Metal organic frameworks derived porous lithium iron phosphate with continuous nitrogen-doped carbon networks for lithium ion batteries

Metal organic frameworks derived porous lithium iron phosphate with continuous nitrogen-doped carbon networks for lithium ion batteries

Journal of Power Sources 304 (2016) 42e50 Contents lists available at ScienceDirect Journal of Power Sources journal homepage: www.elsevier.com/loca...

3MB Sizes 3 Downloads 29 Views

Journal of Power Sources 304 (2016) 42e50

Contents lists available at ScienceDirect

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

Metal organic frameworks derived porous lithium iron phosphate with continuous nitrogen-doped carbon networks for lithium ion batteries Yuanyuan Liu a, Junjie Gu a, b, Jinli Zhang a, Feng Yu c, Lutao Dong a, Ning Nie a, Wei Li a, * a

School of Chemical Engineering and Technology, Tianjin University, Tianjin, 300072, PR China Department of Mechanical and Aerospace Engineering, Carleton University, Ottawa, K1S 5B6, Canada Key Laboratory for Green Processing of Chemical Engineering of Xinjiang Bingtuan, School of Chemistry and Chemical Engineering, Shihezi University, Shihezi 832003, PR China b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 3D porous LFP/N-CNWs was synthesized using MIL-100(Fe) as template and raw material.  LFP/N-CNWs composite possess a high BET surface area of 129 m2 g1.  LFP/N-CNWs shows excellent capacity and rate performance comparing with LFP/CNWs.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 June 2015 Received in revised form 28 October 2015 Accepted 5 November 2015 Available online xxx

Lithium iron phosphate (LiFePO4) nanoparticles embedded in the continuous interconnected nitrogendoped carbon networks (LFP/N-CNWs) is an optimal architecture to fast electron and Liþ conduction. This paper, for the first time, reports a reasonable design and successful preparation of porous hierarchical LFP/N-CNWs composites using unique Fe-based metal organic framework (MIL-100(Fe)) as both template and starting material of Fe and C. Such nitrogen-doped carbon networks (N-CNWs) surrounding the lithium iron phosphate nanoparticles facilitate the transfer of Liþ and electrons throughout the electrodes, which significantly decreases the internal resistance for the electrodes and results in the efficient utilization of LiFePO4. The synthesized LFP/N-CNWs composites possess a porous structure with an amazing surface area of 129 m2 g1, considerably enhanced electrical conductivities of 7.58  102 S cm1 and Liþ diffusion coefficient of 8.82  1014 cm2 s1, thereby delivering excellent discharge capacities of 161.5 and 93.6 mAh$g1 at 0.1C and 20C, respectively. © 2015 Elsevier B.V. All rights reserved.

Keywords: Lithium iron phosphate Carbon networks Template method Metal organic frameworks

1. Introduction With the rapid depletion of traditional fossil fuels and

* Corresponding author. E-mail address: [email protected] (W. Li). http://dx.doi.org/10.1016/j.jpowsour.2015.11.022 0378-7753/© 2015 Elsevier B.V. All rights reserved.

continually worsened environmental pollution, there is a greatly increased demand for the clean and efficient energy storage devices. Because of the excellent safety performance, long cycle life, high operational voltage and energy density, lithium ion batteries (LIBs) have already been used in many portable electronic products over other energy storage technologies in recently years [1,2]. As a preferred cathode materials for LIBs, the lithium iron phosphate

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

(LiFePO4) has attract considerable attention owing to its merits of high theoretical discharge capacity (~170 mAh$g1), acceptable operating voltage (3.42 V), low price, environmental benignity and superior safety [3,4]. However, the drawbacks involving extremely low electronic conductivity (~109 S cm1) and Liþ conductivity (~1.8  1014 cm2 S1), which result in poor electrochemical performance and hider the commercial application of LiFePO4 [5]. Therefore, many efforts have been directed towards increasing the purity and crystallinity [4,6], incorporating supervalent ions, modifying the crystal structure [7], morphology [8,9] and particle size of LiFePO4 [10,11], or creating a fast electron conductive surface, such as phosphorus-doped carbon layers [12] aiming to enhance the electronic conductivity and reduce the Liþ diffusion resistance. It is widely accepted that the power capability of LiFePO4 depends on the kinetics during the Liþ insertion/extraction process, which involves three steps: (i) electron and Liþ diffusion between the electrode and electrolyte, (ii) charge transfer reactions in the cathode-electrolyte interface, and (iii) Liþ diffusion within the electrolyte. Among these reported approaches, fabricating a threedimensional (3D) porous LiFePO4/C architecture with host LiFePO4 nanoparticles embedding in the guest carbon frameworks (CNWs) is the most efficient one to fast electronic and Liþ conduction [13e17]. That is because the carbon networks can bridge the embedded LiFePO4 nanoparticles to overcome the low electron transfer, and the adequate porosity of carbon framework promotes the rapid movement of Liþ for high-rate capabilities. Currently, a variety of synthetic technologies have been applied to synthesize porous LFP/CNWs composites. Template approaches is one of the well-studied and easy methodology to tune the morphology and structure of LiFePO4 with desired architecture. Both soft and hard templates, including tartaric acid [17], carbon monolith [18], colloidal crystal [19], and triblock copolymer [20], have already been utilized to synthesize porous carbon frameworks for loading LiFePO4 nanoparticles. Metal-organic frameworks (MOFs), a fantastic porous crystalline compound which compose of metal ions connected by bridges of organic molecules, have already been attracted considerable interests not only because of their large surface area, controlled pore size and distribution, but also because of their potential applications in hydrogen storage, gas adsorption and separations, catalysis, sensing, as well as photonics and magnetic [21e23]. At present, MOFs have been attempted to be utilized as both a template and starting materials for the preparation of tuneable framework composites, carbon nanotubes, and some other inorganic materials. For instance, well interconnected 3D hierarchically porous networks (AS-ZC-800) were successfully formed using microporous Zn-based ZIF-8 particles as template [24]. The prepared unique ASZC-800 strategy with a large BET surface area ~2972 m2 g1 exhibits high performance as a super capacitor electrode. In addition, a HPCNeS with 3D hierarchically porous nanostructure was prepared by encapsulating sulfur (S) into the porous carbon nanoplates (HPCN) which derived from the pyrolysis reaction of metal-organic frameworks (MOF-5). The resultant HPCNeS composites with a high specific capacity deliver an initial discharge capacity of 1177 mAh$g1 and final 730 mAh$g1 after 50th cycle at 0.1C [25]. Herein, for the first time, we report a reasonable design and a successful preparation of hierarchically porous LFP/CNWs composites using unique Fe-based metal organic framework MIL100(Fe) as a porous template and the starting material of Fe and C. The metal organic framework MIL-100(Fe) has already been utilized as a template to synthesize g-Fe2O3/C [26] and magnetic CoFe2O4 [27] previously. Thereafter, the prepared LFP/CNWs composites were mixed with melamine to synthesize nitrogenmodified LFP/N-CNWs composites, expecting to further increase

43

the electrical conductivity and consequently enhance the electrochemical performance of LiFePO4 cathodes materials. It is worthwhile noting that the XRD, SEM and TEM characterizations indicate that both of the LFP/CNWs and LFP/N-CNWs composites exhibit a 3D porous architecture with LiFePO4 nanoparticles uniformly embed in the continuous carbon framework. Compared with the LFP/CNWs sample, nitrogen-doped LFP/N-CNWs sample show superior electrochemical performance in terms of high discharge capacity, good rate capability and excellent capacity retention which is promising for the applications in LIBs. 2. Experimental 2.1. Materials Fe(NO3)3$9H2O, LiH2PO4 NH4F and C3N3(NH2)3 were purchased from Guang Fu Fine Chemical Research Institute (Shanghai, China). Benzene trimesic acid (BTC) was purchased from SigmaeAldrich Co. LLC (Shanghai, China). All of the materials were of analytical grade and were used without further purification. Demonized water was used throughout the process. 2.2. Synthesis of MIL-100(Fe) MIL-100(Fe) was synthesized according to the literature of Seo et al. [28]. The entirely dissolved Fe(NO3)3$9H2O aqueous solution was loaded into a vessel, followed by benzene trimesic acid (BTC) powder. The resultant composition of the reactant mixture was 1 Fe(NO3)3$9H2O: 0.67 BTC: x H2O (x ¼ 55e280). After stirred for 1 h at ambient temperature, the reactant mixture was shifted into a Teflon-lined pressure crystallizer and then heat treated at 160  C for 12 h. After reaction, the obtained suspension was filtered, and then washed using deionized water at 80  C for 3 h and hot ethanol at 65  C for 3 h until no detected colored impurities in the filtrate. To further purification, a 70  C NH4F aqueous solution was utilized to purified MIL-100(Fe) for 3 h. The final solids were then dried under air at 80  C for 12 h. 2.3. Synthesis of LFP/CNWs and LFP/N-CNWs The synthetic procedure of LFP/CNWs and LFP/N-CNWs is shown in Scheme 1. In a typical synthesis of LFP/CNWs, 6.51 g of the as-prepared MIL-100(Fe) was dispersed in 50 mL ethanol. Next, 4.08 g LiH2PO4 was dissolved in 100 mL ethanol and added into the MIL-100(Fe) turbid liquid under stirring. The molar ratio of Fe/Li is controlled as 1:1. The mixed ethanol solution was evaporated at 80  C for 4 h after been stirred for 24 h at room temperature. The resulting solid product was transferred into a tubular furnace and heat to 600  C for 5 h at the heating speed of 2  C min1 in reducing atmosphere (10% H2 þ 90% Ar). The nitrogen-doped LFP/N-CNWs was prepared via carbonization of the nitrogen-containing C3N3(NH2)3 using the following procedure: 0.4672 g C3N3(NH2)3 is mixed with 4 g of the as-prepared LFP/CNWs in 60 mL ethanol solvent, and then the mixtures were stirred at 60  C for 24 h. The final solids were dried under vacuum at 80  C for 10 h before calcination under a N2 atmosphere at 600  C for 4 h. 2.4. Materials characterization The crystal structure was identified by X-ray diffraction (XRD, Rigaku D/max 2500 V/PC) with a Cu-Ka radiation source. The diffraction date were collected in a 2q range from 5 to 70 with each step of 0.017. The surface morphology of the crystal was observed by a scanning electron microscopy (SEM, Hitachi Ltd., Japan, Model S-4800) with an X-ray energy dispersive spectroscope

44

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

Scheme 1. Schematic representation of the synthesis procedure for LFP/CNWs and LFP/N-CNWs.

(EDS). The surface texture and morphology of the crystal were observed by a transmission electron microscope (TEM, JEM-2010 FEF, Japan). The specific surface area was measured by a N2 adsorptionedesorption method using a BrunauereEmmetteTeller (BET) analyzer, a Nove 2000e (Quantachrome Instruments, USA). Xray photoelectron spectroscopy (XPS) measurement was carried out by a PHI5000 Versa Probe spectrometer. The analyzer pass energy for survey and detailed elemental scans were set to 187.85 eV and 46.95 eV, respectively. The electronic conductivity for the synthesized LiFePO4/C particles was measured by a San Feng SB120 four-point probe measurement system at room temperature. The conductivity was measured three times at different positions along the compacted LiFePO4 column, which was 10 mm in diameter and 1.2 mm in thickness, and the average electronic conductivity was obtained. Thermogravimetric analysis (TG, SHIMADZU TG-50, Japan) was performed from 20  C to 700  C at a heating speed of 2  C min1 under dry air. The total carbon amount of the LiFePO4/C composite was calculated according to the different weight gains observed in the TG curves between LiFePO4 and LiFePO4/C [29].

cut-off voltages from 2.2 V to 4.2 V at ambient temperature. The cyclic voltammetric (CV) tests and the electrochemical impedance spectroscopy (EIS) tests were all carried out at room temperature with electrochemical workstation (ZAHNER Zennium, Germany). The CV curves were recorded in the potential range of 2.2 Ve4.2 V at a scan rate of 0.1 mV s1. The EIS tests were recorded with the frequency from 101 Hze106 Hz at the testing potential of 3.4 V vs. Liþ/Li with approximate 50% SOC. 3. Results and discussion 3.1. Structural characterization The crystallinity and phase information for the as-synthesized products were studied by X-ray diffraction (XRD), as shown in Fig. 1. The good agreement between the XRD pattern of the asprepared MIL-100(Fe) template and the simulated ones reported in the literatures [30,31], confirmed the successful synthesis of the metal-organic framework. Previous studies have reported that a

2.5. Electrochemical tests Electrochemical performance of the synthesized cathode materials was measured by coin-type cell using lithium metal as the anode. The electrode mixture was prepared through blending the LFP/CNWs or LFP/N-CNWs composites, conducting additives Super P-MMM carbon and vinylidene fluoride (PVDF) with the mass ratio controlled to be 8:1:1. The prepared electrode film was then cut into a round shape with the diameter and thickness of 1.3 cm and 28 mm, respectively. The average weight loading in the active material was calculated to be 2.2 mg cm2. This prepared electrode disk was then assembled into a coin-type cell with a lithium metal anode and a polymer separator (Celgard 2400). The electrolyte is a mixture with 1.0 M LiPF6 dissolved in three solvents including dimethyl carbonate (DMC), diethyl carbonate (DEC) and ethylene carbonate (EC) (1:1:1, by volume). All of the assembly process was carried out in a glove box (Mikrouna Co., Ltd., China, Model Super 1220/750) under Ar atmosphere with water content and oxygen content less than 0.1 PPm. The electrochemical performances of the cathode materials were evaluated by an automatic charge/ discharge equipment (Land Co., China, Model CT2001A) with the

Fig. 1. XRD patterns of the as-synthesized template MIL-100(Fe) and its derived LFP/ CNWs and LFP/N-CNWs.

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

pure LiFePO4 phase can be obtained at 539  C through the carbothermal method [32], so the formative temperatures of LiFePO4 was selected as 600  C in this report. The XRD patterns of the products synthesized from pyrolysis of the MIL-100(Fe) template with the precursor LiH2PO4 are listed in Fig. 1. All of the diffraction peaks from the LFP/CNWs and LFP/N-CNWs samples correspond well to the orthorhombic olivine-type LiFePO4 with a space group of Pnma (ICDD PDF No. 81-1173), illustrating that pure-phase LiFePO4 can be successfully prepared using the metal-organic framework MIL100(Fe) as both a template and a Fe precursor. This finding also reveals that the addition of the element nitrogen into the CNWs has no influence on the crystal structure of LiFePO4. Details of the structural parameters for LFP/CNWs and LFP/N-CNWs composites derived from XRD data are summarized in Table 1. As shown in Table 1, the lattice parameters and cell volume for the assynthesized LFP/CNWs are a ¼ 10.315 Ǻ, b ¼ 5.998 Ǻ, c ¼ 4.713 Ǻ and volume ¼ 292.05 Ǻ3, which are similar for the sample LFP/NCNWs (a ¼ 10.319 Ǻ, b ¼ 6.001 Ǻ, c ¼ 4.717 Ǻ and volume ¼ 292.10 Ǻ3). These results confirm that the crystal structure for the synthesized LiFePO4 samples is stable. In addition, no peaks corresponding to carbon deriving from the carbonization of the organic ligands in the MIL-100(Fe) framework can be observed in the XRD patterns, indicating that the generated carbon exists in an amorphous state. X-ray photoelectron spectroscopy (XPS) technology was employed to determine the composition and chemical state of the atoms in the synthesized LFP/CNWs and LFP/N-CNWs composites. The wide XPS spectrum in Fig. 2a lists the binding energy at approximately 56.4 eV, 132.7 eV, 285.5 eV, 398.4 eV and 531.2 eV, respectively corresponding to Li 1s, P 2p, C 1s, N 1s and O 1s [33]. Compared with the survey spectra of the LFP/CNWs sample, a new binding energy at approximately 399 eV for the LFP/N-CNWs sample can be observed clearly, which can be ascribed to N 1s, suggesting the presence of N in the carbon networks. Fig. 2b shows the high resolution XPS spectrum of N 1s for the LFP/N-CNWs. It can be resolved into three components centered at 398.3, 399.5 and 401 eV, corresponding the pyridinic, pyrrolic and graphitic types of N atoms, respectively [34,35]. In addition, the composition contents and the relative atomic ratios of N and C for the as-synthesized LFP/ CNWs and LFP/N-CNWs materials were analyzed using the XPS spectra, and the results are listed in Table 2. As shown in Table 2, the LFP/CNWs sample without nitrogen doping consists of 61.73% carbon and 16.84% oxygen. While for the sample LFP/N-CNWs with nitrogen doping, besides the similar contents of C and O for the LFP/ CNWs sample, there is a small quantity of doped nitrogen with a content of 1.26%. The atomic ratio of N and C for the LFP/N-CNWs sample is calculated to be 2 atm %. According to the surface testing results, it can be concluded that the nitrogen from melamine was successfully doped onto the surface of LiFePO4/C material, which is beneficial to the electron transport and consequently improves the electrochemical performance of LiFePO4/C materials. To illustrate the thermal decomposition mechanism of the MIL100(Fe) temple under reducing atmosphere, the decomposition of the as-synthesized MIL-100(Fe) sample was analyzed using a thermogravimetric (TG) method. The TG pattern, as shown in Fig. 3, demonstrates that the MIL-100(Fe) sample first exhibits a weight Table 1 Lattice parameters and cell volume for the as-synthesized LFP/CNWs and LFP/NCNWs materials. Sample

a (Ǻ)

b (Ǻ)

c (Ǻ)

V (Ǻ3)

LFP/CNWs LFP/N-CNWs

10.315 10.319

5.998 6.001

4.713 4.717

292.05 292.10

45

loss of ~23 wt% as the temperature increases to 250  C, corresponding to the departure of the adsorbed water molecules inside or outside the pores. The second drastic degradation of MIL-100(Fe) takes place between 400  C and 500  C (~44.24 wt%), which is related to the decomposition of organic ligands and the reduction of ferric iron in the MIL-100(Fe) framework. The product that remained above 500  C at the end of the TG experiment should be composed mainly of FeO/C, corresponding to a weight reserve of 32.76 wt%. Thereafter, the carbon mass contents in the LFP/CNWs and LFP/N-CNWs composites were also measured through the TG method, and the corresponding curves are provided in Fig. 3. In the typical decomposition process of LiFePO4 under air, LiFePO4 can be oxidized to Li3Fe2(PO4)3 and Fe2O3 as the temperature rises from 200 to 500  C. The corresponding reaction is shown in equation (1), and the weight gain at this stage is calculated to be 5.07 wt%. LiFePO4 þ 1/4O2 ¼ 1/3Li3Fe2(PO4)3 þ 1/6Fe2O3

(1)

However, for the carbon coating LiFePO4, this equation will be changed: LiFePO4 þ xC þ 1/4O2 þ xO2 ¼ 1/3Li3Fe2(PO4)3 þ 1/6Fe2O3 þ xCO2

(2)

where x relates to the carbon content in LiFePO4/C. At above 600  C, the decomposition of LiFePO4/C composite remains stable, suggesting the oxidization reactions are finished. The mass of carbon in LiFePO4/C is calculated based on the difference in weight gain measured between LiFePO4 and LiFePO4/C. According to the TG curves in Fig. 3, the carbon contents are calculated to be 11.68 wt% and 11.80 wt% for LFP/CNWs and LFP/N-CNWs composites, respectively (Table 3). 3.2. Morphological characterization The morphology of the MIL-100(Fe) template, LFP/CNWs and LFP/N-CNWs products are investigated by SEM. SEM micrographs and the N2 adsorption-desorption method in Fig. 4a and d both reveal that the MIL-100(Fe) particles have a size of 3.2 mm with an octagonal structure, giving a high surface area of 1690 m2 g1. Porous LFP/CNWs composites are shown in Fig. 4b. Although the structure of the MIL-100(Fe) framework is completely destroyed during the pyrolysis process, the 3D LFP/CNWs networks basically maintain the integrated morphology and size parameters of MIL100(Fe). The high magnified SEM images in the inserted sections of Fig. 4b reveal that the host LiFePO4 nanoparticles with an average size of ~100 nm are uniformly incorporated into the interlaced guest carbon networks. The LiFePO4 primary particles seem to be formed at the secondary building units (SBU) of MIL-100(Fe) during the pyrolysis process after the Liþ and (PO4)3þ were impregnated into the cavities and channels of the MOFs. The carbon networks derived from the carbonization of the organic ligands in MIL100(Fe) are polymerized in situ on the surface of the generated LiFePO4 particles, which could impede their growth during the heat treatment. At the same time, numerous micropores were generated throughout the interior of the LiFePO4 particles, constructing a 3D interconnected pore system. After the subsequent modification during the heat treatment, the as-obtained LFP/N-CNWs composite still retained the 3D porous micro-nano-structured architectures (Fig. 4c). Such a porous 3D structure gives a relatively large specific surface area of 126 and 129 m2 g1, and a BJH adsorption average pore diameter of 40.5 nm and 42.3 nm for the LFP/CNWs and LFP/NCNWs respectively (Fig. 4d). This reduced particle size and high surface area result in a shortened Liþ diffusion path, increasing the contact area between the electrode materials with the electrolyte,

46

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

Fig. 2. XPS curves of the synthesized LFP/CNWs and LFP/N-CNWs composites.

Table 2 The composition contents and relative ratios for the as-synthesized LFP/CNWs and LFP/N-CNWs materials. Sample

LFP/CNWs LFP/N-CNWs

atm%

N:C (atm%)

N

C

O

e 1.26

61.73 62.91

16.84 17.16

e 2

dispersed and homogeneously embedded in the porous continuous carbon framework. These synthesized LiFePO4 particles have a small size of ~100 nm because of the restrictive effect of the carbon framework, which can facilitate the Liþ diffusion cause the pathlength is remarkably shortened. Simultaneously, these LiFePO4 nanoparticles are bridged together through the conductive carbon walls. These carbon interconnects favor the particle-to-particle electron contact and make the electron transfer between the LiFePO4 and the current collector fast and efficient, especially for the N-CNWs. Additionally, the well-developed porosity of both the CNWs and N-CNWs gives the LiFePO4 surfaces good access to the electrolyte, thus increasing the electrode-electrolyte interface and decreasing the Liþ diffusion distance (Fig. 5c and d). The homogeneous distribution of the host LiFePO4 lodgers and the guest carbon networks in the LFP/CNWs and LFP/N-CNWs composites is further illustrated by the corresponding EDS elemental mappings of Fe, P, O, C and N (Fig. S1a). Notably, the uniform distribution of N on the surface of the LFP/N-CNWs in Fig. S1b is consistent well with that of Fe, P, O and C, indicating the homogeneous dispersion of nitrogen atom in the N-CNWs. This N-CNWs structure may further improve the electronic conductivity, because the nitrogen atom may contribute additional electrons and provide electron carriers to the conduction band. Improvement in the electron conductivity of the composite cathodes was also demonstrated using a four-point probe method. The four-point probe results indicate that the electrical conductivity for the LFP/N-CNWs composite is 7.58  102 S cm1, which is higher than that of the LFP/CNWs composite (6.01  102 S cm1), suggesting that the increased conductivity of LFP/N-CNWs is induced by the nitrogen doping.

Fig. 3. TG curves of the as-synthesized template MIL-100(Fe) under reducing atmosphere and its derived LFP/CNWs and LFP/N-CNWs under air atmosphere.

3.3. Electrochemical characterization Table 3 BET specific surface areas and carbon content for the as-synthesized MIL-100(Fe), LFP/CNWs and LFP/N-CNWs materials. Sample

SBET (m2$g1)

Carbon content (wt%)

MIL-100(Fe) LFP/CNWs LFP/N-CNWs

1690 126 129

e 11.68 11.71

hence improving the electrochemical performance of LiFePO4. To further investigate the contact between the LiFePO4 nanoparticles and the surrounding carbon walls, the synthesized LFP/ CNWs and LFP/N-CNWs powders were examined by TEM. As shown in Fig. 5a and b, the numerous LiFePO4 particles are highly

The electrochemical performance and properties of the MOFs derived LFP/CNWs and LFP/N-CNWs were tested from 0.1C to 20C rate (1C ¼ 170 mAh$g1) with voltage between 2.2 V and 4.2 V vs. Li/Liþ. The rate performance corresponding to the different current densities are listed in Fig. 6a and b. The specific discharge capacities for the LFP/CNWs composite are 158.1, 93.8 and 80.4 mAh$g1 at the rates of 0.1C, 10C and 20C, respectively, which increase to 161.5, 110.8 and 93.6 mAh$g1 for the LFP/N-CNWs composite. Even as the discharge current rate increases as high as 20C, the LFP/NCNWs sample still exhibits discharge capacities of 93.6 mAh$g1, which confirm the excellent rate capabilities with 58.0% of the maximum capacity (161.5 mAh$g1 at 0.1C). This outstandingly high-rate discharge capacity can be attributed to the refined LiFePO4 nanoparticles in combination with the nitrogen-doped

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

47

Fig. 4. (aec) SEM images and (d) pore size distribution and N2 adsorption-desorption isotherms (inset) for the as-synthesized template MIL-100(Fe) and its derived LFP/CNWs and LFP/N-CNWs composites, respectively.

Fig. 5. (aeb) TEM images and (ced) HRTEM of the as-synthesized LFP/CNWs and LFP/N-CNWs composites.

48

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

Fig. 6. Electrochemical properties of LFP/CNWs and LFP/N-CNWs composites: (aeb) charge and discharge profiles measured at various C-rates; (c) the comparison of the discharge capacities at various C-rates; (d) cyclic voltammograms between 2.2 and 4.2 V at a scan rate of 0.1 mV s1; (e) specific charge energy and discharge energy; (f) energy efficiency at various C-rates, i.e., the ratio of specific discharge energy to specific charge energy.

carbon networks. Nanometer-sized LiFePO4 particles can significantly shorten the diffusion path length for Liþ and thereby present high energy storage capacities. In addition to taking advantage of the LiFePO4 nanoparticles, the constructed nitrogen-doped carbon framework simultaneously contributes to the superior electrochemical performance as follows: (1) the N-CNWs act as a continuous conductive framework for the embedded LiFePO4 nanoparticles, thus overcoming the dominant drawback of sluggish electron transfer between LiFePO4 particles; (2) the abundance of porosity inside the carbon networks permits the penetration of the electrolyte, which to guarantees intimate contact between the immersed LiFePO4 nanoparticles and the electrolyte for the facile transfer of Liþ. Cyclic voltammetry (CV) was also carried out with a scan speed of 0.1 mV s1 to further study the excellent electrochemical properties of the LFP/CNWs and LFP/N-CNWs samples, and the result is shown in Fig. 6d. Both the curves show one pair of redox peaks associated with the Fe2þ/Fe3þ couples, corresponding to the insertion and extraction process of Liþ in the olivine structure LiFePO4. Compared with the CV curve of the LFP/CNWs, the LFP/NCNWs sample shows sharper redox peaks, implying higher electrode kinetics. In addition, potential separation between the cathodic and anodic peaks is also a key factor to evaluate the reversibility of the reaction. According to the curves, the oxidation and reduction peaks are centered at 3.35/3.51 V for the LFP/CNWs

sample, whereas they are 3.36/3.49 V for the LFP/N-CNWs sample. Therefore, the LFP/N-CNWs sample exhibits a smaller difference in potential of 0.13 V than that of the LFP/CNWs sample (0.16 V), suggesting better reversibility of the electrode reaction and a high discharge rate capability, which is consistent well with its high-rate electrochemical performance results, as mentioned above. Notably, the polarization of the LiFePO4/C composites not only affects the potential discharge voltage, but it also influences the specific energy and energy efficiency. As shown in Fig. 6e, the charge energy for the samples LFP/CNWs and LFP/N-CNWs at 0.1C are 546.0 Wh$kg1 and 560.5 Wh$kg1, respectively. And the corresponding discharge energy for the samples LFP/CNWs and LFP/NCNWs at 0.1C are 516.5 Wh$kg1 and 522.3 Wh$kg1, respectively. The energy efficiency shown in Fig. 6f, i.e., the ratio of discharge energies to charge energies at 0.1C rate, is approximately 94.3% for the LFP/N-CNWs sample, which is larger than that of the LFP/CNWs sample (92.6%). As the rate increases to 10C and 20C, the LFP/N-CNWs sample still perform high energy efficiency of 70.9% and 66.6% with discharge energy of 297.3 Wh$kg1 and 241.6 Wh$kg1, respectively. Because of the unique porous micron-sized sphere-like morphology, the tap densities for the prepared LFP/CNWs and LFP/ N-CNWs composites reach to 1.11 g cm3 and 1.09 g cm3, respectively. The volumetric specific capacity values are calculated by multiplying their gravimetric capacity by the tap density of the

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

active materials. To further compare the volumetric capacities, the rate performances for the prepared LFP/CNWs and LFP/N-CNWs composites, which were measured using the gravimetric and volumetric capacities, are shown in Fig. 7a and b. As shown in Fig. 7, the volumetric specific discharge capacity for the LFP/CNWs sample are 175.5 mAh$g1, 106.3 mAh$g1 and 87.0 mAh$g1 at the current rates of 0.1C, 10C and 20C, respectively. While for the sample LFP/ N-CNWs, the volumetric specific discharge capacity are 176.0 mAh$g1, 120.8 mAh$g1 and 104.2 mAh$g1 at the current rates of 0.1C, 10C and 20C, respectively. A much higher volumetric specific capacity is clearly achieved for the LFP/N-CNWs materials. This enhanced volumetric performance can be attributed to the porous sphere-like morphology and the highly conductive N-CNT network, indicating that the LFP/N-CNWs material is promising for high-performance LIBs with enhanced volumetric energy density and power capability. Furthermore, the comparison between the gravimetric and volumetric capacities in Fig. 7 also indicates that the LiFePO4/C composites with poor gravimetric specific discharge capacity still fail to achieve satisfactory volumetric discharge capacity. The optimal balance between the tap density and the gravimetric specific discharge capacity remains an important factor in synthesizing the LiFePO4/C composite with high volumetric energy density. Fig. 8 shows the Nyquist plots of the positive electrodes for the synthesized LFP/CNWs and LFP/N-CNWs samples with the frequency from 101 Hze106 Hz. The mass loading of the active material to each electrode for the AC impedance measurement is approximately the same. The impedance measurements were carried out at a discharge state of approximately 50% SOC (Li0.5FePO4), and the results were analyzed by Zview software. Both of the impedance spectra curves consist of an intercept in the high

49

Fig. 8. Nyquist plots of the discharge state of approximately 50% SOC with the frequency range of 101e106 Hz for the LFP/CNWs and LFP/N-CNWs samples.

frequency region, a depressed semicircle at medium frequency, and an inclined line at low frequency (Fig. 8). The high frequency intercept impedance is attributed to the impedance between the electrode and electrolyte (RU). The depressed semicircle in the medium frequency region is related to the resistance of charge movement (Rct) at the electrolyteecathode interface. The inclined line at low frequency associated with the Warburg impedance (Zw), representing the Liþ bulk diffusion inside the LiFePO4 particle. The simplified equivalent circuit model in the inserted section of Fig. 8 was constructed to analyze the impedance spectra, and the experimental data fit the equivalent circuit date very well. The charge transfer resistance value of Rct for LFP/N-CNWs is approximately 51.7 U, which is smaller than that of the LFP/CNWs (62.4 U). It is believed that the charge impedance in the electrochemical reaction is mediated by the unique architecture of the composites, which is consistent with their electrochemical properties. These results indicate that using the unique Fe-based metal organic framework (MIL-100(Fe)) as both the template and starting material to synthesize porous hierarchical LFP/N-CNWs composites is a successful technology to obtain LiFePO4 with enhanced electrochemical performance. 4. Conclusions A porous LiFePO4/C composite with nano-sized LiFePO4 particles embedded in an interconnected carbon network was successfully synthesized using a unique Fe-based metal organic framework MIL-100(Fe) as both the porous template and the starting material of Fe and C by a carbothermal reduction reaction. The porous LiFePO4/C secondary particles with a size of 3.2 mm are composed of LiFePO4 primary nanoparticles and continuous porous carbon frameworks which were derived from the carbonization of organic ligands in the MOFs. This optimal structure of LiFePO4/C with high BET surface area of 126 m2 g1 facilitates the fast diffusion of Liþ and electrons for high-rate capabilities, which delivers excellent discharge capabilities of 158.1 and 80.4 mAh$g1 at the current rates of 0.1C and 20C, respectively. Such a facile MOFs-template strategy may also be used to prepare other porous materials for advanced batteries. Acknowledgments

Fig. 7. Comparison of (a) gravimetric capacity and (b) volumetric capacity for the sample LFP/CNWs and LFP/N-CNWs.

This work is financially supported by the Special Funds for Major State Basic Research Program of China (2012CB720300), NSFC

50

Y. Liu et al. / Journal of Power Sources 304 (2016) 42e50

(21476158). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.11.022. References [1] F. Yu, L.L. Zhang, Y.C. Li, Y.X. An, M.Y. Zhu, B. Dai, RSC Adv. 4 (2014) 54576e54602. [2] F. Yu, L.L. Zhang, M.Y. Zhu, Y.X. An, L.L. Xia, X.G. Wang, B. Dai, Nano Energy 3 (2014) 64e79. [3] X.Q. Ou, S.Z. Xu, G.C. Liang, L. Wang, X. Zhao, Sci. China Ser. E 52 (2009) 264e268. [4] X.Q. Ou, L. Pan, H.C. Gu, Y.C. Wu, J.W. Lu, J. Mater. Chem. 22 (2012) 9064e9068. [5] F. Yu, J.J. Zhang, Y.F. Yang, G.Z. Song, J. Power Sources 189 (2009) 794e797. [6] X.Q. Ou, H.C. Gu, Y.C. Wu, J.W. Lu, Y.J. Zheng, Electrochim. Acta 96 (2013) 230e236. [7] J.L. Yang, J.J. Wang, X.F. Li, D.N. Wang, J. Liu, G.X. Liang, M.L. Gauthier, Y.L. Geng, D.S. Li, R.Y. Sun, X. Liang, J. Mater. Chem. 22 (2012) 7537e7543. [8] B. Pei, H.X. Yao, W.X. Zhang, Z.H. Yang, J. Power Sources 220 (2012) 317e323. [9] Q.Z. Song, X.Q. Ou, L. Wang, G.C. Liang, Z.R. Wang, Mater. Res. Bull. 46 (2011) 1398e1402. [10] S.L. Yang, X.F. Zhou, J.G. Zhang, Z.P. Liu, J. Mater. Chem. 20 (2010) 8086e8091. [11] Y.Y. Liu, J.J. Gu, J.L. Zhang, F. Yu, J. Wang, N. Nie, W. Li, RSC Adv. 5 (2015) 9745e9751. [12] J.L. Zhang, J. Wang, Y.Y. Liu, N. Nie, J.J. Gu, F. Yu, W. Li, J. Mater. Chem. A 3 (2015) 2043e2049. [13] X. Yang, Y.L. Xu, H. Zhang, Y.A. Huang, Q. Jiang, C.J. Zhao, Electrochim. Acta 114 (2013) 259e264. [14] J.Q. Zhao, J.P. He, J.H. Zhou, Y.X. Guo, T. Wang, S.C. Wu, X.C. Ding, R.M. Huang, H.R. Xue, J. Mater. Chem. C 115 (2011) 2888e2894.

[15] F. Yu, S.H. Lim, Y.D. Zhen, Y.X. An, J.Y. Lin, J. Power Sources 271 (2014) 223e230. [16] F. Yu, L.L. Zhang, L.F. Lai, M.Y. Zhu, Y.C. Guo, L.L. Xia, P.R. Qi, G. Wang, B. Dai, Electrochim. Acta 151 (2015) 240e248. [17] F. Yu, J.J. Zhang, Y.F. Yang, G.Z. Song, J. Power Sources 195 (2010) 6873e6878. [18] C.M. Doherty, R.A. Caruso, B.M. Smarsly, P. Adelhelm, C.J. Drummond, Chem. Mater 21 (2009) 5300e5306. [19] H.R. Xue, J.Q. Zhao, T. Wang, H. Guo, X.L. Fan, J.P. He, RSC Adv. 4 (2014) 39400e39407. [20] S.J. Sun, C.M. Ghimbeu, R. Janot, J.-M. Le Meins, A. Cassel, C. Davoisne, C. Masquelier, C. Vix-Guterl, Microporous Mesoporous Mater. 198 (2014) 175e184. [21] K.G.M. Laurier, F. Vermoortele, R. Ameloot, D.E. De Vos, J. Hofkens, M.B.J. Roeffaers, J. Am. Chem. Soc. 135 (2013) 14488e14491. [22] F. Jeremias, A. Khutia, S.K. Henninger, C. Janiak, J. Mater. Chem. 22 (2012) 10148e10151. [23] C.X. Yang, C. Liu, Y.M. Cao, X.P. Yan, RSC Adv. 4 (2014) 40824e40827. [24] A.J. Amali, J.K. Sun, Q. Xu, Chem. Commun. 50 (2014) 1519e1522. [25] G.Y. Xu, B. Ding, L.F. Shen, P. Nie, J.P. Han, X.G. Zhang, J. Mater. Chem. A 1 (2013) 4490e4496. [26] J.D. Xiao, L.G. Qiu, X. Jiang, Y.J. Zhu, S. Ye, X. Jiang, Carbon 59 (2013) 372e382. [27] F.X. Qin, S.Y. Jia, Y. Liu, X. Han, H.T. Ren, W.W. Zhang, J.W. Hou, S.H. Wu, Mater. Lett. 101 (2013) 93e95. [28] Y.-K. Seo, J.W. Yoon, J.S. Lee, U.H. Lee, Y.K. Hwang, C.-H. Jun, P. Horcajada, C. Serre, J.-S. Chang, Microporous Mesoporous Mater 157 (2012) 137e145. [29] X.M. Lou, Y.X. Zhang, J. Mater. Chem. 21 (2011) 4156e4160. [30] P.-J. Kim, Y.-W. You, H. Park, J.-S. Chang, Y.-S. Bae, C.-H. Lee, J.-K. Suh, Chem. Eng. J. 262 (2015) 683e690. [31] W. Chen, Z.A. Zhang, W.Z. Bao, Y.Q. Lai, J. Li, Y.Q. Gan, J.J. Wang, Electrochim. Acta 134 (2014) 293e301. [32] F. Yu, J.J. Zhang, Y.F. Yang, G.Z. Song, Electrochim. Acta 54 (2009) 7389e7395. [33] B. Wang, W. Al Abdulla, D.L. Wang, X.S. Zhao, Energy Environ. Sci. 8 (2015) 869e875. [34] Z.S. Wu, W.C. Ren, L. Xu, F. Li, H.M. Cheng, ACS Nano 5 (2011) 5463e5471. [35] J.-P. Jegal, K.-C. Kim, M.S. Kim, K.-B. Kim, J. Mater. Chem. A 2 (2014) 9594e9599.