Prelithiated silicon oxide, chemical methods for synthesizing the prelithiated material, and long cycling lithium ion batteries formed with the material

Abstract

A particulate lithium-silicon oxide composite material is described comprising lithium incorporated into a silicon oxide material and having a Cu Kα diffractogram with 2θ peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3. Methods are described for forming these materials. Also, electrodes and lithium-ion cells incorporated this material are described. Good cycling results are obtained.

Description

[0001] PRELITHIATED SILICON OXIDE, CHEMICAL METHODS FOR SYNTHESIZING THE PRELITHIATED MATERIAL, AND LONG CYCLING LITHIUM ION BATTERIES FORMED WITH THE MATERIAL

[0002] CROSS REFERENCE TO RELATED APPLICATIONS

[0003] This application claims priority to copending U.S. provisional patent application 63 / 546,889, filed November 1, 2023, to Gandhi et al., entitled "Prelithiated Silicon Oxide, Chemical Methods for Synthesizing the Prelithiated Material, and Long Cycling Lithium Ion Batteries Formed With the Material," incorporated herein by reference.

[0004] FILED OF THE INVENTION

[0005] The invention relates to the preparation of silicon based material, in particular silicon oxide composite material, for use in a lithium ion secondary cell. The invention further relates to methods of prelithiating the silicon oxide based materials.

[0006] BACKGROUND OF THE INVENTION

[0007] Lithium ion cells are widely used in consumer electronics and are a growing enabling component of the rapidly expanding electric vehicle market. Lithium is a desirable material due to its relatively high energy density and low mass. For some current commercial batteries, the negative electrode material can be graphite, and the positive electrode materials can comprise lithium cobalt oxide (LiCoO2), lithium manganese oxide (LiMn2O4), lithium iron phosphate (LiFePO4), lithium nickel oxide (LiNiO2), lithium nickel cobalt oxide (LiNiCoO2), lithium nickel cobalt manganese oxide (LiNiMnCoO2), lithium nickel cobalt aluminum oxide (LiNiCoAlO2) and the like. Graphite can be limited by its relatively low theoretical capacity, and silicon based materials are a promising alternative to graphite due to a much higher theoretical capacity for lithium incorporation. The successful cycling of silicon-based materials in a lithium ion cell can be challenging.

[0008] The successful cycling of a lithium ion cell relies on structural / compositional changes that take place during the first few electrochemical cycles of the cell. Silicon-based active materials are particularly sensitive to these initial changes in the cell, although these changes are also significant for graphite cells. The ability to commercialize lithium ion cells using significant amounts of silicon-based active materials can provide improved cell performance for significantly growing market segments. SUMMARY OF THE INVENTION

[0009] In a first aspect, the invention pertains to particulate lithium- silicon oxide composite material comprising lithium incorporated into a silicon oxide material and having a Cu Kadiffractogram with 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

[0010] In a further aspect, the invention pertains to a method for chemically prelithiating a silicon suboxide-based material. Generally, the method comprises associating lithium-radical anion complex with the silicon suboxide-based material in a solution wherein the lithiumradical anion complex is dissolved and the silicon suboxide-based material is dispersed to form a lithium deposited silicon suboxide. The deposited silicon suboxide can be annealed. In particular, the lithium deposited silicon suboxide can be annealed at a temperature from about 350°C to about 600°C for at least 1 hour to form a prelithiated silicon suboxide. The annealing can be performed to form the material with the x-ray diffractogram peaks of the above paragraph.

[0011] In another aspect, the invention pertains to a method for forming a lithium deposited silicon suboxide, in which the method comprises associating lithium-radical anion complex with silicon suboxide-based material in a solution wherein the lithium-radical anion complex is dissolved and the silicon suboxide-based material is dispersed, wherein the solution comprises a total lithium amount that is at least about 25 mole% greater than a molar quantity of radical anion and its neutral corresponding compound, and wherein a lithium deposited silicon suboxide is formed.

[0012] In other aspects, the invention pertains to an electrode comprising a blend of a polymer binder, electrically conductive particulates, and an active material comprising particulate lithium-silicon oxide composite material comprising lithium incorporated into a silicon oxide lattice associated with amorphous carbon having a first cycle irreversible capacity loss of no more than about 20 % relative to the first cycle charge capacity evaluated at C / 20 and having a Cu Kax-ray diffractogram of the lithium deposited silicon suboxide composite material with 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

[0013] BRIEF DESCRIPTION OF THE DRAWINGS

[0014] FIG. 1 A is an exploded view of a pouch battery with a battery core separated from two portions of the pouch case.

[0015] FIG. IB is a perspective lower face view of the assembled pouch battery of FIG. 1A.

[0016] FIG. 1C is a bottom plan view of the pouch battery of FIG. IB. FIG. ID is a depiction of an embodiment of a battery core comprising an electrode stack.

[0017] FIG. 2 is a process flow diagram for preparation of prelithiated silicon suboxides.

[0018] FIG. 3 shows x-ray diffraction patterns for, from top to bottom, unlithiated silicon suboxide, prelithiated silicon suboxide prepared by soaking in Li-Bp / THP complex solution and annealing at 650°C for 2 hours, and lithium deposited silicon suboxide prepared by soaking in Li-Bp / THP complex solution but not annealed. X-ray diffraction peaks for Li2SiO3 and silicon are also shown.

[0019] FIG. 4 shows a series of x-ray diffraction patterns for prelithiated silicon suboxides prepared by annealing at 650°C for 2 hours, where the molar ratio of silicon suboxide:Li ranges from 1:1 to 10:1.

[0020] FIG. 5 is a combination plot of specific discharge capacity and percent Initial Coulombic Efficiency (ICE), each as a function of molar ratio of silicon suboxide:Li for half cells including negative electrodes formulated with prelithiated silicon suboxides prepared by annealing at 650°C for 2 hours.

[0021] FIG. 6 is a plot of the number of cycles at 80% capacity retention as a function of molar ratio of silicon suboxide:Li for full cells including negative electrodes formulated with prelithiated silicon suboxides prepared by annealing at 650°C for 2 hours.

[0022] FIG. 7 is a plot of specific discharge capacities, at 1C and 5C discharge rates, as a function of molar ratio of silicon suboxide:Li for full cells including negative electrodes formulated with prelithiated silicon suboxides prepared by annealing at 650°C for 2 hours.

[0023] FIG. 8 shows a series of x-ray diffraction patterns for prelithiated silicon suboxides prepared by annealing for 2 hours at temperatures ranging from 350°C to 950°C with the topmost pattern labelled with different phases of lithium silicates. The x-ray diffraction pattern for prelithiated silicon suboxide is included, as are reference X-ray diffraction peaks of silicon.

[0024] FIG. 9 shows x-ray diffraction patterns for prelithiated silicon suboxide prepared by annealing at 650°C for 2 hours. X-ray diffraction peaks for Li2SiO3 and silicon are also shown.

[0025] FIG. 10 is a plot of percent retention of specific capacity as a function of cycle number for full cells including negative electrodes formulated with prelithiated silicon suboxide prepared by annealing at 650°C for 2 hours.

[0026] FIG. 11 shows XRD patterns for prelithiated silicon suboxide annealed at 450°C.

[0027] FIG. 12 shows superimposed XRD patterns of FIG. 11.

[0028] FIG. 13A shows XRD patterns for prelithiated silicon suboxide annealed at 550°C.

[0029] FIG. 13B shows superimposed XRD patterns of FIG. 13A. FIG. 14 shows XRD patterns for prelithiated silicon suboxide annealed at 650°C.

[0030] FIG. 15 is a plot of cycle number for full cells as a function of annealing temperature used to prepare prelithiated silicon suboxides.

[0031] FIG. 16 shows cycling performance for full cells where the number of cycles reached to obtain 80% capacity is plotted as a function of annealing temperature used to prepare prelithiated silicon suboxide.

[0032] FIG. 17 shows specific capacity as a function of annealing temperature for two sets of the full cells with one set cycled with a 4C charge / 1C discharge rate and the other with a 1C charge / lC discharge rate.

[0033] FIG. 18 shows specific discharge capacity for half cells as a function of annealing temperature used to prepare prelithiated silicon suboxide.

[0034] FIG. 19 shows percent irreversible capacity loss (IRCL) for half cells as a function of annealing temperature used to prepare lithiated silicon suboxide samples.

[0035] FIG. 20 shows a plot of pH of silicon suboxide:Li slurries as a function of annealing temperature used to prepare prelithiated silicon suboxides.

[0036] FIG. 21 shows a plot of peel strength for lithiated silicon suboxide samples on copper current collector, as a function of annealing temperature used to prepare prelithiated silicon suboxides.

[0037] FIG. 22 shows a plot of peel strength as a function of distance for a prelithiated silicon suboxide prepared by annealing at 450°C for 24 hours and for a commercially available lithiated SiO sample.

[0038] FIG. 23 is a plot of voltage as a function of specific capacity for full cells prepared with the prelithiated silicon suboxides and unlithiated silicon suboxide formulated with aqueous and nonaqueous binders.

[0039] FIG. 24 shows a series of x-ray diffraction patterns for prelithiated silicon suboxides prepared by soaking for 2 hours or 24 hours, at temperatures of 40°C, 70°C or 90°C, prior to annealing for 24 hours at 450°C.

[0040] FIG. 25 shows cycling performance for full cells where the number of cycles reached to obtain 80% capacity is plotted as a function of soaking temperature as described for FIG. 24.

[0041] FIG. 26 shows specific capacity retention as a function of cycle number for full cells cycled at room temperature at 1C charge / 1C discharge rates with data for soaking times of 2 and 24 hours, at temperatures of 40°C, 50°C, 70°C or 90°C, as described for FIG. 24.

[0042] FIG. 27 shows x-ray diffraction pattern for lithium. FIG. 28 shows x-ray diffraction pattern for LiiSii .

[0043] FIG. 29 shows x-ray diffraction pattern for Li2SiO3.

[0044] FIG. 30 shows x-ray diffraction pattern for Li2Si20s.

[0045] FIG. 31 shows x-ray diffraction pattern for Li4SiO4.

[0046] FIG. 32 shows x-ray diffraction pattern for LieSi2O7.

[0047] FIG. 33 shows x-ray diffraction pattern for LisSiOe.

[0048] DETAILED DESCRIPTION OF THE INVENTION

[0049] Chemical prelithiation procedures have been discovered that result in an improved prelithiated powder active material that can maintain high capacity for reversible lithium incorporation and release while achieving good cycling properties. The development of these improved materials is based on an elucidation of the crystalline states and balance of crystallization to appropriate phases with maintenance of appropriate amorphous character. These understandings are contrary to previous understandings in the field. Synthesis methodology has been developed to accordingly produce the desired materials. The synthesis methods use radical anion organic species, such as biphenyl anions, to form effective lithium salts that facilitate the lithiation of the active material. The improved materials reduce the amounts of inactive phases of material that do not contribute to usable capacity while at the same time stabilize the material to achieve improved cycling. The best cycling material have a low degree of crystallinity, for example, as evaluated by elemental crystalline silicon scattering peaks, and an additional unidentified phase can be seen in the x-ray scattering data. The use of prelithiated materials simplifies the manufacturing process incorporating high capacity silicon based materials. The improved cycling achievable with these materials provides Applicant with another tool in their achievement of commercially useful cells with high components of silicon based active materials. The use of prelithiated silicon suboxide promises simplified manufacturing and simpler formation cycles for cell preparation.

[0050] The significance of supplemental lithium for lithium ion batteries was found to be particularly valuable for appropriate positive electrode and negative electrode active materials. See, for example, U.S. patent 9,166,222 to Amiruddin et al. (hereinafter the '222 patent), entitled "Lithium Ion Batteries With Supplemental Lithium," and U.S. patent 9,190,694 to Lopez et al., entitled "High Capacity Anode Materials for Lithium Ion Batteries," both of which are assigned to Applicant and are incorporated herein by reference. This significant work was closely followed by similar efforts from Amprius Corporation, see U.S. patent 8,846,251 to Cui et al., entitled "Preloading Lithium Ion Cell Components With Lithium," incorporated herein by reference. For silicon-based electrodes, Applicant has incorporated supplemental lithium into their development pipeline for the high performance lithium in cells with good cycling, as described further below.

[0051] As explained in the '222 patent, there are a range of approaches for introducing the supplemental lithium into the cells. Many of these approaches have been explored. Applicant has commercialized the use of association of elemental lithium with the anode during assembly of the cell. The elemental lithium can be provided as stabilized lithium metal powder, which is commercially available, lithium pellets, or as sheets or foils of lithium metal which can be laminated to the electrode. Due to the reactivity of the lithium metal, it can spontaneously react with the negative electrode (anode) active material spontaneously with sufficient degree of contact or as facilitated by electrolyte. The occurrence of this reaction is controlled to some degree by stabilization coatings over the elemental lithium, the configuration, and the presence of electrolyte. In any case, these reactions to incorporate elemental lithium are presumed to be completed by the end of formation cycles of the cell, which can be one or a few initial chargedischarge cycles.

[0052] Prelithiation can be performed in various ways, such as electrochemical, mechanical, such as high energy milling, or chemical. The reactions involved in the prelithiation process are complex. These reactions can involve both changes to the structure and composition of the active material through the direct formation of lithium compounds as well as the restructuring of the initial materials and the potential formation of coating of the active material, such as a solid electrolyte interphase (SEI) layer when electrolyte is present. The nature of the prelithiation process can result in the formation of distinct materials at the end of the lithiation process. As noted above, Applicant has found the best performance based on the introduction of elemental lithium at the point of cell assembly, thereby avoiding prelithiation prior to cell assembly. But the reactions involved in incorporation of supplemental lithium during cell assembly are equally complex.

[0053] From a process point of view, it is desirable to perform the prelithiation with respect to the powder active material prior to incorporation into an electrode structure. In this way the prelithiated material can be incorporated into a preexisting manufacturing line with less need for capital equipment changes, and the handling of moisture sensitive lithium metal can be confined to material processing context. This approach has to date not provided an equally effective active material in comparison to previous work involving the supply of supplemental lithium at the time of cell assembly. The present work is directed to significant improvement in that regard. As used herein, silicon oxide refers to SiOx, where x is 0.5 < x< 1.4, which can alternatively be referred to as silicon suboxide. SiO can be referred to as silicon monoxide, but in the solid state, this does not seem to be a stoichiometric material with a single unique, homogenous structure. Silicon oxide can have a complex structure with domains of essentially silicon, which may be relatively crystalline, with other amorphous and potentially crystalline domains. Commercial silicon oxide (approximating SiO) are available, for example, from Shin-Etsu, Sigma-Aldrich, Shanshan, Osaka Titanium Technologies Co., Nano structured & Amorphous Materials, Inc., Posco, BTR, Zhide New Energy Materials Co., Ltd., and TNJ Chemical (China).

[0054] Silicon oxide is not conductive. Silicon dioxide, which may be referred in various fields to as silicon oxide or silica, is a well-known dielectric, and elemental silicon is a semiconductor with a low conductivity if undoped. Doping of silicon oxide (suboxide) materials has not been particularly useful as an active material for a cell. Of course, cycling in the cell involves conduction of electrons between the active material and the current collector. To improve the conductivity of the silicon oxide active material, a composite can be formed of the silicon oxide with elemental carbon. Elemental carbon can be moderately electrically conductive in an appropriate form. In particular, amorphous carbon is moderately electrically conductive and graphite / graphene sheets can be conductive along appropriate planes. On the other hand, diamond-like carbon is insulating. In the composite form, the carbon can be intimately associated with the silicon oxide, and low amounts of carbon coatings can significantly improve electrical conductivity without significantly impacting insertion and removal of lithium from the material. This direct association can be achieved, for example, through high energy mechanical milling. The milling can disrupt any initial crystallinity of the carbon in the process of its association with the silicon oxide. An effective way to intimately form a carbon composite with silicon oxide involves the formation of pyrolytic carbon in association with the silicon oxide. Various organic sources can provide the reactant for formation of the pyrolytic carbon. Pyrolytic carbon is generally amorphous with domains of sheets having a graphite structure and with moderate electrical conductivity. Various approaches for forming a carbon composite with silicon oxide is described in U.S. patent 9,601,228 to Deng et al. (hereinafter the '228 patent), entitled "Silicon Oxide Based High Capacity Anode Materials for Lithium Ion Batteries," incorporated herein by reference.

[0055] Silicon based active materials are known to have a high irreversible capacity loss (IRCL) during the formation cycle(s) after the lithium ion battery is assembled. For example, silicon undergoes large volume changes with alloying and dealloying with lithium that is believed to contribute to irreversible capacity loss for silicon. An irreversible capacity loss at the negative electrode corresponds to consumption of lithium that is subsequently not available for cycling. In traditional lithium ion battery designs, the lithium for cycling is supplied in the initial positive electrode active material, and if portions of this capacity of lithium for cycling is irreversibly lost in initial charges of the cell, there is significant mass of positive electrode material that is not used during cycling since there is insufficient lithium to discharge into the positive electrode active material with respect to its full capacity. An initial impetus for supplying supplemental lithium in addition to lithium from the positive electrode active material was to directly provide additional lithium for the IRCL to avoid needing extra positive electrode active material capacity that would not be used for cycling. Additional advantages were discovered for providing even greater amounts of supplemental lithium. The processing described herein is effective to lower the IRCL relative to the material prior to prelithiation, while maintaining a desirable discharge specific capacity and good cycling stability.

[0056] Evidence suggests that at least a significant amount of the IRCL is involved in irreversible changes to the active material structure. Since lithium is evidently consumed in the process, it is appropriate to assume that lithium is incorporated into phases in which the lithium is not extractable under the conditions in the cell. On the other hand, the material retains a significant capacity for extractable lithium, so there presumably are distinct lattice sites in the material attributable to extractable and non-extractable lithium, which is consistent with significant inhomogeneity of the material. Since silicon oxide has proven to be a promising, high capacity silicon-based material for cells with good cycling, significant attention has been devoted to silicon oxide materials in the context of prelithiation, see the '228 patent cited above. It is highly desirable for a prelithiation process to significantly lower or eliminate the IRCL for the prelithiated material since a major objective of the prelithiation process is to avoid the impetus for supplying lithium to, at least in part, compensate for IRCL associated losses of cycling lithium.

[0057] Lithium is known to form various phases of lithium silicates. The prelithiation process has been confirmed to form various silicates depending on the reaction conditions, and these reactions are discussed below in the context of the processing described herein. The reaction conditions can correspondingly be established to control the structure of the resultant composite material. In general, the material is believed to have complex structures with microdomains within the individual particles. Earlier efforts with particle level prelithiation has been based on a proposal that the lithium silicate crystallites should stabilize the structure to provide improved cycling stability. Actual results suggest the exact opposite, as described in detail herein. The active cycling material presumably still undergoes significant volume changes with cycling, and embedding certain lithium silicate crystallites through the active material is found to still result in rapid capacity fade with cycling, while maintaining few of the benefits found in prelithiating with elemental lithium placed in contact with the negative electrode during cell assembly. The processing described herein is designed to result in a lower degree of crystallinity and seems to form another phase of material identifiable by its x-ray scattering peaks. The unidentified phase may be a lithium silicate phase with a distinct stoichiometry relative to more well-known lithium silicate phases.

[0058] An initial effort to incorporate lithium into silicon oxide powder directly was the work of a silicon suboxide supplier. Initial effort was directed to developing prelithiated silicon oxide formed as a particulate active materials. See, U.S. patent 7,776,473 to Aramata et al. (hereinafter the '473 patent), entitled "Silicon-Silicon Oxide-Lithium Composite, Making Method, and Non-Aqueous Electrolyte Secondary Cell Negative Electrode Material," incorporated herein by references. The '473 patent involved high energy milling to form a highly crystalline product. The '473 refers to the formation of lithium silicate, Li4SiO4. While the '473 patent refers to limiting crystallinity, their product material appears to be relatively crystalline, and no other phases of lithium silicate other than Li4SiO4 are clearly observed in x- ray diffractograms. Assignee Shin-Etsu markets a lithium doped SiO.

[0059] Direct chemical pre-lithiation was suggested in the '222 patent. This was exemplified in certain embodiments in published U.S. patent application 2015 / 0115206 to Fujii et al. (hereinafter the '206 application), entitled "Predoping Method for Lithium, Lithium-Predoped Electrode, and Energy Storage Device," incorporated herein by reference. The '206 application does not show any x-ray diffractograms of other spectroscopic characterizations of the product material. Also, the product material maintains significant IRCL.

[0060] Another effort at prelithiating silicon suboxide was claimed by SK Innovations. See, published U.S. patent applications 2021 / 03673020 to Park et al„ entitled "Lithium-Doped Silicon-Based Oxide Negative Electrode Active Material, Method of Preparing the Same, and Negative Electrode and Secondary Battery Including the Same," and 2022 / 0037644 to Park et al., entitled " Lithium-Doped Silicon-Based Oxide Negative Electrode Active Material, Method of Preparing the Same, and Negative Electrode and Secondary Battery Including the Same," both of which are incorporated herein by reference. Hereinafter, these references are referred to as Park applications. The Park application have a process based on mixing silicon oxide with lithium hydroxide and then heating the blend at temperature from about 550 °C to 600°C. The reduction of the lithium would seem to require further oxidation of the silicon oxide. Park comments on the suppression of a Si crystalline phase, which is consistent with an oxidation event from reduction of lithium as well as their x-ray diffractograms.

[0061] The present improved synthesis are based on an understanding of targeting specific nature of the product composite material that involves contrary understanding of the material that provides the best performance. In particular, the material with improved cycling performance maintains a higher degree of amorphous character and avoids certain lithium silicate phases. The synthesis technique improves on some earlier work by Yan et al. (hereinafter the Yan article), "Enabling SiOx / C Anode with High Initial Coulombic Efficiency through a Chemical Pre- Lithiation Strategy for High Energy Density Lithium-Ion Batteries," ACS Applied Materials and Interfaces, DOI: 0.1021 / acsami.0c05153, 21 May 2020 (Web), incorporated herein by reference. Through the understanding of revised target composition, the method is correspondingly modified to produce the desired target materials with significantly improved performance. The improved performance has desirable parameters for a silicon based material with prelithiation in a powder form.

[0062] The method facilitates the lithiation reaction with the silicon oxide active material by forming a salt with a radical anion (RA), such as a biphenyl anion, and Li+cation in an aprotic polar solvent. Transfer of a lithium atom to the active material may result in the deposition of the ion pair (Li+RA‘) into the silicon oxide and / or in the oxidation of the radical anion back to a polycyclic aromatic to deposit lithium in some form into the silicon oxide with the release of the neutral molecule relating to the radical anion. A heat treatment is performed to anneal the material with the deposited lithium regardless of its form. The annealing step is significant with respect to obtaining a final material with desired properties. Through control of the anneal step, a material can be formed that is significantly different from materials obtained in the Yan article.

[0063] In the '473 patent, the discussion is built around the idea that certain phases of lithium silicate functioned to stabilize the material during cycling. But the idea of stabilization can have multiple dimensions. As noted above, when silicon cycles in a lithium ion battery, the material undergoes large volume changes. The substitution of silicon with silicon suboxide reduces somewhat these volume changes at the price of some capacity, but long cycling stability has proven to be a continuing challenge. Initial stabilization of the material may facilitate obtaining a higher discharge capacity since changes in the material structure may inhibit structural changes that raise the energy needed to extract the lithium. On the other hand, gradual structural changes over cycling correspondingly result in capacity fade over cycling that is manifest as a gradual drop in capacity over cycling. The concept of chemical prelithiation was built on the concept of forming a stable crystal phase that could reduce mechanical stresses in the material as a result of cycling. The results herein though demonstrate that highly crystalline domains destabilize the material with respect to cycling stability as well as reducing the specific discharge capacity.

[0064] To stabilize the material, the '473 patent teaches formation of a Li4SiO4 lithium silicate phase. Similarly, the Park applications teach formation of Li2Si20s and Li2SiOa lithium silicate phases. The initial materials generally comprise nanocry stalites of silicon embedded in the material, the prelithiation process may modify the crystalline silicon phases to varying degrees ranging from elimination of silicon diffraction peaks to increasing the crystallinity of the silicon diffraction peaks. Applicant's improved materials maintain a significant presence of crystallites of silicon as established by the presence of an x-ray diffraction (XRD) peak without significantly increasing their crystallinity, as determined by width of the XRD peaks.

[0065] The process of dissolving effectively the lithium into a solution with a negative anion forming compound in a suitable aprotic solvent provides for effective delivery of the lithium into the material. The solvated lithium ion / radical anion compound, such as Li / biphenyl, can deliver the lithium to the active material where it is energetically favorable up to a certain level of lithium deposition to imbibe into the active material. The Yan article suggests that the lithium ion-radical anion complex is absorbed into the silicon oxide, but the results below suggest that this likely is not the full picture of what is happening. Alternatively, lithium atoms can be incorporated into the silicon oxide while taking an electron from the radical anion to reform the neutral species. The radical anion is a strong reducing agent. It is plausible that a mixture of processes are occurring. Regardless, this process seems somewhat different though from the lithiation process that occurs in a cell in which the electron is provided from a current collector. To obtain a more useful structure following delivery of the lithium atoms to the silicon oxide, a anneal step is performed. It has been discovered how to control the anneal step to obtain different material structures that have favorable properties for battery use. Applicant's results indicate that it is important to avoid achieving a high degree of crystallinity, which is observed at high anneal temperatures.

[0066] Lithium- silicon-oxide materials have several different lithium silicate crystal structures available: including well known phases Li2Si20s, Li2SiO3, Li4SiO4, and LisSiCL. All of these lithium silicate phases can be placed on a three component phase diagram along a tie line connecting SiCL and Li2<3 in increasing ratios - 2:1, 1:1, 1:2 and 1:4. All of these can be observed in the prelithiation procedure for various anneal temperatures. The relative amount of lithium in these structures increases in the order presented. In addition to these well known lithium silicate phases, the phase diagram can have other phases that may be more difficult to form. For example, the existence is known of a phase LieSi2O7 (2 SiC : 3 LiC ). The complexity of these materials is evident from the growth of a plurality of these phases simultaneously in various processing contexts, which may be driven in part due to the relative amounts of silicon and oxygen. In particular, it seems significant to avoid formation of significant amounts of a crystalline Li2Si20s phase to achieve desirable electrochemical properties. This phase has the closest stoichiometery to SiCh of identified lithium silicates.

[0067] Applicant's current commercial cell designs are based on association of elemental lithium with the negative electrode at cell assembly. It has been believed that the chemical prelithiation is at least completed after the addition of electrolyte. So after assembly of the cell, near the end of the process electrolyte fluid is added. Once the electrolyte is added, the prelithiation process can begin to occur at a greater rate. Applicant has observed prelithiation reactions occurring even prior to the addition of electrolyte is the elemental lithium is allowed to directly contact the negative electrode. The prelithiation reactions occurring prior to addition of electrolyte fluid can result in reactions that are difficult to control and that produce significant amounts of heat that is not easily dissipated, which can create a significant risk from fire. The prelithiation of the active material powder simplifies the cell assembly process.

[0068] Once the active materials are prelithiated, the powders can be incorporated into processing for negative electrodes. The design of the negative electrodes with silicon-based active material is important to get a desired level of cell performance. Applicant has made dramatic progress in this regard. In addition, electrolyte selection is similarly significant with respect to obtaining a desired level of performance. Applicant's cell designs with silicon oxide based electrodes are able to achieve both high energy and high power, which can be particularly significant for certain high performance applications. In general, the negative electrode can be paired with any reasonable positive electrode suitable for a lithium ion cell, and the capacities can be appropriately adjusted for a desired balance, taking into account the reduced irreversible capacity loss.

[0069] Applicant has achieved very good cycling of silicon-based electrodes using previously developed electrolytes as described in published U.S. patent application 2020 / 0411901 to Dong et al. (hereinafter the '901 application), entitled "Lithium Ion Cells with High Performance Electrolyte and Silicon Oxide Active Materials Achieving Very Long Cycle Life Performance," incorporated herein by reference. While the '901 application achieved important breakthroughs in cycle life performance for silicon-based cells, the results herein are directed to achievement of high discharge power and high rate charging, while maintaining high energy and good cycling performance. Electrolytes particularly suitable for high rate function while maintaining good cycling with silicon oxide based electrodes is described in published U.S. patent application 2023 / 0085778 to Hays et al., entitled "Lithium Ion Cells With High Rate Electrolyte for Cells With Silicon Oxide Active Material Achieving Long Cycle Life," incorporated herein by reference.

[0070] Significant interest has been directed to high capacity negative electrode active material based on silicon. Silicon based active materials generally have not achieved suitable cycling stability for automotive use for batteries containing significant quantities of silicon. With silicon-based active materials, Applicant has demonstrated successful cycling suitable for consumer electronics applications and the like with cycling up to around 200 - 300 cycles at values of at least 80% initial capacity, see published U.S. patent application 2015 / 0050535 to Amiruddin et al., entitled "Lithium Ion Batteries With High Capacity Anode Active Materials for Consumer Electronics," incorporated herein by reference. Applicant has had particular success with respect to cycling stability has been achieved using materials primarily based on silicon oxide composites. With previous electrolyte formulations, Applicant has used effective electrode designs that can be successfully cycled for more than 800 cycles without a drop in capacity below 80% with cycling over a large voltage range at a reasonable rate, see the '901 application cited above.

[0071] Based on improved prelithiation processing of silicon oxide in combination with Applicant's electrode and electrolyte formulations, good cycling performance can be achieved with prelithiated silicon oxide as the dominant negative electrode active material. Nevertheless, to achieve even longer cycling, which can be desirable for some applications such as vehicle use, the prelithiated silicon oxide can be blended with graphite active material. The blended active material can result in an electrode that provides cycling for 800 or more cycles without capacity of the cell dropping below 80% of the initial capacity. The negative electrode may comprise at least about 25 weight percent prelithiated silicon oxide. Further discussion of electrode design is presented below. In any case, the prelithiated silicon oxide is well suited for production of commercial lithium ion cells for a range of applications, such as vehicles or consumer electronics.

[0072] Negative Electrode Active Materials and Prelithiation Process

[0073] The current negative electrodes of interest are based on silicon suboxide active material.

[0074] The material and resulting electrodes can be charactered with respect to a specific discharge capacity, a first cycle irreversible capacity loss (IRCL) and cycling stability. The prelithiation is directed to reducing the IRCL, with the objective of achieving a high specific discharge capacity and good cycling. As noted above, direct use of supplemental lithium has been very effective at lowing or eliminating IRCL as well as improving cycling stability. Lowering the IRCL is desirable to make more efficient use of the cathode active material to improve energy densities of the cells. The results herein display a tradeoff between cycling stability and IRCL, so process conditions can be selected to achieve appropriate balancing of these parameters. In any case, the resulting materials exhibit significant improvement in performance from present commercial materials as well as other known powder active materials.

[0075] While generally the silicon suboxide is represented by SiOx, 0.1<x<1.8, materials of interest have an x close to 1, such as SiOx, 0.9<x<l.l, which also is believed consistent with commercial sources. These materials can be formed in a composite with carbon to improve electrical conductivity. The amount of carbon is generally a few percent, e.g., about 0.5 wt% to about 10 wt%, and is generally amorphous. The carbon component of the composite can be introduced through high energy mechanical milling, pyrolyzing a carbon source or the like.

[0076] The anode active materials generally have a specific discharge capacity of at least about 800 mAh / g, in further embodiments at least about 900 mAh / g, in additional embodiments at least about 950 mAh / g, in some embodiments at least about 975 mAh / g and in other embodiments from about 1000 mAh / g to about 1300 mAh / g when cycled at a rate of C / 10 against lithium metal from 0.005V to 1.5V. A person of ordinary skill in the art will recognize that additional ranges of specific discharge capacity within the explicit ranges above are contemplated and are within the present disclosure. As this implies, the specific capacity of negative electrode active material can be evaluated in a cell with a lithium metal counter electrode, which is generally referred to as a half cell configuration. However, in the batteries described herein, the negative electrodes can exhibit reasonably comparable specific capacities when cycled against high capacity lithium metal oxide positive electrode active materials. In the battery with non-lithium metal electrodes, the specific capacity of the respective electrodes can be evaluated by dividing the battery capacity by the respective weights of the active materials of the electrodes.

[0077] The irreversible capacity loss (IRCL) can be evaluated in a half cell, where the IRCL = first cycle specific charge capacity- first cycle discharge specific capacity. Full cell IRCL can similarly be evaluated, and the full cell values can include multiple contributions, although the results are mainly consistent with the silicon oxide anode half cell results. The IRCL after prelithiation generally ranges from about 5% to about 20% of the initial charge capacity, in further embodiments from about 6% to about 17%, in some embodiments from about 7% to about 15%, and in other embodiments from about 8% to about 13%. The lower limits of these IRCL ranges can be mixed any of the upper limits of these ranges to form alternative IRCL ranges. A person of ordinary skill in the art will recognize that additional ranges of IRCL within the explicit ranges above are contemplated and are within the present disclosure. As noted above, the properties involve a balancing of various performance parameters. Cycling stability depends on the electrode design. Applicant has an enormous amount of experience in silicon based electrode design, and electrode design and cycling are described further below.

[0078] The prelithiation or lithium doping is performed starting with a powder of SiOx / C, which is notation to indicate an amorphous carbon composite of silicon suboxide. While the amorphous carbon component can be considered optional, from a practical perspective, it is generally desirable. The lithium doping process described herein involves dissolving lithium metal in a solution with an aromatic compound that reacts with lithium metal to form a radical anion. The resulting dissolved "salt" or complex has a strongly reducing species. The lithiumradical anion dissolved complex effectively deposits the lithium in association with the silicon oxide. The form of the deposited lithium is unclear and may be in the complex itself, as elemental metal, possibly in some other form, or some mixture of forms. Following initial delivery of lithium metal into the active material, an anneal step incorporates the lithium into the particles in alternative forms. Specifically, the lithium can form various lithium silicates with evidence of multiple phases embedded in the complex material as well as possibly lithium silicides, and the resulting character of the material determines the resulting electrochemical behavior. The particular reactants seem to determine the amount of lithium that can be deposited into the SiOx / C material. The limits on the amount of lithium that can be deposited seems to correspondingly limit the reduction of IRCL resulting from the prelithiation process.

[0079] As described herein, desirable cycling results can be obtained with a combination of a silicon based active material preloaded with lithium in particulate for assembly into an electrode. By controlling the prelithiation conditions, the material can be maintained in a low crystallinity state with little or no amounts of certain lithium silicates. The low crystallinity prelithiated materials provide surprisingly desirable performance. Through the performance of the prelithiation at the powder level, the product materials can be incorporated into cell construction processes as a substitute for active materials that were not prelithiated.

[0080] Significant dependence is found on the annealing conditions. In particular, the annealing temperature significantly influences the resulting material structure. Higher annealing temperatures results in increased crystallinity and alteration of the particular crystalline forms. The anneal temperatures also alter the electrochemical properties, which presumably track the different crystal forms that are produced. Desired electrochemical properties, in particular cycling stability, result from processing to form a less crystalline material, which is generally contrary to conventional wisdom.

[0081] SiOx / C

[0082] In general, silicon suboxide can be formed from high energy milling of particulate elemental silicon and particulate silicon dioxide / silica (S i O2), through the reduction of silica or the oxidation of Si. Regardless of the approach for forming SiOx, the silicon suboxide generally has domains of elemental silicon, which can be nanocrystalites, with various degrees of crystallinity. The resulting material can be milled and possibly sieved to obtain particulate size distributions desired. Carbon coatings can be used to impart desired electrical conductivity.

[0083] As noted above, commercial SiOx / C are available from a significant group of different material suppliers. Applicant has previously explored various approaches for forming SiO- carbon composites. See, U.S. patent 9,601,228 to Deng et al., entitled "Silicon Oxide Based High Capacity Anode Materials for Lithium Ion Batteries," incorporated herein by reference. Composites in which the carbon takes the form of graphene sheets is described in U.S. patent 10,886,526 to Anguchamy et al., entitled "Silicon-Silicon Oxide-Carbon Composites for Lithium Battery Electrodes and Methods for Forming the Composites," incorporated herein by reference. In general, it is believed that the prelithiation processes described herein should be effective for any silicon suboxide carbon composite.

[0084] The results herein do not explore performance differences which may result from use of different composite starting material. Applicant's experience suggests similar results using different commercial composite of silicon suboxide with associated conductive carbon deposit (SiOx / C) for processing. Commercial material tends to use an x ~ 1, which can be written as SiO for simplicity. The use of these suboxides have been useful for achieving good cycling properties, which may be due to a balance of specific capacity and changes resulting from the alloying with lithium. The particulate silicon oxide generally has properties not particularly sensitive to particle size, and the powder can have an average particle size from about 1 microns to about 10 microns and in further embodiments from about 1.5 microns to about 8.5 microns. The powder generally have from about 0.5 wt % to about 5 wt% carbon and in further embodiments form about 0.8 wt% to about 4 wt% carbon. A person of ordinary skill in the art will understand that additional ranges within the explicit ranges above for average particle size and carbon context are contemplated and are within the present disclosure. Lithiated Material

[0085] Lithium is associated with the silicon oxide to reduce the irreversible capacity loss and to stabilize cycling. The nature of the resulting material is found to depend significantly on the processing conditions. Lithium metal reacts spontaneously with silicon oxide in an exothermic reaction. The rate of the reaction though can be slow without facilitation of the reactions. Bare lithium metal is dissolved into a solution with aromatic species suitable for forming radical anions due to reactions with the lithium, and a resulting lithium salt solution is formed in a reducing environment. Selection of the aromatic species can influence the amount of lithium that can be deposited, and the amount of lithium deposited in turn results is differing reductions of the IRCL resulting from the pre-lithiation step. Suitable aromatics for forming radical anions include, for example, acenaphthalene, naphthalene, anthracene, pyrene, perylene, biphenyl, 2- methyl biphenyl, 3,3-dimethyl biphenyl, 4,4-dimethyl biphenyl, 3,3,4,4-tetrramethyl biphenyl and mixtures thereof. In general, the biphenyl species provide stronger reducing strength in their interactions with silicon oxide, and the methyl substituted biphenyls can have somewhat stronger reducing power relative to biphenyl. The reducing power of the radical anion may influence the amount of lithium that can be successfully delivered to the silicon oxide. The amount of lithium delivered can be correspondingly associated with the reduction of IRCL.

[0086] The lithium deposited in association with the silicon oxide using the radical anion solution should be annealed following deposition to provide for prelithiation reactions that allow for stabile material for electrode use. If not annealed, the deposited lithium has some characteristics of lithium metal and in any case remains reactive with respect to the silicon oxide active material. Applicant has discovered that the annealing step needs to balance the corresponding material changes resulting from the anneal step. In particular, to achieve desired cycling properties, the prelithiated silicon oxide should not be heated to a high temperature, such that the material can remain more amorphous.

[0087] Lithium silicates are known to exist with at least five different crystal structures, each of the five with a specific stoichiometery, between SiCL and Li2<D. These five crystal structures correspond with the following stoichiometries: Li2Si20s, Li2SiOa, Li4SiO4, LieSi2O7, and LixSiOe. A three component phase diagram for lithium silicon oxide is presented in Sivonxay et al., entitled " The lithiation process and Li diffusion in amorphous SiCL and Si from first- principles," Electrochemica Acta 331, (2020) 135344, pages 1-9, incorporated herein by reference. Four lithium silicate crystal phases are plotted on a tie line in the phase diagram between S 1O2 and Li2O. As the anneal temperature increases, the crystallinity correspondingly increases, and specific changes in the structure and electrochemistry is discussed further below. Unidentified crystal peaks are found in the Examples below. Also, the degree of crystallinity is discussed.

[0088] Unless indicated otherwise, the material processing is generally performed over an inert atmosphere, such as dry nitrogen or argon, to avoid water or oxygen which may be reactive with some of the materials under the processing conditions. To perform the prelithiation process, lithium metal is dissolved in a solution of aromatic radical ion forming compound in an aprotic polar solvent. Suitable solvents can include, for example, tetrahydro pyran (THP), tetrahydro furan (THF), methyl tetrahydro furan, dimethyl sulfoxide (DMSO), acrylamide, acetone, ethyl acetate, mixtures thereof and the like. Suitable aromatic compounds for forming radical anions are described above. The aprotic polar solvents can be characterized by a dielectric constant of at least about 5. The dissolved lithium is believed to form a salt Li+RA", where RA" is the radical anion. The radical anions generally are strong reducing agents.

[0089] The Li+RA" salt is an effective way to deliver lithium equivalents to the silicon oxide active material. The annealing process completes the lithium incorporation. An understanding if the mechanism is not required to appreciate the results and is not completely clear. Applicant's results though suggest certain aspects of the processing that are somewhat distinct from the suggested mechanism in the Yan article. While the radical anion provides a mechanism to dissolve solid lithium metal for solvation in the solvent, the form of lithium in the Silicon oxide material prior to annealing is not unambiguous, and potentially can be in more than one alternative form. The Yan article suggests that LiBP, where the radical anion is biphenyl (BP), is absorbed or otherwise coated onto the silicon oxide. The results presented herein suggest that excess lithium relative to the radical anion can be effective to further lithiate the silicon oxide. While it is not clear how this occurs, an association of reduced lithium with the silicon oxide in some form would free the RA to dissolve a further lithium atom. On the other hand, there seems to be a limit to the amount of lithium that can be effectively used, and the state of the lithium in association with the silicon oxide is not clear, although lithium metal may alloy with some components of the composite silicon oxide.

[0090] To perform the association step, the solution generally has a concentration of radical anions (assumed to equal the chemical species that are reduced to form them) from about 0.5M to about 2.5M and in further embodiments from about 0.75M to about 2M. Also, the solution can have a lithium concentration from about 0.5M to about 3 M, in further embodiments from about 0.75M to about 2.75M lithium, in additional embodiments from about 0.9M to about 2.5M, and in some embodiments from about IM to about 2.25M, where the lithium concentration refers to any form of lithium in the solution whether or not solvated These components are presumed to form a salt of a radical anion and lithium cations. The solutions have a dispersion of silicon oxide, and the amount of silicon oxide can be described with respect to the moles of SiO or equivalently as a molarity even though not dissolved. The examples describe results using a mole ratio of SiO to RA of 3:1, but other ratio values can be used. While lithium is believed to dissolve in a stoichiometric basis with the radical anions, such as forming Li+RA", it is observed that inclusion a higher ratio of lithium to radical anion, e.g., BP, results in improved prelithiation of the silicon oxide. So the mole ratio of lithium to radical anion can be from about 1:1 to about 5:1, in further embodiments from about 1.25:1 to about 5 : 1 and in other embodiments from about 1.5 : 1 to about 4.5:1. As suggested above, the lithium may or may not be dissolved initially or at any stage of the processing, and the lithium may or may not be associated with the silicon oxide. A person of ordinary skill in the art will recognize that additional ranges of concentration and concentration ratios within the explicit ranges above are contemplated and are within the present disclosure. While not wanting to be limited by theory, to the extent that lithium metal is not immediately dissolved but remains in particulate form, this initially particulate lithium or some fraction thereof presumably may further gradually dissolve as lithium is deposited into the silicon oxide to free a radical anion which can then dissolve another lithium atom.

[0091] To perform the processing, silicon oxide is dispersed in the solution and mixed during processing. The exemplified processing used a molar ratio of silicon oxide to RA of about one to one. The examples in the Yan article always maintained a molar ratio of lithium to BP of 1:1, but the article exemplified a molar ratio from about 0.067 to about 0.6 BP / Li to silicon oxide, with most of their results presented for a molar ratio of about one to three of BP / Li to silicon oxide. The results in Yan show a decrease in discharge specific capacity with increasing lithium amounts along with a decrease in IRCL (IRCL = 1 - ICE (Initial Coulombic Efficiency). Applicant observes more complicated behavior with adjustment of lithium, but the exemplified results maintain the molar ratio of silicon oxide to RA at 1 to 1.

[0092] It has been found that the initial reaction conditions for the incorporation of lithium into the silicon oxide active material also influences the resulting crystal structure of the material following annealing even with the same annealing conditions. Experimental results are presented in the Examples. These results are also consistent with the best cycling performance observed for a low crystallinity material.

[0093] The powder of silicon oxide is then blended with the prelithiation solution with dissolved lithium and radical anion salt, although the order of blending reactants may not be significant. This solution can be heated for at least a portion of the deposition period and stirred over a deposition period to result in the deposit of elemental lithium with the silicon oxide. Commercial mixers are available to perform the mixing during the lithium deposition into the silicon oxide and can match the scale of the reaction from laboratory scale to commercial scale. The incubation temperatures can be from room temperature to about the boiling point of the solvent, in some embodiments from about 25 °C to about 80 °C, in further embodiments from about 30 °C to about 65 °C, in additional embodiments from about 30 °C to about 55 °C and in other embodiments from about 33 °C to about 52°C. In some embodiments, incubation can be performed over different temperatures for selected times, such as at heated temperatures for a first time and room temperature for a second time, or other desired divisions. The Examples have some results using an initial soak time at a heated temperature followed by a longer soak time at room temperature, as used in the Yan article, but improved results were obtained with soaking / incubation at a single temperature for a selected period of time. Total incubation times can be from about 10 minutes to 48 hours, although longer times can be used as desired. In further embodiments, incubation times can be from about 15 minutes to about 40 hours, in other embodiments from about 25 minutes to about 35 hours and in other embodiments from about 30 minutes to about 30 hours. Result herein suggest for the heated incubation that longer soak times at lower temperatures can be beneficial, although at a shorter soak time, results improve with somewhat higher reaction temperatures. As noted above, the total incubation times can be divided into different heating increments at a distinct temperature each over a fraction of the total time. A person of ordinary skill in the art will recognize that additional ranges of incubation temperatures or soak times within the explicit ranges above are contemplated and are within the present disclosure. The amount of lithium deposited with the silicon oxide powder generally reaches a steady state value, and more lithium is not observed to be deposited if provided to the solution. Correspondingly, the deposited amount of lithium generally drives the observed reduction in IRCL, although this is also determined by the anneal conditions as explained below.

[0094] A review of x-ray diffractograms (XRD) for the final prelithiated material that are incubated for initial lithium reaction for different times at different temperatures show that similar materials, as characterized by crystal structure, are found that exhibit similar performance properties. All of these materials were annealed at 450 °C for 24 hours. For example, material incubated at 40 °C for 24 hours has a similar ultimate crystal structure after an anneal step with material incubated at 70 °C for two hours to initially lithiate the silicon oxide, and these two materials have similar cycling stability. In particular, the x-ray structures show the introduction of a low degree of crystallinity, which is characterized by a broad silicon oxide amorphous band and broadened peaks extending from the amorphous background corresponding with overall a low degree of crystallinity. These results are consistent with the overall view of a low degree of crystallinity being desirable, but increased degree of lithium silicate crystallinity leading to decreased cycling stability. As discussed further in the context of the annealing temperatures, the crystal peaks that show up with annealing at 450 °C are not aligned with any expected lithium silicates structures, while annealing at higher temperatures result in XRD peaks identifiable. as known crystalline lithium silicates.

[0095] After completing the incubation or soaking process, the powder with deposited lithium is separated from the solution. Suitable separation can be based on filtering or centrifugation and decanting. The separated powder can be washed and reseparated to obtain desired powder separated from the reactants of the soaking process, such as solvent, aromatics and / or lithiumradical anion salts.

[0096] To further drive the lithium into the silicon oxide material and complete the reaction of the lithium, an anneal step is performed. In particular, after the soak / incubation step, annealing is found to provide for desirable material properties in resulting cells. But as noted above, the anneal step is desirably performed at a temperature that is not too high, as explained in more detail in the following. The materials with the desirable performance have broad x-ray scattering peaks that do not correspond with known lithium silicate peaks. Thus, the results are consistent with achieving a particular, low degree of crystallinity, while avoiding a high degree of crystallinity.

[0097] The process conditions significantly influence the structure of the product material as determined by x-ray diffraction. In particular, anneal temperatures are explored between 350 °C and 950 °C, with a control having no anneal step. Scattering from silicon crystallites is consistent through the samples. The anneal time does not seem to alter the observed crystal phases, as determined by the prominence of the crystal peaks in the x-ray diffractograms (XRD). At 350 °C and at 450 °C, small scattering peaks can be observed, which do not correspond with known lithium silicate peaks, and these peaks are currently unidentified, as discussed further below. At 550 °C and higher, scattering peaks are observed that correspond with known lithium silicates.

[0098] At 550 °C, the most prominent lithium silicate peaks correspond with Li2SiO3. at 650 °C and higher multiple crystal structures of lithium silicate can be seen and Li2Si20s makes contributions at these higher temperatures. At an anneal temperature of 550 °C, the contribution from Li2Si20s is insignificant with an XRD peak height a small fraction of the Li2SiO3 XRD peak height and not readily resolved to have a clear peak identity. Applicant's results are consistent with the Ei2Si20s contributing to a loss of cycling performance, although it is not suggested that a simple explanation can account fully for the results. Results also suggest that the pre-anneal lithiation process influences the resulting ultimate crystal structure. With annealing at 750 °C, Applicant's results include contributions form Li2SiOa and Ei2Si20s, while the Yan article finds only finds contributions from Li2Si20s. The Yan article presents XRD with some remaining silicon oxide amorphous background. It seems possible this is a result of a lower degree of lithium incorporation. The observed IRCL though is roughly comparable.

[0099] The material providing the desirable cycling results have XRD peaks that are not identifiable with known lithium silicate phases. The most prominent of these peaks are at Cu- Ka29 values (degrees): 21-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3. As explained in the examples, some of these peaks have similarities to LiSi alloy. LiSi alloys though can have varying amounts of lithium. The XRD spectra for Li2iSis are similar, see Iwamura et al., "Li-Rich LiSi Alloy As A Lithium-Containing Negative Electrode Material Towards High Energy Lithium-Ion Batteries," Scientific Reports 5:8085, page 1-8 (January 2015, DOI: 10.1038 / srep08085), incorporated herein by reference. In contrast, the XRD spectra for LiisSi4 has significantly different reported peaks, see Obrovac et al., "Structural Changes in Silicon Anodes during Lithium Insertion / Extraction," Electrochemical and Solid-State Letters, 7 (5) A93-A96 (March 2004), incorporated herein by reference. It is plausible, but not clear, that some or all of the unidentified peaks are associated with Li-Si alloying.

[0100] Visually, it is clear that annealing at higher temperatures increases the crystallinity. In part, this can be described in terms of what would appear to be a broad amorphous baseline at lower values of 29, which is reminiscent of amorphous silica. Also, the crystalline peaks change and sharpen. Broadening of the XRD peaks can be associated to small crystallite sizes, which can be approximated with the Scherrer equation. The Scherrer equation evaluates crystallite size according to a measure of the full- width at half maximum of a broadened XRD peak. Since silicon crystallite peaks are observed through all of the relevant spectra, we look at the full width half maximum (FWHM) for the silicon peaks. In particular, the silicon peak at 2theta of about 28.4 degrees is the strongest peak. While a broad baseline of seeming amorphous character complicates peak resolution, the more amorphous samples have values of FWHM in 2theta of at least about 1.25 degrees, and ranges including intermediate crystallinity samples have values of FWHM of at least about 1.5 degrees, and further embodiments have values of FWHM of at least about 1.75 degrees. A person of ordinary skill in the art will recognize that additional ranges of values for FWHM are contemplated and are within the present disclosure.

[0101] Generally, as the crystallinity increases, the specific discharge capacity is observed to decrease, but the IRCL also decreases. To some degree, these two changes balance somewhat, but the dominant effect is a loss of cycling stability at higher anneal temperatures. Cycling stability is significant, so high anneal temperatures are found to be undesirable. This result is contrary to the conventional wisdom. It has been thought that the structure of the lithium silicate crystals would stabilize the material with respect to cycling. The loss of specific capacity with increased crystallinity indicates that removal of lithium from the alloy is not facilitated, and this may be due to formation of specific irreversible phases, but this is not completely consistent with the loss of cycling stability. The loss of cycling stability suggests that removal of lithium from the more crystalline material results in irreversible lattice damage, which seems to get more significant as cycling continues.

[0102] In summary, the material that provides the desirable cycling performance is processed at lower annealing temperature. This material has unidentified XRD peaks diagnostic of phases indicative of a new composite material having been formed. At least some of these peaks may be from a lithium- silicon alloy, but that is not clear. The material also has a lower degree of crystallinity, which can be characterized by the prevalence of relatively prominent lithium silicate phases, diminished broad background at lower scattering angles and sharper XRD peaks, which are characterized specifically with respect to the silicon crystallite peaks.

[0103] General Battery Features

[0104] The negative electrode and positive electrode structures can be assembled into appropriate cells. As described further below, the electrodes are generally formed in association with current collectors to form electrode structures. A separator is located between a positive electrode and a negative electrode to form a cell. The separator is electrically insulating while providing for at least selected ion conduction between the two electrodes. A variety of materials can be used as separators. Some commercial separator materials can be formed from polymers, such as polyethylene and / or polypropylene that are porous sheets that provide for ionic conduction. Commercial polymer separators include, for example, the Celgard® line of separator material from Celgard, LLC, a subsidiary of Asahi Kasei (Japan). Also, ceramic -polymer composite materials have been developed for separator applications. These ceramic composite separators can be stable at higher temperatures, and the composite materials can reduce the fire risk. Polymer-ceramic composites for lithium ion battery separators are sold by Celgard® as well as under the trademarks Separion® by Evonik Industries, Germany and Lielsort® by Teijin Lielsort Korea Co., Ltd. Also, separators can be formed using porous polymer sheets coated with a gel-forming polymer. Such separator designs are described further in U.S. patent 7,794,511 B2 to Wensley et al., entitled "Battery Separator for Lithium Polymer Battery," incorporated herein by reference. Suitable gelforming polymers include, for example, polyvinylidene fluoride (PVDE), polyurethane, polyethylene oxide (PEG), polypropylene oxide (PPO), polyacylonitrile, gelatin, polyacrylamide, polymethylacrylate, polymethylmethacrylate, polyvinylacetate, polyvinylpyrrolidone, polytetraethylene glycol diacrylate, copolymers thereof, and mixtures thereof.

[0105] Electrolyte provides for ion transport between the anode and cathode of the battery during the charge and discharge processes. The electrolytes for lithium ion batteries incorporate non-aqueous solvents and lithium salts. Suitable electrolytes for silicon based electrodes are described further below. The electrolytes generally are infused into the cell prior to sealing the case.

[0106] The electrodes described herein incorporating prelithiated silicon oxide can be assembled into various commercial cell / battery designs such as prismatic shaped batteries, wound cylindrical cells, coin cells, or other reasonable cell / battery designs. The cells can comprise a single pair of electrodes (possibly wound or folded) or a plurality of pairs of electrodes assembled in parallel and / or series electrical connection(s). Electrode stacks can have an extra electrode to end the stack with the same polarity as the other end of the stack for convenience in placement in a container.

[0107] In some embodiments, the positive electrode and negative electrode can be stacked with the separator between them, and the resulting stacked structure can be rolled into a cylindrical or prismatic configuration to form the cell structure. Appropriate electrically conductive tabs can be welded or the like to the current collectors, and the resulting jellyroll structure can be placed into a metal canister or polymer package, with the negative tab and positive tab welded to appropriate external contacts. Electrolyte is added to the canister, and the canister is sealed, possibly after formation cycling, to complete the cell. Some presently used rechargeable commercial cells include, for example, the cylindrical 18650 cells (18 mm in diameter and 65 mm long) and 26700 cells (26 mm in diameter and 70 mm long), although other cell / battery sizes can be used, as well as prismatic cells and foil pouch cells / batteries of selected sizes.

[0108] Pouch batteries can be particularly desirable for various applications, including certain vehicle applications, due to stacking convenience and relatively low container weight. A pouch battery design for vehicle batteries incorporating a high capacity cathode active material is described further in U.S. patent 8,187,752 to Buckley et al, entitled "High Energy Lithium Ion Secondary Batteries” and U.S. patent 9,083,062B2 to Kumar et al., entitled "Battery Packs for Vehicles and High Capacity Pouch Secondary Batteries for Incorporation into Compact Battery Packs," both incorporated herein by reference. While the pouch battery designs are particularly convenient for use in specific battery pack designs, the pouch batteries can be used effectively in other contexts as well.

[0109] A representative embodiment of a pouch battery is shown in Figs. 1 A to ID. In this embodiment, pouch battery 100 comprises pouch enclosure 102, electrode core 104 and pouch cover 106. An electrode core is discussed further below. Pouch enclosure 102 comprises a cavity 110 and edge 112 surrounding the cavity. Cavity 110 has dimensions such that electrode core 104 can fit within cavity 110. Pouch cover 106 can be sealed around edge 112 to seal electrode core 104 within the sealed battery, as shown in Figs. IB and 1C. Terminal tabs 114, 116 extend outward from the sealed pouch for electrical contact with electrode core 104. Fig. C is a schematic diagram of a cross section of the battery of Fig. IB viewed along the C-C line. Many additional embodiments of pouch batteries are possible with different configurations of the edges and seals.

[0110] Fig. ID shows an embodiment of an electrode core 104 that generally comprise an electrode stack. In this embodiment, electrode stack 130 comprises negative electrode structures 132, 134, 136, positive electrode structures 138, 140, and separators 150, 152, 154, 156 disposed between the adjacent positive and negative electrodes. The separator can be provided as a single folded sheet with the electrode structures placed in the separator folds to form a particularly convenient stack structure. Negative electrode structures 132, 134, 136 comprise negative electrodes 160, 162, negative electrodes 164, 166 and negative electrodes 168, 170, respectively, disposed on either side of current collectors 172, 174, 176. Positive electrode structures 138, 140 comprise positive electrodes 180, 182 and positive electrodes 184, 186, respectively, disposed on opposite sides of current collectors 188, 190, respectively. Tabs 192, 194, 196, 198, 200 are connected to current collectors 172, 188, 174, 190, 176, respectively, to facilitate the connection of the individual electrodes in series or in parallel. For vehicle applications, tabs are generally connected in parallel, so that tabs 192, 196, 200 would be electrically connected to an electrical contact accessible outside the container, and tabs 194, 198 would be electrically connected to an electrical contact as an opposite pole accessible outside the container. Electrode stacks can have an extra negative electrode such that both outer electrodes adjacent the container are negative electrodes. The end electrode structures can have an electrode only on one side of the current collector to face the interior of the cell and adjacent positive electrode. In some embodiments, a battery with stacked electrodes of the dimensions described herein have from 5 to 40 negative electrode elements (current collector coated on both sides with active material) and in further embodiments from 7 to 35 negative electrode elements with corresponding numbers of positive electrode elements being generally one less than the negative electrode elements. A person of ordinary skill in the art will recognize that additional ranges of electrode numbers within the explicit ranges above are contemplated and are within the present disclosure.

[0111] As noted above, wound electrodes can be correspondingly used for either a cylindrical battery or a roughly prismatic shaped battery. Wound cells for cylindrical lithium ion batteries are described further in U.S. patent 8,277,969 to Kobayashi et al., entitled "Lithium Ion Secondary Battery," incorporated herein by reference. Prismatic shaped batteries with wound electrodes are described in U.S. patent 7,700,221 to Yeo (the '221 patent), entitled "Electrode Assembly and Lithium Ion Secondary Battery Using the Same," incorporated herein by reference. The Kobayashi '969 patent and the Yeo '221 patent do not describe how to achieve reasonable cycling or a high energy density with silicon based active materials. Designs for prismatic shaped batteries with wound electrodes are described further, for example, in the '221 patent cited above. A particular design of either a stacked set of electrodes or a wound cell can be influenced by the target dimensions and the target total capacity of the battery.

[0112] The improved negative electrodes can be used for a range of applications and cell / battery designs. For electrode stacks, the areas of the electrodes can be selected reasonably based on the volume and design constraints for the particular application. The following discussion focuses on larger cells generally designed for vehicle applications, such as drones, automobiles, trucks, or other vehicles. However, the improved negative electrodes described herein can be effectively used for consumer electronics applications, which can be based on smaller cell formats. Also, it should be noted that vehicles can use smaller consumer electronics cells, and Tesla cars presently are famous for using thousands of small cells in their battery packs. Generally, larger format cells / batteries can achieve larger energy densities within certain ranges. It may be desirable to select positive electrode active materials based on the particular application to balance various considerations, such as energy densities and voltage ranges. With the selection of electrode parameters, the design of high gravimetric energy density cells can incorporate a balance of factors including electrode area, the number of electrode structures, and battery capacity. The electrode area refers to the spatial extent of one of the electrodes along one side of a current collector. Fig. 1A depicts the length “L”, and Fig. 1C depicts the width “W” of an electrode. As shown in the Figures, the area of an electrode can be defined as L x W. In some embodiments, the area of each electrode can be similar such that the dimensions of a battery comprising an electrode stack can have a length and width that are similar to the length and width of each electrode in the stack.

[0113] Electrode Structures

[0114] The electrodes of the cell comprise the active material along with a binder and conductive additives. The electrodes are formed into a sheet, dried and pressed onto a current collector to achieve a desired density and porosity. The electrode sheets are generally formed directly on a metal current collector, such as a metal foil or a thin metal grid. For many cell structures, electrode layers are formed on both sides of the current collector to provide for desirable performance in the assembled cell or battery. The electrode layers on each side of the current collector can be considers elements of the same electrode structure since they are at the same potential in the cell, but the current collector itself, while part of the electrode structure is not generally considered part of the electrode since it is electrochemically inert. Thus, references to the physical aspects of an electrode generally refer to one layer of electrode composition within the electrode structure. An electrically conductive current collector can facilitate the flow of electrons between the electrode and an exterior circuit.

[0115] In some embodiments, when the positive electrode or negative electrode uses a high loading level, the density of the electrode can be reduced to provide good cycling stability of the electrode. The density of the electrodes generally is a function, within reasonable ranges, of the press pressures. The density of the electrodes cannot be arbitrarily increased without sacrificing performance with respect to loading levels while achieving desired cycling performance and capacity at higher discharge rates.

[0116] In some embodiments, a current collector can be formed from nickel, aluminum, stainless steel, copper or the like. An electrode material can be cast as a thin film onto a current collector. The electrode material with the current collector can then be dried, for example in an oven, to remove solvent from the electrode. The electrode material on the current collector is not necessarily subjected to a high pressure and may be simply pressed or calendered to form a cohesive structure. In some embodiments, a dried electrode material in contact with a current collector foil or other structure can be subjected to a pressure from about 2 to about 10 kg / cm2(kilograms per square centimeter), which would generally densify the electrode. The current collector used in the positive electrode can have a thickness from about 5 microns to about 30 microns, in other embodiments from about 10 microns to about 25 microns, and in further embodiments from about 14 microns to about 20 microns. In some embodiments, the positive electrode uses an aluminum foil current collector. The current collector used in the negative electrode can have a thickness from about 2 microns to about 20 microns, in other embodiments from about 4 microns to about 14 microns, and in further embodiments from about 6 microns to about 10 microns. In some embodiments, the negative electrode uses copper foil or nickel foil as current collector. A person of ordinary skill in the art will recognize that additional ranges of current collector thicknesses within the explicit ranges above are contemplated and are within the present disclosure.

[0117] Negative Electrodes

[0118] The basic electrode design comprises one or a blend of active compositions, polymer binder, and an electrically conductive diluent. As noted above, in some embodiments, improved electrode designs can involve a polymer binder blend and optionally a blend of active compositions as well as nanoscale conductive additives, such as conductive carbons or inert metal particulates. While the active material can be solely a silicon based material or composite, an active material blend can comprise in some embodiments a majority of silicon based active material, such as a silicon oxide composite, and a selected amount of electroactive graphite.

[0119] Also, it has been discovered that stabilization of the electrode cycling with silicon oxide based active materials can obtained with a binder that is solvent based or water based. While any graphite can provide electrical conductivity to the electrode, it has also been found that in some embodiments a quantity of distinct nanoscale conductive additives, such as carbon, nevertheless can be significant toward the ability to produce a long cycling negative electrode. In relevant embodiments, the nanoscale conductive carbon is not believed to be electrochemically active while graphite can be added to be electrochemically active. These electrode design aspects are then incorporated into negative electrodes with the improved prelithiated silicon oxide active material.

[0120] Significant interest has been directed to high capacity negative electrode active material based on silicon. Applicant has achieved good cycling properties with silicon oxide based active materials with high energy and high power for a range of applications while using significant quantities of silicon oxide in the electrodes. With suitable electrolyte formulations, Applicant has used effective electrode designs that can be successfully cycled for more than 800 cycles without a drop in capacity below 80% with cycling over a large voltage range at a reasonable rate, see the '901 application cited above. The present work is directed to using electrodes with prelithiated silicon oxide powder that can be incorporated into more conventional cell manufacturing lines without the need to introduce supplemental lithium into the cell manufacturing process.

[0121] The use of the prelithiated silicon oxide can adopt different approaches. In one limit, the prelithiated silicon oxide can be the sole anode active material in the cell. At the opposite extreme, the active material in the anode can be primarily carbon based with a lesser amount of prelithiated silicon oxide. For cells based on primarily or exclusively prelithiated silicon oxide, the cell designs, including, for example, electrode design, electrolyte selection, cell stack design, and the like, can be based on Applicant's previous cell development using silicon based material, as cited above. If the active material is primarily carbon based active materials, the cell design can be based primarily on conventional designs with the prelithiated silicon oxide incorporated in to boost anode capacity. In this context, the range of prelitiated silicon oxide as a fraction of the anode active material can be from about 15 wt% to about 100 wt%, in further embodiments from about 20 wt% to about 95 wt%, and in additional embodiments from about 25 wt% to about 90 wt%. Exemplified cells have 95 wt% prelithiated silicon oxide and 5 wt% active graphite or 70 wt% prelithiated silicon oxide and 30 wt% active graphite active material. Applicant is developing improved cell designs for about 15 wt% to about 45 wt% prelithiated silicon oxide. Correspondingly, the remaining portion of the anode active material if it is not 100 % prelithiated silicon oxide can be active graphite. In some embodiments, the blended active material can comprise from about 1 wt% graphite to about 85 wt% graphite, in further embodiments from about 2 wt% graphite to about 80 wt% graphite, in additional embodiments from about 5 wt% graphite to about 75 wt%, and in other embodiments from about 10 wt% graphite to about 70 wt% graphite. Desirable graphites are described below. A person of ordinary skill in the art will recognize that additional ranges of percentages of prelithated silicon oxide or graphite as a fraction of the anode active material are contemplated and are within the present disclosure.

[0122] As described herein, good cycling results are obtained with a blended active composition with a silicon based active material and active graphitic carbon. Cycling results are presented for 70 wt% prelithiated silicon oxide and 30 wt% active graphite and for 95 wt% prelithiated silicon oxide and 5 wt% active graphite. For these cells, the primary contribution to the capacity is from the prelithiated silicon oxide active material. Generally, the specific discharge capacity of the prelithiated silicon oxide active material can be at least about 750 mAh / g, in further embodiments at least about 900 mAh / g, in additional embodiments at least about 1000 mAh / g, in other embodiments at least about 1100 mAh / g, and in some embodiments at least about 1200 mAh / g cycled against lithium metal from 5 millivolts (mV) to 1.5V at a rate of C / 3. The prelithiated silicon oxide active material can also be characterized by the irreversible capacity loss (IRCL), as described above. Following prelithiation, the silicon oxide based active material can result in a silicon oxide based active material with an IRCL of no more than 25% of the initial charge capacity, in further embodiments no more than about 21% and in some embodiments no more than about 18% of the initial charge capacity. The IRCL values can be generated at a rate of C / 20 or less. The specific discharge capacity and IRCL can be evaluated in a half cell format against a lithium metal foil counter electrode. The IRCL values reported in the Examples involve either just prelithiated silicon oxide as the active material or a blend of silicon based active material and graphite active material. A person of ordinary skill in the art will recognize that additional ranges of specific discharge capacity and IRCL within the explicit ranges above are contemplated and are within the present disclosure.

[0123] As noted above and described in detail below, suitable silicon based active materials can comprise a composite with a carbon component. Silicon based active materials are discussed in detail in the following section. A composite refers to a particulate material with components that are intimately combined into an integral material with effective uniformity over appropriate scales, in contrast with blends that involve mixtures held together with a polymer binder. Composite components that can comprise, for example, silicon, oxygen, carbon and the like. While not wanting to be limited by theory, it is not generally believed that a carbon component of a composite with silicon is active in electrochemistry and generally not graphitic, although the activity is an abstract concept given the intimate combination in the composite and the crystal structure may be extremely complex and difficult to evaluate. In any case, the carbon component of a composite material is readily understood by a person of ordinary skill in the art to be distinguishable from the distinct graphite not in a composite in active material blends. The examples below are based on a commercial composite composition believed to be comprising primarily of silicon suboxide with some amounts of elemental silicon crystals and elemental carbon in a combined composite particulate material.

[0124] Graphite is available commercially in natural and synthetic forms, and suitable graphite includes either natural or synthetic graphite or the like. Graphite is a crystalline form of carbon with covalently bonded carbon in sheets. As used herein, graphite refers to graphitic carbon without requiring perfect crystallinity, and some natural graphite materials can have some crystalline impurities. But the graphite refers generally to a material dominated by a graphitic structure, as would be recognized in the art. Graphite is electrically conductive along the plane of the covalent carbon sheets that are stacked in the crystal. The crystalline carbon in graphitic forms can intercalate lithium, so that it is an established electrochemically active material for lithium ion batteries, although the graphite particle morphology can influence the efficacy of the graphite for lithium intercalation.

[0125] Graphite particles can have average particle diameters from about 1 micron to about 30 microns, in further embodiments from about 1.5 microns to about 25 microns, and in other embodiments from about 2 microns to about 20 microns. In general, it is desirable for the graphite to not include particles greater than the electrode thickness to avoid a bumpy electrode surface, and graphitic particles with a size significantly less than a micron can be less crystalline. In some embodiments, the graphitic carbon can have a D50 (mass median diameter) from about 5 microns to about 50 microns, in further embodiments from about 7 microns to about 40 microns and in additional embodiments from about 10 microns to about 8 microns to about 30 microns. Also, in some embodiments the BET surface area of graphitic carbon active material (which can be evaluated according to ISO 4652) can be from about 1 m2 / g to about 50 m2 / g, in further embodiments from about 1.5 m2 / g to about 35 m2 / g and in additional embodiments from about 2 m2 / g to about 25 m2 / g. A person of ordinary skill in the art will recognize that additional ranges of particle size and surface area for graphitic carbon active materials are contemplated and are within the present disclosure. In comparison, electrically conductive carbon blacks or the like (which have been referred to as paracrystalline) generally have surface areas of at least roughly 40 m2 / g to 1000 m2 / g or greater.

[0126] With respect to the polymer binder, Applicant has obtained reasonable cycling of silicon based cells using high tensile strength binders, e.g., polyimide binder. See U.S. patent 9,601,228 to Deng et al. (hereinafter the '228 patent) entitled: “Silicon Oxide Based High Capacity Anode Materials for Lithium Ion Batteries,” incorporated herein by reference. In some embodiments to obtain longer cycling stability, it has been surprisingly found that a polymer binder blend further stabilizes cycling. In particular, a second polymer or combination of polymers providing a lower elastic modulus (corresponding with greater elasticity) can be blended with high tensile strength polyimide. The binder blend generally comprises at least about 50 wt% polyimide, in further embodiments at least about 55 wt% and in other embodiments from about 60 wt% to about 95 wt% polyimide. Similarly, the binder blend generally comprises at least about 5 wt% polymer with a lower elastic modulus, in further embodiments at least about 10 wt%, and in other embodiments from about 12 wt% to about 40 wt% lower elastic modulus polymer, as specified further below. A person of ordinary skill in the art will recognize that additional ranges of polymer quantities within the explicit ranges above are contemplated and are within the present disclosure. The polymers of the blend can be selected to be soluble in the same solvents. Polymer blends effective to achieve good cycling of silicon-based anodes are summarized below and are described further in U.S. patent 11,094,925 to Venkatachalam et al. (hereinafter the '925 patent), entitled "Electrodes with Silicon Oxide Active Materials for Lithium Ion Cells Achieving High Capacity, High Energy Density and Long Cycle Life Performance," incorporated herein by reference.

[0127] Polyimides are polymers based on repeat units of the imide monomer structure. The polyimide polymer chain can be aliphatic, but for high tensile strength applications, the polymer backbone generally is aromatic with the polymer backbone extending along the N- atom of the polyimide structure. Lor silicon-based anodes that exhibit significant morphological changes during cycling, thermally curable polyimide polymers have been found desirable for high capacity negative electrodes, which may be due to their high mechanical strength. Table 2 provides suppliers of high tensile strength polyimide polymers, and names of corresponding polyimide polymers.

[0128] TABLE 2

[0129] The polyimide polymers can have a tensile strength of at least about 60MPa, in further embodiments at least about lOOMPa and in other embodiments at least about 125 MPa. Some commercial polyimides with high tensile strength can also have relatively high elongation values, which is the amount of elongation tolerated before the polymer tears. In some embodiments, the polyimides can have an elongation of at least about 40%, in further embodiments at least about 50% and in other embodiments at least about 55%. Tensile strengths and elongation values can be measured according to procedures in ASTM D638-10 Standard Test Method for Tensile Properties of Plastics or ASTM D882-91 Standard Test Method for Tensile Properties of Thin Plastic Sheeting, both of which are incorporated herein by reference. Based on values reported by commercial suppliers, the results from these alternative ASTM protocols seem similar to each other for polyimides. A person of ordinary skill in the art will recognize that additional ranges of polymer properties within the explicit ranges above are contemplated and are within the present disclosure.

[0130] Suitable more flexible polymer components can be selected to be inert with respect to the electrochemistry of the cell and to be compatible with processing with the polyimide. In particular, suitable more flexible polymer components include, for example, PVDF, carboxy methylcellulose (CMC), styrene-butadiene rubber (SBR), lithiated polyacrylic acid (LiPAA), or mixtures thereof. With respect to polymer properties, some significant properties for high capacity negative electrode application are summarized in Table 3.

[0131] TABLE 3

[0132] PVDF, CMC, and SBR are available commercially from many sources. LiPAA can be made from LiOH and commercial poly aery lie acid (PAA). For example, a stoichiometric amount of LiOH can be added to a solution of PAA with one mole of LiOH per monomer unit of PAA. The formation and use of LiPAA is described further in Li et al., "Lithium polyacrylate as a binder for tin-cobalt-carbon negative electrodes in lithium-ion batteries," Electrochemica Acta 55 (2010) 2991-2995, incorporated herein by reference.

[0133] The elongation refers to the percent elongation prior to tearing of the polymer. In general, to accommodate the silicon-based materials, it is desirable to have an elongation of at least about 30%, in some embodiments at least about 50%, and in further embodiments at least about 70%. For the polymer binder blend, it can be desirable for the more elastic polymer binder component to have an elastic modulus (alternatively referred to as Young's modulus or tensile modulus) of no more than about 2.4 GPa, in further embodiments no more than about 2.25 GPa, in other embodiments no more than about 2 GPa, and in additional embodiments no more than about 1.8 GPa. A person of ordinary skill in the art will recognize that additional ranges of more elastic polymer component properties within the explicit ranges above are contemplated and are within the present disclosure.

[0134] To form the electrode, the powders can be blended with the polymer in a suitable liquid, such as a solvent for dissolving the polymer. Polyimides and PVDF can generally be processed in N-methyl pyrrolidone (NMP), although other suitable organic solvents may be used. Water processable polyimides are commercially available, and these water processable polyimides are suitable for blending with a wider range of other polymers. The particulate components of the electrode, i.e., the active material and nanoscale conductive carbon, can be blended with the polymer binder blend in the solvent to form a paste. The resulting paste can be pressed into the electrode structure.

[0135] While the polymer binders described above have facilitated achievement of impressive cycling stability for silicon-based negative electrodes, water-based polymer binders have been developed that can achieve comparable or better cycling performance. In some embodiments, a polymer binder that is water processable is a copolymer of acrylic acid salt monomers (S- PAA), where the cation (S) can be a metal cation, and acrylamide monomers. Desirable polymer binders are found to have both good adhesion with respect to the current collector and good particle to particle cohesion. The S-PAA monomers can contribute good cohesion, and the acrylamide co-monomer can contribute good adhesion. Binders based on a copolymer of S-PAA and acrylamide can be provided with appropriate proportions of the monomers to result in excellent cycling with silicon based negative electrode active materials. Water-based polymer blends effective to achieve excellent cycling of silicon-based anodes are summarized below and are described further in published U.S. patent application 2022 / 0006090 to Hays et al. (hereinafter the '090 application), entitled "Lithium Ion Cells With Silicon Based Active Materials and Negative Electrodes with Water-Based Binders Having Good Adhesion and Cohesion," incorporated herein by reference.

[0136] The simultaneous achievement of good adhesion and good cohesion is found to be significant for achieving improved cycling. The monomer units of the copolymer binder are an acrylamide and a salt of polyacrylic acid (S-PAA). The salt cation can be a metal cation, such as lithium (LiPAA) or sodium (NaPAA) or potassium (KPAA), although other metals can be used or non-metal cations, such as ammonium NH4+. A mixture of counter ions can be used if desired. S-PAA polymers are found to contribute strong cohesion to the corresponding electrodes. Cohesion can be evaluated for the electrode structure on the current collector with bending around a mandrel with a particular diameter. Adhesion is evaluated using commercial testing equipment with forces applied in a controlled fashion to evaluate the forces to pull the electrode from the current collector. The ratio of monomer units can be selected to achieve a desired balance of adhesive and cohesive stability.

[0137] The copolymer can be synthesized following published methods. Specifically, the acrylate groups can be formed from a nucleophilic substitution of amide groups by carboxylate groups with the extent of the reaction controlling the degree of acrylamide groups in the final polymer. Alternatively, the copolymer can be formed directly through the polymerization reactions. With the direct synthesis the relative amounts of carboxylate groups and acrylamide groups can be controlled. Also, the molecular weight of the polymers can be appropriately controlled or selected.

[0138] The molar ratio of acrylamide moieties to S-PAA moieties can range from about 5:95 to about 95:5, in further embodiments from about 10: 90 to about 90:10, in additional embodiments from about 20:80 to about 80:20, in other embodiments from about 25:75 to about 75:25, and in some embodiments from about 30:70 to about 70:30. With respect to average molecular weight, the copolymer can have in some embodiments an average molecular weight from about 50,000 Daltons to about 5,000,000 Daltons, in further embodiments from about 75,000 Daltons to about 2,000,000 Daltons, and in other embodiments from about 100,000 Daltons to about 1,000,000 Daltons. A person of ordinary skill in the art will recognize that additional ranges of moiety ratios and average molecular weight within the explicit ranges above are contemplated and are within the present disclosure.

[0139] While desirable cycling results have been achieved with the copolymers alone as the electrode binders, Applicant has had success in improving binder performance using polymer blends. The copolymers described herein may be useful also in polymer blends. Suitable polymer blends would generally include at least 25 weight percent poly(acrylamide-co-M- PAA), in further embodiments at least about 35 wt%, and in other embodiments from about 40 wt% to about 90 wt%. A person of ordinary skill in the art will recognize that additional ranges within the explicit polymer blend rations above are contemplated and are within the present disclosure. Commercial aqueous electrode binders are sold by Sumitomo Seika Chemicals Co., as Aquacharge®. The active material loading in the binder can be large. In some embodiments, the negative electrode has from about 75 to about 94 wt% of negative electrode active material, in other embodiments from about 77 to about 93 wt% of the negative electrode active material, and in further embodiments from about 80 to about 92 wt% of the negative electrode active material. In some embodiments, the negative electrode has from about 4 to about 20 wt% polymeric binder, in other embodiments about 5 to 19 wt% polymeric binder, and in further embodiments from about 6 to 18 wt% polymeric binder. Also, in some embodiments, the negative electrode comprises from about 1 to about 7 wt% nanoscale conductive carbon, in further embodiments form about 1.5 to about 6.5 wt%, and in additional embodiments from about 2 to about 6 wt% nanoscale conductive carbon. A person of ordinary skill in the art will recognize that additional ranges of polymer loadings within the explicit ranges above are contemplated and are within the present disclosure.

[0140] For improved cycling negative electrodes, nanoscale conductive additives or combinations thereof, which are particulates, have been found to be particularly desirable. Nanoscale conductive carbon refers generally to particles of high surface area elemental carbon having at least two dimensions of the primary particles being submicron. Suitable nanoscale conductive carbon includes, for example, carbon black, carbon nanotubes and carbon nanofibers. In some embodiments, the nanoscale conductive carbon additive used in the negative electrode can comprise carbon nanotubes, carbon nanofibers, carbon nanoparticles (e.g., carbon black), or combinations thereof. Other nanoscale conductive additives include, for example, metal nanoparticles, metal nanofibers, metal nanowires, other metal nanoparticulates, and combinations thereof, such as silver nanoparticles, silver nanowires and the like. In some embodiments, to achieve improved performance a conductive additive can have a conductivity of at least about 40 S / cm, in some embodiments at least about 50 S / cm, and in further embodiments at least about 60 S / cm. A person of ordinary skill in the art will recognize that additional ranges of particles loadings and conductivities within the explicit ranges about are contemplated and are within the present disclosure.

[0141] Electrical conductivity, which is the inverse of resistivity, can be reported by distributors, and the conductivity is generally measured using specific techniques developed by the distributors. For example, measurements of carbon black electrical resistance is performed between two copper electrodes with Super P™ carbon blacks, see Timcal Graphite & Carbon, A Synopsis of Analytical Procedures, 2008, www.timcal.com. Suitable supplemental electrically conductive additives can also be added to contribute to longer term cycling stability. Alternatively, some suppliers describe the conductive carbon concentrations to achieve the conductive percolation threshold.

[0142] Carbon black refers to synthetic carbon materials and can alternative be referred to as acetylene black, furnace black, thermal black or other names suggesting the synthesis approach. Carbon black generally is referred to as amorphous carbon, but there are suggestions of small domains with short or medium range order corresponding to graphite or diamond crystal structure in at least some forms of carbon black, but for practical purposes the material can be considered amorphous. Under ISO Technical Specification 80004-1 (2010) carbon black is a nano structured material. The primary particles of carbon black can be on the order of tens of nanometers or less, but the primary particles are generally hard fused into chains or other aggregates, and the smallest dispersible units can be considered between about 80 nm and 800 nm, which is still submicron. Carbon blacks are available commercially that have been synthesized to provide a desirable level of electrical conductivity, such as Super-P® (Timcal), Ketjenblack® (Akzo Nobel), Shawinigan Black® (Chevron-Phillips), and Black Pearls 2000® (Cabot).

[0143] Carbon nanofibers are high aspect ratio fibers that generally comprise graphene layers in plates, cones or other forms, which carbon nanotubes comprise graphene sheets folded into tubes. Carbon nanofibers can have diameters of 250 nm or less and are commercially available, for example, Pyrograf® carbon nanofibers (Pyrograf Products, Inc.) or from American Elements, Inc. Carbon nanotubes have been found to be a desirable conductive additive that can improve cycling performance for either a positive electrode or a negative electrode. Single wall or multiwall carbon nanotubes are also available from American Elements, Inc. (CA, USA), Cnano Technologies (China), Fuji, Inc. (Japan), Alfa Aesar (MA, USA) or NanoLabs (MA, USA).

[0144] The negative electrode used in the cells described herein can have high active material loading levels along with reasonably high electrode density. For a particular active material loading level, the density is inversely correlated with thickness so that an electrode with a greater density is thinner than an electrode with a lower density. Loading is equal to the density times the thickness. In some embodiments, the negative electrode of the battery has a loading level of negative electrode active material that is at least about 1.5 mg / cm2, in other embodiments from about 2 mg / cm2to about 8 mg / cm2, in additional embodiments from about 2.5 mg / cm2to about 6 mg / cm2, and in other embodiments from about 3 mg / cm2to about 5 mg / cm2. In some embodiments, the negative electrode of the battery has an active material density in some embodiment from about 0.5 g / cc (cc = cubic centimeters (cm3)) to about 2 g / cc, in other embodiment from about 0.6 g / cc to about 1.5 g / cc, and in additional embodiments from about 0.7 g / cc to about 1.3 g / cc. Similarly, the silicon oxide based electrodes can have an average dried thickness of at least about 15 microns, in further embodiments at least about 20 microns and in additional embodiments from about 25 microns to about 75 microns. The resulting silicon oxide based electrodes can exhibit capacities per unit area of at least about 3.5 mAh / cm2, in further embodiments at least about 4.5 mAh / cm2and in additional embodiments at least about 6 mAh / cm2. A person of ordinary skill in the art will recognize that additional ranges of active material loading level and electrode densities within the explicit ranges above are contemplated and are within the present disclosure.

[0145] Silicon Oxide Active Material Performance

[0146] When developing a new active material, it can be desirable to characterize the electrochemical performance in a way relatively independent of other design features of the cell. In the present context, Applicant has developed and well characterized several negative electrode designs that provide good performance for realistic commercial cell designs. Two electrode designs are tested in the Examples. As seen in the Examples, these electrode designs provide excellent performance for silicon oxide based materials based on supplemental lithium. Half cells are used to interrogate the basic electrochemical properties, such as specific capacity and IRCL, while full cells are used to explore cycling properties.

[0147] Three electrode designs are tested in the Examples, although only two have been cycled. One electrode design has 95 wt% silicon oxide based active material with 5 wt% active graphite, and this design is clearly dominated by the silicon oxide active material. A second electrode design has 70 wt% silicon oxide based active material and 30 wt% graphite active material. Due to the lower silicon oxide contribution to the active material for the second electrode design, this electrode design tends to have lower specific capacity. This second electrode design exhibited somewhat longer cycling stability. It may be desirable for electrodes to comprise even small proportions of silicon oxide active material to potentially boost other performance features, such as cycling stability. The test values of IRCL and specific discharge capacity without influences of the active graphite material, cells were tested with 100% prelithiated silicon oxide using two binder polymers, although these cells did have an amount of conductive carbon common to all of the electrodes.

[0148] Thus, the observed negative electrode specific capacity depends significantly on the formulation of the active material. In embodiments of most interest, the electrodes comprise at least about 15 wt% silicon based active material, in further embodiments at least about 20 wt%, in other embodiments from about 25 wt % to about 100 wt%, and in other embodiments from about 30 wt% to about 95 wt % silicon oxide based active material. The negative electrode specific capacity can be at least about 425 mAh / g, in some embodiments at least about 500 mAh / g, in further embodiments at least about 600 mAh / g, and in additional embodiments from about 650 mAh / g to about 2000 mAh / g. The IRCL can be determined from the half cell measurements, and the IRCL also depends on the composition of the active material composition of the electrode. Since the IRCL of the silicon oxide active material is generally greater than the IRCL of graphite active material and since the specific discharge capacity of the graphite is significantly lower than the specific discharge capacity of the silicon oxide based active material, the IRCL of the electrode with a blend of active materials can take a value over large range.

[0149] Cycling stability is determined in a full cell configuration. Testing in a coin cell generally provides a lower limit for cycling stability since for equivalent electrodes, a pouch cell format generally can provide a greater number of cycles without dropping below 80% of the initial capacity, although proper assembly of a pouch cell provides challenges. Cycling stability generally also depends to some degree on the electrode compositions, electrolyte, cell design, and overall cell structure, in addition to the active material composition. Nevertheless, the cycling data can be particularly helpful for comparison of different active materials assembled into equivalent cells. The exemplified cells involve realistic electrode designs that can provide commercial levels of energy density. The results in the Examples allow for comparison of different embodiments of the prelithiated silicon oxide powders in equivalent electrodes assembled into equivalent cells. So even though the cycling stability may not in some sense be optimized or focused mainly on the active materials in isolation, it provides useful comparisons between the different embodiments of the prelithiation products. These results also indicate commercial utility for the materials in appropriate applications, with particular interest for vehicle use.

[0150] EXAMPLES

[0151] Preparation of Prelithiated Silicon Suboxide

[0152] FIG. 2 shows a process flow diagram for preparation of prelithiated silicon suboxides. A lithium-biphenyl complex solution, referred to as Li-Bp / THP complex solution, was prepared by combining lithium powder and biphenyl (or biphenyl derivative) dissolved in tetrahydropyran (THP). The concentration of lithium powder was about IM and for biphenyl ranged from IM to about 5M. The Li-Bp / THP complex solution was formed by mixing for a time and at a temperature as described for each example.

[0153] Commercially available SiOx(about 1g) was obtained from several commercial sources, and the results seemed independent of the source. The SiOxhad an x value of about 1 (SiO), and the material had an amorphous carbon coating (SiO / C). An amount of Li-Bp / THP complex solution was combined with silicon suboxide such that the molar ratio of silicon suboxide to lithium varied as described for each example. After incubation for the specified time, the resulting silicon suboxide / Li-Bp / THP mixture was separated and the supernatant decanted. The remaining solid comprising lithium deposited silicon suboxide was washed with THP and separated, then dried for 2 hours 80°C.

[0154] Dried lithium deposited silicon suboxide was annealed in an Argon atmosphere for times ranging from 2 hours to 24 hours and at temperatures ranging from 350°C to 950°C. The annealed mixture was sieved at 44 micron to remove larger granules to give prelithiated silicon suboxide as a final powder.

[0155] X-Ray Diffraction Analysis

[0156] X-ray diffraction (XRD) analysis (Cu Ka) was performed on each of the lithiated silicon suboxide final powders and several commercially available lithiated and unlithiated silicon suboxide samples. Plots of intensity in counts per seconds (cps) as a function of scattering angle are shown and some plots include Si peaks for a reference sample of crystalline silicon.

[0157] Preparation of Coin Cells

[0158] General methods and materials are described in the '090 and ‘925 patents cited above. Lithiated silicon suboxide samples were tested through their incorporation into coin cells constructed as half cells with a lithium metal electrode or as full cells with positive electrode including an active material comprising nickel-rich lithium nickel manganese cobalt oxide.

[0159] The active material for the positive electrodes was a commercially available nickel-rich lithium nickel manganese cobalt oxide having the approximate formula LiNio.8Mno.1Coo.1O2. The positive electrodes had a loading of active material from about 93 wt% to 97.5 wt% blended with 1 wt% to 4 wt% PVDF binder, and 1 wt% to 3 wt% nanoscale carbon. The cathode material was blended with NMP solvent, spread onto an aluminum foil current collector, pressed, and dried. The loading level of positive electrode active material ranged from about 18 to about 25 mg / cm2, The active materials for the negative electrodes included lithiated silicon suboxide prepared as described above as well as commercially available lithiated silicon suboxide. Unlithiated commercial silicon oxide samples were also evaluated. Negative electrodes had a loading of active material from about 70 wt% to 95 wt% lithiated or unlithiated silicon suboxide blended with 30 wt% to 5 wt% of electroactive graphite KS 6 synthetic graphite from Imerys S.A unless otherwise noted. The blend of negative electrode active material was mixed thoroughly with an electrically conductive carbon additive such as acetylene black (Super P® from Timcal, Ltd., Switzerland) and / or carbon nanotubes to form a homogeneous powder mixture.

[0160] To form the negative electrode, the active material and conductive carbon additive were combined with a polymeric binder to form a negative electrode composition. An aqueous acrylic binder (Aquacharge® from Sumitomo Seika Co.) was used at from 7 wt% to 15 wt% of the composition unless otherwise noted. Some negative electrode compositions included a blend of polyimide and a lower elastic modulus binder, with the blend at from 1 wt% to 7 wt% of the composition. The blend was formed by mixing the two polymers in NMP (Sigma- Aldrich) overnight to form a polymer binder-NMP solution. The lower elastic modulus binder to polyimide binder weight ratio was 1 to 0.714. The active material, electrically conductive additive and the binder were mixed for about 2 hours to form a homogeneous slurry. The slurry was applied onto a copper foil current collector to form a thin, wet film and the laminated current collector was dried in a vacuum oven. The dried laminate contained from 2 to 20 wt% binder with the remainder of the electrode contributed by the powders. The density of negative electrode active material ranged from about 3.5 to about 6.5 mg / cm3.

[0161] To form the full cells, sections of negative and positive electrodes were cut to size along with separator comprising Celgard® porous polymer or ceramic hybrid membrane. The electrodes with the separator between them was placed in a coin cell enclosure. Half cells were prepared similarly with lithium foil used instead of the positive electrode.

[0162] An electrolyte was placed in the cell and the cell was sealed. The electrolyte included lithium salt LiPFe at a concentration of from about IM to about 2M and non-aqueous solvent including from about 5 wt% to about 40 wt% fluoroethylene carbonate and from about 95 wt% to about 60 wt% selected from one or more of ethylmethyl carbonate, dimethylcarbonate and propylene carbonate. Testing and Evaluation of Coin Cells

[0163] The coin cells were cycled over a relevant voltage range to evaluate performance. Full cells were cycled by charging from the open circuit voltage to a target voltage of 4.2V and discharging between 4.2V and 2.5 in the first formation cycle and between 4.2V and 2.5V for subsequent cycles. Half cells with lithium foil counter electrode were cycled similarly between 1.5V and 0.005V. With the lithium foil counter electrode used for testing purposes, the electrode with the silicon based material functions as the positive electrode for these batteries, but the electrode with the silicon based material may still be referred to as the “negative electrodes” for simplicity since in a commercial battery these electrodes would be used as negative electrodes with a lithium intercalation composition in the positive electrode. The batteries were discharged at a rate of C / 20, C / 10, C / 5, and C / 3 for the 1stcycle, 2nd cycle, 3rd and 4th cycles, and for subsequent cycles, respectively.

[0164] The resulting coin cell batteries were tested with a Maccor cycle tester to obtain chargedischarge curve and cycling stability over a number of cycles. Half cells were electrochemically characterized in the range of 0.005 to 1.5V (1 cycle at C / 20, 1 cycle at C / 10, 2 cycles at C / 5, and the rest at C / 3).

[0165] Initial Coulombic Efficiency (ICE) was calculated as a percentage of dilithiation capacity / lithiation capacity from the first charge and discharge. ICE = 1-IRCL, where IRCL is the irreversible capacity loss.

[0166] Example 1 - Comparison of Lithiation Methods

[0167] Phases formed for prelithiated silicon suboxide prepared by the annealing process described above versus a soaking method were compared. A first lithium deposited silicon suboxide with silicon suboxide:Li molar ratio of 1:1.5 was prepared by soaking an amount of silicon suboxide in Li-Bp / THP complex solution at 70°C for 120 minutes followed by a room temperature incubation for 16 hours. The sample was removed from the solution, rinsed with THP and after drying at room temperature was annealed at 650°C for 2 hours to form a first prelithiated silicon suboxide. A second lithium deposited silicon suboxide was prepared the same as the first and was not annealed.

[0168] Plots obtained from XRD diffraction analysis are shown in FIG. 3. The topmost plot is for unlithiated SiO / C, followed by the first lithium deposited silicon suboxide and the second lithium deposited silicon suboxide. The bottommost plot is the pattern for Li2SiO3 and crystalline Si with known Si peaks identified. Example 2 - Different Ratios of Li and Silicon Suboxide

[0169] This examples shows the effect of silicon suboxide:Li molar ratio on the formation of different phases, and the effect of the different phases on cell performance.

[0170] Prelithiated silicon suboxides were prepared with silicon suboxide:Li molar ratios ranging from 1 : 1 to 10: 1. For the purposes of this Example, it is assumed that all of the lithium introduced into the reaction is alloyed with the silicon suboxide. The complex solution and silicon oxide were mixed for 2 hours at 70°C followed by 16 hours at room temperature. Each of the samples was annealed for 2 hours at 650°C. Plots obtained from XRD diffraction analysis are shown as a series in FIG. 4 with each panel labeled with the silicon suboxide:Li molar ratio of the corresponding sample. Peaks corresponding to Li4SiO4 along with smaller peaks attributed to Li2SiO3 are identified for the sample with 1:1 molar ratio. Li4SiO4 peaks appear to decrease substantially for samples with reduced lithium going from 1:1 to 2:1 molar ratio. Peaks corresponding to Li2SiO3 are identified for the sample with 2:1 molar ratio. Li2SiO3 peaks are observable at a 1:1 molar ratio, and intensities decrease as molar ratio increases. Peaks corresponding to Li2Si20s are identified for the sample with 3:1 molar ratio. Li2Si20s peaks are observable at 3:1 molar ratio and remain predominant as molar ratio increases to 10:1.

[0171] Half cell performance for the prelithiated silicon suboxides with a high loading of SiO formulated with a small amount of electroactive graphite (90-98 weight percent SiO) and electrolyte was evaluated by measuring specific discharge capacity at the first and ninth cycles as a function of silicon suboxide:Li molar ratio. Initial Coulombic efficiency (ICE) as a function of silicon suboxide:Li molar ratio was also evaluated. Results are shown in FIG. 5. The data show that specific discharge capacity increases as the amount of Li decreases. ICE increases as the molar ratio increases from 1:1 to 2:1 but then decreases as the amount of Li increases. Low discharge capacity and low ICE at silicon suboxide:Li 1:1 molar ratio may be due to formation of irreversible phase Li4SiO4.

[0172] Full cell performance for the prelithiated silicon suboxides formulated with a high loading of SiO with a small amount of electroactive graphite (90-98 weight percent SiO) and electrolyte was evaluated by determining the number of cycles at 80% capacity retention as a function of silicon suboxide:Li molar ratio. Results are shown in FIG. 6 and include data measured for three cells. The worst cycle life was exhibited at 1:1 molar ratio which may be attributed to large amounts of irreversible phase Li4SiO4. Specific discharge capacities at 1C and 5C discharge rates, as a function of silicon suboxide:Li molar ratio are shown in FIG. 7. The lowest specific discharge capacity was exhibited at 1 : 1 molar ratio which may be attributed to large amounts of irreversible phase Li4SiO4. Except for 1:1 molar ratio with irreversible phase Li4SiO4, an increase in Li improves both cycle life and capacity.

[0173] Example 3 - Effect of Anneal Temperature

[0174] This example shows the effect of annealing temperature on the formation of different phases for prelithiated silicon suboxides and the corresponding effect the different phases have on cell performance.

[0175] Prelithiated silicon suboxides were prepared with silicon suboxide:Li molar ratio of 2: 1. The complex solution and silicon oxide were mixed for 2 hours at 70°C followed by 16 hours at room temperature. Each of the samples was annealed for 2 hours at temperatures from 350°C to 950°C. Plots obtained from XRD diffraction analysis of the annealed powder are shown as a series in FIG. 8 with each plot labeled with the annealing temperature. Only peaks corresponding to phases are identified for the topmost plot (sample annealed at 950°C). The bottommost plot labelled silicon suboxide is for silicon suboxide with peaks corresponding to Si identified.

[0176] Peaks corresponding to crystalline Li2SiO3 and potentially other lithium silicate phases emerge at 550°C, and intensity and sharpness of the peaks increase as temperature increases to 950°C. Peaks corresponding to Li2Si20s, LisSiOe and Li4SiO4 become more visible at 650°C and intensity increases to 950°C, but all 3 sets of peaks continue to have less intensity than those for Li2SiO3. The silicon peaks also sharpen with a corresponding decrease in full- width at half-maximum as the anneal temperature is increased. For this molar ratio of lithium, the Li4SiO4 lithium silicate phase is less prominent even at high anneal temperatures.

[0177] XRD analyses of pristine unlithiated SiOxand commercially available lithiated SiOxare shown in FIGS. 9 and 10, respectively. The plot obtained for prelithiated silicon suboxide annealed for 2 hours at 650°C is included in FIG. 9. FIG. 9 shows the presence of Li2SiO3 in both samples, but the sample formed by annealing for 2 hours at 650°C shows significant presence of Li2Si20s. For both FIG. 9, the bottommost plot is the pattern for Li2SiO3 and crystalline Si with known Si peaks identified.

[0178] Full cell performance for prelithiated silicon suboxide prepared by annealing at 650°C was evaluated by determining capacity retention as a function of cycle number. Full cell performance for two commercially available lithiated SiO samples, CAI and CA2, were evaluated as was unlithiated commercial SiO. All negative electrodes were formulated with high loading of SiO and a small amount of electroactive graphite (90-98 weight percent SiO). Results are shown in FIG. 10. The sample formed by annealing for 2 hours at 650°C reached 80% capacity after about 410 cycles which was greater than that reached by either commercially prepared sample by at least 125 cycles.

[0179] Example 4 - Effects of Anneal Times and Temperatures

[0180] This example shows the effect of annealing time and temperature on the formation of different phases for prelithiated silicon suboxide.

[0181] Prelithiated silicon suboxides samples were prepared with silicon suboxide:Li molar ratio of 3:1. The complex solution and silicon oxide were mixed for 2 hours at 70°C followed by 16 hours at room temperature. Eight samples were prepared and annealed as follows: three samples at 2, 12 and 24 hours at a temperature of 450°C, three samples at 2, 12 and 24 hours at a temperature of 550°C, and two samples at 2 and 12 hours at a temperature of 650°C. Plots obtained from XRD diffraction analysis are shown as series in FIGS. 11, 12, 13A, 13B and 14. FIGS. 11, 13A and 14 show positions of Si peaks corresponding to crystalline Si extending from and perpendicular to the x-axis of the plot along with additional peaks. As the anneal temperature is increased to 550 °C and 650 °C, lithium silicate peaks appear. The peaks at 450 °C in addition to the silicon peaks are not identified, but these seem to belong to a phase that is consistent with more stable cycling. These unidentified peaks are visible in the XRD spectrum at 550 °C along with the silicon peaks and emerging lithium silicate peaks.

[0182] FIG. 11 shows XRD patterns for prelithiated silicon suboxides annealed at 450°C. FIG. 12 shows the XRD patterns superimposed on each other.

[0183] FIG. 13A shows XRD patterns for prelithiated silicon suboxides annealed at 550°C. FIG. 13B shows the XRD patterns superimposed on each other.

[0184] FIG. 14 shows XRD patterns for prelithiated silicon suboxides annealed at 650°C.

[0185] Overall, silicate phases were found to form at 450°C which increased crystallinity as the the annealing temperature increased. Full cell performance of the prelithiated silicon suboxides was evaluated by determining specific capacity retention as a function of cycle number. Cells were cycled at room temperature at 1C charge / 1C discharge rates. These samples were formulated with SiO dominant active material having > 55 wt% SiO and the remainder electroactive graphite. Results are shown in FIG. 15.

[0186] FIG. 16 shows cycling performance for the full cells where the number of cycles reached to obtain 80% capacity is plotted as a function of annealing temperature. Data for annealing times of 2, 12 and 24 hours are shown. For samples annealed at 450°C for 12 and 24 hours, 80% capacity was reached at about 700 cycles. In comparison, commercially available lithiated silicon oxide samples reached 80% capacity at about 450-500 cycles. For electrochemical prelithiated SiO electrode evaluated as a coin cell, 80% capacity was reached at about 900-1000 cycles. Electrochemical prelithiation was performed in a half cell with a charge to slightly greater than the IRCL, and the disassembled negative electrode after prelithiation was assembled into a full cell. Selected data are shown in Table 4.

[0187] TABLE 4

[0188] FIG. 17 shows specific capacity as a function of annealing temperature for two sets of the full cells with one set cycled with a 4C charge / 1C discharge rate and the other with a 1C charge / 1C discharge rate. Data for annealing times of 2, 12 and 24 hours are shown for each combination of charge / discharge rates. Specific capacities of the prelithiated silicon suboxides were generally comparable to specific capacities obtained for negative electrodes prepared with either commercially available lithiated silicon suboxide or directly lithiated electrodes prepared by adding lithium metal powder to the surface of similarly formulated SiO electrodes. Selected data are shown in Table 5. Overall, silicate phases are good for improving the IRCL, but affects the cycling.

[0189] TABLE 5

[0190] Half cell performance of the prelithiated silicon suboxides was evaluated by determining specific discharge capacity as a function of annealing temperature. Data for annealing times of 2, 12 and 24 hours are shown in FIG. 18. All samples were formulated with SiO dominant active material > 55 wt% SiO and the remainder electroactive graphite. Specific discharge capacities for selected cells are shown in Table 6. For comparison, half cells similarly formulated with commercially available lithiated SiO and with electrochemically lithiated SiO (prelithiation performed in a half cell with Li foil with charging to slightly greater than the IRCL) are included.

[0191] TABLE 6

[0192] Half cell performance of the prelithiated silicon suboxides was evaluated with respect to IRCL for the first cycle. Data for annealing times of 2, 12 and 24 hours are shown in FIG. 19. For lithiated samples prepared by annealing at 450°C for 12 and 24 hours, the percent IRCLs for the 1stcycle were between 10 and 12%. Similar results were obtained for a half cell similarly formulated with commercially available lithiated silicon suboxide.

[0193] Effects of annealing time and temperature on pH and peel strength of negative electrode compositions were evaluated for the prelithiated silicon suboxides. Slurries of negative electrode compositions were prepared with SiO dominant active material with > 55 wt% SiO and the remainder electroactive graphite. FIG. 20 shows a plot of pH as a function of annealing temperature for samples annealed at 2, 12 and 24 hours. The pH values ranged from about 11.5 to about 12.8. The pH values decreased as annealing time increased. The lowest pH was observed for the prelithiated silicon suboxide annealed at 450°C for 24 hours. For comparison, slurries prepared with commercially available lithiated SiO and unlithiated SiO had pH values of about 12 and 8, respectively.

[0194] The slurries were coated on copper current collector foils and dried in a vacuum oven at a temperature of 80°C to 120°C to give coatings having a thickness of about 40 to 80 microns. Peel strength at 180 degrees was measured using an Ametek Chatillon TCD225 Series force measurement instrument. Testing was carried out as described for FIG. 7 of the '090 application cited above. Generally, a 70 mm by 25 mm section of double-sided tape was adhered to a glass plate which was then fixed to a lower clamp of the instrument. The negative electrodes were compressed face-down on the double- sided tape such that the coating and tape were in contact. The substrate was connected to the other side of the instrument and peel strength was determined by pulling on the substrate at a rate of 2 mm per minute for 20 mm.

[0195] FIG. 21 shows a plot of peel strength as a function of annealing temperature for samples annealed at 2, 12 and 24 hours. Peel strengths ranged from about 0.015 to about 0.06 poundforce (LbF). For comparison, commercially available lithiated SiO and unlithiated SiO had peel strengths of 0.043 LbF and 0.06 LbF, respectively. The highest peel strength was observed for the lithiated silicon suboxide sample annealed at 450°C for 24 hours. FIG. 22 shows a plot of peel strength as a function of distance for this sample and also shows peel strength for a commercially available lithiated SiO (CA).

[0196] Example 5 - Effect of Different Binders

[0197] Prelithiated silicon suboxides samples were prepared with silicon suboxide:Li molar ratio of 2:1. The complex solution and silicon oxide were mixed for 2 hours at 70°C followed by 16 hours at room temperature. Samples were annealed at 650°C for 2 hours. Negative electrodes were prepared by combining the prelithiated silicon suboxide with the aqueous binder and with the NMP-based nonaqueous binder described above. A negative electrode including unlithiated silicon suboxide was also prepared. FIG. 23 is a plot of voltage as a function of specific capacity for half cells prepared with the negative electrodes. The Coulombic efficiency for the negative electrode prepared with aqueous binder was 89% with an IRCL of 11%. Coulombic efficiency for the negative electrode prepared with the NMP- based nonaqueous binder was 82% with an IRCL of 18%. Coulombic efficiency for the negative electrode prepared with unlithiated silicon suboxide was 73% with an IRCL of 27%.

[0198] Silicon oxide samples were soaked in Li-Bp / THP complex solution with silicon oxide:Li molar ratio of 1.5:1 and a BP concentration of IM. The samples were soaked for 24 hours at a temperature of 40°C. Each of the soaked samples was annealed for 24 hours at 450°C. Half cells were prepared using only the prelithiated silicon oxide as an active material, and the first lithiation and first delithiation capacities were measured along with the first cycle Coulombic efficiency. Results are shown in Table 7. TABLE 7

[0199] Example 6 - Effect of Soaking Time, Temperature and Li Concentration

[0200] This examples shows the effect on the formation of different phases for prelithiated silicon suboxide when silicon suboxide is soaked in Li-Bp / THP complex solution prior to annealing.

[0201] Silicon oxide samples were soaked in Li-Bp / THP complex solution with silicon oxide:Li molar ratio of 1.5:1 and a BP concentration of IM. The samples were soaked for 2 hours or 24 hours, at temperatures of 40°C, 70°C or 90°C. Each of the soaked samples was annealed for 24 hours at 450°C. Plots obtained from XRD diffraction analysis are shown as a series in FIG. 24 with each plot labeled with the annealing time and temperature.

[0202] Negative electrodes were prepared by formulating the annealed samples with SiO dominated electroactive material with > 55 wt% SiO and the remainder electroactive graphite. Full cells were prepared. FIG. 25 shows cycling performance for the full cells where the number of cycles reached to obtain 80% capacity is plotted as a function of soaking temperature. Data for soaking times of 2 and 24 hours are shown. FIG. 26 shows specific capacity retention as a function of cycle number for full cells cycled at room temperature at 1C charge / lC discharge rates. Data for soaking times of 2 and 24 hours, at temperatures of 40°C, 50°C, 70°C or 90°C are shown.

[0203] Silicon suboxide samples were soaked in Li-Bp / THP complex solution having a range of molar ratios as shown in Table 8. A silicon suboxide sample was soaked in Li-Bp / THP complex where of 4,4'-dimethyl biphenyl was used instead of biphenyl. The molar ratio of SiO to BP was 3:1. After soaking the samples for 2 -24 hours, the samples were annealed for 24 hours at 450°C. Half cells were prepared and the first lithiation and first delithiation capacities were measured along with the first cycle Coulombic efficiency. Results are shown in Table 8. TABLE 8

[0204] 1. 4, 4'-Dimethyl Biphenyl

[0205] The embodiments above are intended to be illustrative and not limiting. Additional embodiments are within the claims. In addition, although the present invention has been described with reference to particular embodiments, those skilled in the art will recognize that changes can be made in form and detail without departing from the spirit and scope of the invention. Any incorporation by reference of documents above is limited such that no subject matter is incorporated that is contrary to the explicit disclosure herein. To the extent that specific structures, compositions and / or processes are described herein with components, elements, ingredients or other partitions, it is to be understand that the disclosure herein covers the specific embodiments, embodiments comprising the specific components, elements, ingredients, other partitions or combinations thereof as well as embodiments consisting essentially of such specific components, ingredients or other partitions or combinations thereof that can include additional features that do not change the fundamental nature of the subject matter, as suggested in the discussion, unless otherwise specifically indicated. The use of the term "about" herein refers to expected uncertainties in the associated values as would be understood in the particular context by a person of ordinary skill in the art.

Claims

What is claimed is:

1. A particulate lithium- silicon oxide composite material comprising lithium incorporated into a silicon oxide material and having a Cu Kadiffractogram with 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

2. The particulate lithium- silicon oxide composite material of claim 1 comprising an aromatic radical anion.

3. The particulate lithium- silicon oxide composite material of claim 2 wherein the aromatic radical anion comprises a radical anion of a multicyclic or polycyclic hydrocarbon.

4. The particulate lithium- silicon oxide composite material of claim 2 wherein the aromatic radical anion comprises the radical anion of biphenyl or an alkyl substituted biphenyl.

5. The particulate lithium- silicon oxide composite material of any one of claims 1-4 wherein the silicon oxide material comprises SiOxwhere 0.9<x<l.l.

6. The particulate lithium- silicon oxide composite material of any one of claims 1-5 wherein the silicon oxide material comprises domains of elemental crystalline silicon.

7. The particulate lithium- silicon oxide composite material of any one of claims 1-6 wherein the silicon oxide material comprises from about 0.5 wt % to about 5 wt% of the amorphous carbon.

8. The particulate lithium- silicon oxide composite material of any one of claims 1-7 wherein the material is a powder that passes through a 44 micron sieve.

9. The particulate lithium- silicon oxide composite material of any one of claims 1-8 having a low degree of crystallinity as determined by Cu-Ka XRD peaks for silicon crystallites having a full- width at half maximum for 2theta of at least about 1.25 degrees.

10. The particulate lithium- silicon oxide composite material of any one of claims 1-9 wherein the material comprises one or more lithium silicates with the predominate lithium silicate comprising Li2SiO3.

11. The particulate lithium-silicon oxide composite material of claim 10 having an x-ray diffractogram substantially free of x-ray scattering peaks associated with Li2Si20s12. An electrode comprising a blend of a polymer binder, electrically conductive particulates, and an active material comprising the particulate lithium-silicon oxide composite material of any one of claims 1-11.

13. The electrode of claim 12 having a first cycle irreversible capacity loss of no more than about 20 % relative to the first cycle charge capacity evaluated at C / 20.

14. A method for chemically prelithiating a silicon suboxide-based material, the method comprising: associating lithium-radical anion complex with the silicon suboxide-based material in a solution wherein the lithium-radical anion complex is dissolved and the silicon suboxidebased material is dispersed to form a lithium deposited silicon suboxide, and annealing the lithium deposited silicon suboxide at a temperature from about 350°C to about 600°C for at least 1 hour to form a prelithiated silicon suboxide.

15. The method of claim 14 wherein a Cu Kadiffractogram of the prelithiated silicon suboxide has 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

16. The method of claim 14 or claim 15 wherein comprises one or more lithium silicates with the predominate lithium silicate comprising Li2SiOa and is substantially free of x-ray scattering peaks associated with Li2Si20s.

17. The method of any one of claims 14-16 wherein the lithium-radical anion complex comprises a complex of lithium cation and a radical anion of a multicyclic or polycyclic hydrocarbon.

18. The method of claim 17 wherein the multicyclic or polycyclic hydrocarbon comprises biphenyl or an alkyl substituted biphenyl.

19. The method of any one of claims 14-18 wherein the solution comprises an aprotic solvent having a dielectric constant greater than about 5.

20. The method of any one of claims 14-19 wherein the solution comprises an aprotic solvent comprising one or more ethers.

21. The method of any one of claims 14-20 wherein the silicon suboxide-based material comprises SiOxwhere 0.9<x<l.l.

22. The method of any one of claims 14-21 wherein the silicon suboxide-based material comprises domains of elemental crystalline silicon.

23. The method of any one of claims 14-22 wherein the silicon suboxide-based material is a composite comprising amorphous carbon.

24. The method of any one of claims 14-23 wherein a molar ratio of the silicon suboxidebased material to the lithium-radical anion complex is from about 1:0.7 to about 1:5.

25. The method of any one of claims 14-24 further comprising isolating the prelithiated silicon suboxide as a powder passing through a 44 micron sieve.

26. The method of any one of claims 14-25 wherein the lithium deposited silicon suboxide is annealed for no more than about 48 hours.

27. The method of any one of claims 14-25 wherein the lithium deposited silicon suboxide is annealed at a temperature from about 400°C to about 575°C.

28. The method of any one of claims 14-27 wherein the associating is performed while mixing and heating to at least about 35°C for at least about 15 minutes.

29. The method of any one of claims 14-28 wherein the lithium-radical anion complex comprises a complex of lithium cation and an aromatic radical anion dissolved in an aprotic polar solvent, wherein the lithium-radical anion complex is formed by combining lithium and an aromatic compound at a molar ratio of from about 1.25:1 to about 5:1, and wherein a concentration of the lithium-radical anion complex in the solution is from about 0.5M to about 2.5M.

30. An electrode comprising a blend of a polymer binder, electrically conductive particulates and the prelithiated silicon suboxide prepared according to any one of claims 14- 29.

31. A method for forming a lithium deposited silicon suboxide, the method comprising: associating lithium-radical anion complex with silicon suboxide-based material in a solution wherein the lithium-radical anion complex is dissolved and the silicon suboxide-based material is dispersed, wherein the solution comprises a total lithium amount that is at least about 25 mole% greater than a molar quantity of radical anion and its neutral corresponding compound, and wherein a lithium deposited silicon suboxide is formed.

32. The method of claim 31 wherein the lithium-radical anion complex comprises a complex of lithium cation and an aromatic radical anion.

33. The method of claim 32 wherein the aromatic radical anion comprises the radical anion of biphenyl or an alkyl substituted biphenyl.

34. The method of any one of claims 31-33 wherein the solution comprises an aprotic polar solvent having a dielectric constant greater than about 5.

35. The method of claim 34 wherein the aprotic solvent comprises a cyclic ether, an acyclic ether, a glycol-based ether, tetrahydrofuran, methyl tetrahydrofuran, tetrahydropyran, dimethoxyethane or mixtures thereof.

36. The method of any one of claims 31-35 wherein the lithium-radical anion complex is formed by combining lithium and an aromatic compound at a molar ratio of from about 1.25 : 1 to about 5:1.

37. The method of any one of claims 31-36 wherein a concentration of the lithium-radical anion complex in the solution is from about 0.5M to about 2.5M.

38. The method of any one of claims 31-37 wherein the silicon suboxide-based material comprises SiOxwhere 0.9<x<l.l.

39. The method of any one of claims 31-38 wherein a molar ratio of the silicon suboxidebased material to the lithium-radical anion complex is from about 1:0.7 to about 1:5.

40. The method of any one of claims 31-39 wherein the lithium-radical anion complex is formed by combining lithium metal with an aromatic compound suitable for forming a radical anion in an aprotic polar solvent.

41. The method of any one of claims 31-40 wherein associating comprises mixing the silicon suboxide-based material in the solution comprising the lithium-radical anion complex.

42. The method of claim 41 wherein the solution is heated to at least about 35°C for a time from about 30 minutes to about 48 hours.

43. The method of claim 41 wherein mixing is performed for a time from about 1 hour to about 48 hours and the solution is maintained at a temperature of from about 40°C to about 100°C for at least a portion of the soaking time.

44. The method of any one of claims 31-43 further comprising annealing the lithium deposited silicon suboxide at a temperature of at least about 350 °C for at least about 30 minutes to form a prelithiated silicon suboxide.

45. The method of claim 44 wherein a Cu Kadiffractogram of the lithium deposited silicon suboxide has 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

46. An electrode comprising a blend of a polymer binder, electrically conductive particulates, and an active material comprising particulate lithium-silicon oxide composite material comprising lithium incorporated into a silicon oxide lattice associated with amorphous carbon having a first cycle irreversible capacity loss of no more than about 20 % relative to thefirst cycle charge capacity evaluated at C / 20 and having a Cu Kax-ray diffractogram of the lithium deposited silicon suboxide composite material with 29 peaks at 21.0-21.5, 31.3-31.8, 33.5-34.1, 42.7-43.3.

47. The electrode of claim 46 wherein a Cu Kadiffractogram of the lithium deposited silicon suboxide has a low degree of crystallinity and is substantially free of x-ray scattering peaks associated with Li2Si20s.

48. The electrode of claim 46 or claim 47 wherein the particulate lithium-silicon oxide composite material comprises one or more lithium silicates with the predominate lithium silicate comprising Li2SiO3.

49. The electrode of any one of claims 46-48 wherein the particulate lithium-silicon oxide was formed by associating silicon oxide with a lithium-radical anion solution in an aprotic polar solvent to form a lithium deposited silicon suboxide, and annealing the lithium deposited silicon suboxide at a temperature of at least about 350 °C for at least about 15 minutes.

50. The electrode of any one of claims 46-49 wherein the active material comprises a blend of graphite and from about 60 wt% to about 98 wt% of the particulate lithium-silicon oxide composite material.

51. The electrode of any one of claims 46-50 wherein the active material comprises a blend of graphite and from about 15 wt% to about 45 wt% of the particulate lithium-silicon oxide composite material.

52. The electrode of any one of claims 46-51 wherein the active material comprises a blend of the particulate lithium-silicon oxide composite material and from about 45 wt% to about 85 wt% graphite active material.

53. The electrode of any one of claims 46-52 wherein the electrically conductive particulates comprise carbon nanotubes, carbon nanofibers, carbon nanoparticles, carbon black or combinations thereof.

54. The electrode of any one of claims 46-53 wherein the active material has a specific discharge capacity of at least about 800 mAh / g when cycled at a rate of C / 10 against lithium metal from 0.005V to 1.5V.

55. The electrode of any one of claims 46-54 wherein the polymer binder comprises an aqueous binder comprising a copolymer of acrylic acid.

56. The electrode of any one of claims 46-55 wherein the polymer binder comprises at least about 50 wt% polyimide and at least about 5 wt% of a polymer with a lower elastic modulus than the poly imide.

57. The electrode of any one of claims 46-56 wherein the electrode has an active material density of from about 0.5 g / cc to about 2 g / cc.

58. A lithium ion cell comprising: a positive electrode comprising a metal oxide powder, electrically conductive particles and a polymer binder, an electrode of any one of claims 46-57 as a negative electrode, and a separator between the positive electrode and the negative electrode.

59. The lithium ion cell of claim 58 having at least about 800 mAh / g negative electrode specific discharge capacity at a rate of C / 3, a positive electrode specific discharge capacity of at least about 145 mAh / g at a rate of 1C, and a discharge capacity at 500 cycles that is at least about 80% of the discharge capacity at the 5th cycle when cycled at a charge and discharge rate of 1C between the 5th cycle and the 500th cycle.

Images