WO2013133906A2 - Lithium all-solid-state battery - Google Patents

Abstract

An all-solid-state lithium battery, thermo-electromechanical activation of Li₂S in sulfide based solid state electrolyte with transition metal sulfides, and electromechanical evolution of a bulk-type all-solid-state iron sulfur cathode, are disclosed. An example all-solid-state lithium battery includes a cathode having a transition metal sulfide mixed with elemental sulfur to increase electrical conductivity. In one example method of in-situ electomechanical synthesis of Pyrite (FeS₂) from Sulfide (FeS) and elemental sulfur (S) precursors for operation of a solid-state lithium battery, FeS + S composite electrodes are cycled at moderately elevated temperatures.

Description

LITHIUM ALL-SOLID-STATE BATTERY

GOVERNMENT RIGHTS

[0001] This invention was made with government support under grant number FA8650-08-1-7839 awarded by the Air Force Research Laboratory. The government has certain rights in the invention,

PRIORITY CLAIM

[0002] This application claims the benefit of U.S. Provisional Application No. 61/585,098 filed January 10, 2012 and titled "Lithium alS-solid-staie battery," and U.S. Provisional Application No, 61/590,494 filed January 25, 2012 and titled "Thermo-electrochemical activation of solid state electrolyte," both of Yersak, et ai., and each incorporated by reference as though fui!y set forth herein.

BACKGROUND

[00033 Use of traditional rechargeable lithium-ion batteries in consumer products is a concern due to safety issues. Such safety issues include, but are not limited to, solvent leakage and fiammabsiity of commercial grade liquid electrolytes. Traditional solid-state batteries (composed of glass ceramic rather than liquid electrolytes) do not pose such safety risks and offer high reliability for end-users. However, traditional solid-state batteries suffer from poor rate capability, iow ionic conductivity, interfacial instability, and low loading of active materials.

00043 Molten salt solid-state batteries require high operating temperatures (e.g., 400°C and higher), so research was abandoned in search of room temperature lithium-ion and lithium-polymer technologies. Iron disulfide has been successfully commercialized in high energy density primary cells. Unfortunately, sulfide conversion chemistries are irreversible at ambient to moderatel elevated temperatures, making these unusable for rechargeable batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 is a high-level depiction of an exampie iithium or lithium-ion all-solid-state battery structure.

[0006] Figure 2 shows for an exampie battery; (a) a field emission scanning electron microscope (FESE ) micrograph of synthetic FeS that confirms cubic structure with 2-3μηι cubes, and (ø} X-ray diffraction of synthetic pyrite.

[0007] Figure 3 shows as an example of FeSa cycled at ambient temperature {about 3Q°C) and moderately elevated temperature {60^eC) in a liquid coin cell and in an ail-solid-state configuration for: (a) solid-state at 3G°C, (b) solid-state at 60^eC, (c) liquid coin cell at 30X, (d) a !iquid coin eel! at 60^eC, (e) capacity retention comparison of cells cycled at 30 ^*C, and (f) capacity retention comparison of cells cycled at 60°C.

[00083 Figure 4 shows an exampie DFT simulation, where (a) is a so-called "baSi-and -stick" representation of Li_xFeS2 along a charging cycle, and (b) illustrates the average Fe-Fe distance {d^e-Fe) at each state in comparison with the Fe bulk vaiue,

[0009] Figure 5 shows a) Coulometric titration results for a solid-state ceil, b) shows dQ/dV of a solid-state cell, and c) shows deconvoluiion of the dQ/dV peaks.

[00103 Figure 6 shows electrode material from the soiid-siate ceil cycled at 60°C recovered after the twentieth charge for transmission electron microscopy (TEM), where (a) is a bright field TE image, and (b) is a high resolution (HR-) TEM image,

[00113 Figures 7-9 are plots showing actual data obtained for Exampie 2, [0012J Figure 10 is a plot illustrating that activation of U₂S exhibits ionic and electronic conductivity, as weii as added thermal energy.

[0013J Figure 11 is a plot showing cells cycled at C/5 and C/5 charge and discharge rates.

[00143 Figures 12{aHd) are piots of an example rate study.

[001 SJ Figures 13(a)-{b) are plots showing cycling data for a lithium metal cell.

[00163 Figure 14 is an x-ray diffraction (XRD) spectra. [0017] Figure 15 shows (a) x-ray diffraction (XRD) spectra and (b) a TE image.

f0018| Figure 16 shows (a) cyclic stability of a FeS + S/Li battery, and (b)-(e) are voltage profiles of the same battery.

|0019| Figure 17 shows (a) dQ/dV profiles for bulk-type all-solid-state Li metal batteries, and (b)-(e) dQ/dV profiles of an example battery.

[0020] Figure 18 shows (a) XRD of FeS before and after milling, (b) an FESEM micrograph of FeS, and (c) an FESEfvl micrograph of Fe^S after mechanical milling.

|0021| Figure 19 shows (a) voltage profiles for two initially overcharged

FeS/Liln batteries, and (b) XRD measurements normalized with the (100) reflection of the beta-Be sample window.

OO223 Figure 20 shows voltage profiles of a FeS/Liln battery.

[00233 Figure 21 shows (a) specific discharge capacity of a battery as a function of applied current, and (b) voltage profiles of the same battery as a function of applied current.

DETAILED DESCRIPTION

[0024] Many advanced battery technologies are vying to be the successor of today's conventional Li-ion batteries. A strong argument can be made that bulk- type ail-solid-state lithium batteries (ASSLB) hold a competitive edge in this technological race because they are inherently safe, have excellent shelf life, perform stab!y at high temperatures, and enable the reversibility of high capacity conversion battery materials like Fe¾. However, the energy density of high power ASSLBs must be improved. The success of the ASSLB architecture can be realized with energy dense ail-solid-state composite cathodes.

[00253 Examples of an ambient temperature, reversible solid-state cathode are disclosed. An exampie implementation is in a lithium (Li) metal configuration. The battery may be constructed using a sulfide glass-ceramic solid electrolyte, and is implemented in an all-solid-state celt architecture. In an example, the battery may be characterized as *S_y + zS where = Fe. Co, Mo, y = 0,1 , 2, 3 and z = 0, Ya, 1 , and so forth. This nomenclature is intended to include at least the following systems: FeS_2i FeS, FeS + S, and may also include other suitable substitutes as wi!l be understood by those having ordinary skill in the art after becoming familiar with the teachings herein, it is noted that the electrochemical synthesis of metal nano-particles maintains the electrochemical activity of U₂S. Accordingly, the battery addresses issues previously associated with rapid capacity fade at ambient temperature. The eiectrochemicaiSy driven synthesis of orthorhombic-FeS2 (marcasite) can be at least partially achieved at ambient temperatures,

[0026] The design of an ambient temperature transition metal plus sulfide batteries is based at least in part on management of electro-active species formed upon full charge (2.5V versus Li+/U) and full discharge (1.0V versus Lf /Li). Two example species are elemental iron (Fe°) and polysulfides {S_n ^2~). To reduce or altogether prevent diffusion and agglomeration of Fe° nanoparticles in conventional cells, a variety of polymer electrolytes have been employed with limited success.

[0027] A similar approach may be applied to the confinement of intermediate polysulfides in conventional Li-S batteries. Example methods for addressing intermediate poSysu!fsde dissolution and Li₂S irreversibility include polysulfide adsorption on high surface area C -3 nano-porous carbon electrodes, polymer electrolytes, and poiyacrylonit ile-surfur composites.

[0028] Another approach is to limit the upper and/or lower voltage bounds of the FeSs cells, for example, to about 2,2V and 1.3V respectively. The formation of Fe° and Sr,^2" is inhibited by avoiding full discharge and charge.. However, limiting the cell voltage range diminishes achievable energy density and subjects the cells to the risk of over-charge or over-discharge.

[0029] The basic nature of an all-solid-state cell architecture allows for the confinement of electro-active species. For example, F $₂ and LI₂FeS2 can both be utilized reversibly as an ail-solid-state anode. An all-solid-state architecture reduces or altogether prevents Fe^tJ dissolution and agglomeration. Sulfide- based, glass-ceramic solid electrolytes and other materials that are stable at elevated temperatures demonstrate higher conductivities at ambient temperatures. Accordingly, lithium metal anodes can be safely used with solid electrolytes because ceil failure does not precipitate thermal runaway. A lithium metal electrode has a theoretical capacity of about 3876 mAh g^"\ is non~ polarizable and has a low operating voltage that increases achievable ceil energy density. An ail-solid-state architecture not only enables the safe use of a lithium metal anode, but also enables the reversible full utilization the cathode material.

[0030] Further examples herein disclose in-situ electrochemical formation of high capacity conversion battery materials like FeS₂ and reversible utilization of a giass or other stable electrolyte for higher overall electrode energy density, in an example, the best performing composite electrode compositions are composed of no more than about 25% S or LkS by weight. Suifur's high theoretical specific capacity of about 1872 mAh g^"1 offsets poor active material mass loading so that high overall electrode energy densities can be achieved for ASSLBs. To increase the overall energy density of a composite electrode without changing the composition, the techniques described herein reverstb!y electrochemically utilize the glass-ceramic electrolyte. For example, by incorporating pm-Cu powder acetylene black, and 80Li₂S:20P₂S5 glass-ceramic electrolyte into a composite electrode, the the L S component of glass-ceramic electrolyte electrochemical can be utilized. For more effective electrolyte activation and better electrode reversibility, the active metal can be provided by in-situ electrochemical reduction.

[0031] The examples described herein may be further optimized by utilizing a mechanochemicaliy prepared active material nano-composiie of high capacity conversion battery materials like FeS and S. This material provides an alternative to the expensive solvothermal!y synthesized cubic-FeSs (pyrite) based cathode. The precursors {e.g., FeS and S) are comparatively inexpensive and can be obtained in much higher purities than natural pyrite. The mechanical milting process also provides material much mor readily than the soivothermal method.

[0032] During testing, the rapidly increasing specific capacity of the nano- composite electrode (e.g., FeS + S) quickly exceeded its theoretical capacity by about 94% in testing. The excess capacity is a result of a dramatic utilization of the glass electrolyte in the composite electrode without a degradation of ceil performance. At its maximum, an example electrode exhibited an energy density of about 1040 Wh kg^"1 which is the highest energy density achieved for a bulk- type ail-solid-state electrode. With extended cycling, the electrochemistry of the composite electrode (e.g., FeS + S) evolves a redox chemistry based primarily on that of only sulfur. The results show that eiectrochemieally structured interfaces between conversion active materials and the glass electrolyte can be utilized to increase energ density of ASSLBs, while maintaining good rate performance.

|0033| ^ is noted that examples are described herein with respect to specific materials and process parameters for purposes of illustration only, and are not intended to be limiting. Other examples will be understood by those having ordinary skill in the art after becoming familiar with the teachings herein, and are also intended to be included within the scope of the claims.

[0034] Before continuing, it is noted that as used herein, the terms "includes" and including" mean, but is not limited to, Includes" or "including" and Includes at least" or "including at least." The term "based on" means "based on" and "based at least in part on."

Uthium Aii^oild-State Battery

[0035] Figure 1 shows a high-level depiction of a lithium a!l-soi id -state battery structure 100. The battery structure 100 is shown including a cathode 101, solid state electrolyte 102, and an anode 103. The arrows are illustrative of charge and recharge cycles. The anode may include a lithium metal, graphite, silicon, and/or other active materials that transfer electrons during charge/discharge cycles. In an example, the lithium reservoir may be a stabilized Li metal powder.

[0036] The cathode 101 is shown in more detail in the exploded view 101', wherein the white circles labeled 110 represent a transition metal sulfide (e.g., FeS₂ or an mixture of FeS plus S). The gray circles labeled 111 represent solid state electrolyte (SSE) particles. The SSE particles promote ionic conductive pathways in and out of the cathode 101. The black circles iabeled 112 represent a conducting additive, such as acetylene black.

[00373 in example, the transition metal sulfide (e.g., FeS plus S) may be mechanically mixed (e.g., using ball milling or other mechanical processes), and are not chemically combined. The chemistry of the cathode approximates FeS₂ on the first cycle. After 10 or more charge cycles, the cathode exhibits a behavior that is similar to that of the first cycle.

[0038] Although the lithium al!-soiid-state battery structure 100 may be implemented using any suitable transition metal sulfide pius sulfide combination, the following discusses a specific battery structure based upon synthetic iron disulfide,

00393 Synthetically prepared FeS₂ is characterized with Field emission scanning electron microscopy (FESEM) and x-ray analysis. Figure 2 shows (a) a FESEM micrograph of synthetic FeSa that confirms cubic structure with wide 2-3pm cubes, and (b) X-ray diffraction of synthetic pyrite. The FESEfvl micrograph shown in Figure 2(a) reveals cubic FeS₂ particles with 3pm wide faces. The X-ray analysis of synthetically prepared FeS2 shown in Figure 2(b) shows diffraction peaks that match well with indexed peaks.

[0040] Synthetic FeSs was tested in both an all-solid-state and liquid cell configuration. To achieve full utilization of FeS_¾ the ceils are cycied to full discharge (1.0V) and full charge (3.0V), The results of cycling at ambient and moderate temperatures are shown in Figure 3. For purposes of illustration, Figure 3 shows FeSa cycied at ambient temperature (about 30°C) and moderately elevated temperature (about 60°C) in a liquid coin eel! and in an all- solid-state configuration for; (a) solid-state at 30°C_! (b) solid-state at 80°C, (c) liquid coin ceil at 30°C₍ (d) a liquid coin celt at 60°C, (e) capacity retention comparison of ceils cycied at 30°C, and (f) capacity retention comparison of ceils cycied at 60 °C. Al! ceils except for the 30°C solid-state ceil were cycied at a current of 144μΑ which corresponds to a rate of C/10 for charge and discharge. The 30° C solid-state cell was cycled at rate of C/10 for the first cycie and C/20 (72μΑ) for all subsequent cycles. [0041] Both solid-state cells are observed to have good capacity retention and a high degree of FeSa utilization. The gradual increase in capacity with cycling is observed to be a result of better FeS₂ utilization. This conciusion is supported by differential capacity (dQ dV) analysis. By the twentieth cycle, the ceil tested at 3GX exhibits a discharge capacity of nearly 750 mAh while the cell tested at SOX exhibits a theoretical discharge capacity of about 894 mAh g^" \ !t is likely that the temperature dependence of solid electrolyte's conductivity contributes to the full FeS₂ utilization at 60X, but not at 3GX. At 60X the conductivity of the 77.5Li2S-22.5P₂S5 solid electrolyte increases to 4.4x1 tT³ D^"1 cm^"1 from 9.17x10^'4 Ω^"1 cm^"1 at 30X. In addition, a higher operating temperature increases the Li⁺ diffusivify in pyrite particles. More efficient Li* insertion into cubtc-FeS2 is also likely to result in better FeSa utilization.

[0042] in liquid cells, the discharge capacity rapidly fades upon cycling. By the twentieth cycle, the liquid cell tested at 3QX exhibits a discharge capacity of only 190 mAh g^*1 while the cell tested at 60 X exhibits no discharge capacity. Decomposition processes are accelerated at 60X leading to such a fast rate of capacity fade that negligible capacity is observed after the second cycle. On the other hand, we have just shown that cycling a solid-state FeS₂ cell at 6QX only improves its performance.. At 6QX, it is possible to achieve a reversible, four electron utilization of Fe$₂. It is noted that many traction battery packs are designed to operate at temperatures near about 60X. The superior performance of all-solid-state batteries described herein at higher temperatures may reduce the need for extensive thermal management systems.

0043] ossbauer spectroscopy and near-edge X-ray absorption spectroscopy (XA ES) show that the products of FeSa reduction are elemental iron (Fe°) and U₂S. The initial discharge of FeS₂ proceeds in two steps;

(1) FeS₂ + 2Lf + 2e^" <→ Li₂FeS₂

(2) Li₂FeS₂ +2U⁺ + 2e^* <→ 2Li₂S + Fe⁰

[0044] Each reaction can occur at one voltage or two, depending at least in part on the kinetics of the system. 3

[0045] Discharge profiles were observed having one p!ateau when the ceil is cycled at 30°C, as can be seen in Figures 3(a) and (c), and two plateaus at 80°C, as can be seen in Figures 3(b) and (d). At Iower iemperatures the reactions are limited by the Sow diffusivity of Li* into the micron sized FeS₂ particles. At 30X, reactions corresponding to equations (1 ) and (2) shown above proceed simultaneously at 1.5V versus It ill At moderately elevated iemperatures, equations (1 ) and (2) can proceed at 1 .7 and 1 ,5V respectively due to the higher diffusivity of If into the micron sized FeS₂ particles. The shoulder at 1 .3V in the ambient temperature liquid cell, as can be seen in Figure 3(c), may be attributed to an irreversible side reaction of FeS₂ intermediaries with the organic electrolyte. It is noted that the initial discharge profile of the FeS2 cells is much different than subsequent discharge profiles. The variation in discharge profiles may be due to the improved reaction kinetics of charge products compared to that of cubic-MmFeS₂. Better kinetics may be due to changes in F S2 particle morphology and the electrochemical formation of a different phase of stoichiometric of FeS₂ like orthorhombtc-FeSa (marcasite).

[0046} Superior performance observed in the solid state is believed to be due to the confinement of electro-active species. The confinement of Fe° by solid electrolyte partially explains the better performance. Fe° takes the form of super-paramagnetic atoms or small aggregates of atoms of about 3.6nm in diameter. Nano-particies of Fe° have a high reactivity which is related to the nano-particle^'s large surface area. Should Fe° particles agglomerate into larger particles with smaller overall surface area, then these particles will have a lower reactivity. Without meaning to be limited by the theory, it may be the high reactivity of the Fe° nano-particles that maintains the electro-activity of U₂S. Of course, other theories are aiso possible,

00473 But Fe° is susceptible to continuous agglomeration upon cycling. Agglomeration of Fe° results in the isolation of Li₂S species and the observed capacity fade when ceils are discharged to low voltages. An all-solid-state architecture can reduce or altogether prevent the agglomeration of Fe° nana- particles. The atomic proximity of Fe° nanoparticles with L S maintains the eieciro-aciivity of LisS without the excessive amount of conductive additive needed in S/Li₂S based batteries.

{0048] An a!i-soiid-state architecture is also successful at confining polysutfides S„²~ formed when the electro-active species present at full charge are reduced. At ambient to moderate temperatures, FeS₂ is not regenerated by the four electron oxidation of Fe° and LbS. Bui the same is not true for molten salt FeS₂ cells, which operate reversib!y at temperatures in excess of 400°C.

[00493 Generally, subsequent charge and discharge cycles may proceed according to the following reactions:

(3) Fe° + Li₂S «··► Li₂FeS₂ + 2Li⁺ + 2e^"

(4) LiaFeSj *-* Li₂._xFeS₂ + xLf + xe^" (0.5 < x < 0.8}

(5) Li₂.xFeS₂ < - FeS_y + (2-y)S + {2-x)U⁺ + (2-x)e^"

However equation {5} may be better represented by equation (8) based on the results to be outlined below:

(6) L -xFeSs <--> 0.8ortho-FeS₂ + 0.2FeS_6t7 + 0.175S + (2-x)lf + (2-x)e~

[00S0] The direct reduction of sulfur by Lf upon subsequent discharge therefore introduces intermediate po!ysulfides (S„^2") into the system, in a liquid cell, polysutfides dissolve into the electrolyte and participate in a parasitic "shuttle" mechanism which causes rapid capacity fade and self-discharge. The "shuttle" mechanism is the primary degradation process occurring in sulfur- based cells. Polysutfides cannot dissolve into the solid electrolyte. Therefore, the confinement of polysutfides in an all-solid-state ceil inhibits the "shuttle" mechanism,

|0G51J Charge products at about 30~60°C are likely a multi-phase mixture of nano-particles of orfhorhombic-FeS₂, non-stoichtometric FeS_y phases like pyrrhotite and elemental sulfur, in any case, the electrochemicalSy active products resulting from sequential charge cycles simulate the FeS₂ chemistry as well as provide electrical conductivity within the electrode thus reducing the amount of conductive additive required. This conclusion is supported by the results of a DFT simulation shown in Figures 4-6.

[OOS23 Figure 4 shows an example DFT simulation of iithiated UxFeS₂ indicating material amorphization and Fe agglomeration for x - 4, where (a) is a so-called "ball-and-siick" representation of Li_xFeS₂ along a charging cycle from x-4 to x-0, and (b) shows the average Fe-Fe distance (dF_S._f¾) at each state in comparison with the Fe bulk value.

[0053] Figure 5 shows: a) Coulomethc titration results for a solid-staie ceil titrated at 60°C compared with ihe first second, and tenth discharge profiles for a solid-state ceil cycled at BO'C, b) shows dQ/dV of a solid-state ceil cycled at 30°C, and c) shows deconvolution of the dQ/dV peaks at 2.1 and 2.2V with fitted peaks and residual.

[0054] Figure 6 shows electrode material from the solid-state ceil cycled at 6CTC recovered after the twentieth charge for transmission electron microscope (TEM) analysis, where: (a) is a bright field TE image of the twentieth cycle sample with darker areas corresponding to nano-crystalline orthorhombic-FeSs and lighter areas corresponding to an amorphous region composed of FeS_y and elemental sulfur, and (b) is a high resolution (H -) TEivl of the twentieth cycle sample. FFT analysis matches with orthorhombic-FeSa along the [-110] zone axis.

[0055] Contrary to previous assumptions, subsequent discharges largely follow the same initial reaction path, instead, the difference between the initial and subsequent discharge profiles is likely to changes in particle morphology and the formation of the more open orthorhombic-FeS2 (marcasite). Thus, equation (6) more accurately describes the chemistry of subsequent cycles.

[0056] A study used coulomethc titration to indicate that cubic-FeS2 is not produced electrochemicaliy. However, the time needed for the FeS₂ electrode to reach equilibrium is much longer than the 24 hours allowed in that study. When an FeS₂ eel! is allowed up to about 144 hours to establish equilibrium during initial discharge, the open circuit voltage (OCV) of the cell approaches the voltage of a subsequent discharge at th appropriate reaction coordinate (x) as shown in Figure 5{a). This result indicates that the difference between the initial discharge profile and subsequent discharge profiles can be explained by kinetics. The difference is not a result of a different reaction pathway. The diffusion of Li⁺ into the inner core of 2-3 pm pyrite cubes severely limits the reaction rate of FeS₂ reduction, if electrochemical!y produced e$2 particles are nano-crystaliine, then the greatly increased interfacia! surface area facilitates a fast reaction rate despite poor Li^"" diffusivity. The diffusivity of Li^÷ may also be improved by regenerating a phase other than cubic-FeS₂. For example, orthorhombic-FeSa has a more open structure than cubtc-FeSj. The formation of orthorhombfc-FeS2 instead of eubic-FeSa may result in faster Li* diffusion, thus further increasing the reduction reaction kinetics.

£00573 High resolution transmission electron microscopy (HR-TEM) can be used to support this understanding through direct observation of orthor ombic- FeS₂ nano-particSes upon charge. In an example, electrode materia! was recovered from the solid-state cell cycled at about 80°C upon completion of the twentieth charge as can be seen inFigure 3(b). This ceil exhibits full utilization of Fe$₂ so it is unlikely that a significant mass of electrochemica!ly inactive synthetic cubic-FeS^ remains in the ceil by the twentieth charge.

|QG58J Figure 6(a) shows a bright field (BF) TBM image of the 20th cycled charged FeSs solid-state electrode. This image depicts nano-crystaliine domains (darker) of 10Q-200nm in diameter encased by an amorphous material (lighter). Fast Fourier transform (FFT) analyses of HR-TEM images matches well wit orthorhombic-FeS2 a!ong the [- 10] zone axis as can be seen in Figure 8(b).

[00591 High resolution TEM imaging indicates a large amount of amorphous materia! encasing the crystalline F Sa domains. To explore this issue further, the differential capacity of the ail-solid-state ceil cycled at 30°C was examined, as can be seen in Figures 5(b) and (c). The peaks shown in Figure 5(b) correspond to the oxidation of Li₂S, and the reduction of S in an al!-solid-state sulfur cell. The purple peaks correspond to reaction plateaus observed in the so!id-state FeS₂ cells cycled at 30°C. [0060] When the solid-state FeS₂ ceil is charged, no peaks are observed corresponding to the oxidation of U₂S. However, upon discharge a peak is observed at 2.2V, which corresponds to the direct reduction of sulfur to Li₂S. This result indicates that discharge and charge of a FeS₂ ail-solid-state cell at higher voltages at least somewhat follows equation (5), above. It indicates that suifur is eSectrochemically produced by the disproportionation of Li₂FeS₂. For this reason, the amorphous region is likely a mixture of elemental suifur and non-stoichiometric FeS_y.

[0061] To quantify the amount of elemental sulfur produced upon charging, all elemental sulfur is said to be directly reduced to Li₂S at 2.2V. The solid-state cell cycled at about 30°C exhibited a discharge capacity of about 737 mAh g^' upon the ninth discharge. If the peaks at 2.1 and 2.2 V correspond to the reaction of charge products with two electrons, then integrating the dQ/dV curve between about 1.6 and 2.5 V yields a capacity 368 mAh g^{" "} for this cell. When these two peaks are de-convoluted and fitted with a Voigt profile, the calculated total area gives a capacity of about 342.2 mAh g^"1 , as can be seen in Figure 5(c). This value matches well with the expected capacity of about 368 mAh g^"1.

[O0S2J The peak at about 2.2 V has an area of about 57.14 mAh g^"\ while the peak at about 2.1V has an area of about 285.79 mAh g^"1. if (2-y)S is directly reduced to Li₂S, then the remaining capacity may be attributed to FeS_y. The value of y can be determined to be about 0.085. If subsequent discharges follow equation (5), then the chemical formula of FeS_y is about FeS^- if FeS_y primarily takes the form of Fe₇Se (pyrrhoiite), then the chemistry of subsequent cycles likely follows equation (8).

[00633 The charge products are likely a multiple phase mixture of nano- crystailine orthorhombic-FeS₂, sulfur deficient phases of FeS_y and elemental sulfur. Accordingly, the charge products are believed to be nano-crystailine orthorhomb!C-FeS₂ encased in amorphous suifur deficient FeS_y and sulfur (see Figure 6(a)).

[00643 Coulometric titration indicates that the initial discharge is kineiical!y limited, and subsequent discharges follow a similar reaction path (see Figure 5(a)). Nano-crystailine orthorhombic-FeSs enables faster reaction rates. [0065] Sulfur reduction was observed at about 2.2V upon charge, but not Li₂S oxidation upon charge, as can be seen in Figure 5(b). Fe$_y phases and elemental sulfur may explain the amorphous domains observed in the high resolution JEM images. The presence of sulfur deficient FeS_y phases and the observation of orthorhombic-FeS₂ is consistent, Orthorhombic-FeSa exhibits very weak temperature independent paramagnetism. For this reason, it is likely that ⁵⁷Fe Mossbauer spectroscopy used in previous studies was not capable of distinguishing orthorhombic-FeS₂ from other strong temperature dependent paramagnetic phases like FeS_y and cubic- FeS₂. By way of example, Fe Sg is ferrimagnetic while FeS is paramagnetic. It is noted that ferrimagnetism is a much stronger effect.

[0066] The results of the DFT analysis shown in Figure 4 indicate the formation of highly reactive atomic particles of Fe° The fully-discharged, amorphous-like L^FeSa model (Figure 4(a)) shows nanoscale separation of a Fe° nanocluster from U2S. The average Fe-Fe interatomic distance {d_f¾.F_e) at full discharge, x ~ 4, is much shorter than that of Fe in the bulk (Figure 4(b)). A shorter indicates that Fe° should be very catalytically active. Results also indicate the presence of elemental sulfur as a charge product. At x = 0, our atomic model depicts some degree of FeS₂ crystallization with a rather open structure (Figure 4a). The x = 0 model also depicts the presence of a S2 dimer. It is the presenc of elemental suifur that inhibits the full crystallization of FeS_z in our simulation. It is iikely that the observed FeSa-y nano-c!usters could crystalize into orthorhombic~FeS2 rather than cubio-FeS2 because of the former's lower density.

00S7| Before continuing, it shouid be noted that the description of example ambient temperature, reversible metallic lithium iron sulfide (FeSa) solid-state batteries given above and further described below with reference to specific Examples is provided for purposes of illustration, and is not intended to be limiting. Other devices and/or device configurations using these and/or other materials may be utilized as will be readily apparent to one having ordinary skill in the art after becoming familiar with the teachings herein. Example 1

[0068] In this example, the batteries discussed above were made for laboratory scale analysis using commercially available polyvinylpyrrolidone (PVR 0,000) and FeCI₂ ^*4H₂0 (>99%) obtained from Sigma Aidricb, ethylene glycol (99%) obtained from altinckrodt Baker Inc., and sulfur obtained from Fischer Scientific. HPLC grade water, analytical grade MaQH, and absolute eihanol were used: without further purification. The FeS₂ synthetic methodology used solvothermal reaction conditions. Dielectric heating for the reaction was provided with a microwave reactor. Microwave heating was selected because of its high reproducibility and the ability for automation, making this methodology amenable to high throughput syntheses.

[0069] For the reaction 17 mL of ethylene glycol was added to 800 mg of PVP in a 35 ml microwave flask with a magnetic stirbar. Then 127 mg FeCb HsO (0,64 mmol) was introduced, 8 mL of 1 M NaOH was then added, resulting in a dark green color. Finally, 180 mg of sulfur was added. This solution was stirred for 20 minutes while changing color from green to black. Some sulfur remained undissolved during this process. The reaction flask was then capped (70% full) and introduced to the microwave.

[0070] The microwave used for this example was a Discover SP (GEM Inc.). The sample was irradiated with 75 W of power until reaching 190°C, as measured by an infrared detector. The heating took about 7 minutes, and was held at this temperature for 12 hours. Approximately 890 kPa of autogenous pressure was generated. After the reaction was finished, the product was cooled by compressed air,

[0071] The resulting silver colored precipitate was separated by centrifugation and washed three times by sonication in ethanoi. The precipitate was then stored in ethanoi and vacuum dried overnight at 50° C for battery utilization. Synthetic FeSa was characterized by Cu-Κα x-ra diffraction (XRD) measurement, FESEM microscopy fJEOL JS -7401F), and Raman spectroscopy (Jasco NRS-3100).

|0G72J Cell fabrication and cell testing for this example was carried out under an inert argon gas environment. The all-solid-state cells used in this study were 18

based upon the 77.5Li2S-22.5P₂Ss binary soiid-state electrolyte. The composite positive electrode had a 10:20:2 weight ratio mixture of synthetically prepared FeS₂. 77.5ϋ_Ξ$-22.5Ρ₂$5, and carbon biack {Timcal Super C65), respectively. The composite positive electrode was mixed using an agate mortar and pestle. Stabilized lithium metal powder {SLfVlP} was used as the negative electrode (FMC Lithium Corp.). The construction of solid-state cells utilized a tiianium- potyarySetheretherketone (PEEK) test cell die. 2Q0mg of solid electrolyte powder was pressed at i metric ton in the PEEK ceil die. 5mg of composite positive electrode and the stabilized lithium metal powder were then attached to opposite sides of the solid electrolyte layer by pressing at 5 metric tons.

[00733 Liquid cells were fabricated by spreading an electrode slurry with a 8 2:2 weight ratio of synthetic FeSa, polyvinyifiuorine (PVDF) binder (Alfa Aesar) and acetylene black (Alfa-Aesar, 50% compressed) respectively. PVDF binder was first dissolved into -methyl-2-Pyrrolidone (N P) (Alfa-Aesar) solvent. FeS₂ and acetylene black were then stirred into the PVDF binder, A 50 pm thick layer of slurry was spread on onto aluminum foil (ESP! eta!s, 0.001" thick) and dried at SOX in a single waii gravity convection oven (Blue M) for 5 hours. To ensure good electronic contact, the electrode sheet was then calendared with a Durston roiling mill to 75% of the total thickness. 9/18" diameter electrodes were punched and heat treated at 2G0^SC in an Argon environment overnight. Fe$2 electrodes were then assembled into coin ceils with a lithium foi! negative eiectrode (Alfa-Aesar, 0.25mm thick) and 1 M LiPF₄ electrolyte,

|00743 Cells were cycied ga!vanostaticaily using an Arbin BT2100 battery tester at room temperature (30^*C) and elevated temperature (60^*C). Declared C -rates were based upon FeSs's theoretical capacity of 894 mAh g^'\ Reaction equilibrium was studied by use of the galvanostatic intermittent titration technique (GITT),

E amine 2

[00753 Figures 7-9 are plots showing actual data obtained for Example 2. Specifically, Figure 7 shows charge/discharge profiles for FeS₂ in a solid state lithium battery, made from microwave synthesis. Figure S shows charge/discharge profiles for in-situ formation of FeS≥ upon cycling of FeS+S combined by planetary bail milling (cycled at C/10). The top plot in Figure 9 shows the performance of FeS, with significantly lowered capaciiy and a Sack of voltage plateaus correlating to Eq. 7 and 8 below. Upon discharge at room temperature, a slightly Sower voltage is observed due to interna! resistance in the cell, and similarly for charging with a slightly higher voltage. A significant decrease in reversibility is aiso observed, with first cycle couiombic efficiency of both room temperature and elevated temperature FeS cells beSo 75%. The bottom plot in Figure 8 shows a first cycle of FeS at (a) room temperature and (b) elevated temperature.

[00763 'n this example, solid state batteries utilizing highly conducting sulfide based solid electrolytes were used for reversible cycling of FeSg electrodes in both room temperature (25⁵C) and elevated temperature environments (60^*C) as shown in Figure 7. Furthermore, it was shown that there is no capacity loss for increasing rates up to C/8 for elevated temperature testing. The first cycie shows the reaction formation of U₂S and Fe down to 1 .0V. Reversible cycling is allowed by the highly reactive iron allowing successful de-Lithiation of Li₂S, and subsequent formation of various Li-Fe-S compounds according to the following equations which are visible in the charging profile.

3Li + 2FeS₂→ Li₃Fe₂$₄ (2,1!/) {Eq. 7} fci₃Fe₂5₄→ 2Li₂FeS₂ (1.9V) {Eq. 8}

Li₂FeS₂ + 2Li→ Fe + 2Li_zS {Eq. 9}

[00773 After the first cycle, well defined voltage plateaus exist, showing the successful formation of FeS₂ (from discharging} upon charging, and subsequent formation of specific reversible Li-Fe-S phases upon discharge.

[0078] FeS₂ was formed in-situ during the first cycle (and also occurs over the course of many cycles), utilizing stoichiometric combinations of other materials. FeS and S were mixed either by mortar and pestle grinding, or by ball milting to produce an active material that is simply the addition of both, without formation of FeS_a.

[0079] Upon the first discharge, plateaus corresponding to FeS and S were present, and resulted in a similar capacity of FeS₂ made by microwave synthesis. By comparison (Figure 8), it can be seen that there is a difference between discharge profiles for Fe$2 and FeS+S electrodes. However, these both resulted in the same highly reactive iron species that allows the reversible cycling of U₂S. Voltage plateaus corresponding to the similar peaks of FeS₂ occurred for FeS+S electrodes (Figure 8) during charging. Subsequent discharging (2nd cycle) shows that these exhibited high reversibility and reaction plateaus for FeSs, resulting from the FeS/S electrodes. Based on this understanding, it is further noted that another potential system includes electrochemical formation of FeSs from elemental iron and sulfur being ball milted or mixed by hand.

ThermoelQctrochemical activation of soiid state electrolyte

[0080] ThermoeSectrochemica! activation of solid state electrolyte is also disclosed herein. To increase cell energy density, it may be desirable to thermally activate the solid state electrolyte (e.g., Liz ) in sulfide-based solid electrolytes, including but not limited to xLi₂S-(1G0 - x}P₂Ss. Initially charging a celt at an elevated temperature increases the energy density of the cell in one example by over 50%.

[00813 For purposes of illustration, two different composite electrodes were studied: an 80U2S-20P2Ss:aeety!ene black composite with a 20:1 weight ratio respectively and a TiS₂:80Li2S-20P₂Ss:acetylene biack composite with 10.20:1 weight ratio respectively. The cells in this example have an In metal negative eiectrode. However, the claims are not limited to these electrodes, as suitable substitutes may be used as will be understood by those having ordinary skill in the art after becoming familiar with the teachings herein. [0082] Figure 10 is a p!ot illustrating that the activation of Li₂S is based at least in part on both the sonic and electronic conductivity of Ti$₂, as well as added thermal energy (e.g., about 60°C). The specific charge capacities shown in Figure 10 are based on the total mass of TiS₂ solid electrolyte and acetylene black. Eve at elevated temperature the composite with only solid electrolyte shows no capacity. Likewise, the composite with TiS₂ also shows no capacity when charged at room temperature.

[00833 However, the composite electrode with TiS₂ exhibited a charge capacity of 13 mAh g^"1 when charged at an elevated temperature of about 60°C. This corresponds to a specific charge capacity of about 40mAh g^{" 1} based upon the TiS₂ mass. As these cells have a lithium-ion configuration, and T1S2 is already in the charged state, the only source of lithium in these ceils was U₂S. Under an applied current at elevated temperature, it is believed that otherwise inert Li₂S is activated by the highly ionic and electronic conductive character of TiS₂, Other transition metal sulfides have similar material properties as TiS₂, and may be useful i a similar Li₂S activation process.

[0084} Figure 1 shows results of thermo-electrochemtcal activation of Li₂S using nano-LiTiSs. Figure 11 is a piot showing ceils with a 10:20:1 wt% nano- LiTiSs: 80Li₂S-20P₂S₅:ac€tyiene black composite electrode cycled at C/5 and C/5 charge and discharge rates. The bottom data series (squares) shows the celt cycled at 30°C (without thermal-electrochemical activation), and the top data series (triangles) shows the ceSi initially charged at 60 0 before being removed to room temperature. The nano-LiTiS₂ was synthesized using mechano-chemical milling process involving LkU decomposition. The composite electrodes are a 10:20:1 weight ratio mixture of nano-LiTiS₂:80Li₂S- 20P₂S_s:acetyiene black.

100853 The cells in this example have an In metal negative electrode and are cycled at a rate of C/5 for both charge and discharge. However, the initial charge and discharge cycles are both conducted at a rate of C/10. Specific charge capacities presented are based upon the mass of LiTiS₂ initiaily present in the composite electrode. [0086] The first ceil cycled at room temperature exhibits a very stable capacity of about 230 mAh g^"! after about forty cycles. The second ceil undergoes an initial elevated temperature activation charge at about 60°C. !t is then moved to room temperature {about SCTC) for the first discharge and ail cycies thereafter. This ceii exhibits a 345 mAh g^"1 discbarge capacity after about the fortieth cycle. This represents about a 119 mAh g^" (or about a 53%) increase in capacity ove the theoretical capacity of LiTISs of 226mAh g^" .

[00873 The increase in capacity observed with nano-LiTiS₂ is much larger than the 40mAh g^"1 excess capacity achieved with the TiS ~80Li₂S:P2Ss- acetylene b!ack composite. The greater surface area of the nano-UTiS₂ particles is thought to more easily facilitate the activation of LiaS in the solid electrolyte. Supporting the previous finding, the rate performance of nano-LiTiSa composite electrodes is shown in Figure 12,

[00883 Figures 12(a)-{d) are plots of an illustrative rate study of a 10:20:1 wt% LiTiS₂: 801ί₂8-20Ρ₂3₅:3θ€ίγΙβηβ black composite electrode, where (a) and (b) are at 30 , and (c) and (d) are at 6QX. The plots in Figures 12(a)-(d) confirm the repeatability of results shown in Figure 11. The ceil cycled at an elevated temperature exhibits discharge capacities in excess of l eOmAhg^'1 greater than the theoretical capacity of 226 mAhg^'1 for LiTiS₂.

[00893 The cells in this example have a Li metal negative electrode to facilitate fast ion transfer, it is noted that the cell cycled at elevated temperature has a specific discharge capacity of nearly 390 mAh g^" at a rate of C/2 while the celt cycied room temperature only exhibits a capacity of 2 0 mAh g^'1 at C/2. Repeatedly charging at elevated temperature activates U₂S in the solid electrolyte, providing excess capacity.

[00903 Figures 13(a)- b) are plots showing cycling data for a lithium metal cell. The plots show cycling data for a lithium metal cell with a 10:20:1 wt% F Se/ ⁺ S:77.5Li2S-22.5P₂S5:acetyiene black composite cathode cycled at 60°C. At an elevated temperature, the cell gains capacit during the first 10 cycles. The increase in capacity is attributed to the S/Li₂S redox reaction.. The activation of excess sulfur is responsible for the evolution of a voltage plateau at about 2.2V. It can foe seen that the cell has a specific discharge capacity of about 807 mAh g^"' on the first discharge, but a specific discharge capacity of 1341 mAh g^" by the ninth cycle. This represents an activation of about 534 mAh g^"1. The activated capacity can be attributed to the evolution of the S/Li₂S reduction plateau at about 2,2V.

|0091J As stated above, the particular solid electrolyte system in this example is xU₂S~{100~ x)P₂Ss. However, this technique is applicable to any ₂S containing sulfide based electrolyte system is not limited to I^S-GeS PaSs or L S-SiS₂. These solid electrolytes are known as glass ceramics. During electrolyte synthesis (e.g., by melt-quenching or mecha no-chemical milling) Li₂S is incorporated into glass formers not limited to GeS₂, P2S5, and SiS₂. Super- ionicaily conducting crystalline phases can also be precipitated in a glassy matrix upon subsequent heat treatment. It is also possible, that these crystalline phases may decompose and result in some excess capacity.

[00923 Figure 14 are x-ray diffraction (X D) spectra of an example SOLisS- 2OP2S5 solid electrolyte, example 10:20:1 wt% U iS2:80Li2S-20P2S₅;acet !ene black composite electrodes, and indexed spectra for Lh ₀TiS₂ and Li₂S. The XRD spectra of the solid electrolyte in this illustration shows li₂S peaking at about 27 and about 31.2 degrees, indicating that there are some Li₂S domains still present in the solid electrolyte. It is believed that the additional capacity is a result of reacting excess U₂S in the solid electrolyte. It is expected thai the .5Li2S-22.5 2S5 solid electrolyte has less excess L S to participate in activation reactions.

OO933 There is evidence of Cu_yS formation and Li₂S decomposition upon charging at about 25^SC. The cells in this example tend to exhibit initial specific charge capacities of up to about 150 mAh g^"1. Composites without Cu tend to exhibit no capacity, indicating that Cu is a reacting species in the solid electrolyte.

[00943 The process described herein is analogous, but is also somewhat different. That is, Cu reacts to form C _yS, while TiS₂ remains chemically/structurally stable. TiS₂ is an intercalation electrode material, while CuyS is a conversion battery material. TiS₂ succeeds in electrochemical ly activating excess LJ2S because it is both highly ionicaily and electronically conductive. The process described herein is based onan initial charging at elevated temperature, as no excess capacity is observed at room temperature, fOOSS] The iron sulfide (FeS₂, FeS_>;, or FeS* + S) systems disclosed herein are more like the Cu_yS electrodes. During iithiation {reduction), FeS₂ is not chemicaSly/structuraS!y stab!e like TiSa. Instead, FeSz reacts with 4Li⁺ in a conversion reaction to form the completely reduced products of Fe° and 2Li₂S. Fe° acts as a catalyst for the oxidation of Li₂S The products of oxidation include various electronically conducting phases of FeS_x. These phases then help to eiectrochemicaily activate excess Li₂S present in the solid electrolyte.

Electrochemical evolution of a bulk-type all-solid-state iron sulfur cathode

[00963 The lithium aiS-so!id-state battery described above, which may be thermally activated as described above, may also be made using a high capacity conversion battery materials (e.g., FeSa) equivalent. Examples are described in the following discussing as sn-situ electrochemical formation of a FeS2 phase and reversible utilization of a glass electrolyte for higher overall electrode energy density. However, the lithium all-solid state battery is not limited to such an implementation.

[00973 The results described below show that eiectrochemicaiiy structured interfaces between conversion active materials and the glass electrolyte can be utilized to increase energy density of ASSLBs, while maintaining good rate performance,

[00981 For purposes of illustration, synthesis of the iron sulfide based all- solid-state composite electrodes can be by a three step planetary ball milling procedure (Across International, PG-N2), An example 77,5U₂S-22,5P₂Ss (molar ratio) glass electrolyte can be prepared by milling about 0.632g Li₂S (Aidrieh, 99.999%, reagent grade) and about 1.168g P₂S_S (Aidrieh, 99%) in a 500mL stainless steel vial (Across i ternational) with two stainless steel balls (having about a 16mm diameter) and twenty stainless steel balls (having about a 10mm diameter) at about 400 rpm for about 20 hours. [0099] The 1 :1 molar ratio FeS:S active materia! composite (denoted as FeS + S) can be prepared by milling about 0.733g FeS (Aldrich, technical grade) and about 0.267g Sulfur (Aidrich, 99.98%) in a 1 Q0mL agate jar (Across International) with five agate balls (having about a 10mm diameter) and fifty agate baiis (having about a 6mm diameter) at about 400 rpm for about 20 hours,

[00100] The composite electrode can be synthesized by milling a ratio of prepared FeS + S, 77.5Li₂S-22.5P₂S₅₎ and carbon biack conductive additive (Timcai, C85) in a lOOmL agate jar with five agate baiis (having about a 0mm diameter) and fifty agate bails (having about a 6mm diameter) at about 400 rpm for about 18 minutes.

[00101] in an example, cell fabrication and ceil testing was carried out under an inert Argon gas environment, although other environments may also be utilized. The working electrode is about 5mg of the mechanically prepared FeS + S based composite electrode, in this example, about 5mg of stabilized lithium metal powder (SL P) was used as the counter electrode (F C Lithium Corp., Lectro Max Powder 100). The shell of the solid state battery was a titanium- polyarySetheretherketone (PEEK) test celi die. To fabricate each ceil, the giass electrolyte powder was first compressed at about 5 metric tons inside the PEEK cell die to form the separator peliet. In this example, about 5mg of composite positive electrode and the SLMP were then attached to opposite sides of the glass electrolyte pellet with about 5 metric tons force.

|00102] A variety of different batteries were fabricated to aid in the characterization of the FeS + S/LJ battery's electrochemistry. Stiil other examples are contemplated, and the examples discussed herein are merely illustrative. In an example, the FeS in these batteries was prepared by mechanically milling about 2g of FeS in a 100ml agate jar (Across International) with five agate bails (having about a 10mm diameter) and fifty agate baiis (having about a 6mm diameter) at about 400 rpm for about 20 hours. The ceils used to electrochemicaily prepare the cycled XRD samples had a 165 mg FeS composite cathode and an tnli alloy anode. These ceils operate at a lower potential because the InU alloy has a potential of about 0.62V vs. Li⁺/Li. [00103] In these examples, all ceils were cycled under constant current constant voltage (CCCV) conditions using an Arbin BT2000 battery tester at about 60 . Because the overall capacity of the FeS electrode is a moving target, rate performance is described by current and not by C-rate. Unless otherwise noted, specific capacities are given with respect to the total mass of the composite electrode. Materials are characterized by field emission scanning electron microscopy (FESEM, JEOL JS -7401F) and Cu-Κα X-ray (XRD) measurement.

|00104] Figure 15 shows (a) XRD of FeS + 8 composite active material and FeS precursor with indexed reflections for Fe_?Ss and FeS. The FeS precursor material is likely a multiphase mixture of FeS and an iron deficient Fe-_f,_xS phase. After mechanoehemical milling, only reflections for FeS and S are observed which indicates that no solid state reactions occurred to form FeS₂; and (b) is an FESEM micrograph of the FeS + S composite active material.

[00105] XRD measurement presented in Figure 15(a) shows that the FeS precursor is composed partly of FeS (Triolite). However, the precursor also exhibits ferrimagnetism, which indicates that the sample is likely a multiphase mixture of FeS and an iron deficient phase like FerSs (Pyrrhotite). An unidentified peak at 44.64° suggests that other phases may be present as well. 00106] After mechanical milling with S, the reflections for FeS are observed to decrease in intensity and broaden. This is consistent with a decrease in average particle size by mechanical grinding action. Further, no new reflections are observed which suggests that the nano-composite is an intimate mixture of elemental FeS and S. The peak at 37, can be attributed to the strong (317) reflection of the S precursor (JCPDS #832285). Figure 15(b) confirms that the FeS + S nano-composite is comprised of sub-micron sized particles.

00107] The electrochemical behavior of a FeS + S composite electrode over time is complex. Figure 1$ shows (a) cyclic stability of a FeS + S/Li battery. Assuming that the active materia! composition of FeS + S has a theoretical specific capacity of 900 mAh g^*1, the electrode's specific capacity should not exceed 281 mAh g^'\ Excess capacity is evidence for the electrochemical utilization of the 77.5Li2S-22,5P₂S5 glass electrolyte component; and (b)-(e) are voltage profiles for cycles 1 , 2, 10, 20, 30, 40, 80, and 150 of the same battery. The variability of voltage profiles with respect to cycle number suggests an evolving electrochemistry.

[00108] Figure 16(a) shows the cyclic stability of a 5mg electrode cycled at 60°C with a current of 144 μΑ, The electrode exhibits an initial discharge of 320 mAh g or 1020 mAh g^" based only on the mass fraction of the FeS + S nano- composite active material. The theoretical capacity of FeS + S should only be approximately 900 mAh g^'\ however, the other iron deficient phases present in the FeS precursor affect the actual theoretical value. After the initial discharge, the electrode rapidly gains capacity until a maximum capacity of 550 mAh g ¹ and maximum energy density of 1040 Wh kg^"1 are achieved by the 16^ih cycle. This maximum is followed by an extended phase of capacity fade until the electrode's capacity stabilizes at 330 mAh g^{" !} by the sixty-eighth cycle. The initial discharge and the one-hundred-fiftieth cycle both exhibit a specific capacity of very nearly 320 mAh g^' yet the energy densities of these cycles are 470 and 600 Wh kg^'1, respectively. The discrepancy in energy densities despite similar specific capacities is a reflection of voltage profile evolution to higher potentials. Figures 2(b-e) present the voltage profiles of the first and second, tenth and twentieth, thirtieth and fortieth, and sixtieth and one-hundred-fiftieth cycles, respectively.

[00109] Voltage profile evolution can be correlated to the rise, fade and stabilization of the electrode's capacity. To understand the behavior of the FeS + S electrode, dQ/dV analysis was employed to qualitatively identify parallel redox chemistries. Figure 17 shows (a) dQ/dV profiles for bulk-type all-solid-state Li metal batteries with FeS, syn-FeS_2; and S composite cathodes; and (b)-(e) dQ dV profiles for cycles 1 , 2, 10, 20, 30, 40, 60 and 150 of an example FeS + S Li battery. It is observed that the evolution of three parallel redox chemistries accounts for the changing capacity of the battery. Twinning of reduction peaks in voltage ranges between 2,13 - 2.19 and 1.56 - 1.51 V provides evidence for the in-situ electrochemical formation of a FeS₂ phase,

00110] Figure 17a shows the characteristic behavior of FeS, FeS₂, and S in an ASSLB. The FeSs chemistry is characterized by dominant reduction peaks at 28

2.20, 2.14 and 1.49 V, while S is characterized by a reduction peak at 2.2V and FeS by a peak at 1.55V. The dQ/dV profiles in Figure 17b for the first and second cycles show evidence for the reduction of only FeS and S. As the cell's capacity increases, so too does the complexity of the voltage profiles. By the tenth and twentieth cycle, the dQ/dV profiles in Figure 17c show evidence for the reduction of FeS, S and FeS₂. The reduction of FeSs is apparent due to the twinning of reduction peaks in voltage ranges of 2.14 - 2.19 V and 1.51 - 1.56 V. This is the first time evidence for the in-situ electrochemical formation of FeS₂ has been presented.

00111] Next, capacity fade is correlated to the decline of peaks associated with Fe° <---> Fe²⁺ reduction and oxidation. The dQ/dV profiles in Figure 17d for the thirtieth and fortieth cycles during the capacity fade phase show that Fe redox peaks disappear while the S redox peaks remain relatively stable. By the sixtieth cycle the dQ/dV profiles in Figure 17e for the electrode achieve a degree of stability. Extended cycling capacity loss is associated with continued fade of the Fe redox peaks and the fade of S redox peaks.

00112] The electrochemical utilization of the 77.5Li₂S:22.5P₂S5 glass electrolyte can be understood with reference to ex-situ XRD measurement of FeS based electrodes. In this example, FeS was used instead of FeS + S to simplify the analysis. A glass eSectroiyte was used because the absence of ceramic electrolyte diffraction patterns also simplifies the analysis. To mimic the nano-size morphology of the FeS + S active material, the FeS used in this experiment was mechanically milled.

00113] Figure 18 shows (a) XRD of FeS before and after milling indicates that no phase changes occur during milling, (b) an FESE micrograph of FeS as received from the vendor, and (c) an FESEM micrograph of F ^S after mechanical milling. The XRD and FESEM analysis confirm that the mechanical milting provides the needed size reduction. Utilization of nano-FeS helps obtain good contact with the glass electrolyte.

[00114] Figure 19{a) shows voltage profiles for two initially overcharged FeS/Liln batteries. Composite electrodes were collected for XRD after full overcharge (solid) and full discharge (dashed). As seen in Figure 19(a), the FeS composite electrode was recovered after an initial overcharge (so!id) and after full discharge (dashed) (Figure 4a),

100115] in Figure 19(b), all XRD measurements are normalized with the (100) reflection of the beta-Be sample window (dotted). Li₃S reflections are indicated by the dashed lines while Fei-_xS reflections are indicated by the so!id lines. The XRD sample cycled to an initial full discharge exhibits a iow capacity because the electrode composite was mixed manually and not mechanically. Because FeS is already in the fully charged state, these ceiis should have exhibited zero charge capacity. Instead, both cells achieve an overcharge capacity of about 100 mAh g^"1.

00116] ft ^can be seen from the XRD of cycled FeS composite electrodes in Figure 19(b), that this capacity is associated with the utilization of excess Li₂S in the glass electrolyte component of the composite electrode. The diffraction pattern for the glass electrolyte 1950 includes reflections for UaS (dashed). As expected, the diffraction pattern for an uncycled composite electrode 1951 includes reflections for Li₂S and FeS (solid). After an initial overcharge 1952, the reflections for Li₂S disappear. When an electrode is discharged after an initial overcharge 1953, reflections for FeS disappear and weak reflections for L S are once again detected. The reduced intensity of the L\₂S reflections in this sample is believed to be due to the small size of electrochemical^ precipitated LbS particles.

[00117] To determine how reasonable it is to attribute the excess capacity to the utilization of excess Li₂S in the glass electrolyte, the percentage of Li₂S oxidized in the cell presented in Figures 18 and 17 can be estimated by assuming that ali of the FeS + S was electrochemicaiiy utilized and by acknowledging that the highest achieved capacity is 2,8 mAh, A fraction of this capacity can be attributed to utilization of the glass electrolyte at the electrode/glass electrolyte separator interface, as indicated by Figure 21.

[00118] Figure 20 shows the first through seventeenth voltage profiles of a FeS/Liin battery where the cathode is composed entirely of 5mg of nano-FeS. Normally, FeS should reversibly exhibit only a single plateau. However, with cycling, this battery develops a higher voltage plateau at approximaiely 1.4 V versus. Ιηϋ, The evolution of the second plateau can be attributed to the electrochemical utilization of Li₂S in the glass electrolyte at the electrode/glass electrolyte separator interface. By the seventeenth cycle, up to 300 pAh of capacity can be attributed to the utilization of glass electrolyte separator. This interface utilization effect may lead to an overestimation of specific capaciiy and contribute to the observed excess capacity.

[00119] Utilization of the glass electrolyte separator may lead to an overestimation of specific capacity and contribute to the observed excess capacity especially when the electrode is very small like it is in this case. The FeS + S component can account for 1.45 mAh of the maximum capacity. Without accounting for utilization of the giass electroiyte separator, the remaining capacity indicates that 1.16 mg, or 88%, of Li₂S in the giass electrolyte is etectrochemicaSly utilized,

[00120] Such a large percentage of the IkS component is iikeiy not oxidized without decreasing the ionic transport of the composite electrode. Yet, a rate test at 60°C was conducted on another FeS + S electrode after a five cycle activation and good performance was observed.

[00121] Figure 21 shows (a) specific discharge capacity of a battery as a function of applied current, and (b) voltage profiles of the same battery as a function of applied current. The electrode maintains a specific capacity of 200 mAh g^*1 with an applied current of about 2.40 mA. To maintain good ionic transport, it is Iikeiy that the P2S5 component of the glass electroiyte may be electrochemicaily utilized as well. Utilization of P₂Ss also explains why capacity increases during discharge cycles, and not only during the charge cycles when excess L S is initially oxidized to S.

[00122] it is noted that while the initial specific discharge capacity of the ceil presented in Figures 16 and 17 is similar to that presented in Figure 20, the degree and rate of capacity activation of each ceil is much different. By the fifth cycle the first cell exhibits a specific capacity of about 488 mAh g^*\ whiie the celt used in the rate test only exhibits a specific capacity of about 400 mAh g^*\ The maximum capacity of the rate cell is also about 15% lower than that achieved by the first cell. [00123] The electrodes for these two samples were prepared separately which attests to the sensitivity of electrolyte activation to the quality of the glass electrolyte. Electrolyte sensitivit is emphasized by the results of another study that did not observe electrolyte utilization. Like this study, nano-FeS was used as an active material and Li₂S was a precursor for their electrolyte. However, the electrolyte was a different composition, thio-LISfCO Lf₃,₂₅Geo.25Po 5S4, and was prepared by melt quenching instead of mechanochemica! milling. Previous work also did not show evidence for electrochemical activation of the glass electrolyte. In this study, three 5 micron cubes of synthetic FeSa were used as the active material. Such a large particle size results in poor contact between the glass electrolyte and th active material which may inhibit electrolyte utilization. Cubic-FeSa is also a semiconductor with a much lower electronic conductivity than that of the ferrimagnetic Fe^_xS precursor used in this study,

[00124] it is noted that the decision to mechanically combine pm-Cu powder with S or LiaS can be further improved. For example, the conversion materials, FeFa and CuF₂ suggest that a nano-sfructured network of reduced metallic nanoparticles may be employed for good reversibility. A mechanical mixture with micron active metal particles is therefore not ideal for good reversibility or for good electrolyte utilization. Instead, efectrochemicaiiy reduced nano-active metal particles may be used due to the better atomic proximity to other reduced species, as well as to the electrolyte particles, to further enhance reversibility and effective electrolyte utilization.

[00125] The examples shown and described herein are provided for purposes of illustration and are not intended to be limiting. Still other examples are also contemplated.

Claims

1 . An all-solid-state lithium battery, comprising:

a cathode having a transition metal sulfide mixed with elemental sulfur, wherein:

upon full discharge, the cathode undergoes conversion reactions to form a transition metal + lithium sulfide; and

upon full charge, the cathode undergoes conversion reactions to form the transition metal sulfide + lithium + electrons,

2. The battery of claim 1 , wherein the transition metal sulfide is selected from: monosulfides, disulfides, and trisuSfides.

3. The battery of claim 1 , wherein transition metal sulfide Is mechanically combined,

4. The battery of claim 1 , wherein the cathode comprises solid state electrode (SSE) particles, and a conducting additive,

5. The battery of claim 1 , further comprising a solid state electrode (SSE) iayer between the cathode and an anode.

8. The battery of claim 1 , further comprising an anode including lithium metal, graphite, or silicon-based active materials,

7. The battery of claim 1 , wherein the cathode is selected from eSz or FeS2 equivalent.

8, A method of in-situ electrochemical synthesis of pyrite {FeS₂) from iron sulfide (FeS) and elemental sulfur (S) precursors, comprising:

cycling FeS + S composite electrodes at moderately elevated temperature;

wherein charge products are described by the following equation: L(2-xFeS₂ « -> G.8ortho-FeS₂ + 0.2FeS_S;7 + 0.175S + {2~χ)ϋ* + (2-χ)β^'

9. The meihod of claim 8, further comprising producing voltage plateaus indicative of FeS₂ in batter ceiis constructed with FeS + S or as an FeS₂ equivalent.

10. The method of claim 9, wherein the voltage plateaus become more defined upon further cycling.

11. The method of claim 8, wherein the moderately elevated temperature is about 60° C,

12. The method of claim 8, wherein initial discharge of FeSs proceeds in two steps:

(1 ) FeSa + 2LF + 2e^" *→ Li₂FeS₂

(2) Li₂FeS₂ +2Li⁺ + 2e^~ *~* 2Li₂S + Fe°.

13. The method of ciaim 12, wherein subsequent charge and discharge cycles proceed according to the following reactions:

{3} Fe° + Li₂S < Li₂FeS₂ + 2Lf + 2e^"

(4) Li₂FeS2 *→ Li₂._xFeSa + xLi* + xe^" (where, 0.5 < x < 0.8}

(5) Liz-xFeSa → 0.8ortho-FeS₂ + 0.2FeS_Si7 + 0.175S + (2-x)Li^* + (2-x)e\

14. A so!id-state lithium battery, comprising:

a solid state electrolyte; and

an activating agent, wherein the activating agent activates excess L S in the solid state electrolyte to realize an improved charge capacity.

15. The battery of ciaim 14, wherein solid state electrolyte is suSfide-based, 18, The battery of claim 14, wherein activating agent is a transition meta! sulfide such as FeS, TiS_2l Fe$₂ and/or FeS₂ equivalent,

17. The battery of claim 14, wherein the activating agent has a highly ionic and/or electrically conductive character,

18. The battery of claim 17, wherein the highly ionic and electrically conductive character of the activating agent activates the solid state electrolyte.

19. The battery of claim 18, wherein the activating agent activates otherwise inert excess LS2S in the solid state electrolyte,

20. The battery of claim 14, wherein the improved charge capacity is realized after a single charge event at an elevated temperature.

21. The battery of claim 20, wherein the elevated temperature is about BOX.

22. The battery of claim 20, wherein the improved charge capacity is greater than about 50%,

23. The battery of claim 14, wherein the sulfide based solid electrolyte is xLi₂S-(100-x)P₂S5.

24. The battery of claim 14, further comprising a composite electrode,

25. Th battery of claim 24, wherein the composite electrode is 80Li₂S- 20p₂S5'.acet len black.

28. The battery of claim 24, wherein the composite electrode is Τ!3₂:80υ₂8~ 20P₂S₅ acetylene black. 27, The battery of claim 24, further comprising an in metal negative electrode.

28, A method of activation of a solid-state iithium battery, comprising ihermo- electrochemica! activating excess Li₂S in a solid state electrolyte to realize an improved charge capacity.

29, The method of claim 28, wherein the improved charge capacity is realized after a single charge event at an elevated temperature.

30, The method of claim 29, wherein the elevated temperature is about SOX.

31 , The method of claim 29, wherein the improved charge capacity is greater than about 50%.

32. The method of claim 29, further comprising an FeS₂ equivalent cathode.