Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
www.acsami.org
Ordered Mesoporous Graphitic Carbon/Iron Carbide Composites with High Porosity as a Sulfur Host for Li−S Batteries Hao Wei,† Erwin F. Rodriguez,† Adam S. Best,‡ Anthony F. Hollenkamp,§ Dehong Chen,*,∥ and Rachel A. Caruso*,†,∥ †
Particulate Fluids Processing Centre, School of Chemistry, The University of Melbourne, Melbourne, Victoria 3010, Australia CSIRO, Manufacturing, Clayton South, Melbourne, Victoria 3169, Australia § CSIRO, Energy, Clayton, Melbourne, Victoria 3168, Australia ∥ Applied Chemistry and Environmental Science, RMIT University, Melbourne, Victoria 3000, Australia
Downloaded via UNIV OF EDINBURGH on March 26, 2019 at 12:50:33 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
S Supporting Information *
ABSTRACT: The lithium−sulfur battery (LSB) is a promising candidate for future energy storage but faces technological challenges including the low electronic conductivity of sulfur and the solubility of intermediates during cycling. Additionally, current host materials often lack sufficient conductivity and porosity to raise the sulfur loading to over 80 wt %. Here, ordered mesoporous graphitic carbon/ iron carbide nanocomposites were prepared via an evaporation-induced self-assembly process using soluble resol, prehydrolyzed tetraethyl orthosilicate (TEOS), and iron(III) chloride as the carbon, silica (SiO2), and iron precursors, respectively. Graphitization and SiO2 etching were conducted simultaneously via Teflon-assisted, solid-state decomposition at high temperature. A high surface area (∼3100 m2 g−1), large pore volume (∼3.3 cm3 g−1), and graphitized carbon frame were achieved, giving a high sulfur loading (85 wt %) while tolerating volumetric expansion during discharge. Electrochemical testing of a LSB containing the composite/sulfur cathode exhibited a superior reversible capacity exceeding 1300 mAh g−1 at a moderate current (C/10) and a low decay in capacity of 9% after 500 cycles at C/5. The interaction between mesoporous graphitic carbon and sulfur is proposed. KEYWORDS: mesoporous materials, lithium−sulfur battery, long cyclability, graphitic carbon, high porosity
1. INTRODUCTION
the conversion of sulfur to Li2S during discharge may destroy the host scaffold, further accelerating the capacity decay.7,8 To address these issues, various host materials have been investigated to encapsulate sulfur and Li2Sx during cycling while simultaneously delivering enhanced conductivity and providing a fundamentally stable electrode morphology.9−14 Among many possibilities, graphitic carbon (GC) materials, such as graphite,15−17 graphene,18−20 and carbon nanotubes,21,22 exhibit excellent electronic conductivity and mechanical durability. However, because of the absence of strong chemical interactions with sulfur and polysulfides, GCs are not regarded as suitable hosts for the encapsulation and confinement of sulfur species unless the morphology and porosity are optimally designed;23,24 otherwise, Li2Sx are highly exposed to the electrolyte, triggering sulfur loss during cycling.25 On the other hand, highly GC materials tend to agglomerate during synthesis because of the strong van der
Fast-growing commercial markets for electric vehicles and scalable energy storage are driving the search for novel batteries with high energy density and prolonged cyclability.1,2 The development of a lithium−sulfur battery (LSB) is widely anticipated because of four major advantages: a high theoretical capacity (1675 mAh g−1), the natural abundance of sulfur, low cost of sulfur, and environmental friendliness.3,4 However, hindering the practical application of LSBs is the intrinsic low electronic conductivity of both sulfur and the discharged product Li2S, which leads to a low Coulombic efficiency of the cathode.5,6 Apart from these obstacles, highly soluble intermediates, that is, lithium polysulfides (LiPSs, such as Li2Sx, x = 4−6), facilitate sulfur loss through irreversible conversion to less soluble species. Migration of Li2Sx away from the cathode also triggers the so-called “shuttle effect”, where soluble species are partially reduced at the negative electrode and then migrate back to the cathode, thereby introducing inefficiency during cycling. In addition, the appreciable volumetric expansion (∼80%) that accompanies © XXXX American Chemical Society
Received: December 11, 2018 Accepted: March 14, 2019
A
DOI: 10.1021/acsami.8b21627 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces Waals force that may diminish the accessible surface area.26−28 There have been numerous attempts to design carbonaceous host materials with both sufficient porosity and a high degree of graphitization to gain synergistic advantages.29−32 Recently, graphitic carbon nanocages (GCNs) were fabricated with abundant hollow nanostructures uniformly covering a graphene backbone.17 The GCNs had mesopores 3−5 nm in diameter and were wrapped with graphitic shells serving as highly conductive chambers for polysulfides. Although the carbon matrix (composed of graphene and GCNs) facilitated enhanced electron transport, the pore volume (1 cm3 g−1) of the host material was limited, with sulfur loading not exceeding 77 wt %, and there was no tolerance for volumetric expansion during discharge. In another approach to increase the overall sulfur uptake, a three-dimensional hybrid graphene hierarchical macrostructure was obtained from a premade graphene foam nested within a graphene oxide aerogel.18 This porous structure delivered a satisfactory capacity (1000 mAh g−1) but rather poor cyclability (35% decay after 350 cycles) under high sulfur loading (14.36 mg cm−2, 89 wt % S). The suppression of capacity loss after 150 cycles was attributed to saturated polysulfides in the electrolyte. However, the downside of this suppressed capacitance loss was that the Coulombic efficiency gradually decreased, implying continuous sulfur loss from the macroporous structure. The poor electronic conductivity of amorphous carbon channels can hamper electrochemical performance, while in situ graphitization in the presence of the nanocatalyst destroys the mesostructure, leading to lower porosity and higher conductivity.33 Thus, it remains a challenge to prepare graphitic carbon materials with abundant mesoporosity and an elevated degree of graphitization. On the basis of DFT calculations, hierarchical porous materials built by Fe3C nanosheets growing on mesoporous carbon were designed and prepared, exhibiting a low specific surface area of 690 m2 g−1 and high pore volume of 6.5 cm3 g−1. Besides the enhanced adsorption ability of LiPSs, the Fe3C nanosheets also showed an excellent catalytic effect for fast transformation of LiPSs.33,34 Along with attempts to build graphene-oriented microstructures, ordered mesoporous carbon/silica (C/SiO2) hybrid materials, templated by a triblock copolymer, make for reliable carbonaceous porous structures for application as cathodic additives in LSBs.35,36 As a graphitization catalyst, iron oxide (Fe3O4) nanoparticles were successfully introduced into an amorphous C/SiO2 hybrid framework for graphitization, which allowed partially graphitic mesoporous carbon to be obtained, albeit with some loss of periodicity and porosity.37 Hence, ordered mesoporous GC materials, templated by a triblock copolymer, showed a high surface area (3100 m2 g−1) and large pore volume (3.3 cm3 g−1). This material could host sulfur up to 85 wt % while preserving extra space for volumetric expansion and providing accessible pathways for ion diffusion. However, the poor electronic conductivity of amorphous carbon somewhat hampered the actual electrochemical performance. When in situ graphitization was carried out in the presence of a nanocatalyst, this tended to destroy the mesostructure, which led to higher conductivity at the expense of lower porosity. Clearly, it remains a challenge to prepare GC materials with enhanced porosity and conductivity.33 In the present work, starting with an ordered mesostructured polymer/SiO2/Fe3O4 composite, mesoporous GC/iron carbide (Fe3C) nanocomposites were synthesized via an evaporation-induced self-assembly (EISA) procedure com-
bined with Teflon-assisted solid-state decomposition. The breakdown of Teflon at high temperatures creates an environment of highly reactive fluorine compounds that aggressively etch SiO2. In the presence of Fe3C, the residual mesoporous material was highly porous and highly graphitic.38 When incorporated into a typical LSB composite cathode, these mesoporous GC/Fe3C nanocomposites delivered excellent electrochemical performance.
2. EXPERIMENTAL SECTION 2.1. Chemicals. A triblock copolymer, poly(ethylene oxide)block-poly(propylene oxide)-block-poly(ethylene oxide), Pluronic F127 (MW = 12,600, EO106PO70EO106), tetraethyl orthosilicate (TEOS, 98%), ammonium fluoride (NH4F, >98.0%), and sulfur powder (99.98%) were purchased from Sigma-Aldrich. Phenol crystals (AR), 1-butanol (99.8%), absolute ethanol (>99.5%), formaldehyde aqueous solution (25 wt %, AR), and sodium hydroxide (NaOH, AR) were purchased from Chem-Supply. Hydrochloric acid (HCl, 32 wt %) came from Merck, iron(III) chloride hexahydrate (FeCl3·6H2O, AR) from BDH, and carbon disulfide (CS2, reagent grade) from Scharlau. All chemicals were used as received. The water used was collected from a Milli-Q Millipore academic purification system with a resistivity higher than 18.2 MΩ cm−1. 2.2. Synthesis of Resol. The resol precursor (MW < 500) was prepared according to a previous report.36 Typically, phenol (1.20 g) was melted at 42 °C in a flask and mixed with aqueous sodium hydroxide solution (0.25 g, 20 wt %) under stirring. Formaldehyde solution (2.10 g) was added dropwise, and the mixture was stirred for 1 h at 75 °C. After cooling to room temperature, the pH was adjusted to 7 with aqueous hydrochloric acid (HCl, 2 M). The solvent was removed by vacuum evaporation at 45 °C on a rotary evaporator (BUCHI Rotavapor R-215). The final product, denoted as resol solution, was dissolved in ethanol (20 wt %), while precipitated NaCl was collected and removed by centrifugation at 5000 rpm for 15 min (Beckman Coulter Allegra 25R centrifuge). 2.3. Synthesis of Ordered Mesoporous GC/SiO2/Fe3O4 and GC/SiO2/Fe3C Composites. The mesoporous GC/SiO2/Fe3O4 composites were prepared via an EISA process by coassembly of resol, oligomer silicates from tetraethyl orthosilicate (TEOS), and the triblock copolymer F127 template. In a typical synthesis, F127 (5.00 g) was dissolved in ethanol (15 mL), followed by aqueous HCl (0.2 M, 2 mL) and iron(III) chloride hexahydrate (FeCl3·6H2O, 0.108 g), before stirring at 40 °C until a clear solution was formed. TEOS (4.16 g) and resol solution (20.0 g, 20 wt %) were then added. After stirring for 2 h, the mixture was aliquoted into glass Petri dishes (140 mm in diameter) to prepare thin layers (1 mL per dish). Ethanol was evaporated at room temperature for 8 h, followed by thermopolymerization of the resin in an oven at 100 °C for 24 h. The as-made GC/ SiO2/Fe3O4 products were scraped from the Petri dishes and ground into a fine powder. Calcination was carried out at 1000 °C for 6 h under argon flow (99.99%) with a ramp rate of 1 °C min−1 from 25 °C to obtain mesoporous GC/SiO2/Fe3C composites labeled as CSI2. CSI-x denotes the mesoporous GC/SiO2/Fe3C composite, where x represents the Fe/Si molar ratio (100 times the actual value). Four GC/SiO2/Fe3C composites were prepared and labeled as CSI-0, CSI1, CSI-2, and CSI-3, with corresponding Fe/Si molar ratios of 0, 0.01, 0.02, and 0.03. 2.4. Synthesis of Ordered Mesoporous GC. CSI-x (2.0 g) was immersed in aqueous NH4F solution (0.5 M, 50 mL) to which HCl (0.5 M, 100 mL) was added to remove SiO2 and Fe3C. After shaking at 40 °C for 12 h, washing with ethanol/water (1:1 by volume), and drying, the mesoporous GC products were named C-x-HF. 2.5. Synthesis of Ordered Mesoporous GC/Fe3C Composites. The as-made GC/SiO2/Fe3O4 composite powder was mixed with Teflon powder (composite/Teflon, 1/3 by weight). The mixed powders were ground together in a mortar, transferred to an alumina boat, and then heated slowly to 1000 °C (1 °C min−1) for 6 h under argon flow to obtain mesoporous GC/Fe3C composites, denoted as CI-x-Tef. B
DOI: 10.1021/acsami.8b21627 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
samples, which was determined using TGA. Electrochemical impedance spectroscopy (EIS) measurements were performed on a Solartron frequency response analyzer (model 1255b) with the frequency ranging from 10 Hz to 100 kHz. Cyclic voltammetry (CV) was conducted on a Solartron battery test unit (model 1470) at different sweeping rates (0.05−2 mV s−1) between cutoff voltages of 1.7−3.0 V (vs Li/Li+).
CI-2-Tef containing Fe3C (2.0 g) was washed in aqueous HCl (0.5 M, 100 mL) to remove Fe3C. After shaking at 30 °C for 12 h, washing with ethanol/water (1:1 by volume), and drying, the mesoporous GC product was labeled as C-2-Tef. 2.6. Loading Sulfur in Mesoporous GC and GC/Fe3C Composites. To infiltrate sulfur into the mesoporous microspheres, S solution (8 mL, 0.25 g mL−1 in CS2) was added to the mesoporous materials (0.5 g) before the solvent was evaporated with continuous stirring. The dried mixture was sealed in a Teflon-lined autoclave and heated at 155 °C for 24 h to obtain the final composites, which were labeled as C-x-HF/S, CI-x-Tef/S, and C-x-Tef/S. 2.7. Material Characterization. The morphology of the samples was characterized on an FEI Quanta 200F environmental scanning electron microscope (SEM) at an accelerating voltage of 15 kV without Au coating, while energy-dispersive X-ray spectroscopy (EDS) was performed using an INCA SDD X-ray microanalysis system. The sample microstructure was evaluated on an FEI Tecnai F20 transmission electron microscope (TEM) operating at 200 kV. Powder X-ray diffraction (XRD) patterns were acquired on a Bruker D8 Advance Diffractometer using Cu Kα radiation. The diffractometer was set at a 40 kV working voltage and 40 mA working current, with samples scanned from 5° to 80° (2θ). Low-angle (0.5° to 5° in 2θ) X-ray diffraction was recorded on a Bruker D8 Advanced Diffractometer installed with an antiscatter screen over the sample plates (to avoid the parasitic low-angle background scattering). Thermogravimetric analysis (TGA) was conducted on a Mettler Toledo TGA/SDTA851e thermogravimetric analyzer heating from 25 to 900 °C at a ramp of 10 °C min−1 under a flow of oxygen or nitrogen (30 mL min−1). Nitrogen sorption isotherms were measured at −196 °C using a Micromeritics TriStar 3000 Surface Area and Porosity Analyzer. Prior to measurement, calcined samples were degassed at 150 °C for at least 8 h on a vacuum line (
