carbon nanofiber webs for high performance lithium–sulfur batteries

carbon nanofiber webs for high performance lithium–sulfur batteries

Accepted Manuscript Novel interlayer made from Fe3C/carbon nanofiber webs for high performance lithium-sulfur batteries Jian-Qiu Huang, Biao Zhang, Zh...

15MB Sizes 0 Downloads 16 Views

Accepted Manuscript Novel interlayer made from Fe3C/carbon nanofiber webs for high performance lithium-sulfur batteries Jian-Qiu Huang, Biao Zhang, Zheng-Long Xu, Sara Abouali, Mohammad Akbari Garakani, Jiaqiang Huang, Jang-Kyo Kim PII:

S0378-7753(15)00386-9

DOI:

10.1016/j.jpowsour.2015.02.140

Reference:

POWER 20769

To appear in:

Journal of Power Sources

Received Date: 21 December 2014 Revised Date:

7 February 2015

Accepted Date: 25 February 2015

Please cite this article as: J.-Q. Huang, B. Zhang, Z.-L. Xu, S. Abouali Mohammad Akbari Garakani, J. Huang, J.-K. Kim, Novel interlayer made from Fe3C/carbon nanofiber webs for high performance lithiumsulfur batteries, Journal of Power Sources (2015), doi: 10.1016/j.jpowsour.2015.02.140. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Graphical abstract

1

ACCEPTED MANUSCRIPT

Novel interlayer made from Fe3C/carbon nanofiber webs for high performance lithium-sulfur batteries

RI PT

Jian-Qiu Huang, Biao Zhang, Zheng-Long Xu, Sara Abouali, Mohammad Akbari Garakani, Jiaqiang Huang, Jang-Kyo Kim*

SC

Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, PR China. *Corresponding author. E-mail address: [email protected] (J. K. Kim)

M AN U

Abstract

A new freestanding Fe3C/carbon nanofiber (CNF) film is developed using a facile one-pot electrospinning method as an interlayer for high performance lithium-sulfur

TE D

(Li-S) batteries. The interlayer placed between the separator and the sulfur cathode plays many synergistic roles, offering (i) a number of macropores within the nanofiber

EP

web to facilitate ion transport and electrolyte penetration, (ii) nitrogen-containing functional groups that entrap soluble polysulfides by strong interatomic attraction, and

AC C

(iii) much enhanced electron/ion transfer due to the high electrical conductivity of the CNF web. The battery delivers an excellent specific discharge capacity of 893 mA·h/g after 100 cycles, maintaining 76% of its initial capacity of 1177 mA·h/g. These values are among the highest for those reported recently with similar nanocarbon-based interlayers in terms of rate capability and cyclic stability.

1

ACCEPTED MANUSCRIPT Keywords: Lithium-sulfur batteries; Electrospinning; Interlayer; Carbon nanofiber; Nitrogen-containing functional groups

1. Introduction

RI PT

Among many different types of energy storage devices, rechargeable lithium batteries have demonstrated obvious advantages over other types of batteries in terms of their high specific energies and high energy densities [1]. Li ion batteries (LIBs)

SC

have been most-widely studied and practically employed in important applications,

M AN U

such as portable electronics, electric vehicles (EVs) and large-scale grid energy storage. However, the current LIBs are still lacking energy densities and long cyclic life to satisfy demanding applications like next generation EVs, necessitating the need

TE D

for developing new electrode materials with higher capacities. The lithium-sulfur (Li-S) battery is considered one of the most promising alternatives to LIBs because sulfur possesses a high theoretical specific capacity of 1675 mA·h/g − 5 times higher

EP

than the current LIBs − and a remarkable theoretical energy density of 2567 W·h/kg.

AC C

In addition, other advantages, like low cost, natural abundance and nontoxicity, make sulfur very attractive as positive electrodes [2-4]. Despite the aforementioned considerable benefits, however, there are several critical issues that hinder practical application of Li-S batteries. The poor electrical conductivity of sulfur and the creation of reaction products, like Li2S and Li2S2, prevent full utilization of active materials in the electrode. Lithium polysulfide intermediates (Li2Sx, x = 4~8) formed 2

ACCEPTED MANUSCRIPT during the charge/discharge process are soluble in the electrolyte and migrate between the anode and the cathode, causing the so-called ‘shuttle effect’, subsequently leading to capacity fade and low Coulombic efficiencies. Another problem is the volume

RI PT

change of ~80% during charge and discharge, which may lead to cell failure [1-8]. Extensive efforts have been made to address these issues so as to improve the

SC

electrochemical performance of Li-S batteries. Previous studies have focused on

M AN U

developing advanced cathode materials, including composites containing sulfur and conductive polymers [9-11], carbon-based composites [8, 12] and adding adsorptive materials like mesoporous silica or titania in the cathode [12-14]. Another attractive solution to the above issues, especially to mitigate the shuttle

TE D

effect, is to introduce an interlayer between the Li anode/separator and the sulfur cathode [15-18]. Free-standing multi-walled carbon nanotube (MWCNT) papers [15],

EP

microporous carbon papers [16] and networked carbonized membranes [17, 18] have

AC C

been synthesized as the interlayers. They served as a barrier to limit soluble polysulfide diffusion and localize the active material within the cathode side. This in turn facilitated re-utilization of the entrapped active materials in the following cycles, thus improving the capacity retention and cyclic stability of the cells. However, the complicated and expensive synthesis processes and a low polysulfide adsorption ability of the interlayers are still the major constraint to widespread application of the

3

ACCEPTED MANUSCRIPT interlayers. Herein, we report a facile and one-pot synthesis approach to fabricate freestanding, carbon nanofiber (CNF) interlayers containing Fe3C nanoparticles (Fe3C/CNFs). It is demonstrated that the highly porous web structure of CNFs

RI PT

effectively immobilized active materials and hindered the dissolution of polysulfides

SC

for the benefit of enhanced the electrochemical performance of the Li-S battery.

M AN U

2. Experimental 2.1. Materials and preparation of interlayers

The interlayer was prepared by one-pot electrospinning of the mixture of polymer and Fe precursors. The materials and the process employed were essentially similar to

TE D

our recent studies [19, 20]. Typically, 1.0 g polyacrylonitrile (PAN) (Mw=150,000, Aldrich) was dissolved in 20 ml N-dimethylformamide (DMF) and the solution was

EP

vigorously stirred on a hot plate at 80 ℃ for 3 h. 1.5 g Iron (Ⅲ) acetylacetonate was

AC C

added to the solution as the precursor for Fe3C, which was kept stirring overnight. The mixture was used for spinning on an electrospinner (KATO Tech. Co.) at a flow rate of 1.0 ml h-1. A positive voltage of 18 kV was applied between the needle of diameter 1.2 mm and the drum collector of diameter 15 cm rotating at 0.6 m min-1. The electrospun PAN fiber film containing Fe particles was collected from the drum collector, which was subjected to heat treatments. The heat treatment involved

4

ACCEPTED MANUSCRIPT stabilization at 220 ℃ for 3 h in an air-circulated oven, followed by carbonization at 650

for 1 h at a heating rate of 2

min-1 in a tube furnace of N2 atmosphere to

obtain CNF webs containing Fe3C nanoparticles (designated as Fe3C/CNFs). For

SC

procedure without adding Iron (Ⅲ) acetylacetonate.

RI PT

comparison, CNFs without Fe3C (neat CNFs) were also synthesized using the same

M AN U

2.2. Characterization

The phases of CNF webs with and without Fe3C nanoparticles were detected on a high resolution X-ray diffraction (XRD) system (PW1830, Philips) with Cu Kα radiation for 2θ in the range of 10-65°. The nitrogen adsorption and desorption

TE D

isotherms were obtained at 77 K with an automated adsorption apparatus (Micrometritics, ASAP 2020). The structures and morphologies were examined on a

EP

scanning electron microscope (SEM, 6390 and 6700F) and a transmission electron

AC C

microscope (TEM, JEOL 2010). The elemental maps were obtained using a dispersive spectrometer (EDS) linked to SEM 6390. CNF webs mixed with potassium bromide (KBr) by grinding and pressed into pellets were analyzed by Fourier transform infrared spectroscopy (FTIR) (Perkin Elmer Specrum One). X-ray photoelectron spectroscopy (XPS, PHI5600 by Physical Electronics, Inc.) was conducted using a monochromatic Al Kα X-ray at 14 kV. The electrical conductivities of neat CNF and

5

ACCEPTED MANUSCRIPT Fe3C/CNF films were measured on a four-probe resistivity/Hall system (HK5500PC, Bio-Rad).

RI PT

2.3. Electrochemical tests

Electrochemical tests were carried out using CR2032 coin cells prepared from

SC

different CNF films. The cathode slurry was prepared by mixing micro-sized sulfur

M AN U

powders (purum p.a., ≥99.5% Sigma-Aldrich), carbon black (super P) and polyvinylidene fluoride (PVDF) binder with a weight ratio of 7:2:1 in N-methy1-2-pyrrolidone (NMP). After magnetic stirring overnight, the slurry was casted on an aluminum foil and dried at 60

in an air-circulated oven. The cathode

TE D

electrodes were cut into discs of 14 mm in diameter with a high sulfur loading of 2.3-2.8 mg/cm2. The CNF films were used directly as the freestanding interlayer.

EP

CR2032 coin cells were assembled in an argon-filled glove box using the Li foil as

AC C

the anode, the prepared sulfur cathode and the electrolyte (1.0 M lithium bis-trifluoromethane sulfonylimide (LiTFSI) and 1 wt% LiNO3 in 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME) at a volume ratio of 1:1). The interlayers were cut into discs of 15 mm in diameter, 90-120 µm in thickness and 3.9-4.1 mg in weight without any other treatments, and placed between the polyethylene separator (Celgard 2400) and the cathode. Figure 1 shows the schematic cell configuration of

6

ACCEPTED MANUSCRIPT the Li-S battery. The coin cells were charge/discharge cycled between 1.5 and 2.8 V on a LAND 2001 CT battery tester at room temperature. The electrochemical performance was

RI PT

studied on a CHI660c electrochemical workstation. Cyclic voltammetry (CV) tests were performed within the voltage window of 1.5-2.8 V at a scan rate of 0.1 mV/s.

SC

The electrochemical impedance spectra (EIS) were obtained at a constant perturbation

M AN U

amplitude of 5 mV in the frequency range between 0.1 Hz and 100k Hz. To study the changes in morphology and elemental compositions of the interlayers after charge/discharge cycles, the interlayers were collected after 40 and 100 cycles to ex situ examine under SEM and TEM. The cycled cells in the fully charged state at 2.8 V

TE D

were dissembled in the argon-filled glove box and the interlayers were washed with DME to remove the remaining salts and naturally dried in the glove box at room

AC C

EP

temperature overnight [21].

3. Results and discussion 3.1. CNF structures and morphologies The XRD patterns of the CNF webs with and without Fe3C nanoparticles are shown in Figure 2a. The pattern of the neat CNF interlayer revealed a strong and broad peak centered at ~24.9°, showing the amorphous structure. Upon incorporation

7

ACCEPTED MANUSCRIPT of Fe3C nanoparticles in CNFs, a number of peaks related to Fe3C crystals appeared in the range of 2θ = 35-60° with a strongest peak at around 45° due to the (031) plane. Their chemical structures were investigated using the FTIR spectra as shown in

RI PT

Figure 2b. The adsorption band at 1192 cm-1 corresponds to C-C-C (O) while C-N and C=N bands are located at 1284 and 1374 cm-1, respectively. The peak at 1581

SC

cm-1 shows a combination of C=C and C=N vibration [22]. Three different forms of nitrogen atoms and their roles in polysulfide entrapment are discussed in Section 3.2.

M AN U

An additional peak at the wavenumber ~1400 cm-1 in Fe3C/CNF is assigned to the vibration of C-Fe [23].

Figure 3 presents the change in morphology of the Fe3C/CNF interlayer taken

TE D

before and after cycles. The interlayer consisted of beadless nanofibers with an average diameter of ~300 nm (Figure 3a). These randomly intermingled fibers with a

EP

large surface area of 62 m2/g (Figure S1) and macropores in-between allowed the

AC C

electrolyte to penetrate and offered pathways for fast transfer of electrons and ions, while making the electrolyte, active materials and the interlayer in close contact with each other [17]. In addition, the volumetric change arising from the conversion of sulfur to Li2S during the charge/discharge process could be accommodated by the macroporous space. The SEM images of the interlayers taken after cycles (Figures 3b and 3c) presented significant increases in fiber diameter and surface roughness due to

8

ACCEPTED MANUSCRIPT the deposition of polysulfide species on them. The diameters of Fe3C/CNFs after 40 or 100 cycles increased to about 400 nm. The surface roughness of the fibers tended to increase with increasing cycle due to the attachment of more active materials. The

RI PT

contrasting morphologies indicate that the polysulfides diffused from the cathode were entrapped by the interconnected fiber webs during cycles. The uniform

SC

distribution of sulfur on the nanofibers was confirmed by the corresponding SEM

M AN U

elemental maps in Figure S2.

The HRTEM images in Figure 4 further support the above findings. In contrast to the neat CNFs which consisted only of amorphous carbon (Figures 4a, S3a and 3b), the Fe3C/CNF contained homogenously dispersed Fe3C nanoparticles (Figure 4b).

TE D

The selected area electron diffraction (SAED) pattern given in the inset was in agreement with the corresponding planes of the Fe3C crystals. These particles were

EP

surrounded by 8-10 layers of graphitic carbon (Figure 4c) as a result of the catalytic

AC C

effect of Fe promoting graphitization during the carbonization at 650℃ [24, 25]. The d-spacing between the layers was 0.337 nm, corresponding to the (002) plane in graphite and the Fe3C nanoparticle had a lattice distance 0.197 nm, in agreement with the (112) plane of Fe3C. Internal conductive networks were formed by these graphitic carbon layers together with the Fe3C nanoparticles, giving rise to a significantly enhanced electrical conductivity of 0.21 S/cm, more than three orders of magnitude

9

ACCEPTED MANUSCRIPT higher than 8.3×10-5 S/cm of the neat CNF . After cycles, a thin layer of active materials of thickness 20-40 nm is clearly seen on the Fe3C/CNF surface (Figure 4d),

3.2. Mechanisms of polysulfide entrapment

RI PT

consistent with the SEM observations (Figures 3b and 3c).

SC

The mechanisms of how a simple interlayer entrapped polysulfide intermediates

M AN U

were studied using the XPS spectra of the Fe3C/CNF interlayer, as shown in Figure 5. The general spectra obtained before the cyclic tests and after 40 cycles presented four main peaks located at 285.0, 400.0, 532.0 and 708.0 eV which correspond to C1s, N1s, O1s and Fe2p, respectively (Figure 5a). Two prominent new peaks, S2s at 231.0 eV

TE D

and S2p at 168.0 eV, appeared along with weaker intensities of the underlying species like C1s, N1s and Fe2p after cycles, indicating the deposition of active materials on

EP

the interlayer. The small peak F1s at 689.0 eV arose from the remnant electrolyte.

AC C

The deconvoluted N 1s spectra (Figure 5b) presented three nitrogen peaks before cycles which are located at 401.6, 399.8 and 398.3 eV, corresponding to the graphitic N, pyrrolic N and pyridinic N, respectively. Another nitrogen peak N-O appeared at 406.3 eV arising from the chemisorbed nitrogen oxides [26]. The graphitic N was formed when highly coordinated nitrogen atoms were bonded to three carbon atoms in the bulk of a graphene layer (Figure 5d). Embedded in the six-membered ring, the

10

ACCEPTED MANUSCRIPT graphitic N offered extra free electrons to give rise to a higher electrical conductivity of carbon [22]. The strong presence of pyridinic N before cycles means excellent adsorption capability of carbon nanofiber webs with much enhanced interatomic

RI PT

attraction to lithium polysulfides [27, 28]. The pyrrolic N is located in a π-conjugated system by substituting a carbon atom of the five-membered ring, which may also

SC

enhance the surface adsorption although its intensity was not as strong as that of the

M AN U

pyridinic N. After 40 charge/discharge cycles, the binding energies of the main peak down-shifted due to the interactions between the lithium polysulfides and nitrogen functional groups [28, 29]. In addition, the significant reduction in pyridinic N intensity after cycles may be attributed to the coverage of more accumulated

TE D

polysulfides, indicating a stronger interaction of polysulfides with pyridinic N than with pyrrolic N. The N-O groups almost totally disappeared because of the deposition

EP

of polysulfides on the interlayer. The capability of nitrogen-containing functional

AC C

groups to entrap polysulfides has been proposed previously by model calculations, showing much higher binding energies of the Li2Sx/pyridinic N and Li2Sx/pyrrolic N bonds than the Li2Sx/graphitic C bond [30]. The polysulfides diffused from the cathode are captured by these functional groups through the strong interactions, facilitating re-utilization of the entrapped active materials in the following cycles and consequently improving the capacity retention and cyclic stability of the cells. In

11

ACCEPTED MANUSCRIPT summary, both the above XPS analysis and binding energy calculation [30] demonstrate that the nitrogen-containing functional groups present on the Fe3C/CNF

suppressing the dissolution of polysulfides.

RI PT

web interlayer played an important role in immobilizing active materials and thus

The deconvoluted spectrum of S 2p obtained after cycles (Figure 5c) presented

SC

two peaks at 163.7 and 164.8 eV corresponding to the Li-S and S-S bonds of

M AN U

polysulfides, respectively. The additional peaks located between the binding energies 168 and 172 eV is indicative of lithium salt (LiTFSI) present on the interlayer surface [31].

TE D

3.3. Electrochemical performance

The electrochemical test results of the Li-S batteries with and without an

EP

interlayer are shown in Figure 6. The cyclic voltammetry (CV) curves of the cell with

AC C

a Fe3C/CNT interlayer are given in Figure 6a. Two main reduction peaks located at ~2.23 and ~1.92 V correspond to the conversion from elemental sulfur to polysulfides after reactions, Li2Sx where x = 4~8, and the formation of final lithium sulfides, Li2S and Li2S2, respectively. The oxidation peak stabilized at ~2.56 V, suggesting the conversion of Li2S and Li2S2 into higher-order soluble polysulfides [32]. The gradual increase in current at the second catholic peak at ~1.92 V in the following cycles is

12

ACCEPTED MANUSCRIPT indicative of the utilization of more sulfur [33]. The almost identical CV curves for the 10th and 11th cycles are a reflection of the improvement of electrochemical reversibility and cycleability. The discharge curves of the modified cell (Figure 6b)

RI PT

exhibited two plateaus, in accordance with the two typical peaks in the CV profiles. No other plateaus were found in the curves, implying that the interlayers did not react

SC

with the active materials during the charge/discharge process. The slight overpotential

M AN U

disappeared after the 2nd cycle, probably due to the rearrangement of active materials from the original positions to the surface of Fe3C/CNFs [15]. Both the charge/discharge plateau curves in the 2nd and the 100th cycles almost overlapped, also a reflection of sound cyclic stability of the Li-S battery.

TE D

The cyclic performances of three different cells are compared in Figure 6c. Compared to the cell without an interlayer (measured at 100 mA/g), the cells with

EP

both the neat CNF and Fe3C/CNF web interlayers delivered much higher capacities

AC C

and better cyclic stability (at 200 mA/g). A remarkable initial specific discharge capacity of 1177 mA·h/g was achieved for the cell with the Fe3C/CNF interlayer, which was maintained at 893 mA·h/g after 100 cycles with high capacity retention of 76% of the initial value. The corresponding capacities of the cell with a CNF interlayer were 1169 and 711 mA·h/g under the same condition with capacity retention of 61%. The excellent electrochemical performance of these cells is

13

ACCEPTED MANUSCRIPT attributed to the synergies arising from the several important functional modifications by the interlayer, including (i) creation of macropores among the interconnected nanofiber webs providing pathways for ion transport and electrolyte penetration, (ii)

RI PT

nitrogen-containing functional groups present on the CNFs had strong atomic interactions with soluble polysulfide species to entrap them. Evidently, the cell with

SC

the Fe3C/CNF web interlayer had higher capacities and more stable cyclic behavior

M AN U

than the one with the CNF interlayer, especially at high current rates (Figure S4), because of the synergy arising from (iii) the much better electron/ion transfer associated with the higher electrical conductivity of the Fe3C/CNF interlayer containing uniformly-dispersed conductive Fe3C nanoparticles and the surrounding

TE D

graphitic carbon layers (Figure 3c). The Coulombic efficiencies of all these cells were maintained above 95%, probably thanks to the addition of LiNO3 in the electrolyte

EP

[16, 34].

AC C

Furthermore, it is also noted that the cell with the Fe3C/CNF interlayer presented much better high-rate capacities than those with the neat CNF interlayer or without one (Figure 6d). The former cell achieved particularly impressive capacities at high rates of 1.0 and 2.0 A/g where the other two cells presented almost negligible capacities. When the rate was reverted to 100 mA/g, the capacity of the former cell recovered to 951 mA·h/g, indicating excellent reversibility. Apart from the very poor

14

ACCEPTED MANUSCRIPT performance of the cell without an interlayer, the huge difference in rate capability of the two cells with different interlayers may also stem from the large difference in electrical conductivity of the interlayers, i.e. 0.21 vs 8.3×10-5 S/cm. This interesting

RI PT

revelation has a practical implication in that a high electrical conductivity is an important parameter of an interlayer to optimize the electrochemical performance of

SC

Li-S batteries.

M AN U

The electrochemical performance of the cells with and without interlayers was evaluated by EIS and galvanostatic charge/discharge measurements. The EIS spectra in Figure 7 present a series of semicircles in the high frequency region and the corresponding interfacial charge-transfer resistance, Rct, is summarized in Table S1.

TE D

The cell without an interlayer had a high Rct value of 55.3 Ω. After the insertion of a Fe3C/CNF interlayer, the Rct value decreased to 14.2 Ω because the CNF webs

EP

provided conductive pathways for charge transfer. It is interesting to note that the

AC C

impedance further decreased to 10.3 Ω after 20 cycles and then reverted to a higher resistance after further cycles. The initial reduction may be attributed to the rearrangement of the active materials, deposition at more electrochemical sites [35], and gradual activation of the Fe3C/CNF interlayer that nitrogen functional groups on all the fibers work to trap polysulfides. The increase in charge transfer resistance to 13.7 Ω after 100 cycles was expected because thicker insulting active materials were

15

ACCEPTED MANUSCRIPT entrapped on the surface of CNFs, consistent with the SEM observation in Figure 3c. Nevertheless, even after 100 cycles the resistance of the cell with a Fe3C/CNF interlayer was much lower than that of the fresh cell without an interlayer, confirming

RI PT

its beneficial effects.

Apart from entrapping polysulfides, the interlayer also played a role in protecting

SC

the cathode. Without the insertion of interlayers, the cathode would suffer from a

M AN U

thick passivation layer of irreversible Li2S (Figure S5), which may block the utilization of interior active materials and increase the diffusional resistance, consequently leading to the loss of active materials [36]. For the cells with an interlayer, however, CNFs provides massive electrochemical sites, avoiding the

TE D

formation of the insulated passivation layer on the cathode and retaining the high capacity and stability of the battery.

EP

The comparison of electrochemical performance between the current work and

AC C

other carbon-based interlayers is presented in Table 1. It is shown that most advanced Li-S battery systems have achieved a remarkable progress by inserting an interlayer, especially those made of microporous carbon papers [16], which retained over 1000 mA·h/g after 100 cycles at a rate of 1675 mA/g, due to the strong attraction of polysulfides by the microporous carbon. Among these interlayer systems, the Fe3C/CNF interlayer developed in this study were proven to deliver equally excellent

16

ACCEPTED MANUSCRIPT or even better capacities and retention than some earlier studies, with a high capacity retention of 76% and a low degradation rate of 0.24% per cycle at 200 mA/g.

RI PT

4. Conclusions

In this study, we introduced a facile and one-pot approach to synthesize

SC

freestanding electrospun Fe3C/CNF web interlayer. It possessed a very high electrical

M AN U

conductivity of 0.21 S/cm for much reduced interfacial charge transfer resistance. The interlayer was inserted between the separator and the sulfur cathode to enhance the electrochemical performance of Li-S batteries. Many ameliorating effects were revealed from the Fe3C/CNF web interlayer. They include the macropores among the

TE D

interconnected fibers for easy ion transport and electrolyte permeation, much improved electron/ion transfer due to the high electrical conductivity of interlayer

EP

consisting of carbon matrix, Fe3C nanoparticles and graphitic carbon layers, and the

AC C

nitrogen-containing functional groups present on carbon nanofibers. The strong atomic attraction between the nitrogen and polysulfides was crucial to immobilize active materials and constrain the dissolution of polysulfides through the enhanced surface adsorption of carbon fibers. Thanks to these functional advantages, the cells with the Fe3C/CNF interlayer presented an excellent specific discharge capacity of 893 mA·h/g after 100 cycles, maintaining 76% of its initial capacity of 1177 mA·h/g.

17

ACCEPTED MANUSCRIPT

Acknowledgements This project was financially supported by the Research Grants Council of Hong Kong

RI PT

SAR (GRF Project code: 613612). The authors also appreciate the technical assistance

SC

from the Materials Characterization and Preparation Facilities (MCPF) of HKUST.

M AN U

References

[1] P. G. Bruce, S. A. Freunberger, L. J. Hardwick and J. Tarascon, Nat. Mater. 11 (2011) 19-29.

1256-1263.

TE D

[2] M. Barghamadi, A. Kapoor and C. Wen, J. Electrochem. Soc. 160 (2013)

[3] X. Ji and L. F. Nazar, J. Mater. Chem. 20 (2010) 9821-9826.

EP

[4] Z. W. Seh, W. Li, J. J. Cha, G. Zheng, Y. Yang, M. T. McDowell, P. Hsu and Y.

AC C

Cui, Nat. Commun. 4 (2013) 1331. [5] D. Bresser, S. Passerini, and B. Scrosati, Chem. Commun. 49 (2013) 10545-10562.

[6] Y. V. Mikhaylik, and J. R. Akridge, J. Electrochem. Soc. 151 (2004) 1969-1976. [7] S. Lu, Y. Cheng, X. Wu, and J. Liu, Nano Lett. 13 (2013) 2485-2489. [8] A. Manthiram, Y. Fu, and Y. S. Su. Acc. Chem. Res. 46 (2012) 1125-1134.

18

ACCEPTED MANUSCRIPT [9] H. Chen, W. Dong, J. Ge, C. Wang, X. Wu, W. Lu and L. Chen, Sci. Rep. 3 (2013) 1910. [10] W. Zhou, Y. Yu, H. Chen, F. J. DiSalvo, and H. D. Abruña, J. Am. Chem. Soc.

RI PT

135 (2013) 16736-16743.

[11] X. Liang, Y. Liu, Z. Wen, L. Huang, X. Wang and H. Zhang, J. Power Sources

SC

196 (2011) 6951-6955.

M AN U

[12] S. Evers and L. F. Nazar, Acc. Chem. Res. 46 (2012) 1135-1143.

[13] X. Ji, S. Evers, R. Black and L. F. Nazar, Nat. Commun. 2 (2011) 325. [14] S. Evers , T. Yim, and L. F. Nazar, J. Phys. Chem. C 116 (2012) 19653-19658. [15] Y. S. Su and A. Manthiram, Chem. Commun. 48 (2012) 8817-8819.

TE D

[16] Y. S. Su and A. Manthiram, Nat. Commun. 3 (2012) 1166. [17] S. H. Chung and A. Manthiram, Adv. Mater. 9 (2014) 1360-1365.

EP

[18] S. H. Chung and A. Manthiram, ChemSusChem 6 (2014) 1655-1661.

AC C

[19] B. Zhang, Z. L. Xu, Y. B. He, S. Abouali, M. Akbari Garakania, E. Kamali Heidari, F. Kang, J. K. Kim, Nano Energy 4 (2014) 88-96.

[20] Z. L. Xu, B. Zhang and J.K. Kim, Nano Energy 6 (2014) 27–35. [21] M. Wang, Y. Zhang, H, Zhang and H. Zhang, Chempluschem 79 (2014) 919-924. [22] B. Zhang, Y. Yu, Z. L. Xu, S. Abouali1, M. Akbari1, Y. B. He, F. Kang and J. K. Kim, Adv. Energy Mater. 4 (2014) 1301448.

19

ACCEPTED MANUSCRIPT [23] Z. Wen, S. Ci, F. Zhang, X. Feng, S. Cui, S. Mao, S. Luo, Z. He and J. Chen, Adv. Mater. 24 (2012) 1399-1404. [24] J. H. Hafner, M. J. Bronikowski, B. R. Azamian, P. Nikolaev, A. G. Rinzler, D. T.

RI PT

Colbert, K. A. Smith and R. E. Smalley, Chem. Phys. Lett. 296 (1998) 195-202. [25] A. Tanaka, S. H. Yoon and I. Mochida, Carbon 42 (2004) 591-597.

SC

[26] S. Biniaka, G. Szymański, J. Siedlewskia and A. Świa̧ tkowskib, Carbon 35 (1997)

M AN U

1799-1810.

[27] S. Kundu, W. Xia, W. Busser, M. Becker, D. A. Schmidt, M. Havenith and M. Muhler, Phys. Chem. Chem. Phys. 12 (2010) 4351-4359.

[28] F. Sun, J. Wang, H. Chen, W. Li, W. Qiao, D. Long and L. Ling, ACS Appl.

TE D

Mater. Interfaces 5 (2013) 5630–5638. [29] J. Yang, J. Xie, X. Zhou, Y. Zou, J.Tang , S. Wang, F. Chen and L. Wang, J. Phys.

EP

Chem. C 118 (2014) 1800–1807.

AC C

[30] Y. Qiu, W. Li, W. Zhao, G. Li, Y. Hou, M. Liu, L. Zhou, F. Ye, H. Li, Z. Wei, S. Yang, W. Duan, Y. Ye, J. Guo and Y. Zhang, Nano Lett. 8 (2014) 4821–4827.

[31] X. Tao, J. Wang, Z. Ying, Q. Cai, G. Zheng, Y. Gan, H. Huang, Y. Xia, C. Liang, W. Zhang and Y. Cui, Nano Lett. 9 (2014) 5288-5294. [32] G. Zhou, L. C. Yin, D. W. Wang, L. Li, S. Pei, I. R. Gentle, F. Li, and H. M.Cheng, ACS Nano 7 (2013) 5367–5375.

20

ACCEPTED MANUSCRIPT [33] R. Elazari ,G. Salitra, A. Garsuch, A. Panchenko and Doron Aurbach, Adv. Mater. 23 (2011) 5641–5644.

Chem. Commun. 49 (2013) 11107-11109.

RI PT

[34] T. G. Jeonga, Y. H. Moona, H. H. Chunb, H. S. Kimc, B. W. Choc and Y. T. Kima,

[35] S. H. Chung and A. Manthiram. Electrochim. Acta. 107 (2013) 569-576.

SC

[36] S. E. Cheon, K. S. Ko, J. H. Cho, S. W. Kim, E. Y. Chin and H. T. Kim. J.

M AN U

Electrochem. Soc. (2003) 800-805.

[37] G. Ma, Z. Wen, J. Jin, M. Wu, X. Wu, J. Zhang, J. Power Sources 267 (2014)

AC C

EP

TE D

542–546.

21

ACCEPTED MANUSCRIPT Table 1 Comparison of electrochemical performance of different interlayers.

Discharge

Initial

Interlayer

current

discharge

Cycle

reversible

Residual

materials

rate

Capacity

number

capacity

(mA/g)

(mA h/g)

MWCNT

838

1446

100

855

59.1%

Microporous Carbon

1675

1176

100

1000

85.0%

0.15%

[16]

Carbonized Egg Membrane

168

1327

100

1000

75.4%

0.25%

[17]

Carbonized Leaf

336

1320

100

850

64.4%

0.36%

[18]

Acetylene Black

168

1491

50

1062

71.2%

0.58%

[34]

Graphene oxide/carbon Black

168

1260

100

894

71.0%

0.29%

[37]

Fe3C/CNF

200

893

75.9%

0.24%

Current Study

rate

Reference

per cycle

(mA·h/g)

AC C 1

[15]

RI PT

0.41%

SC

M AN U

TE D 100

EP

1177

Degradation Retention

ACCEPTED MANUSCRIPT Figure captions

Figure 1 Schematic cell configuration of the Li-S battery with an interlayer.

RI PT

Figure 2 (a) XRD patterns and (b) FTIR spectra of neat CNF and Fe3C/CNF interlayers.

Figure 3 SEM images of Fe3C/CNF interlayers (a) before, and after (b) 40 and (c)

SC

100 charge/discharge cycles.

M AN U

Figure 4 TEM images of (a) neat CNF and (b) Fe3C/CNF before cyclic test with a SAED pattern in inset of (b); (c) Fe3C nanoparticles enclosed by graphitic carbon layers; and (d) Fe3C/CNF after 40 cycles showing a 20-40 nm thick layer of active

TE D

materials.

Figure 5 XPS spectra of Fe3C/CNF interlayer: (a) general spectra obtained before and after 40 cycles; (b) deconvoluted spectra N 1s obtained before and after cyclic test; (c)

EP

deconvoluted spectra S 2p after cyclic test; and (d) schematic of entrapping

AC C

polysulfides via surface interaction Figure 6 Electrochemical test results of Li-S batteries: (a) cyclic voltammetric profiles at a scan rate of 0.1 mV/s; (b) charge/discharge voltage profiles determined at 200 mA/g; (c) cyclic performance; and (d) rate capacities of cells without and with interlayers made from neat CNF and Fe3C/CNF webs. Figure 7 Electrochemical impedance spectroscopy (EIS) plots of Li-S cells without 1

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

and with interlayer after different number of cycles.

2

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Figure 1 Schematic cell configuration of the Li-S battery with an interlayer.

3

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 2 (a) XRD patterns and (b) FTIR spectra of neat CNF and Fe3C/CNF interlayers.

4

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 3 SEM images of Fe3C/CNF interlayers (a) before, and after (b) 40 and (c) 100 charge/discharge cycles.

5

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

materials.

EP

Figure 4 TEM images of (a) neat CNF and (b) Fe3C/CNF before cyclic test with a SAED pattern in inset of (b); (c) Fe3C nanoparticles enclosed by graphitic carbon layers; and (d) Fe3C/CNF after 40 cycles showing a 20-40 nm thick layer of active

6

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

7

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

TE D

Figure 5 XPS spectra of Fe3C/CNF interlayer: (a) general spectra and (b) deconvoluted spectra N 1s obtained before cyclic test and after 40 cycle; (c) deconvoluted spectra S 2p after 40 cycles; and (d) schematic of entrapping

AC C

EP

polysulfides via surface interaction.

8

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

TE D

Figure 6 Electrochemical test results of Li-S batteries: (a) cyclic voltammetric profiles at a scan rate of 0.1 mV/s; (b) charge/discharge voltage profiles determined at

AC C

EP

200 mA/g; (c) cyclic performance; and (d) rate capacities of cells without and with interlayers made from neat CNF and Fe3C/CNF webs.

9

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 7 Electrochemical impedance spectroscopy (EIS) plots of Li-S cells without and with interlayer after different numbers of cycles.

10

ACCEPTED MANUSCRIPT Highlights

RI PT SC M AN U



TE D



EP



Conductive freestanding Fe3C/CNF webs are synthesized by electrospinning. Nitrogen-containing functional groups confine the dissolution of polysulfides. The cell with a Fe3C/CNF interlayer shows high capacities and stable cyclicality. The cell retains a high capacity of 893 mA·h/g after 100 cycles at 200 mA/g.

AC C



1

ACCEPTED MANUSCRIPT Supplementary data

Novel interlayer made from Fe3C/carbon nanofiber webs for high performance lithium-sulfur batteries

RI PT

Jian-Qiu Huang, Biao Zhang, Zheng-Long Xu, Sara Abouali, Mohammad Akbari Garakani, Jiaqiang Huang, Jang-Kyo Kim*

Department of Mechanical and Aerospace Engineering, The Hong Kong University of

AC C

EP

TE D

M AN U

SC

Science and Technology, Clear Water Bay, Kowloon, Hong Kong, PR China. *Corresponding author. E-mail address: [email protected] (J. K. Kim)

Fig. S1 Pore size distribution and nitrogen adsorption/desorption isotherm curves of neat CNF and Fe3C/CNF interlayers. The Fe3C/CNF interlayer showed a large surface area of 62 m2/g and a pore volume of 0.17 cm3/g, which are 20-30% higher than those of the neat CNF counterpart (48 m2/g and 0.14 cm3/g, respectively), probably due to the diffusive iron crystals at high temperatures and thus the creation of additional pores [1]. 1

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Fig. S2 Elemental maps of Fe3C/CNF interlayer after 40 cycles.

2

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Fig. S3. TEM images of a neat CNF taken at different magnifications.

3

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. S4. Cyclic performance of cells with Fe3C/CNF and CNF interlayers at high

TE D

current rate of 1 A/g.

AC C

EP

Two cycles were tested at 100 mA/g for initial activation of the electrodes before changing to 1 A/g.

4

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

TE D

Fig. S5. SEM images of (a) the fresh cathode, and the cathodes (b) without and (c)

AC C

EP

with an interlayer after 40 cycles at a full charge state.

5

ACCEPTED MANUSCRIPT Table S1 EIS data of the cells without and with a Fe3C/CNF interlayer after different numbers of cycles.

R (Ω)

Without interlayer

Before cycles

4.3

Fe3C/CNF interlayer

Before cycles

5.2

Fe C/CNF interlayer 3

1

4.0

Fe3C/CNF interlayer

20

4.1

Fe3C/CNF interlayer

40

Fe3C/CNF interlayer

100

e

R (Ω) ct

-2

i (mA/cm ) 0

55.3

0.23

14.2

0.91

13.9

0.93

RI PT

Cycles

SC

Materials

1.25

5.1

13.5

0.96

3.9

13.8

0.94

M AN U

10.3

TE D

Re: Sum of the resistance arising from active materials, electrolyte and contact interface between the active materials and current collector.

AC C

EP

Rct: Interfacial charge-transfer resistance. i0: Exchange current density (i0 = RT / nFRct).

6

ACCEPTED MANUSCRIPT Reference

AC C

EP

TE D

M AN U

SC

RI PT

[1] Y. Chen, X Li, X. Zhou, H, Yao, H. Huang, Y. W. Mai and L. Zhou, Energy Environ. Sci. 7 (2014) 2689-2696.

7