Sulfur Diffusion within Nitrogen Doped Ordered ... - ACS Publications

Metal-sulfur batteries represent a potential low cost alternative to Li ion ... how sulfur is infiltrated into the porous carbons.26. Doped carbon ...
22 downloads 0 Views 1MB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Interfaces: Adsorption, Reactions, Films, Forces, Measurement Techniques, Charge Transfer, Electrochemistry, Electrocatalysis, Energy Production and Storage

Sulfur Diffusion within Nitrogen Doped Ordered Mesoporous Carbons Determined by in-situ X-ray Scattering Yanfeng Xia, Chao Wang, Ruipeng Li, Masafumi Fukuto, and Bryan D. Vogt Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01375 • Publication Date (Web): 05 Jul 2018 Downloaded from http://pubs.acs.org on July 6, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Sulfur Diffusion within Nitrogen Doped Ordered Mesoporous Carbons Determined by in-situ X-ray Scattering Yanfeng Xia†, Chao Wang‡, Ruipeng Li§, Masafumi Fukuto§, Bryan D Vogt‡, * †

Department of Polymer Science, Goodyear Polymer Center, The University of Akron, 170 University Circle, Akron, OH 44325, United States ‡

Department of Polymer Engineering, The University of Akron, 250 S Forge St, Akron, OH 44325, United States

§

National Synchrotron Light Source II, Brookhaven National Laboratory, Upton, New York 11973, USA

KEYWORDS: nanopores; Li-S battery; FDU-15; in-situ SAXS

ABSTRACT: The low intrinsic conductivity of sulfur necessitates conductive additives, such as mesoporous carbons, to the cathode to enable high performance metal-sulfur batteries. Simultaneous efforts to address polysulfide shuttling have introduced nitrogen doped carbons to provide both conductivity and suppressed shuttling due to its strong interaction with sulfur. The strength of this interaction will likely impact the ability to fill the mesopores with sulfur via melt infusion. Here, we systematically investigate how nitrogen doping influences the rate that molten sulfur can infiltrate the mesopores and the overall extent of pore filling of highly ordered mesoporous doped carbons using in-situ small angle xray scattering (SAXS). The similarity in electron density between molten sulfur and the soft carbon framework of the mesoporous material leads to a precipitous decrease in the scattered intensity associated with the ordered structure as voids are filled with sulfur. As the nitrogen doping increases from 1 to 20 at%, the effective diffusivity of sulfur in the mesopores decreases by an order of magnitude (2.7 × 10-8 cm/s to 2.3 × 10-9 cm/s). The scattering becomes nearly invariant within 20 min of melt infiltration at 155 °C for all but the most doped carbon, which indicates that sub-micron sized mesoporous carbon particles can be filled rapidly. Additionally, the nitrogen doping decreases the sulfur content that can be accommodated within the mesopores from 95 % filled without doping to only 64% filled with 20 at% N as determined by the residual scattering intensity. The sulfur does not crystallize within the mesopores of the nitrogen doped carbons, which is further indicative of the strong interactions between the nitrogen species and sulfur that can inhibit polysulfide shuttling. in-situ SAXS provides insights into the diffusion of sulfur in mesopores and how the surface chemistry of nitrogen doped carbon appears to significantly hinder the infiltration by sulfur.

INTRODUCTION Metal-sulfur batteries represent a potential low cost alternative to Li ion batteries,1 especially for grid and stationary energy storage.2-4 These sulfur batteries are challenged by a variety of intrinsic issues associated with sulfur,5, 6 its electrochemistry,7, 8 and the electrolyte.9 In particular, the low instrinsic electrical conductivity of sulfur necessitates the dispersion of the sulfur within porous conductor.10, 11 The ideal electrode design is generally a balance between the limited loading of sulfur in small pores and limited sulfur utilization in large pores, especially at high charge/discharge rates.12 In addition to the trade-off between rate and total energy density, the lifetime of room temperature metal sulfur batteries is limited by the polysulfide shuttle effect where the long chain polysulfides in the redox reaction cascade are soluble in the electrolyte.13 These soluble components can be transported from the cathode and deposited on the sodium metal anode, which results in loss of active material.7 A variety of approaches have been taken to minimize the loss of active material on the anode including an interlayer between the separator and

cathode to prevent deposition on the anode,14 polymerization of the sulfur,15 electrolyte engineering to minimize solubility and transport of the polysulfides,16 and components included with the sulfur in the cathode to effectively weakly bind the polysulfides to prevent shuttling while maintaining activity of the sulfur.17 Of particular interest has been the use of doped porous carbons in the cathode to provide both conductive pathways and surfaces that can interact with the polysulfides.18 At high doping (>20 at%), the discharge product has been demonstrated to be limited through binding of the sulfur to nitrogen sites on the carbon.19 These strong interactions that limit polysulfide shuttling may also impact the sulfur loading and its distribution within porous supports, which are key factors in the performance of the metal-sulfur battery. Characterization of the filling and transport of fluids within mesoporous channels is challenged by the requirements to directly visualize the fluids across multiple length scales as the mesopores are generally less than 10 nm, while the transport lengths are generally on the order of microns. Grosso and co-workers utilized the changes in optical properties of mesoporous silica films to

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

visualize the egress of water through the mesopores.20 Gravimetric adsorption and desorption isotherms from water condensation have also been used to provide insights into the transport within the mesopores.21 To better understand the filling of mesopores by condensed gases, small angle neutron scattering has been effectively employed where the condensed gas within the pores matches the neutron scatttering length density of the framework.22, 23 This methodology has afforded new insights into the distribution of porosity within these ordered materials. Analogously, small angle X-ray scattering (SAXS) has been used to probe the filling of anodized alumina with small molecule liquid24 and polymers.25 A similar approach has been used to examine the distribution of sulfur within porous carbon frameworks with SAXS as the electron density of sulfur and carbon is similar.26 These studies provided insight into differences that are dependent on the specifics of how sulfur is infiltrated into the porous carbons.26 Doped carbon supports for metal-sulfur batteries are effective at suppressing polysulfide shuttling.18 However, the influence of this doping on the distribution of sulfur within conductive supports, which is a critical component to determining performance of metal – sulfur batteries, has not been adequately investigated. Here, we systematically investigate how nitrogen doping of well defined ordered mesoporous carbons influences the kinetics of melt infusion of sulfur and the final loading using in-situ SAXS. As the N-doping content in the mesoporous carbon increases, the pores are filled slower and the apparent equilibrium fraction of the mesopores filled decreases under 155 °C melt infusion conditions. The effective diffusivity of sulfur during the infiltration decreases from 10-8 cm/s to 10-9 cm/s as the nitrogen content increases from 1 at % to 20 at %. For the undoped carbon, the pore filling occurs too rapidly to resolve by these measurements. These results provide insight into how nitrogen doping impacts sulfur diffusion in the melt (with the caveat that the microporosity in the walls decreases monotonically with increasing N content and could potentially influence the sulfur filling kinetics)and illustrates that the standard 12 h time frame10 for melt infusion of sulfur into porous carbons is likely much longer than required for homogeneous active materials.

EXPERIMENTAL SECTION Materials Pluronic F127 (Mw = 12,600 g/mol, PEO106-PPO70PEO106), sodium hydroxide (NaOH, >97%), potassium hydroxide (KOH, >85%), ethanol (>99%), acetone (>99%), tetraethyl orthosilicate (TEOS, >98%), phenol (>99%), formaldehyde (ACS reagent, 37 wt% in H2O, contains 10-15% methanol as stabilizer), hydrochloric acid (HCl, ACS reagent, 37%), melamine (>99%), were all purchased from Sigma Aldrich. Sublimed sulfur (S, >99.5%) was purchased from Alfa Aesar. A low molecularweight phenolic resin (resol) precursor was synthesized via base-catalyzed condensation as previously reported.27

Page 2 of 11

Thermal stabilized polyethylene terephthalate (PET) sheets (125 μm thick, Terphane, Inc.) were used as substrates for roll-to-roll processing. Nitrogen (UN1066, >99.95%) was obtained from Praxair. All reagents were used as received unless otherwise noted. Synthesis of Nitrogen Doped Ordered Mesoporous Carbon (NOMC) In a typical preparation, 96 g Pluronic F127 was dissolved in 80 g of 0.2 M HCl and 240 g of ethanol at 45 °C and vigorously stirred for 2 h. 120 g of resol28 solution (50 wt% in ethanol) that was synthesized by NaOH catalyzed condensation of phenol and formaldehyde and 150 g of TEOS were added to the Pluronic Fl27 solution. This solution was stirred for 2 h at 45 °C for hydrolysis of TEOS prior to casting on a PET substrate using a doctor blade casting with an initial wet thickness of 400 μm at 50 cm/min on a roll-to-roll line.27 The residual solvent was evaporated from the as-cast film at 25 °C for 12 h. Subsequently, the resol was crosslinked at 100 °C for 12 h. Stresses generated during the crosslinking enabled the delamination of the self-assembled film from the substrate. The freestanding material was calcined at 350 °C by heating at 1°C/min and holding at 350°C for 1 h under N2 to obtain the mesoporous resol/silica composites. The obtained mesoporous resol/silica composites were mixed with the nitrogen precursor, melamine, and ground into fine powders in 1 min using an electric mortar (MIRA grinder CP-GR-101). One undoped control was synthesized using the same procedure without adding melamine. The mixture was carbonized in N2 at 800 °C for 1 h (heating at 5 °C/min). The mass ratio of the mesoporous resol/silica composite to melamine controls the nitrogen doping level.29 The carbonized powder was immersed into 6 M KOH solution (ethanol:H2O=50:50 v/v) for 24 h at 90°C in order to remove the silica from the framework. The resultant doped meosporous carbon was washed by deionized water >20 times using soxhlet extraction to remove the residual KOH. The washed nitrogen doped ordered mesoporous carbon (NOMC) was dried at 80 °C overnight. Preparation of NOMC/Sulfur Mixture The dried NOMC and sulfur were mixed with an agate mortar and pestle for 5 min in a mass ratio of 1 to 3. This provided more sulfur than theoretically required to completely fill the mesopores. The mixture was transferred into a capillary tube (1.0 mm O.D., 0.01 mm wall thickness, thin-walled capillary tube for X-ray diffraction, Charlessuper Inc.). The capillary tube was sealed using a glass alcohol burner (Grobet File Co of America Inc.). The pores in NOMC remain unfilled during this mixing process as sulfur is solid at room temperature. Characterization The pore filling by sulfur was characterized in-situ using simultaneous small angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS) using the 11-BM CMS beamline at the National Synchrotron Light Source II (NSLS-II) at Brookhaven National Laboratory in Upton,

2 ACS Paragon Plus Environment

Page 3 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

NY. 13.5 keV X-rays (wavelength (λ) = 0.0918 nm) were used for these measurements. The sample-to-detector distance was 1.98 m for SAXS and 0.231 m for WAXS. A Dectris Pilatus 300K CCD detector (pixel size = 172 μm × 172 μm) and a Photonic Science ImageStar 135 mm CCD detector (pixel size = 101.7 μm × 101.7 μm) were used for SAXS and WAXS, respectively. Silver behenate was used to calibrate the scattering vector (q) based on its primary reflection at q = 1.076 nm-1. A temperature controlled, auto-sampler stage was used to cycle through the different NOMC/sulfur samples during the SAXS/WAXS measurements. The temperature was ramped to 155 °C at 15 °C/min and held at 155 °C. Each scattering measurement was around 7 s for SAXS and 17 s for WAXS. With the cycling time through the series of samples (~10 samples per holder), the interval between successive measurements on a given sample was approximately 180 s. 1D scattering/diffraction patterns were obtained by subtracting the background and circularly averaging the data of 2D SAXS and WAXS patterns using the Nika package30 for Igor Pro 6.37. High-resolution transmission electron microscope (HRTEM) micrographs were obtained using FEI Tecnai F20ST/STEM operated at 200 keV. TEM samples were prepared by drop casting from sonicated dispersion of the diluted grounded NOMC in acetone (0.1 mg/10 ml) onto 400-mesh lacy carbon coated copper grids (CF200-CU, Electron Microscopy Sciences). The prepared sample were dried under vacuum for 2 h to remove residual acetone. N2 adsorption-desorption isotherms were measured using a TriStar II (Micromeritics) at 77K. The specific surface area and the pore size distributions were determined from the adsorption isotherms using the Brunauer Emmett and Teller (BET) method31 and Barrett, Joyner and Halenda (BJH) model,32 respectively. The elemental composition of samples was elucidated by X-ray photo-electron spectroscopy (XPS PHI5000 Versa Probe II Scanning XPS Microprobe, ULVAC-PHI, Inc.) with both survey (700 eV to 0 eV) scans and highresolution scans associated with N1s (394eV to 406 eV) with a 0.05 eV step size. All XPS data were recorded using a takeoff angle of 45°. The peaks in the spectra were quantitatively analyzed using Gaussian-Lorentz peak fitting and Shirley background subtraction with the MultiPak Data Reduction software (Physical Electronics). Fourier transform infrared spectroscopy (Thermo Scientific iS50 FT-IR) provided information on the chemical bonds of the nitrogen doped carbon samples. For FTIR, the NOMC was ground into fine powders and measured using powder Praying MantisTM stage (DRP-SAP, Harrick Scientific Products, Inc) in a controlled environment chamber (DRK-3-NI8, Harrick Scientific Products, Inc. The measurement was performed in reflection mode, with a resolution of 4 cm-1 and 256 scans. The particle size of the different NOMCs was determined by dynamic light scattering (DLS, NanoBrook Omni, ZetaPALS, Brookhaven Instruments, Holtsville, NY, U.S.A). The ground powders as described in the fabrication section were dispersed in water (~0.001M) using an ultrasonic bath (B2500A-DYH,

VWR Inc.) for 5 min before measurement. Each sample was measured for 100 s. Effective Diffusion Coefficient of Sulfur in NOMCs The effective diffusion coefficient was calculated from the scattering intensity change during the pore filling process. As the electron scattering length density of carbon (1.38 × 10-5 Å-2) and sulfur (1.55 × 10-5 Å-2) are similar, the intensities of the primary peak will decrease as the pores fill. The standard decoupling approximation as shown in in equation (1) is used to calculate the fraction of the pores that are filled,33, 34

 = 



∙  −   ∙   ∙  ∙  

(1)

where Q is the scattering vector;  is the scattered 

intensity;   is the number density of scatterers; and  are the electron densities of the two phases;  is the scattering volume of particles;  is the form factor and   is the structure factor. In this case, the scattering features are dominated by   related to the ordered mesopores with contrast difference between carbon and air initially and this contrast decreases due to the air being displaced by molten sulfur due to capillary forces that drive the pore filling. As the structure of the mesoporous carbon is not significantly impacted by the sulfur filling, the change in contrast in the system ( −   between carbon-air and carbon-sulfur dominate the changes in I(Q). To compensate for variations in the capillary tubes and packing density of NOMCs, the normalized relative inten  sity ratio ( , instead of absolute intensities, was used 

for analysis, where the scattering at time t, , is normalized by the initial scattering of the samples, 0. In examination of equation (1), the particle volume term,  , can be assumed to be constant as the volume expansion of carbon from room temperature to 155 °C is negligible.35 Similarly,  and   during the pore filling process has been shown for similar materials to be essentially invariant.26, 36 Therefore, the relative intensity ratio can be simplified to equation (2):







=

 !"# $%∙&'()' $*+$%,∙- .  !"# $- /

/

(2)

where the pores are filled by volume fraction of sulfur, φ, with an electron density of 012314 with the remainder of the pores still containing air. This equation represents the assumption that the pores are uniformly filled with a material of electron density, φρsulfur. Alternatively, one could assume that the material consists of two types of regions, one with pores completely filled by S and the other where the pores are completely unfilled (as described in the supporting information, SI). As this later model is less probable than that used for equation (2), we have used the partial filling model to describe the data here.

3 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 11

The electron scattering length densities for carbon, sulfur and air are 1.38× 10-5 Å-2, 1.55 × 10-5 Å-2 and 1.06× 10-8 Å-2. As 564 is much smaller compared with 75489: or 012314 , it is appropriate to assume 564 ; 0. Therefore, equation (2) can be simplified to equation (3) to provide the quantitative relationship between the measured scattering intensity and the fraction of pores filled by sulfur at time t:

? >@

+$=

ABCDBE

FGEHIJ

Here, the scattering length density for nitrogen doped carbon (1.3799× 10-5 Å-2 for 20% nitrogen doped carbon) is very close to the scattering length density for pure carbon (1.3777× 10-5 Å-2). Thus, it is appropriate to assume 75489: ; 75489: . In order to estimate the diffusion coefficient of the sulfur in the NOMCs, a simple mass transfer coefficient approach for porous materials was applied as shown in equation 437: %



= 1 − M $N∙O∙ %K

(4)

where