Operando Studies of Antiperovskite Lithium Battery Cathode Material

Subscriber access provided by NATIONAL UNIV OF SINGAPORE

Article 2

Operando studies of anti-perovskite lithium battery cathode material (LiFe)SO Daria Mikhailova, Lars Giebeler, Sebastian Maletti, Steffen Oswald, Angelina Sarapulova, Sylvio Indris, Zhiwei Hu, Jozef Bednarcik, and Martin Valldor ACS Appl. Energy Mater., Just Accepted Manuscript • DOI: 10.1021/acsaem.8b01493 • Publication Date (Web): 11 Oct 2018 Downloaded from http://pubs.acs.org on October 13, 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 32 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 Energy Materials

Operando studies of anti-perovskite lithium battery cathode material (Li2Fe)SO Daria Mikhailova,1,* Lars Giebeler,1 Sebastian Maletti,1 Steffen Oswald,1 Angelina Sarapulova,2 Sylvio Indris,3 Zhiwei Hu,4 Jozef Bednarcik,5 and Martin Valldor1,* 1

Leibniz Institute for Solid State and Materials Research (IFW) Dresden e.V., Helmholtzstraße

20, DE-01069 Dresden, Germany. 2

Karlsruher Institute for Technology, IAM-ESS, Straße am Forum 8, DE-76131 Karlsruhe,

Germany 3

Karlsruher Institute for Technology, IAM-ESS, Hermann-von-Helmholtz-Platz 1, DE-76344

Eggenstein-Leopoldshafen, Germany. 4

Max Planck Institute for Chemical Physics of Solids, Nöthnitzer Straße 40, DE-01187 Dresden,

Germany. 5

Deutsches Elektronen-Synchrotron DESY, Notkestraße 85, DE-22607 Hamburg, Germany.

KEYWORDS: Li-ion battery, cathode material, anti-perovskite, operando diffraction, operando spectroscopy.

ACS Paragon Plus Environment

1

ACS Applied Energy Materials 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 2 of 32

ABSTRACT: Spectroscopic and X-ray diffraction operando techniques were used to investigate polycrystalline anti-perovskite (Li2Fe)SO as cathode materials in a Li-battery set up. During Li removal, several intermediate, relatively stable phases exist. At low charging, Fe is oxidized from +2 to +3, but, at higher charging, S2‒ is also partly oxidized to elemental sulfur, suggesting a cathode bifunctionality, and both redox processes seem reversible. On cycling (Li2Fe)SO in a battery, spectroscopy data suggest that a part of the Fe atoms irreversibly vacate the high-symmetry positions in the crystal lattice, in line with the broadening of X-ray diffraction peaks. Instead, new, relatively broad reflections appear in the X-ray patterns that might be explained by a crystallographic superstructure, corresponding to a doubling of the cubic unit cell axis; but the peak broadness indicates a lowering in crystallographic symmetry. Using a standard electrolyte and a moderate charging rate of C/10 results in typical capacity loss per cycle but, by using an electrolyte with low sulfur solubility, the (Li2Fe)SO cathode is stabilized and charge densities of more than 200 mAh g‒1 at 1C charging rate is obtained. Additionally, a Li-deficient precursor (Li0.8Fe)SO served as a cathode material in a Na battery providing presumably reversible Na intercalation and removal.

1. Introduction After introducing pyrite-type FeS2 in primary and olivine-type LiFePO4 in secondary lithium batteries, iron compounds have revolutionized commercial market for almost two decades. 1-7 Additionally, iron oxides and fluorides helped to understand fundamental aspects of the conversion mechanism next to typical well-known intercalation.2,8-14 These Fe-based compounds are also environmentally friendly and cost effective in contrast to Co-based materials.15 Recently, a novel


2

Page 3 of 32 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


anti-perovskite (Fe2)SeO with two different orderings of metal vacancies () was discovered.16 By chemically substituting 1 Fe2+ for 2 Li+, the cubic anti-perovskites (Li2Fe)ChO (Ch = S, Se) were synthesized and their preliminary performances as cathodes in batteries account for promising properties.17 However, only macroscopic investigations and ex situ data have been reported so far. Here, we describe the first operando study during charging and discharging of (Li2Fe)SO, as cheaper, more environmentally friendly, nontoxic compound with higher theoretical specific capacity, 227 mAh g‒1 for 1 Li, as compared to (Li2Fe)SeO 163 mAh g‒1. 2. Results and Discussions 2.1. Electrochemical performance. In line with the first report,17 a single-step solid-state synthesis results in X-ray pure (Li2Fe)SO with anti-perovskite structure (Figure 1). The as-prepared polycrystalline sample was integrated with Li-metal as anode into a battery set up, and its performance was investigated. At constant charging/discharging rates (C/10) using LP30 electrolyte, the capacity decreases over 50 cycles from the theoretical value (227 mAh g‒1) down to about 150 mAh g‒1 (Figure S1) with almost ideal Coulombic efficiency, i.e. each charging is similar to the discharging capacity.

Figure 1. Anti-perovskite crystal structure of (Li2Fe)SO.


3


Page 4 of 32

Cyclic voltammetry data reveal that there are several intermediate steps during Li extraction (Figure 2a). Similar but much broader anomalies are visible during the subsequent Li-insertion. After the first charge in the voltammetric measurements at 2.9 V, the process was not complete, and the potential range was extended up to 3 V vs. Li+/Li. In the second and third voltammetric cycles these anomalies are present, although less pronounced; This suggests the presence of relatively stable, intermediate x in (LixFe)SO (x = 2‒0.8), but additional, partial material decomposition cannot be excluded. The open circuit voltage (OCV) was 2 V (Figure 2a) for the first cycle and increase to about 2.5 V for the following cycles, due to changes in macro and micro structures of the material. During the first charging cycle, a large hysteretic effect in the ionic conductivity (Figure 2b) between charging and discharging processes is observed: in the range 2.0‒2.6 V, Li-extraction (battery charging) proceeds at three orders of magnitude higher rate than subsequent Li-insertion (battery discharging). The absolute diffusion rates are in the order of those for LiFePO4,18,19 but recent theoretical work on (Li2Fe)SO suggests a much lower hopping energy barrier for the Li-diffusion of 0.32 eV,20 which is closer to the value of 0.35 eV for anti-perovskite superionic conductor Li3ClO.21 The hysteretic behavior might be explained by a crystal lattice shrinking, due to Li non-stoichiometry, or by an unknown reaction at the grain surfaces that inhibits subsequent Li insertion to some degree.


4

Page 5 of 32 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


Figure 2. (a) Cyclic voltammetry of a Li | electrolyte | (Li2Fe)SO cell, recorded with 0.1mV s‒1 during the first three charging cycles. (b) Li chemical diffusion coefficient (DLi) during the first charging and discharging cycle.

2.2. Microscopy and operando X-ray diffraction. Macroscopic observations by scanning electron microscopy reveal only minor cracking formations during the charging processes (Figure S2). According to operando X-ray diffraction, the long-range atomic ordering is partly broken down already during the first cycle of Li extraction and insertion (Figure 3): main diffraction peaks are strongly broadened and lose absolute intensity, which is due to loss of crystallinity and introduction of atomic disorder on local scale, i.e. ions normally localized on anti-perovskite high-symmetrical positions are randomly off-centered. Further, below about x = 1.85 in (LixFe)SO additional reflections appear at low diffraction angles (Figure 3), which might be either an unknown product from a reaction between carbonate-containing electrolyte and cathode materials, a second phase


5


Page 6 of 32

as product of phase transformation, or as a result of a structural distortion providing super-structure reflections. Analysis of these new reflections shows continuous angular shift during further cell charging and subsequent discharging, hence confirming an electrochemical activity of this unknown structure. After re-inserting Li, the pristine diffraction pattern is not regained related to reflection positions and their intensities, hence pointing irreversibility of structural changes. Note, all the strongest low-angle peaks can be well-indexed as superstructure reflections of a primitive cubic unit cell (Figure 3) with a doubled axis length (a ≈ 7.8 Å). After one completed charging cycle, the lowangle intensities are of the same order as the basic anti-perovskite reflections, which demonstrating broadening and an irregular shape, while low-angle reflections are narrow. From these observations, grave structural transformations in the material upon cycling are concluded. To distinguish between formation of an anti-perovskite superstructure because of a possible partial cation ordering, and a second phase formation, future studies are needed. The initial Li removal causes the cubic unit cell axis (a) to decrease from about 3.914 Å down to 3.875 Å at x ≈ 1.57 (Figure S3), corresponding to a small relative volume change of about 1 %, agreeing with the first report on this material with ex situ data.17


6

Page 7 of 32 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


Figure 3. X-ray diffraction data from a complete battery set up during a single charge and discharge. Possible superstructure reflections are marked and their charge dependent shifts are highlighted with narrow lines at their maximum intensities. Color coding groups reflections from battery cell parts: PTFE is pink, Li-metal is yellow, and Al-metal is light blue.

However, on removing more Li, the unit cell size seems to increase to about 3.90 Å at x ≈ 1.2 (Figure S3), suggesting that more parameters than only the Li content affect the size of the crystallographic unit cell. 2.3. Operando X-ray and Mößbauer spectroscopy. By simultaneously performing X-ray absorption spectroscopy (XAS) on the Fe-edge during the charging and discharging, conclusions on redox reactions can be drawn. The energy position of the Fe-K absorption edge is related to the valence state of 3d transition metals.22 Figure 4a shows the Fe-K XAS spectra of (Li2Fe)SO together with those of the reference materials FeO, Fe3O4, and Fe2O3 with average oxidation states Fe2+, Fe2.67 and Fe3+, respectively for comparison. All spectra are normalized to unity step in the


7


Page 8 of 32

absorption coefficient μ and the appropriate position of the absorption edge is located at  = 0.8.23 The chemical shift between Fe2+ to Fe3+ is about 3 eV in other materials, e.g. silicate glasses.22,24 Upon Li extraction, Fe2+ is oxidized and the Fe3+ content increases according to XAS data. However, the spectroscopically determined Fe oxidation state remains slightly lower than the expected value, as estimated by the total charge in the electrochemical treatment (Figure 4b). This discrepancy between XAS and electrochemical data might correspond to a partial, simultaneous oxidation of S2‒ upon Li-extraction. Note, the reference materials are electronically more correlated as compared to the title compound and the heteroleptic coordination of Fe, by both O and S in (LixFe)SO, might interfere with spectroscopic interpretations; the position of the Fe-K edge may be slightly shifted because of the lower electronegativity of S 2‒ as compared to O2‒. During subsequent Li-insertion into the material, Fe3+ is reversible reduced to Fe2+ (Figure 4a).

Figure 4. Spectroscopy on Fe during first galvanostatic cycling with potential limitations (GCPL) in a Li | electrolyte | (Li2Fe)SO cell. (a) Operando X-ray absorption spectroscopy near the Fe-K edge and (b) the


8

Page 9 of 32 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


estimated Fe oxidation state from electrochemical cycling as compared to spectroscopic results are plotted. (c) From EXAFS, the interatomic distances for Fe were estimated (NN = nearest-neighbor, NNN = nextnearest-neighbor). (d)

57

Fe-Mößbauer spectroscopy at room-temperature on pristine and on

electrochemically treated title compounds, where the colored areas represent the spectral weights of simulated components.

Pre-edge features, like intensity and position, of 3d metals with partially filled d-shells are also very sensitive to distorted metal coordinations.22 Removing Li from the pristine compound, forming (LixFe)SO, causes a strong distortion of the oxygen coordination around Fe, which is reversible upon Li insertion (Figure 4a). The evolution of average Fe‒M distances (up to 3.9 Å) in LixFeSO obtained during Li-extraction and insertion is presented in Figure 3c (Table S1), as obtained by Fourier transforming the k3-weighted Fe-EXAFS data. The cubic symmetry in pristine (Li2Fe)SO provides Fe‒S, Fe‒Fe, and Fe‒Li distances of 2.89 Å, while the Fe‒O distance is shorter (1.93 Å). The Fe‒O distance continuously decreases with the Li content from 1.93 Å to 1.83 Å, agreeing with the oxidation of Fe2+ to Fe3+. At x ≤ 1.8 upon Li extraction, Fe‒S and Fe‒Li(Fe) distances change significantly, indicating strong structural distortions. The Fe‒S distance contracts, what can be caused by oxidation of Fe2+ or of Fe2+ and S2‒. The larger Fe‒Fe distances (3.8 Å) also become smaller because of structural distortions. These changes are mostly reversible, but the crystallographic order does not reappear, as is obvious from broadening and loss of X-ray diffraction intensities (Figure 3).

57

Fe-Mößbauer spectroscopy data on

electrochemically treated materials reveal that Fe in pristine (Li2Fe)SO is a paramagnetic Fe2+ ion with a high-spin 3d6 electronic configuration (S = 2) (Figure 4d). All refined parameters from the modeling of spectra can be found in the supporting information (Table S2). After applying a voltage of 2.45 V, a spectrum is obtained that is well described by two doublets: one typical Fe3+


9


Page 10 of 32

(3d5 high-spin) signal dominates, with a relatively small isomer shift, while the isomer shift of a Fe2+ contribution also diminishes significantly, as compared to the spectrum of the pristine sample. The strong change in the isomer shift with Fe2+ to zero suggests either that the Fe2+ ions are moving to a different lattice position, more suitable for a high-spin 3d6 ion, or that a high-to-low spin-state transition of Fe2+ occurs, or both. At 2.9 V, one Li should have been extracted and only two doublets from two different Fe3+ ions remain. The doublets have different isomer shifts and correspond to two different coordination environments, indicating that a part of the Fe-ions has left the highly symmetrical crystallographic position in the anti-perovskite structure. After discharging the battery, only a broad signal from Fe2+ remains with relatively small isomer shift, inferring that Fe has several different coordination environments and most probably adopts a lowspin, non-magnetic state (3d6 S = 0); magnetic investigations on pristine and on once charge-cycled (Li2Fe)SO suggest that the magnetic signals of both materials are dominated by residual ferromagnetic components (Figure S4), which would correspond to 1‒3 mol% Fe3O4 impurities. However, in the resolution range of Mößbauer data (Figure 4d), all Fe is paramagnetic. Hence, further investigations are needed to conclude on the magnetic nature of the title compound. X-ray photoelectron spectroscopy (XPS) was also performed on Fe (Figure S5) and the observations agree well with XAS data. However, XPS data on S reveal that, at highly charged states, not only Fe but also S is involved in the redox reaction that counter balances the Li extraction (Figure 5). Pristine (Li2Fe)SO exhibits a typical spectrum for a sulfide ion (S2-) suggesting strong electron correlations. At the cell voltage of 2.4 V, corresponding to x = 1.5, the sulfide ion still remains, but already at 2.9 V (x = 1) there is a partial oxidation of S (Figure 5a), additionally to that of Fe (Figure 4): XPS is a highly surface sensitive technique and data suggest that there is elemental S present, at least close to the electrolyte, in the highly charged state of the cathode material. At the


10

Page 11 of 32 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


end of the first complete charge-discharge cycle, S is reduced again to sulfide, but the data at hand do not reveal what type of sulfide is formed; it could be either the reformation of the title phase or formation of a secondary phase, like nanocrystalline or amorphous Li2S, that would not necessarily by detected by X-ray diffraction (Figure 3). Oxygen, on the other hand, remains completely as O2‒ throughout the whole redox reaction (Figure 5b).

Figure 5. XPS data on (a) S and (b) O during one cycle of charging and discharging of a Li | electrolyte | (Li2Fe)SO cell.

2.4. (Li0.8Fe)SO as cathode in Na batteries. By solid-state reactions it was not possible to obtain any Na-containing analogues, (Li2‒yNayFe)SO (y ≤ 2). Instead, electrochemistry was applied to extract about 1.2 Li from (Li2Fe)SO to electrochemically synthesize “(Li0.8Fe)SO”. The delithiated cathode material was washed, combined with a separator plus a Na-metal anode, and subsequently used in an electrochemical cell in the voltage range 2.9‒1.0 V vs. Na+/Na. At relatively low charging rates, the electrochemically produced (Li0.8NazFe)SO (z < 0.7) materials performs surprisingly well (Figure 6b), meaning that Na reversibly is stored in the relatively large cation voids within the deficient anti-perovskite lattice. Although the initial cycles indicate a specific charge capacity of about 60 mAh g1, the following cycles show increasing capacity; this might be due to decreasing average particle size and increasing surface area of the cathode material, which


11


Page 12 of 32

occurs during electrochemical cycling. However, after a maximum of about 110 mAh g‒1 (Li0.8Na0.7FeSO) during the initial 10 cycles, an almost linear decay of capacity is observed over 40 cycles down to about 80 mAh g‒1 (Li0.8Na0.7FeSO).

Figure 6. (a) Specific charging capacity as function of charging cycles of a (Li0.8NazFe)SO (z < 0.7) ‒ Nametal battery at constant charge rate. (b) Two typical charging and discharging curves of the Na battery.

Note, the only indication that the Na content is reversibly changed, and not that of Li, is that the shapes of the charging curves (Figure 6b) are very different from those of the Li charging.

Figure 7. Capacity retention profile of the specific capacity as a function of charging cycles of a Li | (alternative) electrolyte | (Li2Fe)SO cell at changing rates of C/20, C/10, C/5, and 1C.


12

Page 13 of 32 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


2.5. Alternative electrolyte in a (Li2Fe)SO – Li-metal battery. As the choice of electrolyte probably plays an important role and might cause S to leach out,25,26 an alternative carbonate-free electrolyte with less S solvability was used for long-time rate capability tests (Figure 7). The charging capacity is about 275 mAh g‒1 during the first charging cycles at moderate rate of C/10, which means that 1.23 Li has been extracted from the title compound. A markedly high specific charge capacity is still retained at a much higher charge rate: 200 mAh g‒1 at 1C during 10 cycles. This corresponds to 87% of theoretical value for extracting 1 Li per formula unit. Note, this behavior is similar to the case of a LiFePO4 cathode that has been reported to exhibit 157 mAh g‒1 at C/5, corresponding to 92% of the theoretical value, and 140 mAh g‒1 (82%) at 1C.27 Another comparable example is found for S-doped LiFePO4.28 For (Li2Fe)SO, subsequent lowering of the charge rate does increase the charge capacity but the final C/10 charging cycles stay at about the reasonably high charge capacity of 225 mAh g‒1, which is lower than during the initial cycles but yet in agreement with the theoretical value for the extraction of 1 Li per formula unit (223 mAh g‒ 1

). At a charging/discharging rate of C/10 and an open circuit voltage (OCV) of 2.5 V this

corresponds to a specific energy for (Li2Fe)SO of 563 mWh g‒1. Note, this specific energy is already equal to that of LiFePO4: 560 mWh g‒1 (160 mAh g‒1, OCV = 3.5 V),29 so the antiperovskite title compound can easily compete with existing battery technologies as environmentally friendly, cost reducing cathode material.

3. Conclusions Operando investigations on a battery with (Li2Fe)SO as cathode and Li-metal as anode reveal that this anti-perovskite has several rare and unique features. Fe is mainly responsible for the redox


13


Page 14 of 32

reaction at low battery charging but S is involved at higher charging, which shows the bifunctional battery chemistry of (Li2Fe)SO. During battery charging, X-ray diffraction peaks broaden and lose some intensity, which agrees well with movement of Fe from the high-symmetry position in the anti-perovskite structure, as suggested by Mößbauer and X-ray absorption spectroscopy data. Instead, new reflections appearing in X-ray diffraction data can be an atomic superstructure that could correspond to a doubling of the cubic primitive lattice. However, a formation of a second phase cannot be excluded. According to voltammetry data there might be several intermediate compositions, x in (Li2‒xFe)SO, with relatively higher stability. The electrochemically delithiated anti-perovskite (Li0.8Fe)SO can also be well applied in Na batteries. By using an electrolyte with low S solubility, the capacity retention improves significantly, and specific capacity values are still higher than 200 mAh g‒1 even at high charging rates.

4. Experimental Section 4.1. Material synthesis. The polycrystalline sample was prepared from Li2O (Alfa Aesar, 99.5%), Fe (Alfa Aesar, 99.9%), and S (Alfa Aesar, 99.5%) according to the previous report.17 All handling of the sample was done in Ar-filled glove-boxes (MBraun, H2O and O2 < 0.1 ppm). 4.2. Electrochemical measurements. Electrochemical studies were performed with a multichannel potentiostatic-galvanostatic system VMP3 (Biologic, France) in standard Swagelok-type cells with metallic Li as the anode material. For the positive electrode a mixture of the pristine compound, carbon black and polyvinylidene difluoride (PVDF) as polymer binder in an 80:10:10 weight ratio (total mass 12 mg) was pressed on steel-meshes with 10 mm diameter and dried in vacuum at room temperature. For operando experiments, a LP30 electrolyte containing 1M LiPF6 in a mixture of


14

Page 15 of 32 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


ethylene carbonate (EC) and dimethylcarbonate (DMC) (1:1, BASF, total volume 200 L) was used,

while

a

home-made

carbonate-free

electrolyte

containing

1M

lithium

bistrifluoromethanesulfonimidate (LiTFSI) in dimethoxyethane (DME) and dioxolane (DOL) (total volume 250 L) was used for long-time rate capability tests. 4.3 Electron microscopy. Microstructural characterization on the pristine material, after charging up to 2.9 V vs. Li+/Li, and after the first charge-discharge cycle was performed by means of scanning electron microscopy with a Leo 1530 Gemini electron microscope (Zeiss/Leo) equipped with a Bruker InLens detector at an acceleration voltage of 15 kV. After electrochemical treatment, the cells were disassembled in the glovebox; the cathode mixtures were washed with DMC and transferred into the electron microscope with a minimal contact to air. 4.4. X-ray diffraction experiments. Operando X-ray synchrotron diffraction was performed at the synchrotron facility PETRA III/DESY (Hamburg, Germany) at beamline P02.1 in transmission mode using a 16-inch two-dimensional flat panel detector of the XRD 1621 N ES Series (PerkinElmer) with 2048 x 2048 pixels and a pixel size of 200 mm, and a cell setup connected to a VMP multichannel galvanostat.30,31 Data were collected in steps of 0.004° over the 2-range from 0.1 to 15.4° at a wavelength of 0.20718(1) Å, which was determined from the positions of Bragg reflections from a LaB6 reference material. In order to characterize the starting material, a pattern was recorded before the electrochemical process was started. The cell was then successively charged and discharged in galvanostatic mode at a constant current corresponding to the intercalation or deintercalation of 1 Li per formula unit during 10 h (C/10 rate). All diffraction patterns were analyzed by Fullprof implemented into the software package WinPLOTR.32 The Algrid as the current collector on the cathode side served as an internal standard during the


15


Page 16 of 32

measurements, and the refined lattice parameter of Al provided an independent control of the reliability of the obtained model parameters. 4.5. X-ray photoelectron spectroscopy. X-ray photoelectron spectroscopy was applied on the pristine material, on the products with different lithium contents after electrochemical Li extraction, and on the product after completing the first charging-discharging cycle. A PHI 5600 CI system (Physical Electronics) with an Al K 350 W monochromatic X-ray source and a hemispherical analyzer at a pass energy of 29 eV were used for high-resolution spectra. Cells with the same cathode mixture consisting of the pristine compound, carbon black and PVDF with a mass of about 30 mg were charged to 2.4 V and 2.9 V, or charged to 2.9 V and subsequently discharged back down to 1.4 V vs. Li+/Li, and were immediately disassembled in the glove-box. In order to remove electrolyte from the surface, the pellets were washed with DMC and scraped. After drying, the samples were transferred from the glovebox into the XPS system in an Ar-filled transfer chamber. During XPS measurements, when necessary, surface charging was minimized by means of a low-energy electron flood gun. The system base pressure was below 10‒9 mbar. The binding energy scale was calibrated using the O1s peak maximum at 532.7 eV. 4.6. X-ray absorption spectroscopy. Fe-K edge X-ray absorption operando experiments were collected at beamline P65 at PETRA III extension (DESY, Hamburg, Germany) applying the bulk sensitive fluorescence yield method using an energy dispersive 7 pixel HPGe detector. A coin cell holder, coupled with a Biologic Instruments potentiostat, was used to apply the electrochemical procedure. For EXAFS data analysis, the measured spectrum below the pre-edge was fitted linearly while the post-edge background contribution was fitted to a quadratic polynomial. This background 0(E) was subtracted from the absorption spectrum (E) and the resulting data were


16

Page 17 of 32 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


normalized and converted into k space. By weighting (k) with k3, contributions of higher k space were amplified. The resulting k3 (k) was Fourier-transformed into R space, allowing the determination of bond contributions. Least-square fits were performed using the FEFF6 code.33 4.7. Magnetic investigations. Under inert conditions, powder of the title compound was placed in a polycarbonate capsule that was taped to avoid contact with air. The sample was placed in a plastic straw and quickly transferred into the squid magnetometer MPMS-S5 (Quantum Design). 4.8. Ionic diffusion experiments. Chemical diffusion coefficients of Li were determined using galvanostatic intermitted titration technique (GITT) in a Swagelok-type three-electrode cell with a Li negative electrode, the pristine compound as positive electrode, and a metallic Li disk as a reference electrode, following the procedure extensively discussed in the literature.34 A current of C/10 was applied for 10 minutes, where C corresponds to the current needed for the insertion/extraction of 1 Li per formula unit within one hour. Before the next titration step, a rest period of 8 hours followed to allow the determination of an open circuit voltage (OCV) value near equilibrium conditions. 4.9. Mößbauer spectroscopy.

57

Fe Mößbauer spectroscopic experiments were performed in

transmission mode at room temperature using a constant-acceleration spectrometer with a Co(Rh) source. Isomer shifts are given relative to that of -Fe at room temperature. The powder

57

samples were sealed in polyethylene bags in an argon-filled glove box to avoid contact with air during the measurements.

ASSOCIATED CONTENT


17


Page 18 of 32

Supporting Information. Data from charging/discharging experiments of a Li | LP30 electrolyte | (Li2Fe)SO battery, scanning electron micrographs of the cathode, cubic unit cell sizes as function of Li-content (x) in (Li2‒xFe)SO, EXAFS fit parameters, fitted Mößbauer parameters, basic magnetic data of pristine and post-electrochemically cycled (Li2Fe)SO, post-XPS data on Fe in a (Li2Fe)SO – Li-metal battery after applying different voltages. (ae-2018-01493z-SI.docx) This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Authors *Daria Mikhailova, email: [email protected] *Martin Valldor, email: [email protected]. Author Contributions All authors have given approval to the final version of the manuscript. Funding Sources This work was financially supported by the German Science Foundation (DFG) through project VA 831/4-1. Additional funding by European Union and the Free State of Saxony through the NaSBattSy project (SAB grant no. 100234960) is gratefully acknowledged. Notes The authors declare no competing interests.

ACKNOWLEDGMENT


18

Page 19 of 32 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


We are grateful for the opportunity to carry out parts of this research at beamline P02.1 and P65 at PETRAIII/DESY, a member of the Helmholtz Association (HGF), and the support of the beamline staff. REFERENCES 1.

Linden, D.; Reddy T. B. (Eds.), Handbook of Batteries, 3rd ed., McGraw-Hill, New York/USA, 2002.

2.

Daniel, C.; Besenhard, J. O. Handbook of Battery Materials, 2nd ed., Wiley-VCH, Weinheim/Germany, 2011.

3.

Shao-Horn, Y.; Horn, Q. C., Chemical, structural and electrochemical comparison of natural and synthetic FeS2 pyrite in lithium cells. Electrochim. Acta 2001, 46, 2613– 2621.

4.

Fong, R.; Dahn, J. R.; Jones, C. H. W., Electrochemistry of Pyrite‐Based Cathodes for Ambient Temperature Lithium Batteries. J. Electrochem. Soc. 1989, 136, 3206–3210.

5.

Peled, E.; Golodnitsky, D.; Ardel, G.; Lang, J.; Lavi, Y., Development and characterization of bipolar lithium composite polymer electrolyte (CPE)-FeS2 battery for applications in electric vehicles. J. Power Sources 1995, 54, 496–500.

6.

Padhi, A. K.; Nanjundaswamy, K. S.; Goodenough, J. B., Phospho‐olivines as positive‐ electrode materials for rechargeable lithium batteries. J. Electrochem. Soc. 1997, 144, 1188–1194.


19


7.

Page 20 of 32

Whittingham, M. S., Lithium batteries and cathode materials. Chem. Rev. 2004, 104, 4271–4302.

8.

Huggins, R. A. Advanced batteries – materials science aspects, Springer Business + Science Media, NY/USA, 2009.

9.

Cabana, J.; Monconduit, L.; Larcher, D.; Palacín, M. R., Beyond intercalation‐based Li‐ ion batteries: The state of the art and challenges of electrode materials reacting through conversion reactions. Adv. Mater. 2010, 22, E170E192.

10.

Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.–M., Nano-sized transitionmetal oxides as negative-electrode materials for lithium-ion batteries. Nature 2000, 407, 496499.

11.

Taberna, P. L.; Mitra, S.; Poizot, P.; Simon, P.; Tarascon, J.–M., High rate capabilities Fe3O4-based Cu nano-architectured electrodes for lithium-ion battery applications. Nature Mater. 2008, 5, 567576.

12.

Zhang, L.; Wu, H. B.; Lou, X. W. D., Iron‐oxide‐based advanced anode materials for lithium‐ion batteries. Adv. Energy Mater. 2014, 4, 1300958.

13.

Li, H.; Balaya, P.; Maier, J., Li-storage via heterogeneous reaction in selected binary metal fluorides and oxides. J. Electrochem. Soc. 2004, 151, A1878−A1885.

14.

Andre, D.; Kim, S. –J.; Lamp, P.; Lux, S. F.; Maglia, F.; Paschos, O.; Stiaszny, B., Future generations of cathode materials: an automotive industry perspective. J. Mater. Chem. A 2015, 3, 67096732.


20

Page 21 of 32 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


15.

Larcher, D.; Tarascon, J.–M., Towards greener and more sustainable batteries for electrical energy storage. Nature Chem. 2015, 7, 19-29.

16.

Valldor, M.; Wright, T.; Fitch, A.; Prots, Yu., Metal vacancy ordering in an antiperovskite resulting in two modifications of Fe2SeO. Angew. Chem. Int. Ed. 2016, 55, 93809383.

17.

Lai, K. T.; Antonyshyn, I.; Prots, Yu.; Valldor, M., Anti-perovskite Li-battery cathode materials J. Am. Chem. Soc. 2017, 139, 96459649.

18.

Prosini, P. P.; Lisi, M.; Zane, D.; Pasquali, M., Determination of the chemical diffusion coefficient of lithium in LiFePO4. Solid State Ionics 2002, 148, 4551.

19.

Delacourt, C.; Ati, M.; Tarascon, J.–M., Measurement of Lithium Diffusion Coefficient in LiyFeSO4F. J. Electrochem. Soc. 2011, 158, A741A749.

20.

Lu, Z.; Ciucci, F., Anti-perovskite cathodes for lithium batteries. J. Mater. Chem. A 2018, 6, 5185‒5192.

21.

Emly, A.; Kioupakis, E.; van der Ven, A., Phase stability and transport mechanisms in antiperovskite Li3OCl and Li3OBr superionic conductors. Chem. Mater. 2013, 25, 4663– 4670.

22.

Newville, M. Fundamentals of XAFS, University of Chicago, Revision 1.7 July 3, 2004.

23.

Zeng, Z.; Greenblatt, M.; Subramanian, M. A.; Croft, M., Large low-field magnetoresistance in perovskite-type CaCu3Mn4O12 without double exchange Phys. Rev. Lett. 1999, 82, 3164–3167.


21


24.

Page 22 of 32

Berry, A. J.; O’Neill, H. St. C.; Jayasuriya, K. D.; Campbell, S. J.; Foran, G. J., XANES calibrations for the oxidation state of iron in a silicate glass. Am. Mineral. 2003, 88, 967– 977.

25.

Yang, Y.; Zheng, G.; Cui, Y., Nanostructured sulfur cathodes. Chem. Soc. Rev. 2013, 42, 3018–3032.

26.

Manthiram, A.; Fu, Y.; Chung, S. H.; Zu, C.; Su, Y.‒S., Rechargeable lithium–sulfur batteries. Chem. Rev. 2014, 114, 11751–11787.

27.

Yang, L.; Jiao, L.; Miao, Y.; Yuan, H., Synthesis and characterization of LiFe0.99Mn0.01 (PO4)2.99/3F0.01/C as a cathode material for lithium-ion battery. J. Solid State Electrochem. 2010, 14, 1001–1005.

28.

Okada, K.; Kimura, I.; Machida, K., High rate capability by sulfur-doping into LiFePO4 matrix. RSC Adv. 2018, 8, 5848–5853.

29.

Chen, Z.; Dahn, J. R., Reducing carbon in LiFePO4/C composite electrodes to maximize specific energy, volumetric energy, and tap density. J. Electrochem. Soc. 2002, 149, A1184-A1189.

30.

Herklotz, M.; Scheiba, F.; Hinterstein, M.; Nikolowski, K.; Knapp, M.; Dippel, A. –C.; Giebeler, L.; Eckert, J.; Ehrenberg, H., Advances in in situ powder diffraction of battery materials: a case study of the new beamline P02.1 at DESY, Hamburg. J. Appl. Cryst. 2013, 46, 11171127.


22

Page 23 of 32 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


31.

Herklotz, M.; Weiß, J.; Ahrens, E.; Yavuz, M.; Mereacre, L.; Kiziltas-Yavuz, N.; Dräger, C.; Ehrenberg, H.; Eckert, J.; Fauth, F.; Giebeler, L.; Knapp, M., A novel highthroughput setup for in situ powder diffraction on coin cell batteries. J. Appl. Cryst. 2016, 49, 340345.

32.

Roisnel, T.; Rodriguez-Carvajal, J., WinPLOTR: a windows tool for powder diffraction pattern analysis. Mater. Sci. Forum 2001, 378381, 118123.

33.

Rehr, J. J.; Mustre de Leon, J.; Zabinsky, S. I.; Albers, R. C., Theoretical x-ray absorption fine structure standards. J. Am. Chem. Soc. 1991, 113(14), 51355140.

34.

Weppner, W.; Huggins, R. A., Determination of the kinetic parameters of mixed‐ conducting electrodes and application to the system Li3Sb. J. Electrochem. Soc. 1977, 124, 15691578.


23


Page 24 of 32

SYNOPSIS Artwork


24

Page 25 of 32 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


a a

O

a

Li2/3Fe1/3 S



1stnd cycle 2rd cycle 3 cycle

I (mA)

1

0

-1 1

3

2 1st cycle

-10

10

charge discharge -12

10 D (cm2 s-1)

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 26 of 32

10-14 10-16 10

-18

10

-20

2.0

2.2

2.6 2.4 E vs. Li+/Li (V)


2.8

Page 27 of 32

(½00)cub.

Li

Al

l = 0.20718 Å

discharge

charge

Intensity (arb. units)

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


(100)cub.(½10)cub.

(½½0)cub.

PTFE 2

4

8

6

discharge

charge

8

10

12 2 Theta (degr.)


14

16

0.8 0.4


1.9 V 2 (LixFe)SO 2.2 V 1.7 2.3 V 1.6 2.5 V 1.4 charge 2.8 V 1.1 2.9 V 1 3.1 V x =0.8 1s 4p 2.3 V 1 2.2 V 1.1 1.9 V 1.4 discharge 1.8 V 1.6 1.7 V 1.7 1.2 V 2 1s 3d

100 98

a

pristine

Fe2+

96 94

Fe2O3 Fe3O4 FeO

0.0 2.8 2.6 2.4 2.2 2.0 1.8

Page 28 of 32

7100 7110 7120 7130 Energy (eV)

100 98

2.45 V

96

Fe2+/Fe3+

94

b

d

charge discharge

99

2.0 3.0 2.5 Fe valence state - GCPL

3.8 3.6

Fe-NNN(Fe,Li)

3.0

Fe-NN(Fe,Li)

2.9 V

98

Fe3+/Fe3+

97

c

100 99

2.8 Fe-S

2.0 1.8

100

1 cycle

98

Fe2+

Fe-O

97

0.8 1.0 1.2 1.4 1.6 1.8 2.0 x in (LixFe)SO

-2

-1


0 1 2 -1 velocity (mm s )

3

Transmission (%)

Norm. Abs. coeff. m (a.u.) Inter atomic distance (Å) Fe valence state - XANES

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

1.2

Page 29 of 32 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



80

60

(Li0.8NazFe)SO - Na-metal cell

60

40

40

20

charge discharge

20 0

80

0

10

C/10 20 30 Cycle number

40

50

+

a

100

Page 30 of 32

3 E vs. Na /Na (V)

100

120

Specific capacity (mAh g-1)

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

Coulombic efficiency (%)


b 2

7th cycle 8th cycle

1

0


0.2

0.4 0.6 z(Na) in (Li0.8NazFe)SO

Page 31 of 32

300

(C/10)

(C/5)

(1C) (C/20)

(C/10)

100

23 mA g-1

(Li2Fe)SO - Li-metal cell + 1.2 - 3.0 V vs. Li /Li charge discharge

50 0

46 mA g-1

150

230 mA g-1

11.5 mA g

-1

200

-1

250 23 mA g

Specific capacity (mAh g-1)

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


0

10

20

40 30 Cycle number


50

60


Page 32 of 32

S O

1/3 Fe 2/3 Li

2S 2p1/2 0 S 2p3/2 0 S 2p1/2


Fe3+/Fe3+ S2- 2p3/2

Operando Studies of Antiperovskite Lithium Battery Cathode Material

Recommend Documents