Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Article
Reactivity and Diffusivity of Li Polysulfides: A Fundamental Study Using Impedance Spectroscopy Sara Drvari# Talian, Joze Moskon, Robert Dominko, and Miran Gaberscek ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b08317 • Publication Date (Web): 15 Aug 2017 Downloaded from http://pubs.acs.org on August 17, 2017
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 free 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 accessible to all readers and 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.
ACS Applied Materials & Interfaces 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 13
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
ACS Applied Materials & Interfaces
Reactivity and Diffusivity of Li Polysulfides: A Fundamental Study Using Impedance Spectroscopy Sara Drvarič Talian†,§, Jože Moškon†, Robert Dominko†, and Miran Gaberšček*†,§ †
National Institute of Chemistry, Ljubljana, Slovenia
§
Faculty of Chemistry and Chemical Technology, University of Ljubljana, Slovenia
KEYWORDS: polysulfides, reaction mechanism, redox kinetics, diffusion coefficient, disproportionation, impedance spectroscopy ABSTRACT: Polysulfides are central compounds in lithium-sulfur battery cells. However, the fundamental redox and diffusion properties of polysulfides are still poorly understood. We try to fill this gap by performing accurate impedance spectroscopy investigation using symmetrical cells consisting from two planar glassy carbon electrodes separated with catholyte-soaked separator. The catholyte contains a mixture of selected polysulfides with predetermined nominal concentrations. Impedance measurements reveal textbook shapes of spectra for most polysulfide compounds or their mixtures. This allows reliable and accurate determination of the rate constant (exchange current density) for a given redox reaction as well as the diffusion coefficient and diffusion length for the rate-determining polysulfide species. Further, it is confirmed that polysulfides tend to disproportionate with time which significantly changes the chemistry and electrochemistry of the system. Two approaches are proposed for identification of the prevailing redox mechanism in the resulting mixtures. The values of kinetic and transport parameters obtained for different cases of interest are commented in significant detail. The study provides a solid basis for better understanding of the complex processes in polysulfide mixtures.
INTRODUCTION Lithium-sulfur batteries are an emerging battery technology that has a solid potential of replacing the state-ofthe-art lithium-ion systems. They offer, at least theoretically, a significantly higher energy density (between 500 600 Wh/kg) at a relatively lower cost.1–3 In practice however there still exist many issues to be solved such as the transport/reaction problems due to insulating nature of sulfur and Li2S (the final product of discharge),4 the lithium dendrite formation during prolonged cycling5 and, most importantly, the so-called shuttle effect due to solubility of polysulfides species that occur as intermediates during battery operation.6 Although important progress has been made in alleviating the shuttle effect, e.g. by adsorbing/entrapping the polysulfides inside the porous cathode7–10 or by modifying the structure of separator,11,12 there is still a lack of fundamental understanding of chemical and electrochemical behaviour of polysulfides under the conditions met in battery environment. Whereas kinetic,13 transport14 or solubility15,16 studies are occasionally reported, reliable data on fundamental parameters such as rate constants, exchange current density for given redox reaction or diffusion coefficients of polysulfide species are critically missing. In addition, the nature, degree and kinetics of polysulfide disproportiona-
tion17,18 and its impact on battery operation have only partly been explored. In order to get reliable quantitative information about basic properties of a system, one usually needs to employ the most accurate method available. In electrochemistry such a method is impedance spectroscopy. In the case of Li-S system, impedance spectroscopy has already been used successfully to probe the internal resistance of lithium-sulfur cells,19 to study the dissolution and formation of solid reaction products20,21 or to better understand the capacity fading phenomenon.22–24 Occasionally, simulation of impedance response of Li-S cells is also reported.25 However, probably due to the relatively complex electrode and/or cell geometries, those impedance spectrosopy investigations could not provide accurate fundamental data about reactivity and transport properties of polysulfides involved. In this work we employ impedance spectroscopy to get as deep as possible experimental insight into redox reactions involving various polysulfide species of interest. At the same time we try to collect accurate values of the diffusion coefficients of polysulfide species involved. For this purpose a special symmetrical cell with planar electrodes is built – mimicking the theoretical »symmetrical full cell« design used in conventional calculation of electrochemical impedance response.26 During the course of experiments the significance of spontaneous disproportionation of polysulfides is recognized and this effect is
ACS Paragon Plus Environment
1
ACS Applied Materials & Interfaces
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
then used in theoretical analysis of measured responses. Finally, two methods for identification of the actual redox mechanism in a mixture of polysulfides are proposed and tested.
EXPERIMENTAL SECTION Cell setup. Initial impedance spectroscopy tests in this study were conducted on a standard Li-S battery cell. Cathode composite material was prepared from sulfur and ENSACO 350G carbon (Imerys) in a weight mixture 2:1. The material was heated to 155 °C, so the sulfur melted and integrated into the carbon pores. The composite material was mixed in 80:10:10 wt. % ratio with conductivity additive (Printex XE2 from Degussa) and PVdF (Sigma-Aldrich, average molar mass 543.000) in NMP (SigmaAldrich, Chromasolv, >99 %), resulting with the final sulfur content in the cathodes of 53.2 %. The slurry was cast on carbon coated aluminium foil using a Doctor Blade applicator. The cathode composites were dried over night at 50 °C. Mass of active material was 1.0 mg S/cm2, thickness of the cathodes around 150 µm and the carbon mass loading 0.7 mg/cm2. Li foil (FMC, ca. 2 cm2) was used as an anode, Celgard 2400 (thickness 25 µm) as separator. The pouch cell was prepared with 20 µL/mg of sulfur using the base electrolyte - 1 M LiTFSI in TEGDME/DOL 1/1 (v/v). Further tests were done on symmetrical cells, where a pair of 2 cm2 polished glassy carbon (GC) discs were used as the electrodes. The active species were introduced as stoichiometric mixtures representing different lengths of polysulfides, i.e. compounds with general formula Li2SX (x = 4, 6, 7, 8), dissolved in the before mentioned electrolyte in different nominal concentrations. In other words, we prepared a series of symmetrical GC||GC cells with different concentrations of Li2SX as well as different compositions containing two Li2SX members. Due to known instability of Li-polysulfides in the tested catholytes with their tendency to disproportionate the initial concentrations of Li2SX species should not be taken as the actual concentrations during the course of experiment (e.g. impedance measurement) but should rather be considered as the nominal compositions of those species in a system (catholyte)27. The catholyte amount added into the pouch cell depended upon the separator type used. When Celgard 2400 separator was employed, 10 µL of the catholyte mixture was used, while the glassy fibre separator (Whatman, GF/D) was wetted with 60 µL. For the experiment where no separator was applied, a special custom-made Teflon spacer was used. The cell was constructed inside a Swagelok casing. Smaller glassy carbon discs were used (12 mm in diameter). The spacer measured 1 mm in thickness, 13 mm in diameter and had a hole with a cross section of 50.2 mm2. An excess volume of catholyte was added and special care was taken during the assembly in order to achieve that no gas bubbles would be left between the electrodes. All the cells were prepared in an Arfilled glove box, where contents of oxygen and water were maintained at a low level (< 1 ppm). The polysulfides, with different nominal compositions Li2SX, were synthesized in an inert atmosphere of Ar-filled
Page 2 of 13
glove box by mixing the chosen stoichiometric amounts of metallic Li and sulfur in THF at elevated temperature for a few days, until all the solid precursor materials reacted. The compounds were isolated under reduced pressure inside the glove box. Impedance spectroscopy tests and spectra processing. Porous electrode battery was discharged at C/20 until 500 mAh g-1 were reached. 15 minutes of relaxation followed at open circuit (OCV) and after that impedance spectra was measured between the frequencies of 1 MHz and 10 mHz with the input voltage amplitude of 10 mV (rms). For the symmetrical cells that were assembled out of partially pre-cycled initial Li-S cells, two conventional Li-S batteries as identical as possible were prepared. After discharging the two initial Li-S batteries half way (∼50 % DOD), both cells were deconstructed in the glove box. By taking the obtained pair of cathodes, anodes and the two separators, the corresponding cathode||cathode and anode||anode symmetrical cells were assembled. Impedance spectra of the symmetrical glassy carbon cells were measured at OCV with the same voltage amplitude (10 mV, rms). The low-frequency limit was varied and extended down to 0.1 mHz. Electrochemical measurements were done using a Biologic VMP3 or MPG2 galvanostat/potentiostat. Impedance spectra were fitted using ZView version 3.4f. UV-Vis spectroscopy. The UV/Vis pouch cell manufacture and assembly was carried out similarly as described in other works published by our research team.28 For electrodes, glassy carbon discs were replaced with platinum foil, since it allowed for a hole to be punctured, yet it produced similar impedance responses. PEIS measurements from 1 MHz to 1 mHz with the amplitude of 10 mV were conducted by using a Biologic SP-200 galvanostat/potentiostat. At the same time, the symmetrical cells were placed in a Perkin-Elmer Lambda 950 UV/Vis spectrometer, where simultaneously with the impedance measurements the UV/Vis spectra were recorded in reflectance mode every 80 min from 800 nm to 250 nm. Time of the spectra collection was chosen on the basis of the duration of the PEIS measurement. Experimental details of the tests presented in the Supporting information are provided in the Supporting text.
RESULTS AND DISCUSSION Conventional vs. symmetrical cell. A typical impedance response of a conventional lithium sulfur full cell is shown in Fig. 1. Several features can be identified and fitted using an arbitrary equivalent circuit as indicated in the figure. Although the fit may be quite good, an obvious drawback of this approach is that it usually does not offer a reliable mapping between the circuit elements and the underlying physical-chemical processes. In other words, we cannot be sure about the physical meaning of a certain element - which leads to a limited insight into the actual processes probed by impedance spectroscopy. With the aim of determining which contributions to the battery spectrum come from which electrode, symmetrical cells consisting of two previously cycled anodes and two cathodes were constructed, in a similar way as
ACS Paragon Plus Environment
2
Page 3 of 13
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
ACS Applied Materials & Interfaces
reported in literature.29,30 Such experiments were conducted both on freshly prepared cells as well as on cycled cells. Here we focus on typical spectra of partially predischarged cells as these correspond best to the measurements performed on the simplified cell configuration (see in continuation). As evident from Fig. 2, the magnitude of the anode resistance is in the range of a few 10 Ω (half of the resistance of symmetrical Li||Li cell).
These contributions dominate the high frequency part of the battery impedance spectrum. The cathode shows a relatively smaller contribution in high frequency part if compared to the anodic response. By contrast, at lower frequencies the cathode clearly exhibits a large impedance manifested in the quasi-blocking character of impedance spectrum. Since this part clearly dominates the overall Li-S battery impedance, further studies were focused on the electrochemistry and impedance response of polysulfides on carbon surfaces – the main source of the cathodic response. Simplified cell with planar electrodes: theory vs. experiment. In order to decrease ambiguity as regards the physical meaning of recorded spectra, we attempted to build special electrochemical cells with much simpler geometry. Three essential modifications were pursued: (i) instead of the porous carbon electrode a planar glassy carbon electrode (GC) was introduced, (ii) the lithium electrode was removed from the system and replaced with another planar glassy carbon electrode and (iii) active species were added to the cells in the form of Lipolysulfides dissolved in the electrolyte (catholyte).
Figure 1. A typical impedance spectrum of a conventional LiS battery cell based on a porous carbon-sulfur cathode (first cycle, app. 50 % DOD) (white circles). Fit (red circles) with the equivalent circuit model shown above the graph.
Figure 2. EIS symmetrical cells test: measured impedance spectrum of the Li-S battery cell (at ∼50 % DOD) compared to the measured spectra of symmetrical anode||anode and cathode||cathode cells that were assembled by using the electrodes from two partially pre-discharged (∼50 % DOD) Li-S cells.
Figure 3. a) Equivalent circuit for the so-called “symmetrical 26 full cell” (according to terminology in ) with two identical electrodes under supported conditions (in our case addition 26 of a 1 M Li salt); main circuit was adapted from , the expansion in ellipse refers to Eqs(3); the high frequency capacity 26 due to displacement shown in is omitted as it could not be probed in this study. b) Typical impedance response generated by equivalent circuit under (a). The values were selected arbitrarily - only to show the main impedance features. Note that the 2 Warburg features are not due to 2 electrodes (they are assumed identical) but due to two diffusing species, Ox and Red (for further explanation see the main text).
This way, the so-called symmetrical type of GC||GC cell was obtained. A drawback of such a cell is that it cannot
ACS Paragon Plus Environment
3
ACS Applied Materials & Interfaces
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
be charged/discharged. On the other hand, it allows – at least in principle – rather accurate and much less ambiguous measurements of the reaction-diffusion processes of active species contained in the catholyte between both electrodes. Furthermore, the impedance response of symmetrical systems with planar electrodes is very well known,26 which may additionally help identify the actual physical-chemical processes taking place in the system. However, the introduction of glassy carbon electrode in the study of reaction kinetics of polysulfides might raise a question whether this type of electrode promotes the same reaction mechanism as a conventional practical cathode based on a porous carbon matrix. Performing detailed galvanostatic discharge/charge cycling of a GC/catholyte/Li "battery" cell (see SI, Fig. S1) and, additionally, of a modified GC-based cell in which the electrodes were separated with a ceramic Li+ conductive membrane (Fig. S2) we found no important differences in the charge-discharge mechanism with respect to the known mechanisms in conventional porous cathodes. Even more, the same issues were clearly detected: i) only partial electrochemical utilization of sulfur species that are present (dissolved) in electrolyte and ii) irreversible loss of sulfur due to chemical reaction(s) of polysulfides with Li metal anode (Fig.S3). For easier interpretation of the main features of impedance response of a GC||GC cell, let us first briefly recall the general impedance response of a redox system with an oxidized, Ox, and a reduced, Red, species dissolved in an electrolyte enclosed between two symmetrical electrodes. During the measurement two essential processes are taking place: the redox reaction (pOx + ne- = qRed) and the diffusion of Ox and Red to/from each of the electrodes (here migration of both species is neglected as the electric field is supposed to be screened by the charged species in electrolyte with a higher concentration – the so-called supported situation26). This situation yields the Randles-like equivalent circuit shown in Fig. 3a. The physical meaning of the elements in the lower branch (written for a single electrode) is the following:
=
(1)
'( the exchange current, L the effective distance from the electrode within which the species are diffusing (the value is assumed equal for both Red and Ox), A is the geometrical surface area of electrode, ) and ) are the diffu( ( sion coefficients, * and * are the bulk concentrations of the oxidized and reduced species, respectively, j = √−1 and . is the angular frequency of excitation signal. For the special case when the impedance is measured at the equilibrium potential the exchange current can be expressed as: ( 56 789: 34* ( 5; 7: '( = /012 ( 34*
