Solid state ion conducting materials have a number of useful applications, including as materials in a Li-ion all-solid-state battery (ASSB). To commercialize Li-ion ASSBs, a suitable Li-ion conducting solid-state electrolyte is required. A solid-state electrolyte should exhibit a wide electrochemical stability window and ionic conductivity near that of traditional liquid electrolytes. Three compounds with near-liquid-electrolyte conductivity (˜10−2 S cm−1) have been reported: Li10GeP2S12 (LGPS), Li6PS5Br argyrodite, and a Li7P3S11 glass ceramic. Unfortunately, all three discovered electrolytes exhibit electrochemical instability against the Li anode, limiting application in commercial products. Some materials are known or are predicted to have a wide electrochemical stability window, sufficient for resisting electron injection from the Li anode, but suffer from prohibitively low ionic conductivity in their pure or intrinsic form. Accordingly, there is a need for ion conducting solid state materials suitable as solid state electrolytes for battery technologies.
Aspects disclosed herein include materials comprising: a lithium thioborate composition characterized by formula FX1: Li3−z[B+Q]1[S+G]3 (FX1); wherein Q is a first dopant being a substitute for B in the composition and being one or more elements each aliovalent with respect to B; wherein G is a second dopant being a substitute for S in the composition and being one or more elements each aliovalent with respect to S; wherein z is 0 or a number greater than 0 and less than or equal to 0.40, optionally less than or equal to 0.05; and wherein the composition comprises only the first dopant, only the second dopant, or both the first dopant and the second dopant.
Aspects disclosed herein include solid state electrolytes comprising: a lithium solid state electrolyte comprising Li, one or more principal elements, and at least one dopant; wherein the dopant substitutes for a portion of the one of the one or more principal elements of the lithium solid state electrolyte and is aliovalent with the respective substituted principal elements; wherein the ionic conductivity of the lithium solid state electrolyte is greater than or equal to 1·10−5 S/cm at 25° C.
Aspects disclosed herein include doped lithium solid state electrolytes comprising: a doped inorganic composition having at least one dopant; wherein the doped composition has up to 20 at. % of one or more principal elements substituted with the at least one dopant relative to a reference composition of a reference lithium solid state electrolyte; wherein each dopant is one or more elements each aliovalent with the respective substituted principal element; wherein the presence of the one or more dopants provides for an ionic conductivity greater than or equal to 1·10−5 S/cm at 25° C.
Aspects disclosed herein include methods for increasing an ionic conductivity of a reference lithium solid state electrolyte, the method comprising: forming a doped lithium solid state electrolyte having a doped composition; wherein the reference lithium solid state electrolyte has a reference composition, and wherein the doped composition has up to 20 at. % of one or more principal elements substituted with at least one dopant relative to the reference composition; wherein each element of the at least one dopant is aliovalent with respect to the respective substituted principal element; and wherein the doped lithium solid state electrolyte has a greater ionic conductivity than the reference lithium solid state electrolyte by a factor of at least 10.
Without wishing to be bound by any particular theory, there may be discussion herein of beliefs or understandings of underlying principles relating to the devices and methods disclosed herein. It is recognized that regardless of the ultimate correctness of any mechanistic explanation or hypothesis, an embodiment of the invention can nonetheless be operative and useful.
In general, the terms and phrases used herein have their art-recognized meaning, which can be found by reference to standard texts, journal references and contexts known to those skilled in the art. The following definitions are provided to clarify their specific use in the context of the invention.
The term “dopant” is used herein broadly to refer to one or more elements intentionally provided in a material's composition to improve or enhance one or more properties or functionalities of the resulting doped material to compared to the undoped reference form of the material, which is typically intrinsic and stoichiometric. A doped composition is optionally referred to as an extrinsic composition, whereas the undoped composition is optionally referred to as the intrinsic or reference composition. As used herein, the term dopant is used broadly to include low concentration impurities or low concentration additive element(s), such that providing said dopant may be referred to as doping and/or alloying as these terms are known in the art. As used herein, a dopant is a substitute for another or principal element, where the dopant replaces or substitutes for a portion of the amount or concentration of the principal element relative to the amount or concentration the principal element in the undoped composition. For clarity and as a convenient handle, each element of a reference undoped composition may be referred to as a “principal element”. Generally, a principal element is an element identified (or which would be identified by one of skill in a relevant art) in a chemical formula of a composition and is exclusive of impurity elements present in the composition at less than 0.05 at. %, less than 0.05 mol. %, and/or less than 0.05 wt. % (optionally less than 0.04 at. %, less than 0.04 mol. %, and/or less than 0.04 wt. %; optionally less than 0.03 at. %, less than 0.03 mol. %, and/or less than 0.03 wt. %; optionally less than 0.02 at. %, less than 0.02 mol. %, and/or less than 0.02 wt. %; optionally less than 0.01 at. %, less than 0.01 mol. %, and/or less than 0.01 wt. %). Therefore, for example: Li, B, and S are principal elements of the composition Li3BS3; Li, V, and S are the principal elements of the composition Li3VS4; Na, Li, Al, and F are the principal elements of the composition Na3Li3Al2F12; Li and Te are the principal elements of the composition Li2Te; Li, Al, and Te are the principal elements of the composition LiAlTe2; Li, In, and Te are the principal elements of the composition LiInTe2; Li, Mn, and S are the principal elements of the composition Li6MnS4; Li, Ga, and Te are the principal elements of the composition LiGaTe2; K, Li, Ta, and O are the principal elements of the composition KLi6TaO6; Li, Cu, and S are the principal elements of the composition Li3CuS2; etc. Generally, materials disclosed herein comprise one or more dopants that are substitutes for one or more principal elements (optionally, non-Li principal elements) of a composition (i.e., a principal element other than Li of a composition, such as any of those listed in the prior sentence; e.g., a dopant may be a substitute for B or S in Li3BS3, where the B and S are principal elements (optionally, non-Li principal elements) of the composition Li3BS3). As used herein, a dopant is optionally one element being substitute for a principal element of a composition (e.g., Si being a dopant for B in Li3BS3). As used herein, a dopant is optionally two elements being substitutes for a principal element of a composition (e.g., Si and Ge together being a dopant for B in Li3BS3). As used herein, a dopant is optionally three or more elements being substitutes for a principal element of a composition. Herein, the terms “doped” and “substituted” are generally used interchangeably when referring to a material or composition having one or more dopants, as disclosed herein, substituting one or more principal elements, such as one or more principal elements (optionally, non-Li principal elements). A doped composition may have one dopant or more than one dopants. For example, a doped composition may have a first dopant (or, first type of dopant) being a dopant or substitute for a first principal element (optionally, a non-Li principal element) of a composition (e.g., B of Li3BS3) and the same doped composition may have a second dopant (or, second type of dopant) being a dopant or substitute for a second principal element (optionally, a non-Li principal element) of a composition (e.g., S of Li3BS3). As used herein, a dopant element may be present in the structure of the doped composition substitutionally (as a substitutional dopant element), interstitially (as an interstitial dopant element), or both substitutionally and interstitially. As used herein, a dopant element is preferably aliovalent with respect to the principal element it replaces or for which it is a substitute. For example, a dopant element for B (e.g., in Li3BS3) is preferably aliovalent with respect to B, such that said dopant element is a member of an element Group other than Group 13 of the Periodic Table of Elements (e.g., Si, being a member of Group 14). For example, a dopant element for S (e.g., in Li3BS3) is preferably aliovalent with respect to S, such that said dopant element is a member of an element Group other than Group 16 of the Periodic Table of Elements (e.g., Cl, being a member of Group 17). In some aspects, the material or composition thereof having the one or more dopants is characterized as a solid solution. In some aspects, the introduction of one or more dopants to a material or composition thereof obeys Vegard's law where the one or more dopants incorporate into the material's lattice or structure as a solid solution.
The term “amorphizing” refers to a process that reduces grain sizes (average, median, and/or bounds of a 95% confidence interval) of a material (or composition thereof), reduces crystallite sizes (average, median, and/or bounds of a 95% confidence interval) of a material (or composition thereof), increasing amorphous content of a material (or composition thereof), decreasing total crystallinity of a material (or composition thereof), and/or increasing an amount or concentration of defects in a material (or composition thereof). A defect generally refers to a crystallographic defect. As recognized by those skilled in the relevant art such as materials science or crystallography in particular, a defect may be a point defect, a line defect, a planar defect, and/or a bulk defect. A vacancy (or, “vacancy defect”), such as a vacancy of a principal element such as B in Li3BS3, is an example of a defect. A broken or dangling bond, which is optionally but not necessarily a result of a vacancy defect, is another example of a defect. Generally, but not necessarily, amorphizing refers to a process performed on a material after said material is formed or made. Thus, generally but not necessarily, amorphizing does not refer to the process of doping or making a doped composition (although doping may introduce defects such as interstitial defects), but rather to a separate or subsequent processing step performed on a material that has been formed. An example of an amorphizing process is ball milling, or any similar processes. In some aspects, the term amorphizing refers to a process that necessarily increases amorphous content of a material (or composition thereof) and decreases total crystallinity of the material (or composition thereof), while also optionally reducing grain sizes (average, median, and/or bounds of a 95% confidence interval) of the material (or composition thereof), optionally reducing crystallite sizes (average, median, and/or bounds of a 95% confidence interval) of the material (or composition thereof), and/or optionally increasing an amount or concentration of defects in the material (or composition thereof).
The term “ionic conductivity” is intended to be consistent with the term as it is readily known by one skilled in relevant arts, particularly in the art of semiconductors and/or solid state electrolytes, and refers the property of ionic conductivity as it would relate to the performance of a material or composition thereof as an ionically conductive solid state electrolyte in an electrochemical cell such as a battery. In some aspects, ionic conductivity particularly refers to ionic conductivity of Li+ ions in a material or a composition thereof. As used herein, unless explicitly otherwise stated, ionic conductivity of a material (or composition thereof) refers to ionic conductivity within or through the material, such as through a thickness or lateral dimension of the material (e.g., through the thickness of a thin film), instead of a surface ionic conductivity along a film's surface longitudinally. Unless otherwise explicitly stated, the term ionic conductivity refers to a combination of grain boundary transport of ions and bulk ionic conductivity. Preferably, but not necessarily, an ionic conductivity claimed herein is an average ionic conductivity, being an average of at least three repeated measurements. Further to the descriptions in Examples 1A-3 provided herein, useful techniques, assumptions, parameters, calculations, etc., for measuring ionic conductivity in materials disclosed herein is found in P. Vadhva, et al. (“Electrochemical Impedance Spectroscopy for All-Solid-State Batteries: Theory, Methods and Future Outlook”, first published in ChemElectroChem; Volume 8; Issue 11; 2021; Pages 1930-1947; DOI: 10.1002/celc.202100108), which is incorporated herein in its entirety.
The term “electronic conductivity” is intended to be consistent with the term as it is readily known by one skilled in relevant arts, particularly in the art of semiconductors and/or solid state electrolytes, and refers the property of electronic conductivity (conductivity or transport of electrons) as it would relate to the performance of a material or composition thereof as an ionically conductive (and preferably electronically insulating) solid state electrolyte in an electrochemical cell such as a battery. For clarity, electronic conductivity does not refer to nor include ionic conductivity. As used herein, unless explicitly otherwise stated, electronic conductivity of a material (or composition thereof) refers to electronic conductivity within or through the material, such as through a thickness or lateral dimension of the material (e.g., through the thickness of a thin film), instead of a surface electronic conductivity along a film's surface longitudinally. Unless otherwise explicitly stated, the term electronic conductivity may be inclusive of any and all possible mechanisms of electronic transport (i.e., transport/conductivity of electrons; e.g., including Poole-Frenkel emission, hopping conduction, ohmic conduction, space-charge-limited conduction, and/or grain-boundary-limited conduction). Further to the descriptions in Examples 1A-3 provided herein, useful techniques, assumptions, parameters, calculations, etc., for measuring electronic conductivity in materials disclosed herein is found in P. Vadhva, et al. (“Electrochemical Impedance Spectroscopy for All-Solid-State Batteries: Theory, Methods and Future Outlook”, published in ChemElectroChem; Volume 8; Issue 11; 2021; Pages 1930-1947; DOI: 10.1002/celc.202100108), which is incorporated herein in its entirety.
The term “electrochemical cell” refers to devices and/or device components that convert chemical energy into electrical energy or electrical energy into chemical energy. Electrochemical cells have two or more electrodes (e.g., positive and negative electrodes) and one or more electrolytes. For example, an electrolyte may be a fluid electrolyte or a solid electrolyte. In aspects disclosed herein, electrochemical cells comprise at least one solid state electrolyte (optionally but not necessarily also having a fluid electrolyte), the solid state electrolyte comprising a material or composition disclosed herein. The solid state electrolyte is an ionically conductive (e.g., for Li+ ions), and preferably electronically insulating to prevent electrical/electronic shorting between oppositely-charged electrodes within the electrochemical cell or battery. Reactions occurring at the electrode, such as sorption and desorption of a chemical species or such as an oxidation or reduction reaction, contribute to charge transfer processes in the electrochemical cell. Electrochemical cells include, but are not limited to, primary (non-rechargeable) batteries and secondary (rechargeable) batteries. In certain aspects, the term electrochemical cell includes metal hydride batteries, metal-air batteries, fuel cells, supercapacitors, capacitors, flow batteries, solid-state batteries, and catalysis or electrocatalytic cells (e.g., those utilizing an alkaline aqueous electrolyte). In some aspects, an electrochemical cell is a Li-ion or Li-ion based battery.
The term “gravimetric capacity”, consistent with the term as used in the art, particularly in the art of battery devices and electrochemistry, refers to amount of charge that can be stored per unit mass. The units are typically mAh/g or C/g. Generally, in the art, gravimetric capacity is normalized by the mass of active material in a cathode or anode, with the balance-of-plant ignored (carbon, binder, etc.) to allow for comparison between active materials. With respect to a battery, the capacity of the battery is normalized to the entire cell which includes all the “inactive” components like carbons, the current collectors, electrolyte, etc. Solid state electrolytes are normally reported as a way to enable Li metal anodes, which have a much higher gravimetric capacity than commercialized graphite anodes. For example, even though solid-state electrolytes are heavier/denser than liquid electrolytes, a solid-state battery could have higher capacity if the solid-state electrolyte is paired with a Li metal anode.
The term “stability”, as used herein in reference to a solid state electrolyte or a material, or composition thereof, that is a candidate solid state electrolyte generally refers to chemical and electrochemical stability (thermodynamic and kinetic) of the electrolyte or material, or composition thereof, with respect to reduction by a Li metal anode at voltages relevant to the operation of a battery having said electrolyte or material. Further to the descriptions in Examples 1A-3 provided herein, useful background, description, techniques, assumptions, parameters, calculations, etc., for determining stability of solid state electrolytes and materials that are candidate solid state electrolytes is found in H. Park, et al. (“Predicting Charge Transfer Stability between Sulfide Solid Electrolytes and Li Metal Anodes”, ACS Energy Lett. 2021, 6, 1, 150-157; DOI: 10.1021/acsenergylett.0c02372), which is incorporated herein in its entirety.
As used herein, total crystallinity refers to the sum of the wt. % of all crystal phases present in the material or a composition thereof. Optionally in any aspect disclosed herein, a material disclosed herein is characterized by a total crystallinity equal to or less than about 25% by weight (wt. %), optionally equal to or less than about 20 wt. %, optionally equal to or less than about 15 wt. %, optionally equal to or less than about 10 wt. %, optionally equal to or less than about 8 wt. %, optionally equal to or less than about 5 wt. %, optionally equal to or less than about 4 wt. %, optionally equal to or less than about 3 wt. %, optionally equal to or less than about 2 wt. %, optionally equal to or less than about 1 wt. %, optionally equal to or less than about 0.8 wt. %, optionally equal to or less than about 0.5 wt. %, optionally equal to or less than about 0.2 wt. %, optionally equal to or less than about 0.1 wt. %, optionally equal to or less than about 0.08 wt. %, optionally equal to or less than about 0.05 wt. %, optionally equal to or less than about 0.01 wt. %. The total crystallinity of a material or composition thereof can be determined through Rietveld quantitative analysis of X-ray diffraction (XRD) data measured from the material or a representative sample thereof. For example, the XRD may be measured using a sheet, film, pellet, powder, or such, of the material, for example. Optionally, XRD data is collected using a powder x-ray diffraction technique with a scan from 5 to 80 degrees, unless otherwise specified. For example, the Rietveld quantitative analysis method may employ a least squares method to model the XRD data and then determine the concentration of crystal phase(s) in the sample based on known lattice(s) and scale factor(s)s for the identified phase(s). However, it is understood that different methods and instrumentation for determining total crystallinity can also be employed.
In an embodiment, a composition or compound of the invention, such as an alloy or precursor to an alloy, is isolated or substantially purified. In an embodiment, an isolated or purified compound is at least partially isolated or substantially purified as would be understood in the art. In an embodiment, a substantially purified composition, compound or formulation of the invention has a chemical purity of 95%, optionally for some applications 99%, optionally for some applications 99.9%, optionally for some applications 99.99%, and optionally for some applications 99.999% pure.
In the following description, numerous specific details of the devices, device components and methods of the present invention are set forth in order to provide a thorough explanation of the precise nature of the invention. It will be apparent, however, to those of skill in the art that the invention can be practiced without these specific details.
The Li-ion all-solid-state battery (ASSB) is a promising template for next-generation energy storage. To commercialize Li-ion ASSBs, a suitable solid-state electrolyte is required. A solid-state electrolyte must exhibit a wide electrochemical stability window and ionic conductivity near that of traditional liquid electrolytes. Three compounds with near-liquid-electrolyte conductivity (˜10−2 S cm−1) have been discovered: Li10GeP2S12 (LGPS), Li6PS5Br argyrodite, and a Li7P3S11 glass ceramic. All three discovered electrolytes exhibit electrochemical instability against the Li anode, limiting application in commercial products. Li3BS3 is predicted to have a wide electrochemical stability window, sufficient for resisting electron injection from the Li anode.1 However, the ionic conductivity of pure or intrinsic Li3BS3 is prohibitively low (10−7-10−6 S cm−1).
In aspects herein, we present a method to significantly enhance the ionic conductivity of materials such as Li3BS3, which are candidates for solid state lithium ion conductivity electrolytes, through incorporation of Si and subsequent amorphization. Also disclosed here are ionically conductive materials, such as doped or substituted Li3BS3. For example, by substituting up to 5%, for example, of the B sites with Si, Li3−xB1−xSixS3 achieves an ionic conductivity surpassing 10−5 S cm−1 at room temperature. When the doped product is further amorphized, for example through continuous ball milling, the ionic conductivity is further enhanced to above 10−3 S cm−1 at room temperature.
Pure or intrinsic Li3BS3 has been studied and characterized in the past. The pure structure exhibits an ionic conductivity in the range of 10−7-10−6 S cm−1, which is unfavorably low for the material to be useful as a solid state lithium ion conductor. Ionic conductivity of pure/intrinsic Li3BS3 can be enhanced via extended ball milling to near 10−4 S cm−1, 2 aspects disclosed herein include materials and associated methods demonstrating further significant enhancement in ionic conductivity of materials that are candidates for solid state lithium ion conductors (e.g., for use in a solid state electrolyte), such as Li3BS3.
In aspects, substitution of a principal element such as B, S, or both B and S in Li3BS3 with an aliovalent dopant element also reduces the amount of Li in the composition relative to the undoped composition. For example, in aspects, the relative amount of Li is reduced according to formula FX2: Li3−x−yB1−x[Q]xS3−y[G]y, where Q (“first dopant”) is one or more (“first”) dopant elements aliovalent with respect to B, G (“second dopant”) is one or more (“second”) dopant elements each aliovalent with respect to S, x is a number (e.g., selected from range of 0.005 to 0.20) corresponding to the relative amount of substitution of B, and y is a number (e.g., selected from range of 0.005 to 0.20) corresponding to the relative amount of substitution of S. Generally, this reduction in Li occurs to maintain overall charge neutrality of the structure assuming formal oxidation states.
For example, aliovalent substitution of Si into the Li3BS3 lattice may introduce vacancies which can act as charge carrying defects. Aliovalent substitution for 5% of the B, for example, results in Li2.95B0.95Si0.05S3 which exhibits a room temperature ionic conductivity of 1.82·10−5 S cm−1. In aspects, extended amorphization of the Li2.95B0.95Si0.05S3 further improves the ionic conductivity to between 1·10−3 and 3·10−3 S cm−1. Thus, through aliovalent substitution and amorphization, in aspects, the ionic conductivity of Li3BS3 is improved to near that of conventional liquid electrolytes.
In aspects, doped materials and compositions thereof disclosed herein, such as amorphous Li2.95B0.95Si0.05S3 (a-Li2.95B0.95Si0.05S3), offer many unique advantages over most solid-state electrolyte candidates. The synthesis occurs at a relatively low temperature (˜800° C.) and pelletization can occur at room temperature. In some aspects, unlike most oxide candidates, doped materials and compositions thereof disclosed herein, such as a-Li2.95B0.95Si0.05S3, exhibit superb inter-grain conductivity without the need for a high-temperature grain-boundary sintering step. The precursor materials are also relatively inexpensive. For example, in comparing to LGPS, the a-Li2.95B0.95Si0.05S3 swaps Ge (˜$2400/kg) and P (˜$5/kg) for B (˜200/kg) and Si ($1/kg). Additionally, all four constituent elements have a low atomic mass, which is conducive to making ASSBs with high gravimetric capacity.
In some aspects, materials disclosed herein may be useful in a variety of aspects and applications beyond solid-state electrolytes. For example, the doped or substituted materials and compositions thereof disclosed herein, such as Li2.95B0.95Si0.05S3, may be employed as an artificial interphase on the Li anode surface. In such an application, doped materials and compositions thereof, such as Li2.95B0.95Si0.05S3, may facilitate ionic conduction while preventing the Li anode from reducing an adjacent SSE.
In some aspects, materials disclosed herein may also be employed as an additive for electrodes. In some aspects, materials disclosed herein may also be employed in glass electrolyte mixtures. In such applications, for example, doped materials and compositions thereof, such as Li2.95B0.95Si0.05S3, may serve either or both of the following roles: (1) improving electrochemical stability of the electrode/electrolyte and (2) improving ionic conductivity of the electrode/electrolyte.
The substituted or doped compositions disclosed herein generally have a low concentration or a low relative amount of one or more dopants. Preferably, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is less than or equal to 30 at. %, optionally less than or equal to 28 at. %, optionally less than or equal to 25 at. %, optionally less than or equal to 22 at. %, optionally less than or equal to 21 at. %, optionally less than or equal to 20 at. %, optionally less than or equal to 19 at. %, optionally less than or equal to 18 at. %, optionally less than or equal to 17 at. %, optionally less than or equal to 16 at. %, optionally less than or equal to 15 at. %, optionally less than or equal to 14 at. %, optionally less than or equal to 13 at. %, optionally less than or equal to 12 at. %, optionally less than or equal to 11 at. %, optionally less than or equal to 10 at. %, optionally less than or equal to 9 at. %, optionally less than or equal to 8 at. %, optionally less than or equal to 7 at. %, optionally less than or equal to 6 at. %, optionally less than or equal to 5.0 at. %, optionally less than or equal to 4.5 at. %, optionally less than or equal to 4.0 at. %, optionally less than or equal to 3.5 at. %, optionally less than or equal to 3.0 at. %, optionally less than or equal to 2.7 at. %, optionally less than or equal to 2.5 at. %, optionally less than or equal to 2.3 at. %, optionally less than or equal to 2.1 at. %, optionally less than or equal to 2.0 at. %, optionally less than or equal to 1.9 at. %, optionally less than or equal to 1.7 at. %, optionally less than or equal to 1.5 at. %, optionally less than or equal to 1.3 at. %, optionally less than or equal to 1.0 at. %, optionally less than or equal to 0.8 at. %, optionally less than or equal to 0.7 at. %, optionally less than or equal to 0.6 at. %, optionally less than or equal to 0.5 at. %, optionally less than or equal to 0.4 at. %, optionally less than or equal to 0.3 at. %, optionally less than or equal to 0.2 at. %. Preferably, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is greater than or equal to 0.1 at. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 30 at. % (optionally less than or equal to 28 at. %, optionally less than or equal to 25 at. %, optionally less than or equal to 22 at. %, optionally less than or equal to 21 at. %, optionally less than or equal to 20 at. %, optionally less than or equal to 19 at. %, optionally less than or equal to 18 at. %, optionally less than or equal to 17 at. %, optionally less than or equal to 16 at. %, optionally less than or equal to 15 at. %, optionally less than or equal to 14 at. %, optionally less than or equal to 13 at. %, optionally less than or equal to 12 at. %, optionally less than or equal to 11 at. %, optionally less than or equal to 10 at. %, optionally less than or equal to 9 at. %, optionally less than or equal to 8 at. %, optionally less than or equal to 7 at. %, optionally less than or equal to 6 at. %, optionally less than or equal to 5.0 at. %, optionally less than or equal to 4.5 at. %, optionally less than or equal to 4.0 at. %, optionally less than or equal to 3.5 at. %, optionally less than or equal to 3.0 at. %, optionally less than or equal to 2.7 at. %, optionally less than or equal to 2.5 at. %, optionally less than or equal to 2.3 at. %, optionally less than or equal to 2.1 at. %, optionally less than or equal to 2.0 at. %, optionally less than or equal to 1.9 at. %, optionally less than or equal to 1.7 at. %, optionally less than or equal to 1.5 at. %, optionally less than or equal to 1.3 at. %, optionally less than or equal to 1.0 at. %, optionally less than or equal to 0.8 at. %, optionally less than or equal to 0.7 at. %, optionally less than or equal to 0.6 at. %, optionally less than or equal to 0.5 at. %, optionally less than or equal to 0.4 at. %, optionally less than or equal to 0.3 at. %, optionally less than or equal to 0.2 at. %). Any value and range of total dopant concentration (concentration of the one or more dopants in a composition) between 0.1 at. % and 30 at. % is explicitly contemplated and disclosed herein. For example, optionally a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is selected from the range of 0.1 at. % to 20 at. %, optionally selected from the range of 0.1 at. % to 15 at. %, optionally selected from the range of 0.1 at. % to 10 at. %, optionally selected from the range of 0.1 at. % to 5 at. %, optionally selected from the range of 0.1 at. % to 4.0 at. %, optionally selected from the range of 0.1 at. % to 3.0 at. %, optionally selected from the range of 0.1 at. % to 2.5 at. %, optionally selected from the range of 0.1 at. % to 2.0 at. %, optionally selected from the range of 0.1 at. % to 1.5 at. %, optionally selected from the range of 0.1 at. % to 1.0 at. %, optionally selected from the range of 0.1 at. % to 0.95 at. %.
Preferably for some aspects or applications, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is less than or equal to 30 mol. %, optionally less than or equal to 28 mol. %, optionally less than or equal to 25 mol. %, optionally less than or equal to 22 mol. %, optionally less than or equal to 21 mol. %, optionally less than or equal to 20 mol. %, optionally less than or equal to 19 mol. %, optionally less than or equal to 18 mol. %, optionally less than or equal to 17 mol. %, optionally less than or equal to 16 mol. %, optionally less than or equal to 15 mol. %, optionally less than or equal to 14 mol. %, optionally less than or equal to 13 mol. %, optionally less than or equal to 12 mol. %, optionally less than or equal to 11 mol. %, optionally less than or equal to 10 mol. %, optionally less than or equal to 9 mol. %, optionally less than or equal to 8 mol. %, optionally less than or equal to 7 mol. %, optionally less than or equal to 6 mol. %, optionally less than or equal to 5.0 mol. %, optionally less than or equal to 4.5 mol. %, optionally less than or equal to 4.0 mol. %, optionally less than or equal to 3.5 mol. %, optionally less than or equal to 3.0 mol. %, optionally less than or equal to 2.7 mol. %, optionally less than or equal to 2.5 mol. %, optionally less than or equal to 2.3 mol. %, optionally less than or equal to 2.1 mol. %, optionally less than or equal to 2.0 mol. %, optionally less than or equal to 1.9 mol. %, optionally less than or equal to 1.7 mol. %, optionally less than or equal to 1.5 mol. %, optionally less than or equal to 1.3 mol. %, optionally less than or equal to 1.0 mol. %, optionally less than or equal to 0.8 mol. %, optionally less than or equal to 0.7 mol. %, optionally less than or equal to 0.6 mol. %, optionally less than or equal to 0.5 mol. %, optionally less than or equal to 0.4 mol. %, optionally less than or equal to 0.3 mol. %, optionally less than or equal to 0.2 mol. %. Preferably, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is greater than or equal to 0.1 mol. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 30 mol. % (optionally less than or equal to 28 mol. %, optionally less than or equal to 25 mol. %, optionally less than or equal to 22 mol. %, optionally less than or equal to 21 mol. %, optionally less than or equal to 20 mol. %, optionally less than or equal to 19 mol. %, optionally less than or equal to 18 mol. %, optionally less than or equal to 17 mol. %, optionally less than or equal to 16 mol. %, optionally less than or equal to 15 mol. %, optionally less than or equal to 14 mol. %, optionally less than or equal to 13 mol. %, optionally less than or equal to 12 mol. %, optionally less than or equal to 11 mol. %, optionally less than or equal to 10 mol. %, optionally less than or equal to 9 mol. %, optionally less than or equal to 8 mol. %, optionally less than or equal to 7 mol. %, optionally less than or equal to 6 mol. %, optionally less than or equal to 5.0 mol. %, optionally less than or equal to 4.5 mol. %, optionally less than or equal to 4.0 mol. %, optionally less than or equal to 3.5 mol. %, optionally less than or equal to 3.0 mol. %, optionally less than or equal to 2.7 mol. %, optionally less than or equal to 2.5 mol. %, optionally less than or equal to 2.3 mol. %, optionally less than or equal to 2.1 mol. %, optionally less than or equal to 2.0 mol. %, optionally less than or equal to 1.9 mol. %, optionally less than or equal to 1.7 mol. %, optionally less than or equal to 1.5 mol. %, optionally less than or equal to 1.3 mol. %, optionally less than or equal to 1.0 mol. %, optionally less than or equal to 0.8 mol. %, optionally less than or equal to 0.7 mol. %, optionally less than or equal to 0.6 mol. %, optionally less than or equal to 0.5 mol. %, optionally less than or equal to 0.4 mol. %, optionally less than or equal to 0.3 mol. %, optionally less than or equal to 0.2 mol. %). Any value and range of total dopant concentration (concentration of the one or more dopants in a composition) between 0.1 mol. % and 30 mol. % is explicitly contemplated and disclosed herein. For example, optionally a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is selected from the range of 0.1 mol. % to 20 mol. %, optionally selected from the range of 0.1 mol. % to 15 mol. %, optionally selected from the range of 0.1 mol. % to 10 mol. %, optionally selected from the range of 0.1 mol. % to 5 mol. %, optionally selected from the range of 0.1 mol. % to 4.0 mol. %, optionally selected from the range of 0.1 mol. % to 3.0 mol. %, optionally selected from the range of 0.1 mol. % to 2.5 mol. %, optionally selected from the range of 0.1 mol. % to 2.0 mol. %, optionally selected from the range of 0.1 mol. % to 1.5 mol. %, optionally selected from the range of 0.1 mol. % to 1.0 mol. %, optionally selected from the range of 0.1 mol. % to 0.95 mol. %.
Preferably for some aspects or applications, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is less than or equal to 30 wt. %, optionally less than or equal to 28 wt. %, optionally less than or equal to 25 wt. %, optionally less than or equal to 22 wt. %, optionally less than or equal to 21 wt. %, optionally less than or equal to 20 wt. %, optionally less than or equal to 19 wt. %, optionally less than or equal to 18 wt. %, optionally less than or equal to 17 wt. %, optionally less than or equal to 16 wt. %, optionally less than or equal to 15 wt. %, optionally less than or equal to 14 wt. %, optionally less than or equal to 13 wt. %, optionally less than or equal to 12 wt. %, optionally less than or equal to 11 wt. %, optionally less than or equal to 10 wt. %, optionally less than or equal to 9 wt. %, optionally less than or equal to 8 wt. %, optionally less than or equal to 7 wt. %, optionally less than or equal to 6 wt. %, optionally less than or equal to 5.0 wt. %, optionally less than or equal to 4.5 wt. %, optionally less than or equal to 4.0 wt. %, optionally less than or equal to 3.5 wt. %, optionally less than or equal to 3.0 wt. %, optionally less than or equal to 2.7 wt. %, optionally less than or equal to 2.5 wt. %, optionally less than or equal to 2.3 wt. %, optionally less than or equal to 2.1 wt. %, optionally less than or equal to 2.0 wt. %, optionally less than or equal to 1.9 wt. %, optionally less than or equal to 1.7 wt. %, optionally less than or equal to 1.5 wt. %, optionally less than or equal to 1.3 wt. %, optionally less than or equal to 1.0 wt. %, optionally less than or equal to 0.8 wt. %, optionally less than or equal to 0.7 wt. %, optionally less than or equal to 0.6 wt. %, optionally less than or equal to 0.5 wt. %, optionally less than or equal to 0.4 wt. %, optionally less than or equal to 0.3 wt. %, optionally less than or equal to 0.2 wt. %. Preferably, a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is greater than or equal to 0.1 wt. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 30 wt. % (optionally less than or equal to 28 wt. %, optionally less than or equal to 25 wt. %, optionally less than or equal to 22 wt. %, optionally less than or equal to 21 wt. %, optionally less than or equal to 20 wt. %, optionally less than or equal to 19 wt. %, optionally less than or equal to 18 wt. %, optionally less than or equal to 17 wt. %, optionally less than or equal to 16 wt. %, optionally less than or equal to 15 wt. %, optionally less than or equal to 14 wt. %, optionally less than or equal to 13 wt. %, optionally less than or equal to 12 wt. %, optionally less than or equal to 11 wt. %, optionally less than or equal to 10 wt. %, optionally less than or equal to 9 wt. %, optionally less than or equal to 8 wt. %, optionally less than or equal to 7 wt. %, optionally less than or equal to 6 wt. %, optionally less than or equal to 5.0 wt. %, optionally less than or equal to 4.5 wt. %, optionally less than or equal to 4.0 wt. %, optionally less than or equal to 3.5 wt. %, optionally less than or equal to 3.0 wt. %, optionally less than or equal to 2.7 wt. %, optionally less than or equal to 2.5 wt. %, optionally less than or equal to 2.3 wt. %, optionally less than or equal to 2.1 wt. %, optionally less than or equal to 2.0 wt. %, optionally less than or equal to 1.9 wt. %, optionally less than or equal to 1.7 wt. %, optionally less than or equal to 1.5 wt. %, optionally less than or equal to 1.3 wt. %, optionally less than or equal to 1.0 wt. %, optionally less than or equal to 0.8 wt. %, optionally less than or equal to 0.7 wt. %, optionally less than or equal to 0.6 wt. %, optionally less than or equal to 0.5 wt. %, optionally less than or equal to 0.4 wt. %, optionally less than or equal to 0.3 wt. %, optionally less than or equal to 0.2 wt. %). Any value and range of total dopant concentration (concentration of the one or more dopants in a composition) between 0.1 wt. % and 30 wt. % is explicitly contemplated and disclosed herein. For example, optionally a total dopant concentration (concentration of the one or more dopants in a composition) in a material or composition thereof is selected from the range of 0.1 wt. % to 20 wt. %, optionally selected from the range of 0.1 wt. % to 15 wt. %, optionally selected from the range of 0.1 wt. % to 10 wt. %, optionally selected from the range of 0.1 wt. % to 5 wt. %, optionally selected from the range of 0.1 wt. % to 4.0 wt. %, optionally selected from the range of 0.1 wt. % to 3.0 wt. %, optionally selected from the range of 0.1 wt. % to 2.5 wt. %, optionally selected from the range of 0.1 wt. % to 2.0 wt. %, optionally selected from the range of 0.1 wt. % to 1.5 wt. %, optionally selected from the range of 0.1 wt. % to 1.0 wt. %, optionally selected from the range of 0.1 wt. % to 0.95 wt. %.
The substituted or doped compositions disclosed herein generally have only a small amount of a principal element substituted for or replaced with a dopant (the dopant being one or more elements aliovalent with respect to the substituted or replaced principal element). Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is less than or equal to 20 at. %, optionally less than or equal to 19 at. %, optionally less than or equal to 18 at. %, optionally less than or equal to 17 at. %, optionally less than or equal to 16 at. %, optionally less than or equal to 15 at. %, optionally less than or equal to 14 at. %, optionally less than or equal to 13 at. %, optionally less than or equal to 12 at. %, optionally less than or equal to 11 at. %, optionally less than or equal to 10 at. %, optionally less than or equal to 9 at. %, optionally less than or equal to 8 at. %, optionally less than or equal to 7 at. %, optionally less than or equal to 6 at. %, optionally less than or equal to 5.0 at. %, optionally less than or equal to 4.5 at. %, optionally less than or equal to 4.0 at. %, optionally less than or equal to 3.5 at. %, optionally less than or equal to 3.0 at. %, optionally less than or equal to 2.7 at. %, optionally less than or equal to 2.5 at. %, optionally less than or equal to 2.3 at. %, optionally less than or equal to 2.1 at. %, optionally less than or equal to 2.0 at. %, optionally less than or equal to 1.9 at. %, optionally less than or equal to 1.7 at. %, optionally less than or equal to 1.5 at. %, optionally less than or equal to 1.3 at. %, optionally less than or equal to 1.0 at. %, optionally less than or equal to 0.8 at. %, optionally less than or equal to 0.7 at. %, optionally less than or equal to 0.6 at. %, optionally less than or equal to 0.5 at. %, optionally less than or equal to 0.4 at. %, optionally less than or equal to 0.3 at. %, optionally less than or equal to 0.2 at. %. Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is greater than or equal to 0.1 at. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 20 at. % (optionally less than or equal to 19 at. %, optionally less than or equal to 18 at. %, optionally less than or equal to 17 at. %, optionally less than or equal to 16 at. %, optionally less than or equal to 15 at. %, optionally less than or equal to 14 at. %, optionally less than or equal to 13 at. %, optionally less than or equal to 12 at. %, optionally less than or equal to 11 at. %, optionally less than or equal to 10 at. %, optionally less than or equal to 9 at. %, optionally less than or equal to 8 at. %, optionally less than or equal to 7 at. %, optionally less than or equal to 6 at. %, optionally less than or equal to 5.0 at. %, optionally less than or equal to 4.5 at. %, optionally less than or equal to 4.0 at. %, optionally less than or equal to 3.5 at. %, optionally less than or equal to 3.0 at. %, optionally less than or equal to 2.7 at. %, optionally less than or equal to 2.5 at. %, optionally less than or equal to 2.3 at. %, optionally less than or equal to 2.1 at. %, optionally less than or equal to 2.0 at. %, optionally less than or equal to 1.9 at. %, optionally less than or equal to 1.7 at. %, optionally less than or equal to 1.5 at. %, optionally less than or equal to 1.3 at. %, optionally less than or equal to 1.0 at. %, optionally less than or equal to 0.8 at. %, optionally less than or equal to 0.7 at. %, optionally less than or equal to 0.6 at. %, optionally less than or equal to 0.5 at. %, optionally less than or equal to 0.4 at. %, optionally less than or equal to 0.3 at. %, optionally less than or equal to 0.2 at. %). Any value and range of the relative amount, of any principal element of a composition that is substituted with a dopant, between 0.1 at. % and 20 at. % is explicitly contemplated and disclosed herein. For example, optionally, the relative amount of any principal element of a composition that is substituted with a dopant is selected from the range of 0.1 at. % to 20 at. %, optionally selected from the range of 0.1 at. % to 15 at. %, optionally selected from the range of 0.1 at. % to 10 at. %, optionally selected from the range of 0.1 at. % to 5 at. %, optionally selected from the range of 0.1 at. % to 4.0 at. %, optionally selected from the range of 0.1 at. % to 3.0 at. %, optionally selected from the range of 0.1 at. % to 2.5 at. %, optionally selected from the range of 0.1 at. % to 2.0 at. %, optionally selected from the range of 0.1 at. % to 1.5 at. %, optionally selected from the range of 0.1 at. % to 1.0 at. %, optionally selected from the range of 0.1 at. % to 0.95 at. %.
Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is less than or equal to 20 mol. %, optionally less than or equal to 19 mol. %, optionally less than or equal to 18 mol. %, optionally less than or equal to 17 mol. %, optionally less than or equal to 16 mol. %, optionally less than or equal to 15 mol. %, optionally less than or equal to 14 mol. %, optionally less than or equal to 13 mol. %, optionally less than or equal to 12 mol. %, optionally less than or equal to 11 mol. %, optionally less than or equal to 10 mol. %, optionally less than or equal to 9 mol. %, optionally less than or equal to 8 mol. %, optionally less than or equal to 7 mol. %, optionally less than or equal to 6 mol. %, optionally less than or equal to 5.0 mol. %, optionally less than or equal to 4.5 mol. %, optionally less than or equal to 4.0 mol. %, optionally less than or equal to 3.5 mol. %, optionally less than or equal to 3.0 mol. %, optionally less than or equal to 2.7 mol. %, optionally less than or equal to 2.5 mol. %, optionally less than or equal to 2.3 mol. %, optionally less than or equal to 2.1 mol. %, optionally less than or equal to 2.0 mol. %, optionally less than or equal to 1.9 mol. %, optionally less than or equal to 1.7 mol. %, optionally less than or equal to 1.5 mol. %, optionally less than or equal to 1.3 mol. %, optionally less than or equal to 1.0 mol. %, optionally less than or equal to 0.8 mol. %, optionally less than or equal to 0.7 mol. %, optionally less than or equal to 0.6 mol. %, optionally less than or equal to 0.5 mol. %, optionally less than or equal to 0.4 mol. %, optionally less than or equal to 0.3 mol. %, optionally less than or equal to 0.2 mol. %. Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is greater than or equal to 0.1 mol. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 20 mol. % (optionally less than or equal to 19 mol. %, optionally less than or equal to 18 mol. %, optionally less than or equal to 17 mol. %, optionally less than or equal to 16 mol. %, optionally less than or equal to 15 mol. %, optionally less than or equal to 14 mol. %, optionally less than or equal to 13 mol. %, optionally less than or equal to 12 mol. %, optionally less than or equal to 11 mol. %, optionally less than or equal to 10 mol. %, optionally less than or equal to 9 mol. %, optionally less than or equal to 8 mol. %, optionally less than or equal to 7 mol. %, optionally less than or equal to 6 mol. %, optionally less than or equal to 5.0 mol. %, optionally less than or equal to 4.5 mol. %, optionally less than or equal to 4.0 mol. %, optionally less than or equal to 3.5 mol. %, optionally less than or equal to 3.0 mol. %, optionally less than or equal to 2.7 mol. %, optionally less than or equal to 2.5 mol. %, optionally less than or equal to 2.3 mol. %, optionally less than or equal to 2.1 mol. %, optionally less than or equal to 2.0 mol. %, optionally less than or equal to 1.9 mol. %, optionally less than or equal to 1.7 mol. %, optionally less than or equal to 1.5 mol. %, optionally less than or equal to 1.3 mol. %, optionally less than or equal to 1.0 mol. %, optionally less than or equal to 0.8 mol. %, optionally less than or equal to 0.7 mol. %, optionally less than or equal to 0.6 mol. %, optionally less than or equal to 0.5 mol. %, optionally less than or equal to 0.4 mol. %, optionally less than or equal to 0.3 mol. %, optionally less than or equal to 0.2 mol. %). Any value and range of the relative amount, of any principal element of a composition that is substituted with a dopant, between 0.1 mol. % and 20 mol. % is explicitly contemplated and disclosed herein. For example, optionally, the relative amount of any principal element of a composition that is substituted with a dopant is selected from the range of 0.1 mol. % to 20 mol. %, optionally selected from the range of 0.1 mol. % to 15 mol. %, optionally selected from the range of 0.1 mol. % to 10 mol. %, optionally selected from the range of 0.1 mol. % to 5 mol. %, optionally selected from the range of 0.1 mol. % to 4.0 mol. %, optionally selected from the range of 0.1 mol. % to 3.0 mol. %, optionally selected from the range of 0.1 mol. % to 2.5 mol. %, optionally selected from the range of 0.1 mol. % to 2.0 mol. %, optionally selected from the range of 0.1 mol. % to 1.5 mol. %, optionally selected from the range of 0.1 mol. % to 1.0 mol. %, optionally selected from the range of 0.1 mol. % to 0.95 mol. %.
Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is less than or equal to 20 wt. %, optionally less than or equal to 19 wt. %, optionally less than or equal to 18 wt. %, optionally less than or equal to 17 wt. %, optionally less than or equal to 16 wt. %, optionally less than or equal to 15 wt. %, optionally less than or equal to 14 wt. %, optionally less than or equal to 13 wt. %, optionally less than or equal to 12 wt. %, optionally less than or equal to 11 wt. %, optionally less than or equal to 10 wt. %, optionally less than or equal to 9 wt. %, optionally less than or equal to 8 wt. %, optionally less than or equal to 7 wt. %, optionally less than or equal to 6 wt. %, optionally less than or equal to 5.0 wt. %, optionally less than or equal to 4.5 wt. %, optionally less than or equal to 4.0 wt. %, optionally less than or equal to 3.5 wt. %, optionally less than or equal to 3.0 wt. %, optionally less than or equal to 2.7 wt. %, optionally less than or equal to 2.5 wt. %, optionally less than or equal to 2.3 wt. %, optionally less than or equal to 2.1 wt. %, optionally less than or equal to 2.0 wt. %, optionally less than or equal to 1.9 wt. %, optionally less than or equal to 1.7 wt. %, optionally less than or equal to 1.5 wt. %, optionally less than or equal to 1.3 wt. %, optionally less than or equal to 1.0 wt. %, optionally less than or equal to 0.8 wt. %, optionally less than or equal to 0.7 wt. %, optionally less than or equal to 0.6 wt. %, optionally less than or equal to 0.5 wt. %, optionally less than or equal to 0.4 wt. %, optionally less than or equal to 0.3 wt. %, optionally less than or equal to 0.2 wt. %. Optionally, the relative amount of any principal element of a composition that is substituted with a dopant is greater than or equal to 0.1 wt. % (optionally greater than or equal to 0.12%, optionally greater than or equal to 0.14%, optionally greater than or equal to 0.15%, optionally greater than or equal to 0.16%, optionally greater than or equal to 0.17%, optionally greater than or equal to 0.19%, optionally greater than or equal to 0.20%, optionally greater than or equal to 0.21%, optionally greater than or equal to 0.22%, optionally greater than or equal to 0.23%, optionally greater than or equal to 0.25%, optionally greater than or equal to 0.27%, optionally greater than or equal to 0.29%, optionally greater than or equal to 0.30%, optionally greater than or equal to 0.35%, optionally greater than or equal to 0.40%, optionally greater than or equal to 0.45%, optionally greater than or equal to 0.50%, optionally greater than or equal to 0.55%, optionally greater than or equal to 0.65%, optionally greater than or equal to 0.70%, optionally greater than or equal to 0.75%, optionally greater than or equal to 0.80%, optionally greater than or equal to 0.85%, optionally greater than or equal to 0.95%, optionally greater than or equal to 1.0%, optionally greater than or equal to 1.2%, optionally greater than or equal to 1.5%, optionally greater than or equal to 1.7%, optionally greater than or equal to 2.0%, optionally greater than or equal to 2.2%, optionally greater than or equal to 2.5%) and less than or equal to 20 wt. % (optionally less than or equal to 19 wt. %, optionally less than or equal to 18 wt. %, optionally less than or equal to 17 wt. %, optionally less than or equal to 16 wt. %, optionally less than or equal to 15 wt. %, optionally less than or equal to 14 wt. %, optionally less than or equal to 13 wt. %, optionally less than or equal to 12 wt. %, optionally less than or equal to 11 wt. %, optionally less than or equal to 10 wt. %, optionally less than or equal to 9 wt. %, optionally less than or equal to 8 wt. %, optionally less than or equal to 7 wt. %, optionally less than or equal to 6 wt. %, optionally less than or equal to 5.0 wt. %, optionally less than or equal to 4.5 wt. %, optionally less than or equal to 4.0 wt. %, optionally less than or equal to 3.5 wt. %, optionally less than or equal to 3.0 wt. %, optionally less than or equal to 2.7 wt. %, optionally less than or equal to 2.5 wt. %, optionally less than or equal to 2.3 wt. %, optionally less than or equal to 2.1 wt. %, optionally less than or equal to 2.0 wt. %, optionally less than or equal to 1.9 wt. %, optionally less than or equal to 1.7 wt. %, optionally less than or equal to 1.5 wt. %, optionally less than or equal to 1.3 wt. %, optionally less than or equal to 1.0 wt. %, optionally less than or equal to 0.8 wt. %, optionally less than or equal to 0.7 wt. %, optionally less than or equal to 0.6 wt. %, optionally less than or equal to 0.5 wt. %, optionally less than or equal to 0.4 wt. %, optionally less than or equal to 0.3 wt. %, optionally less than or equal to 0.2 wt. %). Any value and range of the relative amount, of any principal element of a composition that is substituted with a dopant, between 0.1 wt. % and 20 wt. % is explicitly contemplated and disclosed herein. For example, optionally, the relative amount of any principal element of a composition that is substituted with a dopant is selected from the range of 0.1 wt. % to 20 wt. %, optionally selected from the range of 0.1 wt. % to 15 wt. %, optionally selected from the range of 0.1 wt. % to 10 wt. %, optionally selected from the range of 0.1 wt. % to 5 wt. %, optionally selected from the range of 0.1 wt. % to 4.0 wt. %, optionally selected from the range of 0.1 wt. % to 3.0 wt. %, optionally selected from the range of 0.1 wt. % to 2.5 wt. %, optionally selected from the range of 0.1 wt. % to 2.0 wt. %, optionally selected from the range of 0.1 wt. % to 1.5 wt. %, optionally selected from the range of 0.1 wt. % to 1.0 wt. %, optionally selected from the range of 0.1 wt. % to 0.95 wt. %.
Various aspects are contemplated and disclosed herein, several of which are set forth in the paragraphs below. It is explicitly contemplated and disclosed that any aspect or portion thereof can be combined to form an aspect. In addition, it is explicitly contemplated and disclosed that: any reference to Aspect 1 includes reference to Aspects 1a, 1b, 1c, 1d, 1e, 1f, 1g, 1h, 1i, 1j, 1k, and/or 11, and any combination thereof; any reference to Aspect 3 includes reference to Aspects 3a, 3b, and/or 3c; and so on (i.e., any reference to an aspect includes reference to that aspect's lettered versions). Moreover, the terms “any preceding aspect” and “any one of the preceding aspects” means any aspect that appears prior to the aspect that contains such phrase (for example, the sentence “Aspect 15: The material, device, electrolyte, or method of any preceding Aspect . . . ” means that any Aspect prior to Aspect 15 is referenced, including letter versions, including aspects 1a through 14b). For example, it is contemplated and disclosed that, optionally, any composition, method, or formulation of any the below aspects may be useful with or combined with any other aspect provided below. Further, for example, it is contemplated and disclosed that any embodiment or aspect described above may, optionally, be combined with any of the below listed aspects.
Li3−z[B+Q]1[S+G]3 (FX1);
Li3−z[B+Q]1[S+G]3 (FX1);
Li3−z[B+Q]1[S+G]3 (FX1);
Li3−z[B+Q]1[S+G]3 (FX1);
The material, device, electrolyte, or method of Aspect 1, wherein z is greater than 0 and less than 0.40 (optionally less than or equal to 0.35, optionally less than or equal to 0.30, optionally less than or equal to 0.25, optionally less than or equal to 0.20, optionally less than or equal to 0.18, optionally less than or equal to 0.16, optionally less than or equal to 0.15, optionally less than or equal to 0.13, optionally less than or equal to 0.11, optionally less than or equal to 0.10, optionally less than or equal to 0.09, optionally less than or equal to 0.08, optionally less than or equal to 0.07, optionally less than or equal to 0.06, optionally less than or equal to 0.05, optionally less than or equal to 0.04, optionally less than or equal to 0.03, optionally less than or equal to 0.025).
Li3−x−yB1−x[Q]xS3−y[G]y (FX2);
Li3−xB1−x[Q]xS3 (FX3);
Li3−yB1S3−y[G]y (FX4); wherein:
wherein:
Li3−z[B+Q]1[S+G]3 (FX1);
Li3−z[B+Q]1[S+G]3 (FX1);
Li3−z[B+Q]1[S+G]3 (FX1);
Li3VS4 (FX5);
Na3Li3Al2F12 (FX6);
Li2Te (FX7);
LiAlTe2 (FX8);
LiInTe2 (FX9);
Li6MnS4 (FX10);
LiGaTe2 (FX11);
KLi6TaO6 (FX12); or
Li3CuS2 (FX13).
Li3VS4 (FX5);
Na3Li3Al2F12 (FX6);
Li2Te (FX7);
LiAlTe2 (FX8);
LiInTe2 (FX9);
Li6MnS4 (FX10);
LiGaTe2 (FX11);
KLi6TaO6 (FX12); or
Li3CuS2 (FX13).
The invention can be further understood by the following non-limiting examples.
Overview of Examples 1-3: Despite ongoing efforts to identify high-performance electrolytes for solid-state Li-ion batteries, thousands of prospective Li-containing structures remain unexplored. Here, we employ a semi-supervised learning approach to expedite identification of superionic conductors. We screen 180 unique descriptor representations and use agglomerative clustering to cluster ˜26,000 Li-containing structures. The clusters are then labeled with experimental ionic conductivity data to assess the fitness of the descriptors. By inspecting clusters containing the highest conductivity labels, we identify 212 promising structures that are further screened using bond valence site energy and nudged elastic band calculations. Li3BS3 is identified as a potential high-conductivity material and selected for experimental characterization. With sufficient defect engineering, we show that Li3BS3 is a superionic conductor with room temperature ionic conductivity greater than 1 mS cm−1. While the semi-supervised method shows promise for identification of superionic conductors, the results illustrate a continued need for descriptors that explicitly encode for defects.
Identifying new materials that could improve solid-state ion battery prospects is an ongoing challenge. The search for an ideal solid-state Li electrolyte is a prime example. Research has focused on eight classes of materials: LISICON-type structures, argyrodites, garnets, NASICON-type structures, Li-nitrides, Li-hydrides, perovskites, and Li-halides1. However, only three compounds with near-liquid-electrolyte conductivity (˜10−2 S cm−1) have been discovered: Li10GeP2S12 (LGPS)2, Li6PS5Br argyrodite3, and Li7P3S11 ceramic-glass1,4. Although promising discoveries, all three high-conductivity structures are unstable against the Li anode5-10. While investigations to limit instability are ongoing11,12, identification of additional superionic structures is desirable. Discovery of new structures that support superionic conductivity improves the odds of identifying or engineering a stable electrode|SSE interface. For example, engineering solutions that fail to stabilize the Li|argyrodite interface may prove more successful when applied to not-yet-discovered superionic conductors. Discovery of new superionic conductors may also enable stable architectures via multi-electrolyte approaches which have been proposed as more promising than single-electrolyte architectures for achieving stability against Li metal and cathode materials13. High-performing structures that enable new battery chemistries may exist outside of the eight classes. However, exploration under the traditional Edisonian approach prioritizes small perturbations to well-known variable spaces.
Machine learning (ML) is a promising tool for expediting the discovery of useful solid-state materials. By describing prospective materials with physically meaningful descriptors, ML models can identify high-dimensional patterns in large datasets that are not readily apparent14-20. Ongoing descriptor engineering21-26 has enabled discovery of battery components27,28, electrocatalysts15,29, photovoltaic components16,30, piezoelectrics31, new metallic glasses14 and new alloys32. However, application of ML for discovery of SSEs and other emerging technologies can be challenging. Supervised ML approaches require empirical data for use as “labels”. For example, graph neural network (GNN) approaches have been successful in many domains but generally require thousands to tens of thousands of labels to avoid overfitting33. By contrast, relatively few SSEs have been experimentally characterized compared to the ˜26,000 known Li-containing structures19,34-36. Characterized materials often exhibit ill-defined properties owing to the variety of synthetic approaches and non-standardized testing methods37. Well-performing materials often contain charge-carrying defects that are not explicitly characterized or reported38. Negative examples, i.e. materials with undesirable properties, are useful for ML models but are seldom reported.
Semi-supervised ML can guide synthetic prioritization of SSEs by overcoming the issues associated with label scarcity. Supervised ML requires labels because it infers correlation functions by mapping the input descriptors to the labels39. Semi-supervised ML prioritizes comparison of descriptors to identify relationships between the descriptors in a dataset36,39. The input compositions are clustered (or grouped) by comparison of descriptors using a similarity metric. The clustering process does not consider labels, and thus circumvents the need for abundant labels. The resultant clusters can be labeled ex post facto to examine correlation between the descriptor and a physical property of interest. For semi-supervised ML, ideal descriptors result in a set of clusters where each cluster has similar labels and thus the label variance is minimized. Promising synthetic targets may then be identified by their membership in clusters that contain desirable labels.
A key insight of this work is that semi-supervised ML can be used to rank descriptors in terms of their correlation to physical properties of interest. Descriptors are representations of the input materials that encode the chemistry, composition, structure, and/or other system properties. An ideal descriptor should be a unique representation, a continuous function of the structure, exhibit rotational/translational invariance, and be readily comparable across all structures in the dataset24-26. Recently, Zhang et al. demonstrated that a modified X-Ray diffraction (mXRD) descriptor lead to favorable clustering for Li SSEs34. By labeling the resultant clusters with experimental room-temperature Li-ion conductivities, they identified 16 prospective fast-ion conductors. However, an ideal descriptor is not known a priori, and no comprehensive descriptor screening has yet been pursued for correlation with SSE properties. Descriptor screening is desirable for both experimentalists and computationalists. For experimentalists, ranking of descriptors affords insight into what aspects of materials are most correlated with target properties. For computationalists, descriptors rankings enable improved regression and supervised learning models by guiding the selection of input representation(s). Descriptor transformations for inorganic structures have been curated in a variety of software packages, including: Matminer24, Dscribe25, SchNet40, and Aenet41
Herein, we employ hierarchical agglomerative clustering to screen many descriptors, without assuming correlation to ionic conductivity. The performance of 20 descriptors is assessed for semi-supervised identification of Li SSEs. Each descriptor is paired with 9 structural simplification strategies, yielding a total of 180 unique representations per input structure. The approach is applied to a dataset of ˜26,000 Li-containing phases, encompassing all Li-containing structures contained in the Inorganic Crystal Structure Database (ICSD—v.4.4.0) and the Materials Project (MP—v.2020.09.08) database (
Using the descriptors, the semi-supervised approach can identify potential fast solid-state Li-ion conductors. By selecting structures in clusters containing high conductivity labels, the ˜26,000 input structures are down selected to just 212 promising structures. Practical considerations, a semi-empirical bond valence site energy (BVSE) method,42 and the Nudged Elastic Band (NEB) method are employed to rank the structures. From the ten highest ranking structures, Li3BS3 is selected for model validation. Synthesis of pure Li3BS3 yields a poor conductor. However, by employing defect engineering strategies we demonstrate that Li3BS3 is a superionic conductor with an ionic conductivity greater than 10−3 S cm−1.
Screening Simplification-Descriptor Combinations:
A set of 20 descriptors is selected for screening the semi-supervised learning approach (Table 1). The descriptors generally encode four types of information: the spatial environment, the chemical bonding environment, the electronic environment, and composition. All descriptors are implemented in Python using the Matminer24 or Dscribe25 libraries. The code is published to a github repository and is available for download (https://github.com/FALL-ML/materials-discovery). Zhang et al. illustrated that structure simplification prior to learning can produce lower variance outcomes34. Their mXRD descriptor was found to work best with removal of all cations, all the anions replaced by a single representative anion, and the structure volume scaled to 40 Å3 per anion. Inspired by the previous success in using structure simplification, we screen eight structure simplifications in addition to the unperturbed structure. For simplifications the following categories of atoms are replaced with a representative specie: (1) Cations are represented as Al, (2) Anions are represented as S, (3) Mobile ions are represented as Li, and (4) Neutral atoms are represented as Mg. Categories of atom are removed as to yield the four simplifications: CAMN (all atoms retained), CAN (mobile ions removed), AM (cations and neutral atoms removed), and A (only anions retained). Four additional simplifications are formed by scaling each lattice volume to 40 Å3 per anion: CAMN-40, CAN-40, AM-40, and A-40.
Agglomerative clustering is performed on all Li-containing structures from the ICSD and MP repositories. Agglomerative clustering is a “bottom-up” approach to clustering where each structure starts in its own cluster of one. Clusters are merged according to Ward's Minimum Variance criterion in Euclidean space, which minimizes the global descriptor variance57.
where nC is the number of clusters in a set, Ck is cluster k, di is a descriptor representation for structure i, and
An ideal simplification-descriptor combination results in clustering where each cluster contains labels with similar σRT values. Ward's minimum variance method is applied to the conductivity labels as a measure of clustering efficacy:
where nC is the number of clusters in a set, Ck is cluster k, and
Using σ25° C. labels, the best semi-supervised ML performance is attained when using the SOAP descriptor. SOAP is a spatial descriptor that employs smeared gaussians to represent atomic positions for each crystal structure25. Predictions using the SOAP descriptor have exhibited similar performance to state-of-the-art graph neural networks (GCNs) on a variety of materials science datasets58. Optimization of SOAP hyper-parameters (radial cutoff, number of radial basis functions, degree of spherical harmonics) is explored in section VI of the supplemental information. SOAP is found to perform best when combined with the CAN structure simplification. That is, the simplification where the mobile Li atoms are removed, and the remaining atoms are simplified into three representative species: cations, anions, and neutral atoms. SOAP outperforms all other descriptors for all depths of clustering. The SOAP descriptor can be modestly improved (2-3% decrease in WQ) by mixing with other descriptors to make a 2nd order SOAP descriptor (see Example 1B: section VI).
Semi-Supervised Identification of Prospective Li-Ion Conductors:
Agglomerative clustering with the 2nd order SOAP descriptor is used to identify prospective ionic conductors. Wσ minimization is prioritized over WEa minimization because Ea alone is not necessarily a good predictor of conductivity; σ25° C. may be affected by properties including the ionic carrier concentration, hopping attempt frequency, and the presence of concerted migration modes59. The agglomerative dendrogram for the 2nd order SOAP is shown in
Candidates for next-generation SSEs can be identified by evaluating clusters that either contain or are near high conductivity labels. Clusters #2, #4, and #7 are promising because they account for 85% of the high σ25° C. labels. However, targeting these clusters would necessitate screening thousands of structures. Instead, we search from the 241st cluster depth, targeting all clusters that contain or are directly adjacent (i.e. the nearest cluster in the Euclidean feature space) to high σ25° C. labels. The promising structures are further screened using calculated stability (E vs. Ehull) and band gap (Eg) properties from the Materials Project, and the BVSE Ea values. We select the structures that have (1) an Ehull of 70 meV or lower,60 (2) an Eg of at least 1 eV, and (3) a BVSE-calculated Ea below a conservative 0.6 eV. We note that while a true Eg value of 1 eV would be problematic for an SSE, the bandgaps reported on Materials Project are typically underestimated by about 40%61. The approach identifies 212 structures as prospective ionic conductors. Climbing image nudged elastic band (CI-NEB) is employed to calculate the Ea for Li-ion hopping on the ten materials with the lowest BVSE-calculated Ea and an Ehull of 0 eV. The CI-NEB functionals and parameters can be found in the supporting information section VII. The top 10 prospective structures are tabulated in Table 2.
The CI-NEB calculations generally agree with the BVSE calculated Ea values, suggesting favorable activation energies (<500 meV). Discrepancies between the two values may arise because BVSE does not allow framework ions to relax during Li+ migration and does not account for repulsive interactions between atoms of the mobile ion species. BVSE also does not capture cooperative conduction mechanisms or those involving the so-called paddlewheel effect. Despite these limitations, we note that the model identifies numerous diverse structures beyond those routinely explored. Table 1 includes four tellurides, a vanadium sulfide, and multiple transition-metal-containing structures. Of the structures in Table 1, 70% avoid the space groups for the best-performing SSEs discovered to date: LPS (62), LGPS (137), the argyrodites (216), and LLZO (230).
Data Processing and Semi-Supervised Learning:
The ˜26,000 input compositions are exported from the Inorganic Crystalline Structure Database (ICSD v. 4.4.0) and Material's Project (MP—v.2020.09.08) as crystallographic information files (.cif). All structures containing Li are imported. Although transition metals could produce undesirable redox activity, transition metal containing structures are not screened out. Some of the best-performing SSEs contain transition metals (e.g. LLZO and LLTO). Entries that existed in both ICSD and MP are merged. Data manipulations and structure simplifications are performed using the Python libraries NumPy (v1.19.1), Pandas (v1.0.5), ASE (v3.19.1), and Pymatgen (v2020.8.3). Descriptor transformations are performed using the Python libraries Pymatgen (v2020.8.3), Matminer (v0.6.3), and Dscribe. Agglomerative hierarchical clustering is performed using the Python library scipy (v1.5.0). All code has been successfully executed on a custom-built CPU with an AMD Ryzen Threadripper 3990x Processor and 256 GB of RAM, in Ubuntu 20.04 running on Windows Subsystem for Linux 2. All code is made available on the github (https://github.com/FALL-ML/materials-discovery).
CI-NEB:
Migration barriers for Li ion hopping are evaluated with the Climbing Image—Nudged Elastic Band (CI-NEB) method as implemented in the QuantumESPRESSO PWneb software package81-84. Density-functional theory (DFT) calculations are performed using the Perdew-Burke-Ernzerfof (PBE) generalized gradient approximation functional and projector-augmented wave (PAW) sets85,86. Convergence testing for the kinetic-energy cutoff of the plane-wave basis and the k-point sampling is performed for each structure to ensure an accuracy of 1 meV per atom. The lattice parameters and atomic positions of the as-retrieved structure are optimized. Supercells are created for each structure that are a minimum of 10 Å in each lattice direction to minimize interactions between periodic images of the mobile ion. To study the migration barrier in the dilute limit, a single Li vacancy is created in the boundary endpoint structures of each studied pathway. A uniform background charge is used to balance excess charge. Each boundary configuration is relaxed until the force on each atom is less than 3×10−4 eV/Å. Images are created by linearly interpolating framework atomic positions between the initial and final boundary configurations. The initial pathway for the mobile ion is generated from the BVSE output minimum energy pathway to promote faster convergence of the NEB calculation. An NEB force convergence threshold of 0.05 eV/Å is used. The calculation is first converged using the default NEB algorithm and then restarted with the CI scheme to allow for the maximum energy of the pathway to be determined.
Section I. Digitized Labels for Lithium-Ion Conductors: RT Conductivity, Activation Energy, and Corresponding ICSD Identifier
Data labels for the semi-supervised learning approach were ultimately digitized from over 300 literature publications. Many more publications were initially examined. The stepwise decision chart below was used as a guide for deciding what data to digitize. Room temperature conductivity data was only digitized if it originated from an equivalent circuit fit (where a blocking feature was clearly present) or if calculated from NMR. DC techniques were categorically discounted because they cannot differentiate between electronic and ionic conductivity.
All of the digitized data is presented in the subsequent tableI. Activation energies were also digitized when available. The activation energies were not used in the manuscript but are still presented here to aid future machine learning endeavors. The digitized data was manually matched with the appropriate ICSD ID, so that the crystallographic information file (.cif) can be downloaded.
<1E−8
Section II. Labels for Comparing all Descriptors-Simplification Combinations
A subset of the digitized labels was used for comparing between the different semi-supervised learning models. In total, the label subset is comprised of 155 structures. The subset is required because not all structures are compatible with all the descriptor transformations. Some descriptor-structure combinations produce coding errors, imaginary values, or infinite values. To directly compare all the descriptors, its necessary to have a common set of labels. The 155 labels that worked for all descriptors is listed in the subsequent table:
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
2E−5
2E−7
<1E−10
<1E−10
<1E−10
1E−8
1E−9
<1E−10
<1E−10
<1E−10
7E−3
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
8E−6
<1E−10
<1E−10
<1E−10
1E−7
<1E−10
Section III. Labels Used for the Final SOAP Model
Once the best-performing descriptor-simplification is identified, an expanded set of labels can be employed. The mathematical transformation for the SOAP descriptor is compatible with most of the ˜26,000 structures. In addition to the 155 labels used for descriptor comparisons, 64 labels were added. The full list of labels is included in the table below:
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
2E−5
2E−7
1E−8
<1E−10
<1E−10
<1E−10
1E−8
1E−9
<1E−10
1E−3
<1E−10
<1E−10
7E−3
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
7E−4
<1E−10
<1E−10
<1E−10
<1E−10
<1E−10
1E−5
<1E−10
<1E−10
8E−6
<1E−10
<1E−10
<1E−10
1E−7
<1E−10
Section IV. Wσ Optimization
Ward's minimum variance method applied to the conductivity labels (Wσ) is used to assess the utility of each descriptor-simplification combination. The Wσ is calculated after agglomerative clustering, for each clustering set:
W
σ=Σk=1n
where nC is the number of clusters in a set, Ck is cluster k, and where
A frozen-state strategy is employed to prevent any label from dropping out of the Wσ calculation. The frozen-state strategy operates by calculating the partial variance (PV) for each label at each clustering depth:
PV
x,C
=[log(σRT)x−
where PVx,C
PV
x,C
=PV
x,C
where Cj denotes the cluster with only one label and Ck denotes the cluster that label x previously resided in. Without the frozen state strategy, poor models will reach desirable Wσ values at sufficient depths of clustering. The artificial depression of the Wσ value occurs because clusters that contain a single label evaluate to 0 (the label mean and cluster mean are the same). Whereas the frozen state strategy effectively “remembers” how well (or poorly) the label was clustered before it drops out.
Hyperparameter tuning was employed for some of the descriptors. At least one Wσ representation exists for each unique combination of structure simplification and descriptor. However, some of the descriptors can be altered by tuning associated hyperparameters, resulting in more Wσ representations. The descriptors with hyperparameter tuning are the global instability index, radial distribution function, smooth overlap of atomic positions (SOAP), and mXRD. A grid search was done over the hyperparameters, for each descriptor, with parameters shown in Table 3.
Ultimately, the SOAP-CAN descriptor-simplification outperforms all other descriptor-simplifications when the averaging hyperparameter is set to ‘outer’. Setting the ‘outer’ hyperparameter results in averaging over the power spectrum of different sites. Whereas the ‘inner’ setting averages over the sites first, before summing up the magnetic quantum numbers. The other three hyperparameters (rcut, nmax, and Imax) are less consequential, with most combinations tested outperforming all other non-SOAP descriptors. To illustrate the point, three different SOAP-CAN outcomes are depicted in
Section V. WEa Optimization
Each clustering outcome is also assessed by labeling with approximate activation energies for ion hopping. The activation energies are calculated using a bond valence site energy (BVSE) method developed by Adams and Rao331,332. The strategy approximates the Ea as the sum of an attractive Morse-type potential term and a repulsive Coulombic interaction term. The Morse-type potential term represents mobile ion interactions with lattice anions. While the Coulombic interaction term represents mobile ion interactions with lattice cations. Relative to DFT-based methods, the BVSE method is a computationally lean approach that can be used to readily assess thousands of structures. However, the BVSE method tends to overestimate activation energies because it (1) does not allow for structural relaxation as the mobile ion moves and (2) does not consider repulsive interactions between mobile ions331,332. The BVSE method has been implemented by He et al. and is available for use through their python API333. Using the BVSE method, we label 6845 structures with activation energies (6845 is the number of structures successfully solved given a computing time cutoff of 20-minutes for each structure). Ward's minimum variance method applied to the activation energy labels (WEa) is calculated in a similar manner to the Wu:
where nC is the number of clusters in a set, Ck is cluster k, and where (Ea,BVSE)k denotes the mean for all labels in cluster k. Each descriptor's WEa results are shown in
For Ea labels, all descriptor-simplification pairings result in better semi-supervised ML performance than randomized clustering. The SOAP descriptor performs well relative to most, but five other descriptors outperform it: CAVD, orbital field matrix-CAN, density, mXRD-CAMN, and the packing efficiency descriptors. The favorable performance of CAVD is anticipated because the BVSE calculation directly uses the CAVD descriptor as a parameter. The favorable performance of the density and packing efficiency descriptors may be explained by their similarity to CAVD: the Voronoi decomposition to encode void space is dependent on the density and packing efficiency of the structure. Similarly, the orbital field matrix descriptor relies on calculation of Voronoi polyhedra to understand the coordination environment for each atom. A mXRD-CAMN descriptor-simplification performs well on the BVSE label set; however, the mXRD representation used by Toyota (mXRD—A40) drops from to 14th best on the Ea label set. The result may suggest that the mXRD—A40 pairing does not generalize well. When comparing the top 10 descriptors for each label set, 6 descriptors are common to both approaches: SOAP, density, mXRD, structure heterogeneity, orbital field matrix, and bond fraction.
Section VI. Second-Order SOAP Descriptor
Semi-supervised ML models may be further improved by merging descriptors and clustering on the union representation. Second order descriptor unions are examined by combining the best-performing descriptors with all other descriptors. The two input descriptor vectors (dA and dB) were combined with a mixing ratio (a) to yield the union representation (dAB):
d
AB
=d
A
∪αd
B
The ideal mixing ratio is unknown for each union and we find that incremental changes to the mixing ratio do not result in continuous changes to the Wσ. Thus, outcomes are manually screened for mixing ratios from 10−6 to 106 (see supplemental information—section VI). Most descriptor unions result in no improvement to the WQ, across all mixing ratios. However, the Wσ for SOAP when mixing with the non-simplified sine Coulomb matrix descriptor (for α=2·10−6-4·10−6) is lowered by 2-3%, with the exact percentage depending on the depth of clustering.
Almost no descriptor combinations are successful in reducing the Wσ. Excluding combinations that include the SOAP-CAN descriptor, no combinations outperform the 1st order SOAP-CAN representation. For combinations that include SOAP-CAN, some mixing ratios with the sine Coulomb matrix and the Ewald energy descriptors resulted in modest improvements in the Wσ. The best improvement is found when mixing SOAP-CAN with the sine Coulomb descriptor for α=2·10−6, 3·10−6, and 4·10−6. All three combinations result in the same improved curve, plotted below in
The agglomerative dendrogram in the main text shows that the 2nd-order SOAP-CAN descriptor facilitates aggregation of high-conductivity labels. In the simplified 9-cluster representation, most of the high-conductivity (σRT>10−5 S cm−1) labels are contained within the 2nd “mega cluster”. The 2nd mega cluster accounts for only 15% of the input structure. By clustering further, increasingly dense representations are found. For example, at the 241st clustering depth, the 21 high-conductivity labels have been sorted into five subclusters (
Section VII. Climbing Image—Nudged Elastic Band
Migration barriers for Li ion hopping are evaluated with the Climbing Image—Nudged Elastic Band (CI-NEB) method as implemented in the QuantumESPRESSO PWneb software package334-337. Density-functional theory (DFT) calculations are performed using the Perdew-Burke-Ernzerfof (PBE) generalized gradient approximation functional and projector-augmented wave (PAW) sets338,339. Convergence testing for the kinetic-energy cutoff of the plane-wave basis and the k-point sampling is performed for each structure to ensure an accuracy of 1 meV per atom. The lattice parameters and atomic positions of the as-retrieved structure are optimized. Supercells are created for each structure that are a minimum of 10 Å in each lattice direction to minimize interactions between periodic images of the mobile ion. To study the migration barrier in the dilute limit, a single Li vacancy is created in the boundary endpoint structures of each studied pathway. A uniform background charge is used to balance excess charge. Each boundary configuration is relaxed until the force on each atom is less than 3×10−4 eV/Å. Images are created by linearly interpolating framework atomic positions between the initial and final boundary configurations. The initial pathway for the mobile ion is generated from the BVSE output minimum energy pathway to promote faster convergence of the NEB calculation. An NEB force convergence threshold of 0.05 eV/Å is used. The calculation is first converged using the default NEB algorithm and then restarted with the CI scheme to allow for the maximum energy of the pathway to be determined.
Section VIII. a-Li2.95B0.95Si0.05S3 Impedance
Electrochemical impedance data for the amorphized Si-substituted Li3BS3 (a-Li2.95B0.95Si0.05S3) suggests the presence of two RC features. The VSP-300 potentiostat can supply a maximum sinusoidal frequency of 3 MHz, sufficient to resolve a partial semicircle in the Nyquist impedance plot (
Section IX. Full List of Promising Structures
Excluding the labeled dataset, there are 50 compounds that are predicted to be stable and to exhibit a Li-hopping activation energy below 600 meV. Ten of the predicted compounds have already been experimentally examined and are hereafter excluded: Li2O, Li2S, LiCl, LiI, LiBr, Li6AsS5I, Li4Ti5O12, Li2InCl3, LiInI4, Li6NiCl8. Another nine are excluded because they are used in cathodes, anodes, or glassy electrolyte formulations: LiFeCl4, Li2CO3, Li2PtO3, Li2NiGe3O8, Li2CrO4, Li2SeO4, LiAlS, Li2Mn3NiO8, LiInSe2. The remaining 31 promising structures are discussed below and plotted by ascending activation energy in
a. Stable Compounds
Each structure is examined in order of ascending activation energy in
b. Quasi-Stable Compounds (Ehull below 15 meV)
Excluding the labeled dataset, there are 34 compounds that are predicted to be within 15 meV of the convex hull (Ehull) and to exhibit a Li-hopping activation energy below 600 meV. Ten of the predicted compounds have already been experimentally examined and are hereafter excluded: Li3SbS4, Li6AsS5I, Li6PS51, Li3ScCl6, Li2MnBr4, Li3N, LiTi2P3O12, Li10SiP2S12, Li2ZnCl4, Li3InO3. Another three are currently being excluded because they are used in cathodes: Li3NbS4, Li3CuS2, Li6VCl8. The remaining 21 promising structures are discussed below and plotted by ascending activation energy in
See
c. Unknown-Stability Compounds (Sans Materials Project Entry)
There are 18 predictions that have no associated Material's Project entry. These structures lack stability data. Seven of the predicted compounds have already been experimentally examined and are hereafter excluded: Li2O, Li2S, Li7Y7Zr9S32, Li4SnSe4O13, Li2MnBr4, Li5AlS4, Li3Fe2P3O12. Another five are currently being excluded because they are used in cathodes: Li2Mn3NiO8, Li2Mn3CoO8, Li5Mn16O32, Li2Mn15AlO32, Li3V2P3O12. The remaining 6 promising structures are discussed below and plotted in order of ascending activation energy in
See
From the ten most promising candidates, Li3BS3 was selected for synthesis and characterization. Li3BS3 stands out because it has been explored experimentally and computationally before. Experimentally, Vinatier et al. previously determined that Li3BS3 has a total DC conductivity of 2.5·10−7 S cm−1 with an activation energy of 700 meV62. The DC measurement was not included in our label set because DC measurements cannot differentiate between ionic and electronic conductivity, so they were categorically discounted from the label set (see supplemental information I for more details on label selection). Although the conductivity and activation energy reported by Vinatier et al. are underwhelming, there are promising theoretical reports. Density functional theory molecular dynamics (DFT-MD) simulations from Sendek et al.63 suggest that Li3BS3 should have a room temperature conductivity between 3.1·10−6 and 9.7·10−3 S cm−1. Our NEB-calculated activation energy for Li3BS3 is 260 meV, corroborating a previous NEB result from Bianchini et al.64. Additionally, Li3BS3 is practically attractive because: (1) Li3BS3 contains no redox-active metals, (2) band edge calculations have suggested stability against metallic Li65, (3) DFT-MD calculations have suggested a kinetic barrier for decomposition against metallic Li63, and (4) the synthesis is reported66. It is simpler to avoid redox active metals in the SSE as they may be reduced and oxidized at electrode interfaces. However, we note that Li0.5La0.5TiO3 is a widely studied SSE that contains redox active Ti67,68 so the compounds we report here that contain Mn, V, and Cu should not be categorically discounted. It is important to note that while studying Li3BS3 as a candidate Li-ion conductor for model validation, Kimura et al. reported that a so-called “Li3BS3 glass” exhibits an ionic conductivity of 3.6·10−4 S/cm−1 at 25° C.69.
Li3BS3 is prepared using solid-state synthesis from Li2S, B, and S precursors. The diffraction and quantitative Rietveld refinement are shown in
where T is the temperature, kB is the Boltzmann's constant, σ0 is the conductivity prefactor and Ea is the activation energy for ionic conductivity. The room temperature ionic conductivity (σ25° C.) is 7.16(±0.21)·10−7 S cm−1 and the activation energy is 400±47 meV. The low conductivity and high activation energy may be due to lack of charge-carrying defects in the Li3BS3 lattice70,71. Although a sufficient carrier concentration is necessary for facile ionic conduction in most materials, the descriptors in the semi-supervised model do not explicitly encode for charge-carrying defects. In the label set, conductivity is likely influenced by the defect concentration but defects are typically not reported. Still, the semi-supervised model may infer a structure's capacity to support conductive defects via correlation with the descriptors.
Li3BS3 Synthesis:
Li3BS3 is synthesized by reaction of Li2S (Alfa Aesar, 99.9%), S8 (Acros Organics, >99.5%), and elemental B (SkySpring Nanomaterials, Inc. 99.99%). The reactants are first mixed stoichiometrically (300 rpm for 1 h) using a planetary ball mill (MSE PMV1-0.4L) in 50 mL ZrO2 jars with ZrO2 balls. Two grams of reactants are always combined with 2 large balls (10 mm diameter), 34 medium balls (5 mm diameter), and 8 grams of small balls (3 mm diameter). Loading of ball mill jars occurs in an Ar-filled glovebox (Mbraun) and the jars are sealed before removal. After the 1 h of milling, the precursor mixture is pumped back into the glovebox and 330-340 mg of the powder is loaded into carbon coated vitreous silica ampoules (10 mm ID×12 mm OD). The ampoules are evacuated (<10 mtorr) prior to sealing. Pure Li3BS3 is obtained via a four-step heating protocol in a Lindberg/Blue furnace: (1) ramp to 500° C. at 5° C. min-1, (2) hold at 500° C. for 12 h, (3) ramp to 800° C. at 5° C. min-1, and (4) hold at 800° C. for 6 h. The hot melt is then quenched from 800° C. into room temperature water. Recovered ingots are typically covered in a C shell. The C shell is either sanded off or the ingot is ground into smaller pieces and the C is manually removed.
To test the hypothesis, we use two strategies to engineer vacancies: aliovalent substitution and amorphization via extended ball milling. Aliovalent substitution has been shown to improve conductivity in Li-argyrodites, -sulfides, and -garnets by introducing vacancies70,71. Similarly, amorphization can introduce defects and vacancies that enable Li+ hopping69,71-73.
Aliovalent substitution of Li3BS3 is achieved by substituting Si for B. The XRD patterns and quantitative Rietveld refinements of Li2.975B0.975Si0.025S3 and Li2.95B0.95Si0.05S3 are shown in
We find that amorphization significantly improves Li-ion conductivity. EIS measurements of a-Li2.95B0.95Si0.05S3 are shown in
To determine if the local structure in the crystalline material is maintained after amorphization, we turn to 7Li and 11B NMR. If the local structure is not altered by amorphization, then it is likely that the ion diffusion pathways are similar. Comparing the ion diffusion pathways is important because the machine learning points to the structure of the crystalline Li3BS3 phase. The 7Li NMR spectra of Li3BS3, Li2.95B0.95Si0.05S3, and a-Li2.95B0.95Si0.05S3 are shown in
Although investigation of interfacial stability is beyond the scope of the model, we note that the Si-substituted Li3BS3 is a promising candidate for future investigations into interfacial stability. Work by Park et al. demonstrated that the (010) facet for Li3BS3 has a conduction band minimum 0.5 eV above the Li/Li+ couple65. Since decomposition of Li3BS3 is likely to be mediated by electron injection from Li, their results suggest that thermodynamic stability can be engineered via orientation. From a kinetic perspective, high-temperature DFT-MD simulations show no mobility for B and S, suggesting large kinetic diffusion barriers63. Since decomposition of Li3BS3 would entail the diffusion of these species, the reaction may be sluggish or wholly precluded. Interfacial stability has been previously demonstrated for a glassy electrode in the Li—B—S—Si—O phase space78. The result may indicate that stability can be engineered into Si-substituted Li3BS3 by partial isovalent substitution of 0 for S. Finally, recently-synthesized Li—B—S—X (X═Cl, Br, I) quaternaries have exhibited promising conductivities79. With similar elemental composition, the Si-substituted Li3BS3 may be a good candidate for a multi-electrolyte architecture with the halide-containing quaternaries13
In addition to our experimental model validation, another of the predicted materials, KLi6TaO6, was recently synthesized with aliovalent Sn-substitution by Suzuki et al80. With a reported ionic conductivity near 10−5 S cm−1, KLi6TaO6 is better than 70% of the SSEs in the semi-supervised labels. Further improvement may be possible via extended amorphization to introduce structural defects, as is observed for Li3BS3.
Substituted Li3BS3:
Aliovalent substitution is accomplished by adding elemental Si (Acros, 99+%) into the precursor mixture prior to the 1 h mix. Si-substitution stoichiometry assumed that each Si atom replaces one Li and B: Li3−xB1−xSixS3. Aside from the addition of Si, all steps are the same as for the synthesis of Li3BS3. Amorphization is accomplished via extended planetary ball milling in Ar of the 5% Si-substituted Li3BS3 (Li2.95B0.95Si0.05S3). Approximately 1 g of Li2.95B0.95Si0.05S3 is combined in a ZrO2 ball mill jar with 3 large balls (10 mm diameter), 51 medium balls (5 mm diameter), and 12 g of small balls (3 mm diameter). The powder is ground in a planetary ball mill (MSE PMV1-0.4L), under Ar atmosphere, for 100 h.
Material Characterization:
Li3BS3 materials are characterized using powder X-ray diffraction (XRD) and electrochemical impedance spectroscopy (EIS). XRD patterns are attained on a Rigaku Smartlab by scanning from 10° to 70° 2θ at 2 degrees per minute. The Smartlab employs a Cu-Kα source with a 20 kV accelerating voltage. For EIS measurements, 50-100 mg of powder is first hot-pressed (100° C., 5 min) into a ¼″ diameter pellet. The pellet faces are polished using diamond lapping powder (Allied High Tech Products Inc.) in sequentially finer grits: 60, 30, 6, 0.5, and 0.1 micron. Au contacts are sputtered (90 s at 40 mA) onto the polished surfaces using a 108 Auto Sputter Coater (Cressington). Pellets are then assembled into a Swagelok ¼″ cell with stainless steel current collectors. After applying pressure with a hand vise (˜100 MPa), EIS data is collected on a VSP-300 with a Biologic low-current channel. All EIS data is collected to an upper frequency of 3 MHz. The lower frequency is case dependent, with a frequency cutoff selected such that the Warburg polarization feature is visible. 7Li and 11B MAS MAS NMR spectra were acquired using a Bruker DSX-500 spectrometer with a 4 mm ZrO2 rotor. The operating frequencies for 7Li and 11B are 190.5 and 160.5 MHz, respectively. The 7Li and 11B spectra were referenced to a 1 M LiCl aq. solution and BF3-OEt2, respectively. A spinning speed of 12 kHz was used, and the spectra were gathered after applying a single 0.5 μs to 15° pulse for both 7Li and 11B.
Further Discussion for Examples 1-3
Identification of functional materials is critical for improving technologies. Here, we show the utility of using semi-supervised learning as a method for guiding next-generation materials discovery in emerging fields. The method's focus on identifying the relationships between descriptors, prior to labeling, enables understanding of compositional spaces where most inputs are unlabeled. We demonstrate how semi-supervised learning can be used to identify descriptors correlated with superionic conductivity in Li SSEs. By analyzing all Li-containing structures from the ICSD and MP database, we identify 212 materials that show promise as SSEs. All 212 structures exhibit a BVSE-predicted Ea below 0.6 eV.
The results illustrate why careful screening of descriptors is useful when identifying new materials. While chemical intuition can be useful for descriptor selection, chemical intuition is often biased to favor previously investigated compositional spaces. For material discovery in emerging fields, use of handpicked descriptors may miss complex phenomena that more generally describe the dataset. Descriptor screening reveals which material properties are correlated to a property of interest to help enhance chemical intuition. In the case of Li SSEs, spatial descriptors excel over compositional, bonding, and electronic descriptors: the Smooth Overlap of Atomic Positions (SOAP), modified X-ray diffraction (mXRD), and general density descriptors are within the top four models. For spatial descriptors, simplification of the input structure tends to improve clustering outcomes. Removing the mobile ions from the structure and simplifying the remaining atoms, i.e. the “CAN” simplification, is most effective. Thus, the placement of framework atoms, but not their precise identity, is most correlated with ionic conductivity. Specifying the mobile ion positions hurts the model performance, suggesting a low correlation of mobile ion positions with ionic conductivity.
Predictions from the semi-supervised method are promising starting points for experimental identification of new superionic conductors but defects must be considered. The proposed materials are diverse, with the top thirty including halides, sulfides, tellurides, nitrides, oxides, and oxyhalides (see Example 1B: section IX). As a structure that falls outside of the eight routinely studied SSE classes, we demonstrate experimental characterization of Li3BS3 to confirm the utility of the approach. However, pure Li3BS3 exhibits poor ionic conductivity. Defects must be introduced into the material to achieve a superionic conductivity above 10−3 S cm−1, a value that surpasses most reported SSEs. We note that the defects are introduced while maintaining the local structure of the crystalline material and thus the ionic conduction pathways are likely similar. The need to introduce defects highlights the paramount importance that defects play when measuring real materials. Many of the highest performing SSEs contain charge-carrying defects that are not explicitly encoded in their structure files. It is likely that some of the descriptors indirectly encode information about defects. By using experimental conductivity values as the evaluation metric, we may be prioritizing descriptors that encode information about a structures ability to support charge-carrying defects. Although Li3BS3 is a poor conductor, it is clearly able to support charge-carrying defects. The large conductivity difference between pristine Li3BS3 and a-Li2.95B0.95Si0.05S3 highlights the importance of these defects. To improve predictive models and enhance chemical intuition, descriptors that explicitly encode defects are needed.
Now developed, the semi-supervised learning approach can serve as a template for material discovery beyond Li SSEs. The code is thoroughly documented following pythonic coding standards and made freely available on Github. Although the present effort focuses on Li SSEs, the approach is applicable to any material discovery space where labels are sparse. Discovery of new Li cathodes could be accomplished by using Li diffusivity, cathode capacity, and metal redox couple voltages as labels. Discovery of divalent SSEs (e.g. Mg2+, Ca2+, Zn2+) could foreseeably be accomplished in a similar manner. The semi-supervised learning strategy may accelerate identification of fast ionic conductors for ion exchange membranes, solid oxide fuel cells, and various sensor applications.
In some aspects, the “first dopant” Q in FX1 (Li3−z[B+Q]1[S+G]3) optionally comprises one or more transition metal elements. On the other hand, in some aspects, the “first dopant” Q in FX1 is free of transitional metal elements due to the propensity of transition metal elements to participate in redox reactions occurring in a solid state battery. Selection of the first dopant element (the one or more dopant element corresponding to first dopant Q) is generally dependent on application-specific particular chemistry, redox reactions, voltages, and other conditions.
In an aspect, a particularly useful material disclosed herein has a composition characterized by formula FX15A: Li3−xB1−xSixS3 (FX15A); wherein x is greater than 0 and less than or approximately equal to 0.05. For example, in an aspect, a furthermore particularly useful material disclosed herein has a composition characterized by formula FX15B: LiaBbSixSc (FX15B); wherein a is approximately 2.95, b is approximately 0.95, x is approximately 0.05, and c is approximately 3.0. For example, in an aspect, a yet furthermore particularly useful material disclosed herein has a composition characterized by formula FX15C: Li2.95B0.95Si0.05S3 (FX15C). It is found that the compositions of FX15A, FX15B, and FX15C may correspond to an optimum or near-optimum doped composition of lithium thioborate with respect to ionic conductivity, optionally with respect to other additional features, where the undoped composition and low-doped composition (e.g., x being less than 0.025) have lower ionic conductivity and higher dopant amounts (e.g., x being greater than or equal to 0.075) cause formation of unfavorable impurities and/or other unfavorable features (e.g., too much disruption of crystallographic structure with respect to that of Li3BS3). It is found that compositions of FX15A, FX15B, and FX15C have high ionic conductivity and low electronic conductivity, especially if the material is amorphized. For example, as also discussed in Examples 1A-3, whereas a room temperature (e.g., 25° C.) ionic conductivity of undoped Li3BS3 without substitution may be approximately 7.2·10−7 S/cm substitution/doping of the composition thereby making Li2.95B0.95Si0.05S3 (FX15C) may result in an increase of ionic conductivity (relative to that of the undoped Li3BS3) by a factor of approximately 25 such as to an ionic conductivity of approximately 1.82(±0.21)·10−5 S cm−1 at room temperature (e.g., 25° C.). Amorphization of the composition Li2.95B0.95Si0.05S3 (FX15C) may further increase the ionic conductivity (relative to that of the undoped Li3BS3) by a factor of at least 1375 such as to an ionic conductivity of approximately 1.07(±0.08)·10−3 S cm−1 at room temperature (e.g., 25° C.) or even about 3·10−3 S cm−1 according to some aspects. In contrast, the electronic conductivity of these doped compositions of FX15A, FX15B, and FX15C have low electronic conductivity, such as, for example, in aspects, less than or equal to about 4·10−10 S/cm as measured by DC polarization at room temperature (e.g., 25° C.).
All references throughout this application, for example patent documents including issued or granted patents or equivalents; patent application publications; and non-patent literature documents or other source material; are hereby incorporated by reference herein in their entireties, as though individually incorporated by reference, to the extent each reference is at least partially not inconsistent with the disclosure in this application (for example, a reference that is partially inconsistent is incorporated by reference except for the partially inconsistent portion of the reference).
The terms and expressions which have been employed herein are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments, exemplary embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims. The specific embodiments provided herein are examples of useful embodiments of the present invention and it will be apparent to one skilled in the art that the present invention may be carried out using a large number of variations of the devices, device components, methods steps set forth in the present description. As will be obvious to one of skill in the art, methods and devices useful for the present methods can include a large number of optional composition and processing elements and steps.
As used herein and in the appended claims, the singular forms “a”, “an”, and “the” include plural reference unless the context clearly dictates otherwise. Thus, for example, reference to “a cell” includes a plurality of such cells and equivalents thereof known to those skilled in the art. As well, the terms “a” (or “an”), “one or more” and “at least one” can be used interchangeably herein. It is also to be noted that the terms “comprising”, “including”, and “having” can be used interchangeably. The expression “of any of claims XX-YY” (wherein XX and YY refer to claim numbers) is intended to provide a multiple dependent claim in the alternative form, and in some embodiments is interchangeable with the expression “as in any one of claims XX-YY.”
When a group of substituents is disclosed herein, it is understood that all individual members of that group and all subgroups, including any isomers, enantiomers, and diastereomers of the group members, are disclosed separately. When a Markush group or other grouping is used herein, all individual members of the group and all combinations and subcombinations possible of the group are intended to be individually included in the disclosure. When a compound is described herein such that a particular isomer, enantiomer or diastereomer of the compound is not specified, for example, in a formula or in a chemical name, that description is intended to include each isomers and enantiomer of the compound described individual or in any combination. Additionally, unless otherwise specified, all isotopic variants of compounds disclosed herein are intended to be encompassed by the disclosure. For example, it will be understood that any one or more hydrogens in a molecule disclosed can be replaced with deuterium or tritium. Isotopic variants of a molecule are generally useful as standards in assays for the molecule and in chemical and biological research related to the molecule or its use. Methods for making such isotopic variants are known in the art. Specific names of compounds are intended to be exemplary, as it is known that one of ordinary skill in the art can name the same compounds differently.
Certain molecules disclosed herein may contain one or more ionizable groups [groups from which a proton can be removed (e.g., —COOH) or added (e.g., amines) or which can be quaternized (e.g., amines)]. All possible ionic forms of such molecules and salts thereof are intended to be included individually in the disclosure herein. With regard to salts of the compounds herein, one of ordinary skill in the art can select from among a wide variety of available counterions those that are appropriate for preparation of salts of this invention for a given application. In specific applications, the selection of a given anion or cation for preparation of a salt may result in increased or decreased solubility of that salt.
Every device, cell, electrolyte, material, composition, and method described or exemplified herein can be used to practice the invention, unless otherwise stated.
Whenever a range is given in the specification, for example, a temperature range, a time range, or a composition or concentration range, all intermediate ranges and subranges, as well as all individual values included in the ranges given are intended to be included in the disclosure. It will be understood that any subranges or individual values in a range or subrange that are included in the description herein can be excluded from the claims herein.
All patents and publications mentioned in the specification are indicative of the levels of skill of those skilled in the art to which the invention pertains. References cited herein are incorporated by reference herein in their entirety to indicate the state of the art as of their publication or filing date and it is intended that this information can be employed herein, if needed, to exclude specific embodiments that are in the prior art.
As used herein, “comprising” is synonymous with “including,” “containing,” or “characterized by,” and is inclusive or open-ended and does not exclude additional, unrecited elements or method steps. As used herein, “consisting of” excludes any element, step, or ingredient not specified in the claim element. As used herein, “consisting essentially of” does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim. In each instance herein any of the terms “comprising”, “consisting essentially of” and “consisting of” may be replaced with either of the other two terms. The invention illustratively described herein suitably may be practiced in the absence of any element or elements, limitation or limitations which is not specifically disclosed herein.
One of ordinary skill in the art will appreciate that starting materials, biological materials, reagents, synthetic methods, purification methods, analytical methods, assay methods, and biological methods other than those specifically exemplified can be employed in the practice of the invention without resort to undue experimentation. All art-known functional equivalents, of any such materials and methods are intended to be included in this invention. The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention that in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.
This application claims the benefit of priority to U.S. Provisional Patent Application No. 63/348,603, filed Jun. 3, 2022, which is hereby incorporated by reference in its entirety.
Number | Date | Country | |
---|---|---|---|
63348603 | Jun 2022 | US |