. Crystal data for SC-1: orthorhombic, a = 18.062(5) Å, b = 35.561(5) Å, c = 55.601(5) Å, α = β = γ = 90°, V = 35713(12) Å3, space group Fddd, Z = 8, 74250 reflections measured, 8014 independent reflections (Rint = 0.1090). The final R1 value was 0.0904 (I > 2σ(I)). The final wR(F2) value was 0.2467 (I > 2σ(I)). The final wR(F2) value was 0.2758 (all data). The goodness of fit on F2 was 0.785. Crystal data for SC-2: orthorhombic, a = 18.238(5) Å, b = 33.694(5) Å, c = 56.174(5), α = β = γ = 90°, V = 34520(11) Å3, space group Fddd, Z = 8, 63667 reflections measured, 7759 independent reflections (Rint = 0.1035). The final R1 value was 0.0684 (I > 2σ(I)). The final wR(F2) value was 0.1659 (I > 2σ(I)). The final wR(F2) value was 0.1978 (all data). The goodness of fit on F2 was 1.102. The hydrogen atoms attached to carbons were generated geometrically. The hydrogen atom positions were fixed geometrically at calculated distances and allowed to ride on the parent atoms. The disordered atoms of compounds SC-1 (C36) and SC-2 (C38) was refined using C atoms split over two sites. The H atoms of the disordered C atoms were not included in the model. Non-hydrogen atoms were refined with anisotropic temperature parameters except C36, C39 and C40 in compound SC-1. Notably, selected bond lengths and angles are listed in Supplementary Tables S1–S2 and the CCDC number for SC-1 and SC-2 is 1567549 and 1567550 separately.

2. Experimental section 2.1. Preparation of SC-1 and SC-2

2.3. Methods and measurements

The bid and btd ligands were synthesized by procedures reported earlier [26]. Other reagents of analytical grade and solvents employed were commercially available and used as received without further purification. Preparation of [Ni2(TATB)2(bid)2(H2O)]·2H2O crystals (SC-1). A mixture of bid (0.041 g, 0.15 mmol), NiSO4·6H2O (0.039 g, 0.15 mmol), H3TATB (0.044 g, 0.1 mmol), DMF (5 mL) and water (10 mL) was placed in a Teflon reactor (20 mL) and heated at 110 °C for 3 days. After the mixture had been cooled to room temperature at a rate of 10 °C·h−1, green crystals of SC-1 were obtained with a yield of 86%. C76H68N18Ni2O13, FT-IR (KBr pellet, cm−1): 3396w, 2297s, 2855 m, 1616s, 1571s, 1503s, 1420s, 1402s, 1347s, 1284w, 1136s,1101w, 1017s, 993 m, 880w, 819s, 771s, 697 m, 504 m. Preparation of Ni2(TATB)2(btd)2(H2O) crystals (SC-2). The preparation of SC-2 was similar to that of SC-1 except that the btd (0.041 g,

FT-IR spectra were recorded from KBr pellets in range 4000400 cm−1 on a Nicolet 380FT-IR spectrometer. Thermogravimetric analyses (TGA) were carried out under atmosphere on a Beijing Hengjiu HTG-1 instrument with a heating rate of 10 °C min−1. The powder Xray diffraction (PXRD) data of the samples was collected on a Rigaku Dmax 2000 X-ray diffractometer with graphite monochromatized Cu Kα radiation (λ = 0.154 nm) and 2 h ranging from 5 to 50°. The electrochemical measurements were carried out by an electrochemical analyzer system, CHI660E (Chenhua Instrument, Shanghai, China) in a three-compartment cell with a platinum wire counter electrode, an Hg/ HgO reference electrode and an above-mentioned working electrode. The electrolyte was a 6.0 M KOH aqueous solution and electrochemical impedance spectroscopy (EIS) measurements of as-synthesized samples 45

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Because the right- and left-handed double homochiral helices arealternately arranged, the whole chirality of the structure is therefore racemic. The diameter of the cyclinder helical channels are approximately 10.71 Å measured by the nearest radial direction Ni atoms along a axis. The Ni atoms in helical channels are further bridged by TATB ligands with one carboxylate group being deprotonation to generate a novel three-dimensional coordination framework (Fig. 1b). In SC-2, each Ni(H2O)Ni cluster links four btd ligands and four TATB ligands, and each btd and TATB ligand links two Ni(H2O)Ni clusters, respectively. Topologically, if the TATB and btd ligands are considered as linkers, the Ni(H2O)Ni cluster is considered as a node, respectively, the three-dimensional framework can be classified as a single 8-connected with Schläfli symbol of 422·54·62 (Fig. S2b). Field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM) were also used to characterize the micro/nano-structure and morphology of the Co-based MOFs crystalline material. Figs. S3a and S3b is SEM image for the crystalline samples after grinding, suggesting that the morphologies for SC-1 and SC-2 both are disordered rodlike structures. It is in good agreement with the isostructural structures nature obtained from single-crystal X-ray analysis. The distinct place is that the width measured from TEM images for SC-1 ranges in 50–160 nm, while 200-500 nm range for SC-2 sample as depicted in Fig. S4. Complexes SC-1 and SC-2 are air-stable and insoluble in water and common organic solvents. The thermogravimetric analysis (TGA) was performed in atmosphere on crystalline samples of complexes SC-1 and SC-2, and the TG curves are shown in Fig. S5 in the Supplementary Information. It is noteworthy that the crystalline samples are isostructural with the single-crystals confirmed by XRD results (Fig. S6). The TG curve of SC-1 indicates the coordinating and lattice water loss (3.7%) occurs in the temperature range 145–330 °C, then organic ligands departure gradually, and the framework begins to collapse at 350 °C. For SC-2, the solvate loss (1.2%) occurs in the temperature range 110–125 °C, which corresponds to the loss of the coordinated water molecules. The framework of SC-2 remains to 320 °C begins to collapse.

were conducted at open circuitvoltage in the frequency range of 100 kHz to 10 mHz. 3. Results and discussion 3.1. Description of SC-1 and SC-2 In previous reports, the TATB (the anion of 4,4′,4″-s-triazine-2,4,6triyl-tribenzoicacid, H3TATB) ligand has been used to construct porous and cage metal-organic frameworks to investgate their potential applications in gas storage and adsorptive separation [30–32]. In this work, by introducing two long flexible N-donor ligands, 1,10-bisimidazoledecane (bid) and 1,10-bistriazoledecane (btd), into TATB-based metal-organic frameworks and thus should produce new structural types in MOF studies. The btd or bid secondary ligand contains multy CC bonds which can rotate freely than TATB. Therefore, the expected structure type is very difference from previous reports, which can further lead to diverse properties. The compounds SC-1 and SC-2 are isostructural, only compound SC2 is described here in detail (the structure description for SC-1 see Supplementary information and Fig. S1). The structure of SC-2 is an unusual 3D framework displaying 1D left- and right-helical channels consisted of octahedral metal centers connecting by the anti- and cisconfigurations ligands. Single crystal X-ray analysis reveals that the asymmetric unit of SC-2 contains one TATB ligand, two half a btd ligand, one NiII ion and half a water molecule (Fig. S2a). The unique Ni1 ion, which is at the center of a slightly distorted octahedron coordination geometry, is six-coordinated by two N atoms from one anticonfiguration and one cis-configuration btd ligands, and four O atoms from three TATB ligands and one water molecule. In SC-2, there are two types of left- and right-handed helical channels in the structure, which results in a sufficient large void space inside the network. The total void value of the channels is estimated (by Platon [29]) to be 6073.9 Å3, approximately 17.6% of the total crystal volume, 34520.0 Å3. In order to realize the nature of this complicated architecture, the TATB ligands can be neglected from the structure. A better insight into the conformation of the helical channels, the arrangement of the btd ligands with cis- and trans-conformationis highlighted in Fig. 1a. Two Ni atoms are linked by one water molecule to form a Ni(H2O)Ni metal oxygen cluster, and the clusters are connected by the trans-configuration btd ligand to generate an infinite one-dimensional chain with the Ni⋯Ni distance of ca. 14.99 Å. Further, the two Ni atoms from two chains are linked by cis-configuration btd ligands to give rise to two types of cyclinder helical channels with opposite chirality. Each channel consists of left- or right-handed double homochiral helical chains.

3.2. Electrochemical performance Except of the stability, in order to testify the applicability of assynthesized single crystals SC-1 and SC-2 (directly used as active materials) as electrochemical capacitor electrodes, their electrochemical performance was firstly studied by cyclic voltammetry (CV) using the classical three electrode method in a 6.0 M KOH electrolyte. Fig. 2a shows cyclic voltammetry motifs of SC-1 at varying scan rates within a potential window of 0.6 V. According to the observed oxidation and

Fig. 1. (a) The Schematic representation of left and right helical channels in different directions. (b) The three-dimensional framework with left- and right-handed helical channels.

46

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Fig. 2. CV curves of as-synthesized SC-1 (a) and SC-2 (b) electrodes at different scan rates separately. (c) CV curves for SC-1 and SC-2 at a scan rate of 10 mV s−1. (d) Variation of specific capacitance of SC-1 and SC-2 as a function of scan rates.

reduction peaks, the electrode behaves as a pseudo-capacitor, unlike the recent reports on Ni-based MOFs Ni3(HITP)2 which displayed electrical double layer capacitive behavior [19]. As compared, SC-2 electrode is also investigated as shown in Fig. 2b. The shape of the CV curves exhibits similar trends with SC-1 electrode and is different from the rectangular ones, which is also typical for pseudocapacitive behavior [33]. For SC-1 and SC-2 electrode, the main redox process can be attributed to the conversion between NiII and NiIII species, which is similar with the redox reaction of [Ni3(OH)2(C8H4O4)2(H2O)4]·2H2O [20]. Both of them have a slight shift of the redox peak with the increase of the scan rate which indicates some kinetic irreversibility in the electrochemical system [34,35]. The difference between them is that the peaks corresponding to cathodic process in SC-1 electrode are centered at two places, ca. 0.43 V and 0.49 V (0.49 V for SC-2, Fig. 2c). Obviously, as shown in Fig. 2c, the peak current of SC-1 electrode is slightly higher than that of SC-2 electrode, indicating that SC-1 has better electrochemical reaction activity and higher capacitive performance. The specific capacitances of as-synthesized Ni-MOFs electrodes at different scan rates are calculated and the corresponding relationships are shown in Fig. 2d. The calculated specific capacitance values for SC-1 are 441, 399, 347, 265, 182 F g−1 and 392, 337, 285, 208, 156 F g−1 at the scan rates of 5, 10, 20, 50, and 100 mV s−1, respectively. In pseudocapacitor electrode materials, the dependence of scan rate (v) with voltammetric current (i) depends on whether the capacitance originates from surface redox reactions or bulk diffusion. In general, i∝v represents for surface redox reaction, and i∝v1/2 for semi-infinite bulk diffusion [36]. Two straight lines corresponding to i∝v1/2 are observed as shown in Fig. 3. Therefore, bulk diffusion occurred during the electrochemical reaction for the MOFs electrodes. The linear relationship further indicates that the diffusion of OH− is probably to control the redox reaction occurring in the electrochemical process. The apparent diffusion coefficient (D) of OH− ion at 25 °C is calculated by employing

Fig. 3. Voltammetric current as a function of square root of scan rate of the Ni-MOFs electrodes in 6.0 M KOH electrolytes.

Randles–Sevcik equation [37]:

i = 2.69 × 105 × n3/2 × A ×

D × C0 ×

v

where n is the number of the electrons transferred, A is the surface area of the electrode, D is the diffusion coefficient, v is the scan rate, i is the peak current and C0 is the proton concentration. For samples SC-1 and SC-2, the n, A, C0 and v can be considered the same. Thus, the diffusion coefficient only depends on the slope of i∝v1/2 (Fig. 3). Evidently, SC-1 electrode has the higher diffusion coefficient, indicating a better electrode reaction activity due to the faster ionic transportation. Galvanostatic charge-discharge curves at various current densities within the potential range 0–0.55 V vs. Hg/HgO of the two MOFs electrodes are illustrated in Fig. 4a and b. The shape of the chargedischarge curves is not a standard triangle, but distorted one. For SC-1 or SC-2 electrode, the charging step shows two stages: the first linear 47

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Fig. 4. Charge-discharge curves for SC-1 (a) and SC-2 (b) electrodes at different current density respectively. (c) CP curves for SC-1 and SC-2 at a current density of 4.0A g−1. (d) Variation of specific capacitance of SC-1 and SC-2 as a function of current densities.

discharge time) from the charge-discharge curves with respect to the applied current density. The specific energy of SC-2 electrode decreases from 28.0 to 10.4 Wh kg−1, while the specific power increases from 456 to 2954 W kg−1 as the current density increased from 1.0 to 10.0 A g−1. As to SC-1 electrode, the energy densities have slightly higher than that of SC-2 electrode and the energy value is determined to be 29.6, 22.8, 19.8, 17.8, 16.6, and 15.0 Wh kg−1at the current density of 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 A g−1 separately (Fig. 5a). Interestingly, however, as shown in Fig. 5b, the specific power for SC-1 electrode remains similar values with SC-2 electrode due to the approximate discharge time and similar energy density. Except of the diffusion coefficient reason, the different electrochemical performance between SC-1 and SC-2 electrodes mainly depends on their crystal structure nature. Although compounds SC-1 and SC-2 are isostructural, they are still different from each other, such as the average Ni-O and Ni-N bond lengths, Ni⋯Ni distances separated by H2O, cis- and trans-configuration btd ligands, channel diameters, dihedral angle between imidazole (or triazole in SC-2) rings in the cisand trans-configuration btd ligands, and total void value of the channels (the discrepancies for SC-1 and SC-2 are listed in Table 1), which also can affect the electron transport and the diffusion of OH−, and thus result in variant electrochemical performance. The secondary factor is the little different surface area. The surface area and pore size distribution of SC-1 and SC-2 were investigated by nitrogen adsorptiondesorption techniques. As shown in Fig. S7, the N2 isotherms for assynthesized compounds belong to type IV, and the hysteresis belongs to type H1. In comparison with the SC-2 (3.6 m2 g−1), the BET surface area of the SC-1 is 5.5 m2 g−1, a little higher than that of crystals. The pore size distribution curve (see Fig. S7b and Fig. S7d in SI) can be plotted by the tested data calculated from the BJH method, which suggests the presence of meso-pores. The average pore diameter for SC1 and SC-2 are determined to be 11.6 nm and 14.5 nm separately. Considering the diffusion of OH− is probably to control the redox

one corresponds to the oxidation process of Ni-based MOFs, whereas the second one represents the charging process itself. Furthermore, the discharging curve displays a sharp negatively-sloped segment of potential drop at first due to the internal resistance followed by a slightly decaying second part, related to the pseudo-capacitive behavior of the electrode material. This result is in good agreement with the CV observations. For comparison, the charge-discharge curves at the same constant current density (4.0 A g−1) are displayed in Fig. 4c. The increase in charging time corresponds to the higher capacitance, demonstrating the higher electrochemical performance for SC-1, in keeping with the CV result. Moreover, the energy deliverable efficiency (η/%) information can be calculated as η = Td/Tc × 100 (Td and Tc are the discharging and charging time respectively) from charge-discharge curves. The η is determined to be 89.5% and 89.3% separately, suggesting the excellent energy deliverable ability. Detailed gravimetric capacitances for the two electrodes at various current densities are determined and depicted in Fig. 4d. The gravimetric capacitance values are calculated to be 705, 542, 472, 424, 394, 357 F g−1 and 666, 406, 342, 306, 282, 248 F g−1 at the current density of 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 A g−1 respectively, approximate 51% and 40% capacitance is still maintained after 10 times increase, demonstrating the good rate capability of Ni-based MOFs electrodes. The electrochemical performance is comparable to that of recent reports on Ni-MOFs-based electrodes and the corresponding Ni-MOFs derived materials (details see Table S3) [13–17,19,38–41]. Two important parameters evaluating the performance of an electrochemical capacitor device, that is, the specific energy (Es) density and specific power (Ps) are well represented with histograms and Ragone plots (Fig. 5a-b). The energy density of SC-1 and SC-2 electrode calculated from the equation of Es = 0.5C(ΔV)2 × 1/3.6 (Wh kg−1, C and ΔV are the specific capacitance and potential window separately) and the power density calculated as P = E/Δt (W kg−1, Δt is the 48

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Fig. 5. (a) The histogram of energy density for SC-1 and SC-2 electrodes at different current densities. (b) Energy and power densities vs. current density of the as-synthesized samples.

electrode, the deviations from the semicircle indicate the existence of dispersion effect. The EIS data can be well fitted by an equivalent circuit (Fig. S10). Rs represents the series resistance including the bulk solution resistance, the contact resistance between the current collector and the active material and so forth, Rct is the charge-transfer resistance and CPE is a constant phase element to account for the double layer capacitance at the electrode/electrolyte interface (Because pseudocapacitance occurs on a large electrode surface, it always takes place along-side double-layer capacitance [3]), while CP corresponds to a pseudocapacitive element from the redox process of Ni-MOFs. According to the point intersecting with the real impedance (Z′) axis at the high frequency region (the inset section in Fig. 7), the Rs value for SC-1 and SC-2 electrode is 0.46 and 0.50 Ω, respectively. Relating the obtained equivalent circuit with the Nyquist plot, the parallel Rct‒CPE configuration accounts for the second semicircle feature and an Rct of 2.0 and 3.5 Ω is deduced and simulated by ZsimpWin software for SC-1 and SC-2 electrode separately. These observations manifest that SC-1 electrode with lower Rct and Rs has more affinity towards the Faradaic reactions. Cycling stability under extreme load of the electrodes is crucial for practical applications of supercapacitors, an endurance test was conducted using CV measurement to examine the cycle life for the Ni-MOFs electrodes, as shown in Fig. 8a and c. The specific capacitance for SC-1 electrode even grow a little larger during the first 750 cycles, which might be due to an electrochemical activation process of electrode that electrolytes in general require a period of time to penetrate the entirely interior space of an active electrode material. Then the capacitance decreased slightly over the next 4250 cycles (Fig. 8b). For SC-2 electrode, the capacity attenuation is very severe in the first 400 cycles and then maintain at an approximate value over the rest of 4600 cycles (Fig. 8d). After 5000 continuous cycles, the capacitance retention for SC-1 and SC-2 is 92.1% and 88.7% respectively, indicating the good cycle stabilities for the as-synthesized Ni-MOFs crystals electrodes. The electrochemical measurement indicates that both of the as-synthesized Ni-MOFs single-crystal electrodes deliver high gravimetric capacitance with good rate capability, excellent energy deliverable ability and cycling stability. This should be ascribed to the following factors. First, the three-dimensional and large frameworks of Ni-MOFs constructed by many C-C bonds create the structural stability and consequently results in the cycling stability. Second and most importantly, two types of 1D helical channel with mean diameter of 11 Å facilitate the diffusion of OH−, which controls the redox reaction occurring in the electrochemical process and is the main origin of the high capacitance according to the i∝v1/2 fitting results. Third, low bulk solution and chargetransfer resistances accelerate the contact between OH− and active species in the electrode, and thus resulting in efficiency Faradaic reaction.

Table 1 The discrepancies between SC-1 and SC-2. Compound

SC-1

SC-2

Average Ni-O bond length (Å) Average Ni-N bond length (Å) Ni⋯Ni distance separated by H2O (Å) Ni⋯Ni distance separated by cis-configuration N-donor ligand (Å) Ni⋯Ni distance separated by trans-configuration N-donor ligand (Å) Channel diameter defined by latest Ni⋯Ni (Å) Dihedral angle between triazole (or imidazole) rings in the cisconfiguration N-donor ligand (°) Dihedral angle between triazole (or imidazole) rings in the trans-configuration N-donor ligand (°) Total void value of the channels is estimated by Platon (Å3) without solvent molecules

2.057 2.038 3.586 18.70

2.052 2.096 3.544 18.34

14.76

14.99

11.62 45.06

10.71 50.76

34.86

34.57

6572.9

6073.9

reaction occurring in the electrochemical process, the concentration of OH− may play a vital role in its diffusion process and consequently affect the resulting performance. Thus, the electrochemical performances for SC-1 and SC-2 in a 1.0 M KOH electrolyte have also been investigated. When decreasing the electrolyte concentration to 1.0 M, the specific capacitances for SC-1 and SC-2 both reduce in various extents (Fig. 6a-d, detailed CV figures and energy histograms see Supplementary Fig. S8 and Fig. S9). The gravimetric capacitance values are calculated to be 601, 594, 527, 446, 410 F g−1 and 230, 220, 204, 156, 140 F g−1 at the current density of 1.0, 2.0, 4.0, 6.0 and 8.0 A g−1 severally. However, the two electrodes still remain superior rate capability (68% and 61% remained after 8 times increase for SC-1 and SC-2 separately) and energy deliverable ability (91.3% and 91.6% for SC-1 and SC-2 respectively). This result confirms that the diffusion of OH− mainly controls and involves in the NiII/NiIII conversion process. In SC1 and SC-2 single crystal system, the higher the concentration of OH− and diffusion coefficient is and the higher the specific capacitance will be. Electrical resistance is another important parameter of an electrochemical capacitor electrode, which can be quantitatively evaluated by electrochemical impedance spectroscopy (EIS) measurements. The Nyquist plots for SC-1 and SC-2 electrode measured in the frequency range of 0.01 Hz–100 kHz are presented, which can be used to further explain the different electrochemical behaviors (Fig. 7). The plot for SC1 is composed of two semicircles located in the high (this one is not apparent and arises from the electron transport process in the surface of electrode) and low frequency (related to Faradaic reactions process) separately (Fig. 7a and inset part). The absence of straight line indicates that the resistance caused by diffusion can be omitted and the electrode process is mainly dominated by kinetic control. For SC-2, there is a similar electrode process, but a semicircle with a larger diameter occurring in the low frequency (Fig. 7b and inset part). For SC-1 and SC-2 49

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Fig. 6. Charge-discharge curves for SC-1 (a) and SC-2 (b) electrodes in 1.0 M KOH electrolyte respectively. (c) CP curves for SC-1 and SC-2 at a current density of 1.0 A g−1 in1.0 M KOH electrolyte. (d) Variation of specific capacitance of SC-1 and SC-2 as a function of current densities in 1.0 M KOH electrolyte.

4. Conclusions

MOFs is 92.1% and 88.7% severally. The specific capacitance and energy for Ni-MOFs is 705, 666 F g−1 and 29.6, 28.0 Wh kg−1 at current density of 1 A g−1 respectively, which are comparable to other electrochemical studies based on MOFs materials. This work opens new channel for the directly application of 3D topological MOFs with large frameworks and especially possessing porous structures in energy storage.

Two topological and isostructural 3D Ni-based metal-organic frameworks with novel 1D helical channel, synthesized by rigidly tripodal ligand (H3TATB) and flexibly secondary linker (bid or btd), are first introduced and directly applied them as SCs electrodes. Both of them behave as pseudo-capacitors and the capacitance originates from bulk diffusion according to the results from CV and charge-discharge curves. Although the different diffusion coefficient and subtle structure discrepancies results in nuanced electrochemical performance, they both deliver high specific capacitance and energy with good rate capability, excellent energy deliverable ability and cycling stability. After 5000 continuous cycles, the capacitance retention for the as-synthesized

Acknowledgements This work was supported by the Fundamental Research Funds for the Central Universities (KYZ201540), the National Natural Science Foundation of China (21371098), the Qing Lan Project of Jiangsu

Fig. 7. Nyquist plot for SC-1 (a) and SC-2 (b) electrodes at room temperature. The inset part represents a magnification image of EIS data at high frequency region.

50

Journal of Power Sources 377 (2018) 44–51

K. Wang et al.

Fig. 8. Endurance tests of 5000 continuous cycles for as-prepared SC-1 (a) and SC-2 (c) electrodes at a nominal scan rate of 35 mV s−1, respectively. The variation plots of the specific capacitance for SC-1 (b) and SC-2 (d) electrodes along with the increase of cycle numbers.

Provice, the Natural (BK20131314), the (2015M570430), the (1401007C) and the Agricultural University

Science Foundation of Jiangsu Province China Postdoctoral Science Foundation Jiangsu Postdoctoral Science Foundation Scientific Research Foundation of Nanjing (050804087).

(2016) 66–73. [18] H. Jasuja, Y. Huang, K.S. Walton, Langmuir 28 (2012) 16874–16880. [19] D. Sherbla, J.C. Bachman, J.S. Elias, C.J. Sun, Y. Shao-Horn, M. Dincǎ, Nat. Mater 16 (2017) 220–224. [20] D. Chen, Q. Wang, R. Wang, G. Shen, J. Mat. Chem. A 3 (2015) 10158–10173. [21] X. Liu, C. Shi, C. Zhai, M. Cheng, Q. Liu, G. Wang, ACS Appl. Mat. Interfaces 8 (2016) 4585–4591. [22] A.M. Spokoyny, D. Kim, A. Sumrein, C. Mirkin, Chem. Soc. Rev. 38 (2009) 1218–1227. [23] M. Oh, C.A. Mirkin, Nature 438 (2005) 651–654. [24] K. Wang, Z. Geng, M. Zheng, L. Ma, X. Ma, Z. Wang, Cryst. Growth Des. 12 (2012) 5606–5614. [25] J.D. Rocca, D. Liu, W. Lin, Acc. Chem. Res. 44 (2011) 957–968. [26] H. Wu, X.L. Lü, C.L. Yang, C.X. Dong, M.S. Wu, CrystEngComm 16 (2014) 992–1000. [27] G.M. Sheldrick, SHELXS-97, Programs for X-ray Crystal Structure Solution, University of Göttingen, Göttingen, Germany, 1997. [28] L.J. Farrugia, WINGX, a Windows Program for Crystal Structure Analysis, University of Glasgow, Glasgow, U. K, 1988. [29] A.L. Spek, J. Appl. Crystallogr. 36 (2003) 7–13. [30] D. Sun, S. Ma, Y. Ke, D.J. Collins, H.C. Zhou, J. Am. Chem. Soc. 128 (2016) 3896–3897. [31] S. Ma, D. Sun, M. Ambrogio, J.A. Fillinger, S. Parkin, H.C. Zhou, J. Am. Chem. Soc. 129 (2007) 1858–1859. [32] S. Ma, X.S. Wang, D. Yuan, H.C. Zhou, Angew. Chem. Int. Ed. 47 (2008) 4130–4133. [33] S.K. Meher, P. Justin, G. Ranga Rao, ACS Appl. Mat. Interfaces 3 (2011) 2063–2073. [34] B.E. Conway, J. Electrochem. Soc. 138 (1991) 1539–1548. [35] S.K. Meher, P. Justin, G.R. Rao, Electrochim. Acta 55 (2010) 8388–8396. [36] I.E. Rauda, V. Augustyn, B. Dunn, S.H. Tolbert, Acc. Chem. Res. 46 (2013) 1113–1124. [37] X. Cao, J. Wei, Y. Luo, Z. Zhou, Y. Zhang, Int. J. Hydrogen Energy 25 (2000) 643–647. [38] C. Liao, Y. Zuo, W. Zhang, J. Zhao, B. Tang, A. Tang, Y. Sun, J. Xu, Russ. J. Electrochem 49 (2013) 983–986. [39] L. Kang, S.X. Sun, L.B. Kong, J.W. Lang, Y.C. Luo, Chin. Chem. Lett. 25 (2014) 957–961. [40] G.C. Li, P.F. Liu, R. Liu, M. Liu, K. Tao, S.R. Zhu, M.K. Wu, F.Y. Yi, L. Han, Dalton Trans. 45 (2016) 13311–13316. [41] Y. Xia, B. Wang, G. Wang, H. Wang, RSC Adv. 5 (2015) 98740–98746.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.jpowsour.2017.11.087. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

P. Simon, Y. Gogotsi, Nat. Mater 7 (2008) 845–854. Y. Gogotsi, Nature 509 (2014) 568–570. P. Simon, Y. Gogotsi, B. Dunn, Science 343 (2014) 1210–1211. Y. Zhu, S. Murali, M. Stoller, K.J. Ganesh, W. Cai, P.J. Ferreira, A. Pirkle, R.M. Wallace, K.A. Cychosz, M. Thommes, Science 332 (2011) 1537–1541. G. Wang, L. Zhang, J. Zhang, Chem. Soc. Rev. 41 (2012) 797–828. T. Lin, I.W. Chen, F. Liu, C. Yang, H. Bi, F. Xu, F. Huang, Science 350 (2015) 1508–1513. X. Huang, X. Qi, F. Boey, H. Zhang, Chem. Soc. Rev. 41 (2012) 666–686. J. Hou, C. Cao, F. Idrees, X. Ma, ACS Nano 9 (2015) 2556–2564. T. Liu, C. Jiang, B. Cheng, W. You, J. Yu, J. Power Sources 359 (2017) 371–378. C. Wu, X. Wang, B. Ju, L. Jiang, H. Wu, Q. Zhao, L. Yi, J. Power Sources 227 (2013) 1–7. J. Liu, X. Wang, J. Gao, Y. Zhang, Q. Lu, M. Liu, Electrochim. Acta 211 (2016) 183–192. O.M. Yaghi, G.M. Li, H.L. Li, Nature 378 (1995) 703–706. J. Yang, P. Xiong, C. Zheng, H. Qiu, M. Wei, J. Mat. Chem. A 2 (2014) 16640–16644. J. Yang, C. Zheng, P. Xiong, Y. Li, M. Wei, J. Mat. Chem. A 2 (2014) 19005–19010. P.C. Banerjee, D.E. Lobo, R. Middag, W.K. Ng, M.E. Shaibani, M. Majumder, ACS Appl. Mat. Interfaces 7 (2015) 3655–3664. P. Wen, P. Gong, J. Sun, J. Wang, S. Yang, J. Mat. Chem. A 3 (2015) 13874–13883. C. Qu, Y. Jiao, B. Zhao, D. Chen, R. Zou, K.S. Walton, M. Liu, Nano Energy 26

51