Conducting Polymer Hydrogel Integrated

Jun 25, 2019 - The free-standing chemically integrated monolithic carbon .... carbon network and the improved charge transfer at the MOF–sulfur ... ...
0 downloads 0 Views 9MB Size
Letter Cite This: Nano Lett. 2019, 19, 4391−4399

pubs.acs.org/NanoLett

Metal−Organic Frameworks/Conducting Polymer Hydrogel Integrated Three-Dimensional Free-Standing Monoliths as Ultrahigh Loading Li−S Battery Electrodes Borui Liu,† Renheng Bo,† Mahdiar Taheri,‡ Iolanda Di Bernardo,† Nunzio Motta,∥ Hongjun Chen,† Takuya Tsuzuki,‡ Guihua Yu,§ and Antonio Tricoli*,†

Downloaded via UNIV OF SOUTHERN INDIANA on July 17, 2019 at 08:44:33 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



Nanotechnology Research Laboratory, Research School of Electrical, Energy, and Materials Engineering and ‡Laboratory of Advanced Nanomaterials for Sustainability, Research School of Electrical, Energy, and Materials Engineering, Australian National University, Canberra, ACT 2601, Australia § Materials Science and Engineering Program and Department of Mechanical Engineering, The University of Texas at Austin, Austin, Texas 78712, United States ∥ School of Chemistry, Physics, and Mechanical Engineering, Queensland University of Technology, Brisbane, QLD 4001, Australia S Supporting Information *

ABSTRACT: The lithium−sulfur (Li−S) system is a promising material for the nextgeneration of high energy density batteries with application extending from electrical vehicles to portable devices and aeronautics. Despite progress, the energy density of current Li−S technologies is still below that of conventional intercalation-type cathode materials due to limited stability and utilization efficiency at high sulfur loading. Here, we present a conducting polymer hydrogel integrated highly performing free-standing three-dimensional (3D) monolithic electrode architecture for Li−S batteries with superior electrochemical stability and energy density. The electrode layout consists of a highly conductive three-dimensional network of N,P codoped carbon with welldispersed metal−organic framework nanodomains of ZIF-67 and HKUST-1. The hierarchical monolithic 3D carbon networks provide an excellent environment for charge and electrolyte transport as well as mechanical and chemical stability. The electrically integrated MOF nanodomains significantly enhance the sulfur loading and retention capabilities by inhibiting the release of lithium polysulfide specificities as well as improving the charge transfer efficiency at the electrolyte interface. Our optimal 3D carbon-HKUST-1 electrode architecture achieves a very high areal capacity of >16 mAh cm−2 and volumetric capacity (CV) of 1230.8 mAh cm−3 with capacity retention of 82% at 0.2C for over 300 cycles, providing an attractive candidate material for future high-energy density Li−S batteries. KEYWORDS: Li−S batteries, high loading, three-dimensional electrodes, conducting polymer hydrogel, metal−organic frameworks cm−1) of sulfur at room temperature;6 and (ii) the formation of soluble high-order lithium polysulfide (LiPS) intermediates (Li2Sx, 4 ≤ x ≤ 8). The former decreases the utilization efficiency of the active materials in the cathode, resulting in low capacity and poor rate capability.7 The latter (LiPS) are responsible for a recurrent shuttle effect during the discharge/ charge cycles, which leads to short battery life and potential safety issues.8,9 In the past decade, significant efforts have focused on overcoming these challenges. Most successful approaches include nano/microstructuring and functionalization of the electrodes and separators with meso/microporous carbon,7 graphene,10 conductive polymers,11,12 carbon nanotubes,13,14 oxides,15−20 metal−organic frameworks,21−23 metal disul-

T

he rapidly increasing demand for portable energy storage in mobile applications such as electrical vehicles and wearable/portable electronics is driving the search toward battery systems and materials with significantly higher energy density than commercial technologies.1−3 Among promising candidates, the lithium−sulfur (Li−S) system has recently attracted significant attention as a powerful alternative to established lithium ion intercalation-based cathode materials such as LiCoO2, LiMn2O4, and LiFePO4, having limited specific capacities of usually less than 200 mAh g−1. The Li−S battery technology relies on the reaction of elemental sulfur (S8) to lithium sulfide (Li2S), which provides a high theoretical gravimetric capacity and gravimetric energy density of 1675 mAh g−1 and 2567 Wh kg−1, respectively. Moreover, the earthabundance, cost effectiveness, and environmental friendliness of sulfur make it also a commercially competitive material, suitable for large-scale production.4,5 Two major challenges still hinder the development and commercialization of Li−S batteries: (i) the very low electrical conductivity (5 × 10−30 S © 2019 American Chemical Society

Received: March 12, 2019 Revised: May 29, 2019 Published: June 25, 2019 4391

DOI: 10.1021/acs.nanolett.9b01033 Nano Lett. 2019, 19, 4391−4399

Letter

Nano Letters

Figure 1. Synthesis and characterizations of the monolithic 3D carbon (CIMC)-MOF electrodes. (a) Schematic illustration of the synthesis processes of the 3D carbon-MOF electrodes; step 1, gelation of SP@PANi CPH on the melamine foams; step 2, carbonization of the melamine foam impregnated with SP@PANi- resulting in a 3D monolithic carbon network; step 3, integration of ZIF-67 or HKUST-1 nanoparticles into the 3D carbon networks. (b) Digital images of as-synthesized SP@PANi CPH and of the impregnated melamine foam cut in the shape of a disk with a diameter of 15 mm. (c) Digital image of the resulting N,P-doped monolithic 3D carbon electrodes with a diameter of 10 mm obtained by carbonization of the impregnated foam disks (b). (d) SEM images of the 3D carbon electrodes. (e) TEM analysis of representative 3D carbon electrodes. Region 1: HRTEM area. Region 2: elemental mapping area. (f) HRTEM image of the meso/microporous ligaments in the 3D carbon network. (g) EDX mapping of C, N, and P elements in the 3D carbon network.

fides,24−30 and their hybrids.23,31−33 However, the areal sulfur mass loading of most Li−S architectures remains in the order of 2 mg cm−2. Considering the typical 4 mAh cm−2 areal specific capacity of conventional Li-ion batteries and their higher operational voltage (3.5 V) than Li−S batteries (2.1 V), a minimum sulfur areal mass loading of 6 mg cm−2 with an average specific capacity of about 1000 mAh g−1 would be required to achieve comparable areal capacities.34,35 This sulfur loading (>6 mg cm−2) is beyond the feasible stable loading capacity of existing Al-foil-based planar electrode architectures for Li−S batteries.36 Recently, the design of three-dimensional (3D) electrode architectures has been increasingly investigated as a potential approach to increase the sulfur loading.35 A variety of innovative 3D electrode architectures have been reported for Li−S batteries. Miao et al. used an artificial pseudo-2D network of cotton threads to achieve a high sulfur loading density of 6.7 mg cm−2 with areal and specific capacities of over 7 mAh cm−2 and 1100 mAh g−1 for 50 dischargingcharging cycles, respectively.37 Yuan et al. fabricated a freestanding hierarchical carbon nanotube (CNT)-sulfur paper electrode with high sulfur loading.28 There, superlong CNTs were employed as the building blocks of the long-range conductive network and 3D scaffold, while short multiwalled CNTs enhanced the short-range electrical conduction. A high sulfur loading of 17.3 mg cm−2 with an areal capacity of 15.1 mAh cm−2 was achieved. It is worth noting that the volumetric energy density (Ev) and volumetric capacity (CV) of these electrodes are essential for numerous mobile applications. The assembly of highly conductive CNTs, carbon nanofibers (CNFs), and graphene with other functional building blocks has been recently explored and demonstrated for the

fabrication of 3D current collectors.30,38−42 Composite 3D electrodes, such as CNF paper with layered WS2 and CNT paper with metal−organic framework (MOF) microparticles, have successfully been implemented, demonstrating their potential for trapping of the soluble polysulfide species and improving the long-term cyclability of Li−S batteries. Overall, incorporation of various MOFs in Li−S cells has resulted in enhanced chemical and electrochemical properties for hundreds of cycles.19,22,33 However, the utilized MOFs, having lattice openings smaller than both S8 molecules (0.68 nm)22 and Li2S6,43 may negatively affect the sulfur loading potential of these electrodes. Furthermore, many 3D electrode architectures have relied on physical stacking methods, such as vacuum filtration, which may result in limited domain interconnectivity, and thus insufficient mechanical and electrical stability.35,42,44 Here, we present a conducting polymer hydrogel-integrated free-standing 3D monolithic electrode architecture for Li−S batteries, offering very high sulfur-loading capacity as well as excellent long-term electrochemical and mechanical stability. Our electrode layout consists of a highly conductive 3D hierarchical network of N,P codoped carbon synergistically complemented by homogeneously grafted MOFs nanodomains. The optimal incorporation of MOFs with cobalt (ZIF-67) and copper (HKUST-1) metal cations increases the sulfur-loading capacity of the bare carbon electrodes from 15 mg cm−2 (tap density, ρ ∼ 1.15 g cm−3) to 17.1 mg cm−2 (ρ ∼ 1.32 g cm−3) and 18.8 mg cm−2 (ρ ∼ 1.45 g cm−3), respectively, while simultaneously enhancing the capacity retention for over 300 cycles with a highest areal capacity of 17.6 mAh cm−2 and volumetric capacity of 1354 mAh cm−3. Detailed electrochemical and physical characterizations reveal 4392

DOI: 10.1021/acs.nanolett.9b01033 Nano Lett. 2019, 19, 4391−4399

Letter

Nano Letters the superior role of the HKUST-1 in inhibiting the LiPS shuttling effect, resulting in the best long-term performance. We believe that this highly performing material architecture and its facile, low-cost fabrication method provide an attractive candidate for the design of the next generation of high-energy density Li−S batteries. Figure 1 shows a schematic representation of the fabrication of the 3D monolithic carbon electrodes with integrated ZIF-67 and HUKST-1 MOFs. First, a conductive polymer hydrogel (CPH, SP@PANi) of Super P carbon black nanoparticles and polyaniline was synthesized (also see Supporting Information) and impregnated into commercially available melamine foams (Figure 1a). The SP@PANi possesses an interconnected 3D hierarchically porous architecture, where the SP nanoparticles serve as a cost-effective electrical conductivity enhancer (Figure 1b and Figure S1). The melamine foam has a microstructured spongelike porous morphology consisting of formaldehyde-melamine-sodium bisulfite copolymer (Figure S2). It provides outstanding mechanical stability and elasticity, even after the subsequent high-temperature pyrolysis step. These SP@PANi impregnated melamine foams were easily tailored into various shapes and dimensions such as a disc of 15 mm in diameter (Figure 1b), while maintaining their mechanical strengths. The free-standing chemically integrated monolithic carbon electrodes (3D CIMC) were fabricated by pyrolysis and eventual carbonization of the SP@PANi impregnated melamine foams at 1000 °C for 3 h in an inert Ar atmosphere. The impregnated foam disks maintained their round shapes while shrinking from 15 to 10 mm in diameter due to the loss of volatile compounds (Figure 1c). X-ray microcomputed tomography (micro-CT) confirmed a continuous monolithic domain with a macro-porosity (≥3 μm) of 6−7% (Figure S3). Scanning electron microscope (SEM) analysis (Figure 1d) provides further insights revealing a significant mesoporosity with pore size in the range of 200−500 nm well distributed through the 3D carbon network. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) imaging show that this 3D carbon network is comprised of meso/microporous ligaments (Figure 1e,f). These monolithic electrodes also possess a very high specific surface area (SSA) of 483 m2 g−1 and a Barrett− Joyner−Halenda (BJH) adsorption cumulative pore volume of 0.213 cm3 g−1 (Figure S4). Notably, this SSA is more than 2 orders of magnitude higher than that (e.g., 1.3 cm2 g−1) commonly reported for 3D battery electrodes.35,44,45 The high SSA of our monolithic carbon electrodes is attributed to the chemical integration of the SP@PANi and melamine foam, which results in a 3D hierarchically porous nanostructure with accessible macro-, meso-, and microporosity.46 Furthermore, upon carbonization the electrical conductivity was greatly enhanced (Figure S5). The N,P codoping of the 3D carbon electrodes was confirmed by energy dispersive X-ray spectroscopy (EDX) (Figure 1g) and X-ray photoelectron spectroscopy (XPS) (Figure 2). Elemental mapping shows a uniform distribution of C, N, and P elements over the nanostructured carbon network. Importantly, the N, P codoping does not only improve the electrical conductivity of the electrodes but also helps to trap soluble lithium polysulfides in the electrode by strong surface polar adsorption. Analysis of the XPS survey spectra (Figure 2a) indicates a major carbon component of 87 atom % and the presence of the P and N dopants with a concentration of 3.5

Figure 2. (a) Survey XPS spectrum of the surface chemical composition of N, P codoped 3D carbon networks. Core level spectra: (b) C 1s, (c) P 2p, and (d) N 1s XPS spectra of the N, P codoped 3D carbon networks.

and 1.5 atom %, respectively. A small fraction of oxygen is also observed and attributed to adventitious adsorbed species, present in the atmosphere, such as CO2 and H2O.52 The highresolution C 1s core level spectrum provides some structural insights on the different chemical environments of the carbon atoms in the 3D structure (Figure 2b). The main, asymmetric peak at 284.7 eV originates from sp2 bonded C, corresponding to the graphitic planar domains, also observed by HRTEM (Figure S6). The components at 285.6, 287.5, and 290.7 eV are assigned to CP, CN, and CO bonds, respectively.46,47 The line shape of P 2p can be deconvoluted into two peaks, each associated with a spin−orbit split 2p doublet (Figure 2c). The lower binding energy component (2p3/2) at 131.9 eV is attributed to PC bonds, further confirming the chemical bonding between dopant and hosting structure, rather than the formation of segregated potassium agglomerates. The higher binding energy component (2p3/2) at 133.1 eV indicates some bonding between potassium and adventitious oxygen. The N 1s core level can be fitted by four synthetic components (Figure 2d), corresponding to N in pyridinic (398.4 eV), pyrrolic (400.4 eV), and graphitic (401.2 eV) configuration, and NO groups (403.1 eV).46,47 The predominance of substitutional nitrogen in graphitic sites indicates the local planarity of the graphitic domains. The minor pyrrolic and pyridinic components instead originate from N hosted in a nonplanar configuration, which involves a rehybridization of the carbon lattice from an sp2 to an sp3 configuration. The latter N species are expected to be predominantly located in proximity of high-curvature regions of the 3D network, where they can facilitate and foster the bending of the graphitic planes. Consequently, the n-type codoping of the 3D electrodes with P and N was confirmed and is expected to provide free excess electrons in the amorphous carbon network, which are expected to improve the 4393

DOI: 10.1021/acs.nanolett.9b01033 Nano Lett. 2019, 19, 4391−4399

Letter

Nano Letters

Figure 3. Morphological and composition analysis of the MOFs-containing monolithic 3D carbon electrodes. SEM images of ZIF-67 containing 3D carbon electrodes at (a) low and (b) high magnifications. (c) TEM image and EDX elemental mapping (C, Co, and N elements) of the ZIF-67 containing 3D carbon electrodes. SEM images of HKUST-1 containing 3D carbon electrodes at (d) low and (e) high magnifications. (f) TEM image and EDX elemental mapping (C, Cu, and N elements) of HKUST-1 containing 3D carbon electrodes.

electrical conductivity and result in higher rate battery performance.48 This highly conductive monolithic carbon network provides an excellent hierarchical 3D matrix for incorporation and electrical interconnection of the MOF nanodomains. The latter are expected to enhance the long-term sulfur retention and loading capacity of the 3D electrodes. Here, ZIF-67 and HKUST-1 nanograins with a MOF opening size of 0.34 and 0.90 nm and thus smaller and larger than that of the sulfur (S8) molecules (0.68 nm), respectively, were incorporated into the 3D carbon electrodes (Figure 1a). Figure 3 shows the morphology and composition of the resulting 3D MOFcarbon monoliths. Upon MOF growth (see experimental section), a homogeneous distribution of ZIF-67 and HKUST-1 nanoparticles with a size of 200−400 nm and 20−100 nm, respectively, were observed in intimate contact with the 3D carbon network, as also confirmed by X-ray diffraction (XRD) analysis (Figure S7). The ZIF-67 formed mostly large

individual cubic or truncated dodecahedron nanoparticles (Figure 3a,b), whereas the HKUST-1 resulted in smaller interconnected agglomerates of quasi-spherical primary particles (Figure 3d,e). The mass fractions of ZIF-67 and HKUST-1 in the 3D carbon electrodes were determined to be about 8 wt % and about 12 wt % by thermogravimetric analysis (TGA) (Figure S8) and directly weighing the samples. The transmission electron microscopy (TEM) analysis and energy dispersive X-ray (EDX) elemental mapping revealed that the ZIF-67 and HKUST-1 MOF domains were welldistributed and intimately embedded in the 3D carbon network structure (Figure 3c,f, Figures S9 and S10), as also confirmed by XRD analysis. Upon ZIF-67 and HKUST-1 incorporation, the BET SSAs and BJH pore volume increased significantly from 483 to 634 and 685 m2 g−1, and from 0.213 to 0.438 and 0.488 cm3 g−1, respectively. The increase in mesopore volume is attributed to the extra nanoscale pores formed by the MOF nanoparticles (Figure S4). 4394

DOI: 10.1021/acs.nanolett.9b01033 Nano Lett. 2019, 19, 4391−4399

Letter

Nano Letters

Figure 4. Electrochemical characterizations of MOFs-containing monolithic 3D carbon electrodes. (a,b) CV and EIS plots of a representative ZIF67-containing electrode on the first, second, third, fifth, and 10th cycles. (c,d) CV and EIS plots of a representative HKUST-1 containing electrode on the first, second, third, fifth, and 10th cycles. (e) Comparison of rate performance of the ZIF-67 and HKUST-1 containing electrodes from 0.05C to 0.6C with a sulfur loading density of about 17.1 mg cm−2 and about 18.8 mg cm−2, respectively. (f) Rate-dependent discharge−charge profiles of the ZIF-67 and HKUST-1 containing electrodes from 1.7 to 2.8 V vs Li+/Li.

which are surprisingly lower than that of the bare carbon monoliths (Figure S15b). This is surprising and important and notable as both ZIF-67 and HKUST-1 feature very poor electrical conductivity.21−23 This effect is tentatively attributed to an excellent electrical connection between the MOF nanodomains and the doped carbon network and the improved charge transfer at the MOF−sulfur interface. One of the main advantages of HKUST-1, which led to the exploration of its performance in our study, is its well suited pore size of 0.9 nm.33 This is, for instance, larger than that of ZIF-67 (0.34 nm) and allows the access of elemental sulfur (0.68 nm) into the HKUST-1 porous structure. The pore size of HKUST-1 is also sufficiently small to hinder the transport of long-chain highly soluble polysulfides (>0.4 nm), and thus it is within the required window for elemental sulfur uptake and hindering of polysulfide shuttling. HKUST-1 features relatively good thermal, chemical, and mechanical stability, making it an attractive candidate for sulfur cathodes.33 Previous studies have observed that HKUST-1 containing battery cells had the same electrochemical performance in typical Li−S battery electrolytes,21 indicating a good chemical stability of HKUST-1. Here, we have conducted XRD of the pure HKUST-1 particles before and after soaking in the Li−S electrolyte to evaluate its chemical stability. The same XRD patterns were observed before and after soaking in the electrolyte (Figure S16a) indicating a stable crystal composition. Furthermore, electron microscope analysis of the electrodes reveals a similar morphology of the HKUST-1 grains in the 3D electrodes, before and after battery cycling (Figure S17), further indicating a stable HKUST-1 nanocomposite structure in the Li−S operation conditions. It should also be noted that while the Cu2+/Cu+ standard redox potential is outside the discharge/ charge voltage window of the Li−S system, the standard electrode table potentials are measured in aqueous electrolytes and cannot be directly applied to the used Li−S electrolyte. We have performed additionally CV measurements of a non-

The suitability and stability of this 3D MOF-carbon hybrid nanostructure for sulfur loading and electrolyte soaking was initially investigated by physical and chemical analysis. After sulfur impregnation with a loading density of over 20 mg cm−2, both ZIF-67 and HKUST-1 containing electrodes show an excellent uptake without any noticeable sulfur segregation. The TEM and EDX elemental mapping (Figure S11−S13) confirm that sulfur was successfully impregnated in both MOF-carbon electrodes (denoted as S/3D ZIF67-CIMC and S/3D HKUST1-CIMC) with no obvious degradation of the composite electrode structure. In line with the electron microscope analysis, upon S-loading the BET SSAs and pore volume of the ZIF-67 and HKUST-1 electrodes decreased to 4.69 and 18.3 m2 g−1, and 0.00589 and 0.0389 cm3 g−1, respectively (Figure S4). These further confirm the uptake of elemental sulfur in the meso/micropores of the 3D structure. The electrochemical performance of these monolithic 3D MOF-carbon electrodes was evaluated for long-term battery cycling with a standard coin cell configuration. Notably, in the electrochemical measurements the achievable sulfur loading capacity increased from 15 mg cm−2 (ρ ∼ 1.15 g cm−3) of the bare 3D carbon electrodes to 17.1 mg cm−2 (ρ ∼ 1.32 g cm−3) and 18.8 mg cm−2 (ρ ∼ 1.45 g cm−3) for the ZIF-67 and HKUST-1 enhanced architectures, respectively. Figure 4a,c shows the cyclic voltammetry (CV) curves of both MOFcontaining electrodes for the first 10 representative cycles. Both compositions show two main cathodic peaks, which are attributed to the conversion of sulfur to long-chain LiPSs at 2.3 V and their further conversion to short-chain Li2S2/Li2S at 2.0 V. Similar peaks were observed on the bare 3D carbon electrodes (Figure S14 and Figure S15a). The sharp peak profiles indicate that even at such high sulfur loading, these monolithic MOF-carbon electrodes maintain an excellent electrical conductivity. The electrical impedance spectroscopy (EIS) analysis (Figure 4b,d) further confirms small charge transfer impedances for both MOF-containing electrodes, 4395

DOI: 10.1021/acs.nanolett.9b01033 Nano Lett. 2019, 19, 4391−4399

Letter

Nano Letters

Figure 5. Cycling performance of MOF-containing 3D carbon electrodes and postcycling characterizations. (a) Comparison of cycling performance of the ZIF-67 and HKUST-1 containing electrodes at 0.2C rate. The sulfur areal loading is about 17.1 and 18.8 mg cm−2, respectively. The dotted lines indicate a typical areal capacity of conventional LIBs, and the shaded area indicates the areal capacity of the bare monolithic 3D carbon electrodes. (b) SEM images show structural variation of a representative 3D carbon electrode before and after cycling. (c) UV−visible absorption spectra of Li2S6 solutions (diluted by 10 times before measurements) by soaking postcycling electrodes (discharged state) in DOL/DME (1:1, vol.) mixture solvent for about 1 h. The 260 nm band is attributed to the S62− or S32− species. The 280 and 340 nm bands are both attributed to the S62− species. The 310 nm band is attributed to the S62− or S42− species. The 420 nm band is attributed to the S42− species.

agree well with the peak current positions in the CV curves. This indicates an improved high-rate capability with outstanding reversibility over the bare 3D carbon electrodes (Figure S15c,d) and other high sulfur-loading content electrodes based on 3D carbon materials such as CNTs and graphene.35,53,54 The long-term cycling stability of these electrodes was evaluated by galvanostatic discharge/charge cycling at 0.2C (Figure 5a). After 300 cycles, the capacity decreased to 9.8 mAh cm−2 (571 mAh g−1, CV ∼ 754 mAh cm−3) and 14.3 mAh cm−2 (757 mAh g−1, CV ∼ 1100 mAh cm−3) for the ZIF-67 and HKUST-1 containing electrodes, respectively. This corresponds to a capacity retention of 61% and 82% and about 0.13% and 0.06% decay per cycle, respectively. Notably, this is a significant improvement over the commercial LIBs (about 4 mAh cm−2) and the bare carbon monolith electrodes (about