Insights into Mg2+ Intercalation in a Zero-Strain Material: Thiospinel

Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Insights into Mg2+ Intercalation in a Zero-Strain Material: Thiospinel MgxZr2S4 Patrick Bonnick, Lauren Blanc, Shahrzad Hosseini Vajargah, ChangWook Lee, Xiaoqi Sun, Mahalingam Balasubramanian, and Linda F. Nazar Chem. Mater., Just Accepted Manuscript • Publication Date (Web): 22 Jun 2018 Downloaded from http://pubs.acs.org on June 22, 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 25 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

Chemistry of Materials

Insights into Mg2+ Intercalation in a Zero-Strain Material: Thiospinel MgxZr2S4 Patrick Bonnick,a‡† Lauren Blanc,a‡ Shahrzad Hosseini Vajargah,a Chang-Wook Lee,b Xiaoqi Sun,a⸕ Mahalingam Balasubramanianb and Linda F. Nazara* a) Department of Chemistry and the Waterloo Institute of Nanotechnology, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada b) X-ray Science Division, Argonne National Laboratory, 9700 S Cass Ave, Lemont, Illinois, 60439, United States ‡ - These authors contributed equally to this work. Abstract The Mg2+ battery cathode material, thiospinel MgxZr2S4 (0 ≤ x ≤ 1) exhibits negligible volume change (ca. 0.05%) during electrochemical cycling, providing valuable insight into the limiting factors in divalent cation intercalation. As demonstrated by Rietveld refinement of XRD data of compositions at various states of charge, coupled with EDX analysis, Mg2+ can be inserted into Zr2S4 at 60°C up to x = 0.7 at a C/10 rate (up to x = 0.9 at very slow rates) and cycled with a high coulombic efficiency of 99.75%. HAADF-STEM studies provide clear visual evidence of Mg occupation in the lattice, whereas XAS studies show that Zr4+ was reduced upon Mg2+ intercalation. Operando and synchrotron XRD studies reveal the creation of two phases during the latter stages of discharge (x > 0.5) as the lattice fills and Mg2+ ions begin occupying tetrahedral (8a) sites in addition to octahedral (16c) interstitial sites. Compared to the isostructural Ti2S4 thiospinel, Zr2S4 presents a slightly larger cell volume and hence an almost ideal zero-strain lattice on Mg2+ insertion. Nonetheless, its four-fold lower electronic conductivity results in a diffusion coefficient for Mg2+ ions (DMg; 10-10 to 10-9 cm2/s) that is more than a factor of ten lower than in Ti2S4. This shows that delocalization of the electron charge carriers in the framework is a very important factor in governing multivalent ion diffusivity in the thiospinel framework and, by extension, in other materials. 1 ACS Paragon Plus Environment

Chemistry of 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 25

Introduction To utilize renewable resources and reduce fossil fuel consumption, energy storage technology is needed that surpasses Li-ion batteries which currently dominate the global market. Magnesium batteries are an exciting alternative due to the abundance, inherent safety, relatively low reduction potential (-2.37 vs. SHE), and exceptionally high volumetric capacity (3,832 mAh/cm3) of the Mg metal negative electrode.1–6 Despite these benefits, magnesium battery technology is plagued by the lack of suitable intercalation electrode materials and electrolytes that are stable at high potentials. Due to its divalent charge, Mg2+ experiences much stronger interactions with its host lattice in comparison to monovalent cations. These interactions often cause electrode materials to exhibit poor Mg2+ solid state diffusion,7,8 and promote a tendency to convert upon insertion9–12 that results in the formation of impenetrable layers on the surface of the electrode material. As a divalent cation, Mg2+ also experiences strong ion pairing and solvent interactions in solution that can lead to a large desolvation penalty at the electrode-electrolyte interface, which serves as an additional barrier to intercalation.13,14 In order to compete with Li-ion cells, the coulombic efficiency of both positive and negative electrodes in Mg2+ battery systems must also be improved to attain cycle life closer to typical Li-ion technology (i.e. > 500 cycles). As an example, thiospinel MgxTi2S4 delivers a reversible energy density of 230 Wh/kg, but with a coulombic efficiency of only 99%. 15 In a cell without excess Mg, such a coulombic efficiency would limit the cell to about 70 cycles before it could only deliver 50% of its initial capacity. Low coulombic efficiencies can be caused by a range of possibilities,16 but one common source is the expansion, shrinkage and consequent cracking of the active material during cycling. Li4Ti5O12 has found commercial success as a Li-ion battery negative electrode due to its lack of volume change during cycling, yielding an impressive cycle life (> 5,000 cycles).17 By minimizing structural changes upon (de)intercalation,

such

“zero-strain”

materials

exhibit

enhanced

stability

and

electrochemical performance. Highly reversible Mg2+ positive electrode (cathode) intercalation materials that have been reported (the metallic Chevrel phase Mo6S8 and electronically conductive layeredTiS2/spinel-Ti2S4)15,18,19 are able to reversibly cycle Mg2+ by utilizing a soft, polarizable 2 ACS Paragon Plus Environment

Page 3 of 25 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


anion-sulfide lattice that exhibits relatively weak interactions with the mobile ions. Density functional theory further reveals that the metallic nature of the Mo6S8 host provides fast, responsive electronic screening that shields the charge of Mg2+, promoting its good mobility in this structure.20 The performance of some of the sulfide materials was improved by operating cells at elevated temperatures to better overcome energetic barriers for insertion into, and diffusion through, the host. Despite these kinetic advantages, sulfides present high mass and low operating voltages, which leads to poor energy densities.21 It is theoretically possible to increase the operating voltage by using transition metals with high theoretical redox potentials such as Cr2S4; however, attempts to do so have failed to produce notable electrochemical reversibility.22 This is thought to be due to significant overlap between the sulfur 3p and transition metal 3d orbitals in these compounds, which effectively means that anion redox (S2-/S-) would govern the electrochemistry. The formation of electron holes in the sulfur 3p states is not favorable, rendering these materials electrochemically inactive.23,24 Cathode materials for magnesium intercalation with high energy densities can theoretically be obtained by using a light-weight and high-voltage oxide lattice. However, these compounds often favor conversion (to form MgO and stable transition metal oxides) over the intercalation of Mg2+ due to the highly polarizing nature of the O2- anion.4,9 After years of research efforts,25–30 corrosion-free electrolytes now exist that can resist oxidation up to 3.2 V (vs. Mg), which is sufficient for an intercalation cathode with an average discharge potential of about 2.5 V. It would be advantageous to use oxides to obtain high voltage cathode materials so that Mg batteries could compete with the energy density of Li-ion batteries; however, Mg2+ solid state diffusion must be improved before oxides can reversibly cycle this divalent ion. By studying systems that are capable of reversible Mg2+ intercalation, it is possible to isolate the key factors that affect Mg2+ diffusivity in solid lattices. These aspects can then be optimized to design electrode materials. Potential factors that limit diffusion rates include 1) restricted physical pathways for ion migration that lead to high activation energy barriers; 2) interactions of the divalent cation with either the anions and/or existing intercalated cations within the lattice; and 3) a lack of screening of the high positive cation charge by itinerant electrons in the host framework which typifies insulators or poor semiconductors. By comparing isostructural 3 ACS Paragon Plus Environment


Page 4 of 25

thiospinels that exhibit essentially the same framework for cation (de)insertion but differ in lattice volume and electronic conductivity, these factors can be deconvoluted as we show here. To this effect, we have synthesized and characterized the new spinel phase MgxZr2S4 to determine whether its larger (but less conductive) lattice exhibits better kinetic performance than MgxTi2S4.7,15 The two structures exhibit similar electrochemical behavior where Mg2+ reversibly intercalates into Zr2S4 at a slightly lower average potential than in Ti2S4, but with similar capacity. Importantly, despite a larger cell volume, the Mg2+diffusion coefficient decreases rather than increases, as measured by the galvanic intermittent titration technique (GITT). The primary advantage of the larger lattice is that MgxZr2S4 does not expand significantly upon Mg2+ insertion. This results in a zero-strain intercalation material which exhibits better coulombic efficiency (99.75%). Its low voltage (0.8 V vs. Mg) means MgxZr2S4 could be suitable as a counter electrode in configurations that explore new positive electrode materials for Mg2+ intercalation. More significantly, our study enables us to unequivocally identify the four-fold lower order of magnitude in electronic conductivity of Zr2S4 as the critical factor that impedes Mg2+ diffusion. This validates the concept that electrostatic interactions of a mobile ion are lowered when the electronic structure of its host framework can screen its charge, which is particularly important for multivalent cations such as Mg2+.31 Experimental Spinel CuZr2S4 was synthesized by mixing the elements in stoichiometric ratios in a mortar and pestle inside an Ar-filled glovebox. The mixture was pelletized, sealed in a quartz tube under vacuum, and heated at 750 °C for 1 day, followed by heat treatment at 450 °C for 3 days. The secondary particle agglomerates were about 10 μm in diameter (Figure S1). Copper was removed from the lattice by initial ion exchange with lithium followed by chemical oxidation in order to preserve the spinel phase of the material. The lattice was lithiated by stirring CuZr2S4 powder in excess n-butyllithium for 7 days at room temperature inside an Ar-filled glovebox due to the highly flammable nature of the reagent used. The product was chemically oxidized with excess I2 to clean Cu from the surface and remove lithium from the lattice.32 The final product, cubic Zr2S4, was collected.

4 ACS Paragon Plus Environment

Page 5 of 25 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 Zeiss field emission scanning electron microscope (SEM) equipped with an energy dispersive X-ray spectroscopy (EDX) detector was used to examine material morphologies and elemental compositions. Samples were transferred to the instrument with minimal exposure to air, and scans were collected at an accelerating voltage of either 15 or 20 keV. Zr2S4 electrodes were prepared in an Ar-filled glovebox by mixing the oxidized material with Super P and polyvinylidene fluoride (PVDF, average Mw ~ 534,000) in an 8:1:1 weight ratio in N-methyl-2-pyrrolidone (NMP, Sigma-Aldrich 99.5%). This mixture was ground for several minutes before the slurry was cast onto Mo foil and then dried at elevated temperature to remove NMP. The theoretical capacity was determined by the amount of active material in the positive electrode assuming a maximum of 1 Mg 2+ can intercalate per Zr2S4 unit.33 This capacity was used for determining the C-rate and x values in MgxZr2S4. The All-Phenyl Complex (APC) electrolyte was synthesized34,35 following a slightly modified literature procedure, with tetrahydrofuran (THF) or tetraglyme (G4) as the solvent. Prior to the preparation of APC in G4, excess solvent was removed from the phenylmagnesium chloride precursor through evaporation under vacuum at room temperature for 12 hours in order to crystallize C6H5MgCl∙6THF. The synthesis of magnesium monocarborane (MMC) in tetraglyme (G4) electrolyte has been described previously.7 These electrolytes were prepared without exposure to air or water to ensure safe handling of the required reagents. All electrochemical characterization were performed on cells assembled/materials stored in an Ar-filled glovebox. Cycling studies were performed by assembling coin cells (2325) using polished Mg as the counter electrode and a glass fiber separator. Galvanic Intermittent Titration Technique (GITT) experiments7 were carried out using ConFlat™ cells with three layers of (dry) glass fiber filter paper as separators and Mg foil as the reference and counter electrodes. All electrochemistry was measured with a Bio-logic VMP3 potentiostat while cells were held at 60°C in a thermostatted oven. Electrical conductivity was measured using a standard DC four-point probe method. Scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) characterization studies were carried out on a FEI Titan cubed 80-300 STEM equipped with a high-brightness field-emission gun (X-FEG), aberration correctors on probe- and image-forming lens systems, a monochromator and a Gatan image filter 5 ACS Paragon Plus Environment


Page 6 of 25

Quantum-965 spectrometer, operated at 200 kV. Atomic-number sensitive (Z-contrast) high-angle annular dark-field (HAADF) images and annular bright-field (ABF) images were obtained at the beam convergence semi-angle of 19.1 mrad using a detector angular range of 79.5-200 mrad and 10-20 mrad, respectively. Multislice HAADF-STEM image simulations were performed using Dr. Probe software36 with the same microscope parameters as the experimental imaging condition. In order to take the effect of a probe spread function into account, all of the simulated images were convolved by a Gaussian with the full-width at halfmaximum of 0.04 nm. The EELS images were recorded with 1 eV dispersion of the spectrometer to include Mg K-edge and Zr L2,3 and S k-edges in a single spectral image. EELS data were acquired with a beam current of 400-500 pA and a pixel dwell time of 20 ms to maximize the signal to noise ratio while minimizing the beam damage artifacts. X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) data were collected in transmission mode at the 20-BM beamline of Argonne's Advanced Photon Source. The incident beam was monochromatized using a Si(111) fixed-exit, double-crystal monochromator. Zr K-edge and Cu K-edge measurements were carried out for various MgxCuyZr2S4 samples. The energy scales were calibrated by defining the first derivative maxima of Zr and Cu foils as 17995.88 and 8990.48 eV, respectively.37 Data reduction was performed by established methods using the ATHENA module in Demeter. Structural parameters were determined by fitting the data to theoretical standards generated with FEFF 6 code, using the ARTEMIS module in Demeter.38,39 Operando X-Ray diffraction (XRD) measurements were performed using a homemade cell assembled in an Ar-filled glovebox. The positive electrode was prepared by casting the cathode slurry onto a glassy carbon electrode. This cathode was paired with a polished Mg metal anode using APC in G4 as the electrolyte. The cell was cycled at C/20 and held at 60 °C while XRD scans on the cathode materal were recorded operando. Scans were collected every 30 minutes which corresponds to Δx ~ 0.025 in MgxZr2S4. These scans were then fitted by the LeBail method to determine the lattice parameter evolution of the spinel phase during cycling. The Mg-intercalated thiospinel crystal structure was also analyzed ex-situ with XRD at the Canadian Light Source (CLS) synchrotron with a wavelength of 0.77492 Å. Rietveld refinement40 was performed by sequentially refining scale factor, zero point, background, lattice parameters, fractional coordinates, occupancies and atomic displacement parameters 6 ACS Paragon Plus Environment

Page 7 of 25 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


using the FullProf suite. The full refinement procedure is included in the Supplementary Information (Tables S1a-g). Results and Discussion The CuZr2S4 lattice was lithiated prior to copper extraction because previous studies have shown that, unlike CuTi2S4, no more than 70% of Cu can be removed from the Zr thiospinel topotactically through direct chemical oxidation.41 Since lithium can be removed using a mild oxidizing agent (e.g. I2 in ACN), this preliminary ion exchange avoids the use of harsher oxidizers that simultaneously oxidize the sulfide lattice during the extraction process.32 After de-copperiation, the resulting gray-brown powder consisted of primary crystallites ranging from approximately 0.3 - 1 μm in diameter (Figure S1), and the lattice parameter decreased from 10.3968(1) Å in pristine CuZr2S4 to 10.3561(2) Å in the oxidized material (Figure 1a,b). A combination of energy dispersive x-ray spectroscopy (EDX) and Rietveld refinement40 of X-ray diffraction (XRD) data (see Table 1) confirmed that the oxidized product maintained a spinel-phase cubic Fd3̅m lattice and contained a minimal amount of residual copper, Cu0.063(6)Zr2S4. Given the small Cu content, even though the lattice was not completely emptied, further references to the oxidized material are simplified to Zr2S4 for clarity and brevity. Atomic resolution HAADF-STEM and ABF-STEM images show the spinel phase obtained after the removal of Cu from the lattice. Images of the oxidized Zr2S4 oriented along the [011] zone axis are also displayed in Figure 1, where projected atomic structures are superimposed onto the experimental images. In the HAADF-STEM image (Figure 1c) Zr columns are clearly visible and the ABF- STEM image (Figure 1d) elucidates S which has a lower atomic number. The electrochemical behavior of Zr2S4||Mg coin cells was measured at 60°C. Figure 2a shows that the first and second cycles of these cells are markedly similar to those of Ti2S4||Mg cells15,33 except that Zr2S4 has a lower average discharge potential of 0.76 V (compared to 1.18 V for Ti2S4). An irreversible capacity loss of about 20 mAh/g occurs after the 1st discharge, indicating some Mg2+ is trapped within the thiospinel lattice (also visible in panels b and c). After the first cycle, Mg2+ reversibly intercalates into the structure to give



Page 8 of 25

Figure 1. XRD Rietveld refinement of (a) CuZr2S4 (blue asterisks mark the reflections that disappear upon removal of Cu) and (b) Zr2S4 that shows experimental data (black crosses), fitted profile (red line), difference map between observed and calculated data (blue line), and Bragg positions of the Fd3̅m spinel phase (green ticks). The insets show the crystal structure of the Fd3̅m spinel phase comprised of ZrS6 (green octahedra); in CuZr2S4 the Cu ions are shown as blue circles. Experimental atomic resolution (c) HAADF-STEM and (d) ABFSTEM images of the oxidized Zr2S4 along the [011] zone axis. Corresponding multislice simulated images are delineated by white and black boxes in the top left corner of each figure, and the projected atomic structures shown in the bottom right corner mark Zr atomic sites with higher and lower number of Zr atoms within the atomic column (dark and light green circles, respectively). S atomic sites (yellow circles) are superimposed onto the experimental images.

Mg0.7Zr2S4 on discharge at a C/20 rate, corresponding to a capacity over 100 mAh/g. Compared to MgxTi2S4, MgxZr2S4 displays a much better coulombic efficiency of 99.75(9)% for its first 50 cycles, as shown in Figure 2b. From cycle 50 to 70, soft short-circuits are evident in the galvanostatic potential profiles (Figure S2) until a complete, resistive, shortcircuit is formed around cycle 70. These features in the electrochemical profile indicate the formation of Mg dendrites, as has been observed in previous studies. 7 Several similar coin 8 ACS Paragon Plus Environment

Page 9 of 25 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. Electrochemical performance of MgxZr2S4||Mg cells at 60°C, (a) Galvanostatic cycling using APC in THF as the electrolyte. The inset shows cycle 1, which had a larger discharge capacity than future cycles. The inset has identical ranges for the x- and y-axes as the larger plot. (b) Cycling behavior using APC in G4 and a C/10 rate. Every 2nd discharge data point is omitted for clarity. (c) Rate capability of MgxZr2S4 electrode material showing linear fits (black lines) at each rate. (d) dQ/dV curve using 0.55 M MMC in G4, cycled at a C/50 rate. The red and blue lines represent the average discharge (0.76 V) and charge (1.02 V) potentials.

cells using one layer of glass fiber separator and either an all-phenyl complex (APC) or magnesium monocarborane30 (MMC) in tetraglyme (G4) electrolyte were cycled but all failed due to Mg dendrites. This emphasizes the need for future research to explore the conditions that support or eliminate Mg dendrite growth. Regardless of issues with the Mg negative electrode, thiospinel Zr2S4 displays superior cycling stability which implies it has utility as a stable counter electrode to probe the properties of high voltage positive electrode materials in a rocking chair configuration. Employing Zr2S4 as a counter electrode could facilitate the 9 ACS Paragon Plus Environment


Page 10 of 25

search for novel cathode materials by allowing the use of oxidation-resistant electrolytes that would otherwise passivate the surface of the traditional Mg negative electrode. Figure 2c shows the rate capability of a Zr2S4|MMC in G4|Mg cell cycled at 60°C. About 70% of the capacity of MgxZr2S4 at a C/20 rate is delivered at a C/5 rate. The fade rate depends on the cycling rate, as emphasized by the slope of the black linear fits drawn in Figure 2b. Since the fade rate is faster at slower cycling rates, the underlying mechanism is likely time dependent. A sustained chemical reaction may occur that slowly degrades the Zr2S4, which might be the breakdown of intercalated MgZr2S4 into MgS and transition metal sulfides. Figure 2d illustrates the differential capacity (dQ/dV) of a typical cycle at C/50. The broad peak at higher potential indicates a single phase process, such as intercalation. The valley between the broad and narrow peaks, at about 0.77 V during discharge, could be due to Mg2+ ordering (i.e. when the octahedral 16c sites are ¼ filled) as postulated by Emly and Van der Ven in the MgxTi2S4 thiospinel;15,42 however, XRD evidence of ordering was not observed for either the Ti2S4 or Zr2S4 thiospinels in our work. In fact, the sharp peak at lower potentials in Figure 2d is commonly indicative of a two-phase reaction, suggesting the valley is instead a transition from one electrochemical process to another as described below. The reversible intercalation of Mg2+ on cycling was confirmed using a combination of EDX, XRD, and Rietveld refinement at various states of (dis)charge (i.e. MgxZr2S4, 0 ≤ x ≤ 1), as summarized in Table 1. Since the Cu content in electrochemically cycled materials was invariant within experimental error, Cu EDX values are omitted from the bottom half of Table 1 for clarity. Refined Mg2+ occupancies and EDX measurements were comparable to the electrochemical Mg2+ content calculated assuming two electrons per Mg2+ ion (more detail is shown in Figures S3 and S4, and Tables S1a-f). This excellent agreement confirms that the observed discharge/recharge capacity can indeed be attributed to Mg2+ intercalation, as opposed to parasitic reactions. XRD analysis of both the partially and fully discharged material showed that Mg0.31(5)Zr2S4 remained single phase with no increase in lattice size during the first stage of intercalation, while interestingly Mg0.74(4)Zr2S4 separated into two distinct phases during the second stage of intercalation (Table 1 and Figure 5).


Page 11 of 25 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


Table 1. Mg2+ and Cu2+ occupancies in MgxCuyZr2S4 as determined by Rietveld refinement of XRD, and electrochemistry, with EDX for comparison. The lattice parameters for the cubic Fd3̅m phases are also included, which emphasize the formation of a larger cell volume phase only in the latter half of discharge. Mg2+ insertion and removal was carried out at a C/20 rate.

Lattice XRD Rietveld Parameter, Refinement a (Å)

ElectroEDX chemical

Pristine, CuZr2S4

10.3968(1)

0.950(6)

-

0.93(12)

Oxidized, Zr2S4

10.3561(2)

0.063(6)

-

0.08(1)

Partially Discharged, Mg0.3Zr2S4

10.3540(3)

0.31(5)

0.39

0.32(5)

Phase 1: 43(2)%

10.3607(4)

0.54(5)

Phase 2: 57(2)%

10.4240(4)

0.89(7)

Total 0.77 0.74(4)

0.76(2)

10.3597(5)

0.08(6)

0.16

0.11(4)

Cuy (top) and Mgx (bottom) content in MgxCuyZr2S4

Cuy

Mgx

Fully Discharged, Mg0.7Zr2S4

Recharged, Mg0.1Zr2S4

To further investigate Mg2+ insertion into the Zr2S4 structure, we analyzed MgxZr2S4 discharged to x = 0.4 and x = 0.6 with electron energy loss spectroscopy (EELS) and atomic resolution STEM. Figures 3a and b display the Mg K-, Zr L2-, Zr L3- and S K-edges EELS spectra for the pristine (x = 0), discharged (x = 0.4 and 0.6) and recharged (x = 0.1) materials. The Mg K-edge peak at 1300 eV appears in the discharged materials and almost completely disappears after charging. The remaining intensity of the Mg K-edge peak is due to residual Mg2+ in the structure. The shift of the Zr L2,3-edges toward higher energy (2228 eV) in the discharged materials - accompanied by the lower L3/L2 ratio - indicates a lower average oxidation state of Zr in comparison to the pristine material. The peak energy of the charged material shifts back to lower energy (~2225 eV), similar to that of the pristine material (2224 eV). In contrast, the peak position and fine structure of the S K-edge is almost identical in all four materials indicating that S is electrochemically inactive throughout (dis)charge. This correlates well with sulfur 2p XPS data that compares charged and discharged MgxZr2S4, which also reveals no S redox activity (Figure S5). To further confirm the intercalation of Mg2+, MgxZr2S4 discharged to x = 0.6 was explored with STEM. Of the six different octahedral Mg2+ sites in the structure, two do not overlap with Zr atoms along the [011] projection axis, 11 ACS Paragon Plus Environment


Page 12 of 25

Figure 3. EELS spectra of the (a) Mg K-edge, (b) Zr L2,3 edge and S K-edge of MgxZr2S4 with pristine x=0 (black line), discharged x=0.4 (green line) and x=0.6 (red line), and recharged x=0.1 (blue line). An expanded region of the spectra is shown in panel (e), indicating a 4 eV shift of the Zr L 3-edge in the discharged material with respect to the pristine. Experimental atomic resolution (c) HAADF-STEM and (d) ABF-STEM images of discharged Mg0.6Zr2S4 along the [011] zone axis with the corresponding multislice simulated images delineated by white and black boxes in the images, respectively. The projected atomic structures are superimposed onto the experimental images, showing Zr atomic sites (green circles), S atomic sites (yellow circles), mixed Zr and Mg columns (blue triangles), and Mg atomic sites (red squares). Mg2+ ions are marked by red arrows on both the simulated and experimental images. Since the Cu content is too low to be detectable by these imaging techniques, Cu atomic sites were not included in the projected atomic structures.

and thus their presence can be used to verify Mg2+ occupation. The multislice HAADF-STEM image simulation performed on Mg0.6Zr2S4 (top-left inset in Figure 3d) clearly shows the columns of octahedrally coordinated Mg2+ ions in the ABF-STEM image along this zone axis. The contrast observed in the experimental ABF-STEM image matches the simulation well, and the Mg ion columns are highlighted by the red boxes in Figure 3d.


Page 13 of 25 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


X-ray absorption spectroscopy (XAS) confirms that Zr4+ is reduced to a formal oxidation state of Zr3+ during discharge and oxidized during charge. However, in a highly covalent lattice, zirconium’s “true” oxidation state will not be equivalent to its formal value. Figure 4a shows the Zr K-edge (XANES) for MgxCuyZr2S4 at various states of x and y, while panel b shows a magnified image highlighting the energy values at half-height of the normalized XANES spectra (E0.5, circles). Zr undergoes reversible redox when Cu ions are removed from the CuZr2S4 thiospinel (gray to black curve), and as Mg2+ is inserted (black to red curve) and removed (red to blue curve).37 Surprisingly, the E0.5 energy of discharged Mg0.76Cu0.08Zr2S4 is higher than pristine CuZr2S4 in Figure 4b, which is not consistent with the commonly assumed oxidation state of +1 for Cu in thiospinels. A comparison between the differences in E0.5 values and the Mg and Cu compositions, as measured by EDX, reveals that the Zr K-edge energies are only consistent if Cu is in a +2 oxidation state (see Table S2). Unfortunately, Cu K-edge XANES experiments comparing Cu2S, CuS and CuZr2S4 were

Figure 4. Zr K-edge XANES spectra of MgxCuyZr2S4 at various values of x and y showing CuZr2S4 (gray line), oxidized Zr2S4 (black line), discharged Mg0.76Zr2S4 (red line) and recharged Mg0.11Zr2S4 (blue line). Panel (b) is a magnified rectangle from Panel (a), showing the points used to conclude that Cu exists in a +2 oxidation state in CuZr2S4 (see Table S2). The estimated error in the energy differences (ΔE) is ±0.05 eV.



Page 14 of 25

inconclusive in directly measuring the copper oxidation state (Figure S6) due to the differing local structure and bonding in these compounds (Table S3).43,44 The local structure of zirconium and copper in MgxCuyZr2S4 was also probed by EXAFS. The main results of the fits to the EXAFS data are given in Table 2; further details can be found in the SI (Figure S7). When copper is removed from CuZr2S4 to form Zr2S4, there is a small but measurable decrease in the Zr-S correlation distance by ~0.015 Å. Likewise, the Zr-Zr correlation distance also decreases by ~0.028 Å. These observations are consistent with the changes expected upon zirconium oxidation. When Mg2+ is inserted to form Mg0.76Zr2S4, the Zr-S bond distance recovers to a value close to that seen for CuZr2S4. This is consistent with the reduction of zirconium on magnesium insertion. Intriguingly, the Table 2. Structural parameters extracted from the Zr K-Edge EXAFS analysis; the coordination number (CN), distance (R), ∆R = Rsample – Roxidized, and mean-squared relative displacement (σ2) are reported. Sample

Correlation

CN (set)

R (Å)

∆R (Å)

σ2 (10-4 Å2)

CuZr2S4 (synthesized)

Zr – S

6

2.584 ± 0.006

0.015

62.9 ± 3.9

Zr – Zr

6

3.706 ± 0.015

0.028

130.7 ± 17.2

Zr – S

6

2.569 ± 0.006

0

68.0 ± 4.3

Zr – Zr

6

3.678 ± 0.009

0

92.2 ± 9.5

Zr – S

6

2.583 ± 0.005

0.014

69.0 ± 3.5

Zr – Zr

6

3.667 ± 0.015

-0.011

154 .0± 18.3

Zr – S

6

2.570 ± 0.006

0.001

65.2 ± 4.1

Zr – Zr

6

3.675 ± 0.009

-0.003

93.1 ± 9.2

Zr2S4 (oxidized) Mg0.76Zr2S4 (discharged) Mg0.11Zr2S4 (charged)

Zr-Zr correlation distance decreases slightly or remains similar to Zr2S4, within the uncertainty of the measurements. On subsequent removal of Mg2+, the Zr-S bond distance returns to a value similar to that seen for Zr2S4, consistent with the oxidation of zirconium. However, again the Zr-Zr correlation distance remains largely invariant. The variations in the Zr-S bond distances after discharge/recharge prove that the redox charge compensation is endured by zirconium, in agreement with the XANES and EELS results. The absence of significant changes in the Zr-Zr correlation distance also confirm that the system behaves as a zero-strain material, even on the medium-range scale. This echoes the beneficial behavior seen in the XRD data (vide infra), which elucidates the response of the average structure. 14 ACS Paragon Plus Environment

Page 15 of 25 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


Operando XRD33 studies were performed on a Zr2S4||Mg cell to better characterize the structural changes taking place within the MgxZr2S4 lattice during each stage of discharge. The results are shown in Figure 5a, which depicts the (004) and (044) reflections that are characteristic of the appearance and disappearance of a 2nd Fd3̅m phase at x ≳ 0.5, as suggested by the dQ/dV plot (Figure 2d). Le Bail fitting was used to track the lattice parameter evolution of the spinel phase(s) upon Mg2+ (de)intercalation (Figure 5b). During the first half of discharge, the lattice parameter does not deviate significantly from its pristine value, indicating a zero-strain lattice as Mg2+ enters the structure up to x ~ 0.5. After this point, the material splits into two separate cubic Fd3̅m phases where Phase 1 (blue squares) maintains a relatively static lattice parameter and Phase 2 (red circles) exhibits a slight increase. The single phase material is recovered on subsequent recharge as Mg2+ exits the lattice. Since strain-induced cracking promotes continual SEI growth, and eventually leads to pulverization, prohibiting this source of capacity fade would further optimize

Figure 5. Operando XRD performed during cycling of a MgxZr2S4|APC in G4|Mg at 60°C and a C/20 rate. a) Sequential XRD patterns from the cell. Every other scan recorded is shown, corresponding to Δx ~ 0.05 in MgxZr2S4. (b) Cubic Fd3̅m cell parameter evolution as a function of x. Phase 1 (blue squares) shows minimal lattice strain and Phase 2 (red circles) appears around x ~ 0.5 during discharge. (c) Cell voltage of the first cycle as a function of x.



Page 16 of 25

electrochemical performance. As such, the lack of significant strain during cycling in MgxZr2S4 likely contributes to the good coulombic efficiencies displayed in Figure 2b. The galvanic intermittent titration technique (GITT) was used to compare the Mg2+ diffusivity between the analogous thiospinels MgxZr2S4 and MgxTi2S4, (Figure S8).33 Figure 6 shows that there is a sharp decrease in the Mg2+ diffusion coefficient during the discharge of MgxZr2S4 that is coincident with the onset of phase separation. In this two phase regime, the standard GITT technique cannot accurately determine DMg and thus we denote data for DMg for x > 0.5 as a dotted line. Interested readers are directed to the work of Zhu and Wang, who propose a technique to deconvolute the ion diffusion and phase boundary movement in two phase regimes.45 Their phase transformation GITT method has shown that the true ion diffusion coefficient in two-phase regions can be several orders of magnitude higher than what is predicted by standard GITT experiments. This suggests that DMg in MgxZr2S4 may not decrease as sharply beyond x ~ 0.5 as indicated in Figure 6. Further investigation is required to elucidate the true behavior in this region, which is outside the scope of this work. We note that previous experiments have indicated that MgxTi2S4 remains a single phase

Figure 6. Mg2+ diffusion coefficients in both spinel Ti2S4 and Zr2S4, measured using GITT at 60°C. MgxZr2S4 separates to form a 2nd phase above x ~ 0.5 that prevents the accurate determination of DMg in the two-phase region using standard GITT methods; DMg in this regime is denoted by dotted lines. Both data sets were prematurely terminated on charge by Mg dendrite-induced short circuits.


Page 17 of 25 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


material throughout discharge,15 but it shows a similar decline in Mg2+ diffusivity past x ~ 0.6 that was assigned to the onset of Mg2+ tetrahedral occupation.7 Most importantly, in comparison to Ti2S4, the diffusion coefficient of Mg2+ in Zr2S4 is about an order of magnitude smaller throughout the single phase region up to x ~ 0.5, which leads to increased overpotential and more limited discharge capacity. Since the slightly larger lattice dimensions of Zr2S4 (a = 10.356 Å) vis a vis Ti2S4 (a = 9.776 Å) do not induce better Mg2+ diffusivity, clearly this factor does not strongly influence divalent ion diffusion in the thiospinel. While too large a pathway dimension can decrease mobility, such an effect would not be expected here given the small change in lattice parameter. A more important parameter is the electronic conductivity of the host. Thöle et al. have suggested that intercalation diffusion rates are also dependent on the ease with which the electron cloud in the host can travel through the lattice to accompany the moving cation;20 theory reveals that in the metallic Chevrel phase, the ionic charges of the sulfide anions are shielded from the highly charged Mg2+ ion by the itinerant electrons of the framework. Consistent with this finding is the fact that our measurements show that Zr2S4 (σe = 1.7 × 10-4 S/cm) is much less electronically conductive than Ti2S4, (σe = 3.54 S/cm). The lower electronic conductivity of CuZr2S4 vs CuTi2S4 has also been reported in the literature,46 and is ascribed to a difference in band structure. We conclude this factor is responsible for the more sluggish Mg2+ diffusion in the poor semiconductor Zr2S4, indeed showing that the diffusivity of divalent ions is very sensitive to the degree of electron delocalization in the lattice. Ex-situ synchrotron XRD (Figure 7) also demonstrates that the thiospinel maintains the cubic Fd3̅m space group throughout Cu removal and Mg2+ cycling as expected, and the structural data obtained from Rietveld refinement are in good agreement with the EXAFS results (Table S4). These structural data were used to confirm that an equally facile diffusion pathway is available in the larger Zr2S4 lattice in comparison to Ti2S4. Since the bottleneck to Mg2+ diffusion in the thiospinel lattice is the migration through the triangular face shared between adjacent octahedral and tetrahedral sites,47 the bond distance between a Mg2+ ion in the center of this face and its closest S neighbor was calculated. This distance is even longer in Zr2S4 (2.149(4) Å) than in Ti2S4 (2.089(3) Å), confirming that the limited Mg2+ diffusivity observed in Figure 5 is not due to a more restricted window in the Zr thiospinel lattice. In Figure 7, the slight decrease in lattice parameter upon oxidation from CuZr2S4 to Zr2S4 17 ACS Paragon Plus Environment


Page 18 of 25

Figure 7. a) Synchrotron XRD (λ = 0.77492 Å) patterns of cubic, Fd3̅m thiospinel CuZr2S4 and MgxZr2S4 at various x values; (b) Enlarged (004) and (044) reflections showing the Bragg peaks of Phase 1 (green asterisks) and Phase 2 (purple asterisks) for the two-phase, fully discharged material; (c) Two-phase behavior of discharged MgxZr2S4 as a function of discharge rate and equilibration time. Mg0.7Zr2S4 was obtained through discharge at C/20 and the material was either equilibrated before cell disassembly (green line) or disassembled immediately upon discharge (non-equilibrated, red line). Increased Mg2+ occupation (Mg0.9Zr2S4) was achieved through slow, C/100 discharge (purple line).

signifies that the lattice shrinks as Cu2+ is removed from the tetrahedral (8a) lattice sites. Rietveld refinement reveals that MgxZr2S4 exhibits minimal (almost zero) strain throughout the first half of discharge as described above in Figure 5, during which Mg2+ only occupies octahedral (16c) sites. The larger phase forms in the second half of discharge when Mg2+ begins occupying tetrahedral (8a) sites as well (Figures S3 and S4, and Tables S1a-f). Since the Zr thiospinel lattice is only forced to expand during the latter step, the tetrahedral site occupancy results in an increase in cell volume and separation of a second Fd3̅m phase observed in the operando XRD studies shown in Figure 5. The experiments shown in Figure 7 demonstrate that as MgxZr2S4 is discharged, Mg2+ ions occupy thermodynamically favored octahedral sites47 from x = 0 to about 0.5, as they migrate toward the core of the particle. The sharp onset of dual site occupancy suggests that the thermodynamic stability of Mg2+ occupation of the tetrahedral sites increases with x until it nears that of the octahedral sites and becomes accessible at x ~ 0.5. However, we cannot rule out the effect of kinetics potentially forcing Mg2+ into tetrahedral sites since the diffusion of Mg2+ will also encounter a higher energy barrier at high Mg2+ concentrations due to the onset of di-vacancy hops,7 as we have previously suggested for MgxTi2S4. We note that the 18 ACS Paragon Plus Environment

Page 19 of 25 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


separate phases that clearly form during discharge in MgxZr2S4 persist when the material is discharged at a slow rate or equilibrated after discharge, as shown in Figure 7c. A visual depiction of the Rietveld refinement results of these experiments is shown in Figure 8 (see the SI for a more in-depth discussion of this figure). In contrast, even though Mg2+ diffusivity in MgxTi2S4 decreases past x ~ 0.6, the fully discharged material appears to remain in a single phase, reaching a final composition of Mg[oct]0.59Mg[tet]0.19Ti2S4.7,15 The fully discharged material likely has a core-shell morphology composed of the larger unit-cell phase (Mg[oct]0.53Mg[tet]0.37Zr2S4) at the surface surrounding a zero-strain core (Mg[oct]0.54Zr2S4). This composition is achieved when the material is discharged at C/20 to a relatively low x value (Mg0.7Zr2S4) and given minimal equilibration time to relax before disassembly of the cell. If a C/20 discharged electrode is equilibrated for several days before disassembly (until dV/dt drops below 0.1 mV/h), then the ratio of Phase 1:Phase 2 in Mg0.7Zr2S4 increases over time, from 43:57 in non-equilibrated electrodes to 55:45 in equilibrated electrodes (Figure 8a vs 8b, and Table S1e). In all C/20 discharged phases, the fraction of Mg2+ occupying octahedral sites is capped below a maximum of x = 0.6, near the sharp decrease of Mg2+ diffusivity in the MgxZr2S4 lattice (Figure 6, Tables S1d,e). The total Mg2+ content can be pushed beyond x ~ 0.7 by discharging the material using slow rates at 60°C. MgxZr2S4 discharged at C/40 and C/100 reached final compositions of Mg0.89Zr2S4 and Mg0.93Zr2S4, respectively (Figure 8c,d). In these cases, the discharged material was mostly comprised of Phase 2 (Mg[oct]0.55Mg[tet]0.40Zr2S4), since the larger lattice forms at high Mg2+ occupancy

Figure 8. Summary of the Rietveld refinement site occupation data for MgxZr2S4 discharged at different rates (see Tables S1d and e). The equilibrated material in panel b was left in the cell at 60°C at open circuit until dV/dt < 0.1 mV/h; the other materials were analyzed by XRD immediately after discharge finished at Ecell = 0.1 V.



Page 20 of 25

(Figure 5). Interestingly, the composition of Phase 1 remaining in these materials showed a significant increase in octahedral site occupation up to Mg[oct]0.9Zr2S4 in the sample discharged at C/100 (Table S1e). This indicates that pushing octahedral occupation beyond x ~ 0.5 is thermodynamically achievable, yet kinetically inaccessible at reasonable discharge rates. Conclusions Synthesis and examination of the new zero-strain Mg2+ intercalation material, thiospinel MgxZr2S4, provides important insight into the factors that control Mg2+ intercalation and diffusion within solid state lattices. Electrochemical, HAADF-STEM, EDX and XRD Rietveld refinement of compositions at various states of charge provide clear evidence of Mg2+ insertion in this material, which exhibits an impressive coulombic efficiency of 99.75% due to its mostly static lattice parameters during cycling. During the initial stages of discharge (x < 0.5), Mg2+ only occupies octahedral (16c) sites, but on deeper discharge Mg2+ ions begin occupying tetrahedral (8a) sites in MgxZr2S4 at about x = 0.5. Tetrahedral site occupation presumably occurs as the energy of the tetrahedral and octahedral sites become equivalent; however, we cannot rule out the role that kinetics may play in determining the average Mg2+ concentration at which this occurs. Tetrahedral site occupation expands the lattice, triggering a phase separation between the zero-strain core where Mg2+ occupies only octahedral sites, and the expanding shell where Mg2+ occupies both octahedral and tetrahedral sites. This process is an example of two thermodynamically distinct solid solutions where the crossover between the phases is driven by local volume changes.48 More importantly, comparison of the two nearly isostructural materials, MgxTi2S4 and MgxZr2S4, allows the identification of factors controlling ion diffusivity. We find that the electrical conductivity of the host material is a crucial factor in determining magnesium ion diffusivity. This is evidenced by the much lower Mg2+ diffusivity of MgxZr2S4 - despite exhibiting a zero-strain lattice and slightly larger unit cell volume - in comparison with MgxTi2S4, which has an electrical conductivity four orders of magnitude higher. This also highlights the challenge in developing purely insulating solid electrolytes for use with 20 ACS Paragon Plus Environment

Page 21 of 25 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


divalent cations.49 In the ongoing search for new Mg-ion positive electrode materials, targeting mixed ion/electronic conductors that exhibit high electronic conductivity of the host is important to unlock a boost in Mg2+ diffusion rates. Associated Content Supporting Information: Further discussion of phase separation, detailed experimental methods, Rietveld refinement methods and results, XANES analysis of Cu oxidation state, EXAFS details (including the raw EXAFS spectra, Fourier transforms and a comparison of the raw data to the fits), SEM images, demonstration of soft short-circuits in electrochemical cycling, sulfur 2p XPS, raw data for the GITT experiments. This material is available free of charge via the Internet at http://pubs.acs.org. Author Information Corresponding Author * Prof. Linda F. Nazar, [email protected] Current Addresses †

Patrick Bonnick – Toyota Research Institute of North America, 1555 Woodridge Ave., Ann

Arbor, Michigan, 48105, USA ⸕

Xiaoqi Sun – Northeastern University, Department of Chemistry, No. 11, Lane 3, WenHua

Road, HePing District, Shenyang, Liaoning, 110819, China Author Contributions ‡ - These authors contributed equally to this work. P.B. and L.F.N. conceived the experiments and study; P.B. and L.B. carried out the materials synthesis, electrochemical studies, and materials characterization/analysis with the help of X.S.. S.H.V. carried out the TEM studies and analyzed the data; C.-W.L. and M.B. performed the XAS study and analysis and contributed discussion to the text; P.B., L.B. and L.F.N. wrote the manuscript.



Page 22 of 25

Funding Sources This work was supported by the Joint Center for Energy Storage Research (JCESR), an Energy Innovation Hub funded by the US Department of Energy (DOE), Oﬃce of Science, Basic Energy Sciences. NSERC is acknowledged by LFN for a Canada Research Chair, and support through the NSERC Discovery Grant program. X-ray absorption spectroscopy measurements were performed at the Argonne National labs at the APS. Sector 20 operations are supported by the DOE Office of Science. Synchrotron XRD measurements were performed using beamline 08B1-1 at the Canadian Light Source, which is supported by the Canada Foundation for Innovation, Natural Sciences and Engineering Research Council of Canada, the University of Saskatchewan, the Government of Saskatchewan, Western Economic Diversification Canada, the National Research Council Canada, and the Canadian Institutes of Health Research. Acknowledgements We are grateful to Dr. Joel Reid at the Canadian Light Source for performing the synchrotron XRD experiments. Electron Microscopy work was carried out at the Canadian Centre for Electron Microscopy (CCEM), a facility supported by NSERC and McMaster University. P.B. would also like to thank Dr. Jung-Soo Kang for his help with XPS. Graphic for Table of Contents

References (1)

Muldoon, J.; Bucur, C. B.; Gregory, T. Fervent Hype behind Magnesium Batteries: An Open Call to Synthetic Chemists-Electrolytes and Cathodes Needed. Angew. Chem. Int. Ed Engl. 2017, 56, 12064–12084. 22 ACS Paragon Plus Environment

Page 23 of 25 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)

(3)

(4)

(5) (6)

(7) (8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16) (17)

Cabana, J.; Monconduit, L.; Larcher, D.; Palacín, M. R. Beyond Intercalation-Based LiIon Batteries: The State of the Art and Challenges of Electrode Materials Reacting Through Conversion Reactions. Adv. Mater. 2010, 22, E170–E192. Yoo, H. D.; Shterenberg, I.; Gofer, Y.; Gershinsky, G.; Pour, N.; Aurbach, D. Mg Rechargeable Batteries: An on-Going Challenge. Energy Environ. Sci. 2013, 6, 2265– 2279. Canepa, P.; Sai Gautam, G.; Hannah, D. C.; Malik, R.; Liu, M.; Gallagher, K. G.; Persson, K. A.; Ceder, G. Odyssey of Multivalent Cathode Materials: Open Questions and Future Challenges. Chem. Rev. 2017, 117, 4287–4341. Muldoon, J.; Bucur, C. B.; Gregory, T. Quest for Nonaqueous Multivalent Secondary Batteries: Magnesium and Beyond. Chem. Rev. 2014, 114, 11683–11720. Saha, P.; Datta, M. K.; Velikokhatnyi, O. I.; Manivannan, A.; Alman, D.; Kumta, P. N. Rechargeable Magnesium Battery: Current Status and Key Challenges for the Future. Prog. Mater. Sci. 2014, 66, 1–86. Bonnick, P.; Sun, X.; Lau, K.-C.; Liao, C.; Nazar, L. F. Monovalent versus Divalent Cation Diffusion in Thiospinel Ti2S4. J. Phys. Chem. Lett. 2017, 8, 2253–2257. Levi, E.; Lancry, E.; Mitelman, A.; Aurbach, D.; Isnard, O.; Djurado, D. Phase Diagram of Mg Insertion into Chevrel Phases, MgxMo6T8 (T = S, Se). 2. The Crystal Structure of Triclinic MgMo6Se8. Chem. Mater. 2006, 18, 3705–3714. Sun, X.; Blanc, L.; Nolis, G. M.; Bonnick, P.; Cabana, J.; Nazar, L. F. NaV1.25Ti0.75O4: A Potential Post-Spinel Cathode Material for Mg Batteries. Chem. Mater. 2018, 30, 121– 128. Sun, X.; Duffort, V.; Mehdi, B. L.; Browning, N. D.; Nazar, L. F. Investigation of the Mechanism of Mg Insertion in Birnessite in Nonaqueous and Aqueous Rechargeable MgIon Batteries. Chem. Mater. 2016, 28, 534–542. Arthur, T. S.; Zhang, R.; Ling, C.; Glans, P.-A.; Fan, X.; Guo, J.; Mizuno, F. Understanding the Electrochemical Mechanism of K-α-MnO2 for Magnesium Battery Cathodes. ACS Appl. Mater. Interfaces 2014, 6, 7004–7008. Ling, C.; Zhang, R.; Arthur, T. S.; Mizuno, F. How General Is the Conversion Reaction in Mg Battery Cathode: A Case Study of the Magnesiation of α-MnO2. Chem. Mater. 2015, 27, 5799–5807. Kundu, D.; Vajargah, S. H.; Wan, L.; Adams, B.; Prendergast, D.; Nazar, L. F. Aqueous vs. Nonaqueous Zn-Ion Batteries: Consequences of the Desolvation Penalty at the Interface. Energy Environ. Sci. 2018, 11, 881–892. Wan, L. F.; Perdue, B. R.; Apblett, C. A.; Prendergast, D. Mg Desolvation and Intercalation Mechanism at the Mo6S8 Chevrel Phase Surface. Chem. Mater. 2015, 27, 5932–5940. Sun, X.; Bonnick, P.; Duffort, V.; Liu, M.; Rong, Z.; Persson, K. A.; Ceder, G.; Nazar, L. F. A High Capacity Thiospinel Cathode for Mg Batteries. Energy Environ. Sci. 2016, 9, 2273–2277. Smith, A. J.; Burns, J. C.; Xiong, D.; Dahn, J. R. Interpreting High Precision Coulometry Results on Li-Ion Cells. J. Electrochem. Soc. 2011, 158, A1136–A1142. Ohzuku, T.; Ueda, A.; Yamamoto, N. Zero‐Strain Insertion Material of Li[Li1/3Ti5/3]O4 for Rechargeable Lithium Cells. J. Electrochem. Soc. 1995, 142, 1431–1435.



Page 24 of 25

(18) Aurbach, D.; Lu, Z.; Schechter, A.; Gofer, Y.; Gizbar, H.; Turgeman, R.; Cohen, Y.; Moshkovich, M.; Levi, E. Prototype Systems for Rechargeable Magnesium Batteries. Nature 2000, 407, 724–727. (19) Sun, X.; Bonnick, P.; Nazar, L. F. Layered TiS2 Positive Electrode for Mg Batteries. ACS Energy Lett. 2016, 1, 297–301. (20) Thöle, F.; Wan, L. F.; Prendergast, D. Re-Examining the Chevrel Phase Mo6S8 Cathode for Mg Intercalation from an Electronic Structure Perspective. Phys. Chem. Chem. Phys. 2015, 17, 22548–22551. (21) Liu, M.; Jain, A.; Rong, Z.; Qu, X.; Canepa, P.; Malik, R.; Ceder, G.; Persson, K. A. Evaluation of Sulfur Spinel Compounds for Multivalent Battery Cathode Applications. Energy Environ. Sci. 2016, 9, 3201–3209. (22) Barahona, P.; Galdámez, A.; Lopez Vergara, F.; Manríquez, V.; Peña, O. Synthesis and Magnetic Properties of Chromium-Based Cu(Cr2-xTix)S4 Thiospinels and Their Deficient Structures Cu1-yOyCr2-xTixS4 Obtained by Copper Extraction. J. Chil. Chem. Soc. 2014, 59, 2294–2298. (23) Rouxel, J. Anion–Cation Redox Competition and the Formation of New Compounds in Highly Covalent Systems. Chem. – Eur. J. 1996, 2, 1053–1059. (24) Rouxel, J. Some Solid State Chemistry with Holes: Anion-Cation Redox Competition in Solids. Curr Sci 19970710, 73, 31–39. (25) Mohtadi, R.; Matsui, M.; Arthur, T. S.; Hwang, S.-J. Magnesium Borohydride: From Hydrogen Storage to Magnesium Battery. Angew. Chem. Int. Ed Engl. 2012, 51, 9780– 9783. (26) Muldoon, J.; Bucur, C. B.; Oliver, A. G.; Sugimoto, T.; Matsui, M.; Kim, H. S.; Allred, G. D.; Zajicek, J.; Kotani, Y. Electrolyte Roadblocks to a Magnesium Rechargeable Battery. Energy Environ. Sci. 2012, 5, 5941–5950. (27) Carter, T. J.; Mohtadi, R.; Arthur, T. S.; Mizuno, F.; Zhang, R.; Shirai, S.; Kampf, J. W. Boron Clusters as Highly Stable Magnesium-Battery Electrolytes. Angew. Chem. 2014, 126, 3237–3241. (28) Muldoon, J.; Bucur, C. B.; Oliver, A. G.; Zajicek, J.; Allred, G. D.; Boggess, W. C. Corrosion of Magnesium Electrolytes: Chlorides – the Culprit. Energy Env. Sci 2013, 6, 482–487. (29) Bucur, C. B.; Gregory, T.; Oliver, A. G.; Muldoon, J. Confession of a Magnesium Battery. J. Phys. Chem. Lett. 2015, 6, 3578–3591. (30) Tutusaus, O.; Mohtadi, R.; Arthur, T. S.; Mizuno, F.; Nelson, E. G.; Sevryugina, Y. V. An Efficient Halogen-Free Electrolyte for Use in Rechargeable Magnesium Batteries. Angew. Chem. 2015, 127, 8011–8015. (31) Whittingham, M. S.; Siu, C.; Ding, J. Can Multielectron Intercalation Reactions Be the Basis of Next Generation Batteries? Acc. Chem. Res. 2018, 51, 258–264. (32) James, A. C. W. P.; Ellis, B.; Goodenough, J. B. Lithium Insertion into the Normal Thiospinel CuZr2S4 and the Defect Thiospinel Cu0.05Zr2S4. Solid State Ion. 1988, 27, 45– 55. (33) Talaie, E.; Bonnick, P.; Sun, X.; Pang, Q.; Liang, X.; Nazar, L. F. Methods and Protocols for Electrochemical Energy Storage Materials Research. Chem. Mater. 2017, 29, 90–105. (34) Mizrahi, O.; Amir, N.; Pollak, E.; Chusid, O.; Marks, V.; Gottlieb, H.; Larush, L.; Zinigrad, E.; Aurbach, D. Electrolyte Solutions with a Wide Electrochemical Window for Rechargeable Magnesium Batteries. J. Electrochem. Soc. 2008, 155, A103–A109. 24 ACS Paragon Plus Environment

Page 25 of 25 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


(35) Duffort, V.; Sun, X.; Nazar, L. F. Screening for Positive Electrodes for Magnesium Batteries: A Protocol for Studies at Elevated Temperatures. Chem. Commun. Camb. Engl. 2016, 52, 12458–12461. (36) Bartehl, J. Dr. Probe - High-resolution (S)TEM image simulation software http://www.erc.org/barthel/drprobe/ (accessed Mar 6, 2018). (37) Kraft, S.; Stümpel, J.; Becker, P.; Kuetgens, U. High Resolution X‐ray Absorption Spectroscopy with Absolute Energy Calibration for the Determination of Absorption Edge Energies. Rev. Sci. Instrum. 1996, 67, 681–687. (38) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: Data Analysis for X-Ray Absorption Spectroscopy Using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537–541. (39) Rehr, J. J. Recent Developments in Multiple-Scattering Calculations of XAFS and XANES. Jpn. J. Appl. Phys. 1993, 32, 8. (40) Rietveld, H. M. A Profile Refinement Method for Nuclear and Magnetic Structures. J. Appl. Crystallogr. 1969, 2, 65–71. (41) Sinha, S.; Murphy, D. W. Lithium Intercalation in Cubic TiS2. Solid State Ion. 1986, 20, 81–84. (42) Emly, A.; Van der Ven, A. Mg Intercalation in Layered and Spinel Host Crystal Structures for Mg Batteries. Inorg. Chem. 2015, 54, 4394–4402. (43) Manceau, A.; Marcus, M. A.; Grangeon, S. Determination of Mn Valence States in Mixed-Valent Manganates by XANES Spectroscopy. Am. Mineral. 2012, 97, 816. (44) Kumar, P.; Nagarajan, R.; Sarangi, R. Quantitative X-Ray Absorption and Emission Spectroscopies: Electronic Structure Elucidation of Cu2S and CuS. J. Mater. Chem. C 2013, 1, 2448–2454. (45) Zhu, Y.; Wang, C. Galvanostatic Intermittent Titration Technique for PhaseTransformation Electrodes. J. Phys. Chem. C 2010, 114, 2830–2841. (46) Matsumoto, N.; Hagino, T.; Taniguchi, K.; Chikazawa, S.; Nagata, S. Electrical and Magnetic Properties of CuTi2S4 and CuZr2S4. Phys. B Condens. Matter 2000, 284–288, 1978–1979. (47) Rong, Z.; Malik, R.; Canepa, P.; Sai Gautam, G.; Liu, M.; Jain, A.; Persson, K.; Ceder, G. Materials Design Rules for Multivalent Ion Mobility in Intercalation Structures. Chem. Mater. 2015, 27, 6016–6021. (48) Liu, J.; Whitfield, P. S.; Saccomanno, M. R.; Bo, S.-H.; Hu, E.; Yu, X.; Bai, J.; Grey, C. P.; Yang, X.-Q.; Khalifah, P. G. In Situ Neutron Diffraction Studies of the Ion Exchange Synthesis Mechanism of Li2Mg2P3O9N: Evidence for a Hidden Phase Transition. J. Am. Chem. Soc. 2017, 139, 9192–9202. (49) Canepa, P.; Bo, S.-H.; Sai Gautam, G.; Key, B.; Richards, W. D.; Shi, T.; Tian, Y.; Wang, Y.; Li, J.; Ceder, G. High Magnesium Mobility in Ternary Spinel Chalcogenides. Nat. Commun. 2017, 8, 1759.


Insights into Mg2+ Intercalation in a Zero-Strain Material: Thiospinel

Recommend Documents