Systems and methods for operating a battery based on changes in electrode crystal structure

Abstract

Systems and methods for operating a battery based on electrode crystal structure changes are provided. A battery includes: an electrode that exhibits a crystal structure change when lithiation exceeds a threshold potential, and a battery management system. The battery management system includes a controller configured to determine the threshold potential while the battery is online, determine a battery operating parameter based on the determined threshold potential, and operate the battery based on the determined battery operating parameter.

Description

[0001] Priority claim

[0002] This application claims priority to U.S. Provisional Application Serial No. 62 / 690,376, filed on June 27, 2018, entitled “Method for Operating Batteries Based on Electrode Crystal Structure Change,” the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] The present disclosure relates generally to batteries, and more particularly to battery management systems for batteries. Background Art

[0004] Several new battery chemistries are entering the market to provide the capabilities necessary in specialized applications. Once upon a time, the lithium-ion battery market was driven by the use of such batteries in portable electronic devices, which required high energy but only limited life and power. Recently, other industries have focused on the use of batteries. As an example, batteries are often incorporated into power tools and certain types of hybrid electric vehicles. Each new industry requires different performance characteristics. Certain applications, such as automotive applications, require battery stability in terms of battery safety for both large packaging and long life (e.g., at least 10 to 15 years).

[0005] Lithium-ion batteries have become the industry standard in both electric vehicles and portable electronic device applications. Lithium-ion batteries operate based on the movement of lithium ions between a negative electrode (also called an "anode") and a positive electrode (also called a "cathode"). The current negative electrode is based on graphite, which intercalates lithium and has a capacity of 372 mAh / g. 石墨 Silicon forms an alloy with lithium and achieves 3579 mAh / g Si Silicon has been identified as a potential negative electrode material due to its ability to provide a high gravimetric density. However, currently, the use of pure silicon as a negative electrode has proven challenging due to the high volume expansion rate that occurs during the lithiation process of pure silicon. Nevertheless, some current batteries incorporate small amounts of pure silicon or silicon-containing materials (such as silicon oxide (SiO) or silicon alloys (SiB3, Si2Fe, TiSi2 and others)) into graphite-based negative electrodes to increase the gravimetric capacity of the negative electrode above the level of pure graphite.

[0006] Lithium-ion batteries are typically coupled to a battery management system (BMS) during operation of the battery. The BMS generally includes a controller that executes program instructions stored in a memory to operate the battery to control the rate at which the battery charges and discharges based on a known model of the battery's operating parameters.

[0007] What is needed, therefore, is an improved way to design BMS strategies based on measurable characteristics in order to improve the cycle life of lithium-ion batteries and reduce defects that may occur due to volume expansion during the lithiation process. Summary of the invention

[0008] A battery includes an electrode that exhibits a crystal structure change when lithiation occurs beyond a threshold potential, and a battery management system. The battery management system includes a controller configured to determine the threshold potential while the battery is online, determine a battery operating parameter based on the determined threshold potential, and operate the battery based on the determined battery operating parameter.

[0009] In an embodiment of the battery, determination of the threshold potential comprises identifying an operating characteristic indicative of an internal state of the electrode.

[0010] In another embodiment, determining the threshold potential includes: charging the battery to a first potential exceeding the threshold potential, discharging the battery from the first potential, and storing a first discharge curve in a memory; and charging the battery to a second potential not exceeding the threshold potential, discharging the battery from the second potential, and storing a second discharge curve in a memory.

[0011] In further embodiments, the identification of the operating characteristic includes identifying at least one feature that is present in the first discharge curve and is absent in the second discharge curve.

[0012] In some embodiments of the battery, the determination of the threshold potential further comprises performing charge and discharge cycles to a plurality of different cut-off potentials, and determining the threshold potential based on corresponding discharge curves from the plurality of charge and discharge cycles.

[0013] In some embodiments, determining the threshold potential based on discharge curves from the plurality of charge and discharge cycles may include selecting as the threshold potential a lowest cutoff potential of a plurality of cutoff potentials where the corresponding discharge curve does not include the at least one feature.

[0014] In one embodiment, the threshold potential is determined to within 2 mV.

[0015] In some embodiments, the determination of the battery operating parameter includes selecting a state of charge curve based on the determined threshold potential and a charge cutoff potential from a most recent charge.

[0016] In yet another embodiment, the determination of the battery operating parameters includes adapting boundary conditions of the charging process based on the determined threshold potential.

[0017] In further embodiments, the determination of the battery operating parameter includes selecting a charging target potential that is within 2 mV of the threshold potential.

[0018] In another embodiment, a method of operating a battery using a battery management system includes, while the battery is online, determining a threshold potential, wherein an electrode of the battery exhibits a crystal structure change when lithiation occurs beyond the threshold potential. The method also includes determining a battery operating parameter based on the determined threshold potential, and operating the battery based on the determined battery operating parameter. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a schematic view of a battery pack according to the present disclosure.

[0020] Figure 2 is an electrode that exhibits a change in crystal structure Figure 1 Schematic view of a battery cell of a battery pack.

[0021] Figure 3 is a graph of lithiation and delithiation versus capacity for a silicon electrode composed of silicon and a conductive additive, showing the difference in the delithiation curves after lithiation to either 10 mV or 60 mV.

[0022] Figure 4 is a graph of lithiation and delithiation versus capacity curves for half-cell tests of silicon-containing electrodes, showing the difference in the delithiation curves after lithiation to either 10 mV or 60 mV.

[0023] Figure 5 is a graph of lithiation and delithiation versus capacity in Ah for a full cell test of a lithium-ion battery having a silicon-containing negative electrode, showing the difference between the delithiation curves after lithiation above the crystal structure change potential and after lithiation below the crystal structure change potential.

[0024] Figure 6 Depicted is a series of three experiments that demonstrate how the potential at which a change in crystal structure occurs can be identified within 2 mV accuracy by visualizing the measured potential versus time.

[0025] Figure 7 Depicted is a flow chart of the model optimization process that uses the potential for crystal structure alterations to identify model parameters for BMS.

[0026] Figure 8A flow chart depicting a process of using detection of crystal structure changes to improve model voltage prediction and SOC estimation accuracy in a BMS. DETAILED DESCRIPTION

[0027] For the purpose of promoting an understanding of the principles of the embodiments described herein, reference is now made to the drawings and the description in the following written specification. No limitation of the scope of the subject matter is intended by reference. The present disclosure also includes any changes and modifications to the illustrated embodiments, and includes additional applications of the principles of the described embodiments as would normally occur to a person skilled in the art to which this document relates.

[0028] The various operations may be described in turn as multiple discrete actions or operations in a manner that is most helpful in understanding the claimed subject matter. However, the order of description should not be interpreted as implying that these operations are necessarily order-dependent. In particular, these operations may not be performed in the order presented. The described operations may be performed in an order different from the described embodiments. Various additional operations may be performed, and / or the described operations may be omitted in additional embodiments.

[0029] The terms "comprising," "including," "having," and the like as used with respect to the embodiments of the present disclosure are synonymous. As used herein, the term "approximately" refers to a value within ±10% of a reference value.

[0030] The embodiments of the present disclosure discussed below are applicable to any desired battery chemistry. For illustrative purposes, some examples relate to lithium-ion batteries. As used herein, the term "lithium-ion battery" refers to any battery that includes lithium as an active material. In particular, lithium-ion batteries include, without limitation, lithium-based liquid electrolytes, solid electrolytes, colloidal electrolytes, and batteries commonly referred to as lithium polymer batteries or lithium-ion polymer batteries. As used herein, the term "colloidal electrolyte" refers to a polymer impregnated with a liquid electrolyte.

[0031] Reference now Figure 1 , the battery pack 100 includes a plurality of battery cells 102 arranged in a pack housing 104. Each battery cell 102 includes a cell housing 106 from which a positive terminal 108 and a negative terminal 112 are exposed. In a parallel arrangement, the positive terminals 108 may be connected to each other through current collectors 116, and the negative terminals 112 may be connected to each other through different current collectors 120. In a series arrangement, the positive terminal 108 may be connected to an adjacent negative terminal 112 through a current collector. The current collectors 116, 120 are connected to respective positive and negative battery pack terminals 124, 128, which are connected to an external circuit 132, which may be powered by the battery pack 100 or may be configured to charge the battery pack 100.

[0032] In addition, the battery pack 100 includes a battery management system (BMS) 140, which includes a controller 144, a memory unit (not shown), and in some embodiments, one or more sensors (not shown). The operation and control of the battery pack 100 are performed with the help of the BMS 140. The controller 144 of the BMS 140 is implemented using a general or dedicated programmable processor that executes programmed instructions. The instructions and data necessary to perform the programmed functions are stored in a memory unit associated with the controller. The processor, memory, and interface circuit configure the controller 144 to operate the battery pack 100 to charge and discharge the battery to the desired charge and discharge thresholds within the desired charge and discharge rates, and to operate the battery pack 100 in other ways. The processor, memory, and interface circuit components may be provided on a printed circuit card or as circuits in an application-specific integrated circuit (ASIC). Each circuit may be implemented using a separate processor, or multiple circuits may be implemented on the same processor. Alternatively, the circuit may be implemented using circuits or discrete components provided in a VLSI circuit. The circuits described herein may also be implemented using a combination of processors, ASICs, discrete components, or VLSI circuits. Further discussion of the BMS and electrochemical model-based BMS may be found, for example, in U.S. Patent No. 8,188,715, issued May 29, 2012, the contents of which are incorporated herein by reference in their entirety.

[0033] Each battery cell 102 includes Figure 2 2, electrode configuration 200 includes positive electrode current collector 204, positive electrode layer 208, separator layer 212, negative electrode 216, and negative electrode current collector 220. In some embodiments, multiple layers of electrode configuration 200 are stacked on top of each other to form an electrode stack. In other embodiments, electrode configuration 200 is wound around itself in a spiral shape to form an electrode configuration known as a "jelly roll" or "Swiss roll" configuration.

[0034] The positive electrode current collector 204 connects the positive terminal 108 of the battery cell 102 with the positive electrode 208 to enable the flow of electrons between the external circuit 132 and the positive electrode 208. Similarly, the negative electrode current collector 220 connects the negative terminal 112 with the negative electrode layer 216. In the illustrated embodiment, the negative electrode layer 216 includes a combination of graphite and one or more of silicon (Si), silicon oxide (SiO), and a silicon alloy such as silicide. In another embodiment, the negative electrode layer 216 includes a different material that undergoes a crystal structure change during lithiation and delithiation.

[0035] When the battery pack 100 is connected to an external circuit 132 powered by the battery pack 100, lithium ions are separated from the electrons in the negative electrode 216. The lithium ions travel through the separator 212 and into the positive electrode 208. The free electrons in the battery create a positive charge in the battery and then flow from the negative electrode 216 to the negative terminal 112 of the battery cell 102 through the negative electrode current collector 220. The electrons are then collected by the pack current collector 120 and transmitted to the pack terminal 128. The electrons flow through the external circuit 132 to provide power to the external circuit 132, and then pass through the positive pack terminal 124, through the positive pack terminal 116, and back into the battery cell 102 via the positive terminal 108, where the electrons are collected by the positive electrode current collector 204 and distributed into the positive electrode 208. The electrons returning to the positive electrode 208 are associated with the lithium ions that have crossed the separator 212. Connecting the battery pack 100 to an external circuit that charges the battery pack 100 results in an opposing flow of electrons and lithium ions.

[0036] When one of the electrodes comprises a material such as silicon or a silicon-based material that undergoes crystal structure changes during lithiation or delithiation, the OCP / SOC or OCP-capacity relationship exhibits different electrochemical properties compared to electrode materials that do not undergo crystal structure changes.

[0037] Figure 3 Lithiation and delithiation curves are plotted for a silicon electrode comprising silicon and a conductive additive. Figure 3 In the graph 300 of FIG. 1 , curve 304a represents the lithiation of the silicon electrode to a cutoff potential of 0.010V, while curve 308a represents the lithiation of the silicon electrode to a cutoff potential of 0.060V.

[0038] like Figure 3 As depicted in , when silicon is lithiated to 10 mV (curve 304a), the delithiation curve 304b has a plateau region 304c at ~0.42 V. When silicon is lithiated to 60 mV or above (e.g., curve 308a), the delithiation curve 308b lacks the plateau region and instead has a continuous slope feature 308c between 0.20 and 0.45 V. This observed change in the delithiation curves 304b, 308b is due to changes in the crystal structure of silicon that occur during lithiation at potentials applied below ~55 mV. The region where the crystal structure changes occur is designated as Region 1 and is marked by circle 312, while the region where the difference between the two delithiation curves is observed is designated in Region 2 and is marked by arrow 316.

[0039] In a similar manner, differences in the delithiation curves may also illustrate the difference between electrodes that have undergone a crystal structure change and electrodes that have not undergone a crystal structure change. Figure 4A "half-cell" test of the delithiation curve of a negative electrode comprising a mixture of graphite and SiO is illustrated. Figure 4 Graph 400 illustrates lithiation and delithiation versus capacity using an electrode comprising a mixture of SiO and graphite, the electrode extracted from a commercially available battery in 18650 format, and the extracted electrode cycled against lithium metal.

[0040] exist Figure 4 In the graph 400 of FIG. 404, curve 404a represents lithiation of the electrode to 1 mV, while curve 404b represents delithiation of the negative electrode after lithiation to 1 mV. Curve 408a represents lithiation of the electrode to 60 mV, while curve 408b represents delithiation of the negative electrode after lithiation to 60 mV. When silicon or silicon-based materials are lithiated below approximately 55 mV, as in curve 404a, the material undergoes a crystal structure change (marked by circle 412). The crystal structure change of the material results in a delithiation curve 404b that is different from silicon or silicon-based materials that are lithiated to a higher cutoff voltage at or above 55 mV (such as the 60 mV lithiation represented by curves 408a and 408b).

[0041] When the negative electrode is lithiated to 1 mV, the delithiation curve 404b has a plateau region 416 at approximately 0.44 V. In contrast, when lithiation of the negative electrode stops at or above 60 mV, the delithiation curve 408b of the negative electrode lacks a plateau region and instead has a continuous slope feature 418 between 0.24 V and 0.5 V. As such, there is a difference between the delithiation curves 404b and 408b as indicated by arrow 420. The plateau region 416 at approximately 0.44 V and the difference 420 between the curves 404b, 408b are due to changes in the crystal structure of silicon, and as will be discussed in detail below, this difference in the delithiation curves can be used to partially verify the electrochemical model of the battery containing silicon-based materials.

[0042] In conventional cells, the potential at which the cell undergoes a change in crystal structure is generally assumed to correspond to approximately 55 mV based on previous literature and initial experimental data. Figure 4 As shown in Figure 1, in the hybrid silicon-graphite negative electrode, the two regions on the delithiation curve are separated from each other by the presence of the graphite material. By varying the cutoff voltage from 70 mV to 50 mV, the half-cell potential (also called ) of the crystal structure change is determined with greater accuracy than the conventional estimate of 55 mV. In some embodiments, the crystal structure change is determined to within 2 mV, while in other embodiments the crystal structure change is determined to within 1 mV.

[0043] Figure 5Graph 500 illustrates a "full cell test" in which a commercial battery is charged to two different potentials at a constant current of 2C. One charging potential results in a measurable signature of delithiation of the battery, demonstrating that the charging potential is high enough to cause crystal structure changes to occur. The test was performed on a commercially available 18650 battery in which the negative electrode comprises graphite mixed with SiO. Figure 5 The full unit test shown in .

[0044] The first curve 504a shows a first charging process, in which the battery is charged to a potential of 4.35V, which causes the negative electrode to be lithiated to a level below approximately 55mV. Figure 5 As seen in FIG. 5 , the charging process of the first curve 504a results in a resting full-cell potential 504c of 4.07V.

[0045] On the other hand, the second curve 508a shows a curve in which the battery is charged to a potential of 3.95 V, which corresponds to the cessation of lithiation of the negative electrode at a potential greater than 55 mV. The charging process depicted by the second curve 508a results in a resting full-cell potential 508c of approximately 3.95V.

[0046] Due to the additional charging of the cell in first curve 504a (marked by circle 512), during the discharge of the cell at a rate of C / 10, the discharge curve 504b of the first charging process exhibits a difference from the discharge curve 508b of the cell charged by the second process (represented by curve 508a). This difference, visible at location 516 in graph 500, can be attributed to the difference in delithiation after the electrodes have been subjected to crystal structure changes. Thus, the charging conditions of first curve 504a cause crystal structure changes in the silicon material, while the charging conditions of second curve 508a do not result in a change in crystal structure changes in the silicon. Using Figure 4 and 5 In the experiments shown in , the full-cell potential at which the crystal structure change occurs can be correlated to the half-cell potential of the negative electrode.

[0047] Parameterization and validation of electrochemical cell models are challenging tasks. Parameterization typically requires electrical testing and specialized electrochemical testing across the operating specifications of the battery. Typically, the signals that can be used to quantify the performance of the battery model are voltage and temperature. Due to the significant large number of parameters necessary for simulating electrochemical models, many parameters are fitted to match model predictions to available experimental data. Challenges in optimization problems include the presence of local minima, insufficient data quality for identifying certain model parameters, and structural challenges in the model that make it difficult or impossible to uniquely identify certain parameters. Therefore, different parameter groups may result in similar voltage and temperature predictions from the model.

[0048] The accuracy of the fitted parameters and the quality of the model can be determined by parameter-specific electrochemical experiments. However, such experiments can be time consuming, and for some parameters, the available methods may not be directly applicable or well refined. An alternative to further evaluate the quality of the model is to measure the internal state of the battery, such as the negative electrode and electrolyte potentials. The acquisition of such measurements typically requires a specialized cell design with multiple reference electrodes. However, in the case of batteries containing active materials that exhibit crystal structure changes, it is possible to partially verify the internal state predictions of the model by detecting features that indicate crystal structure changes, as described above with reference to Figure 3-5 as described.

[0049] Figure 6 is a graph 600 depicting three exemplary charge-discharge cycles 604, 608, 612, where the charge potential is varied with the goal of identifying the charge potential at which the crystal structure change first occurs within 2 mV. Starting from approximately 0% SOC on the first cycle, the cell is charged with a constant current until the cutoff potential is reached , and then discharge the cell at a slow rate (e.g., C / 10) until the cell reaches a potential of 2.5V to observe the presence or absence of features associated with changes in the crystal structure. After the discharge is complete, increase or decrease the cutoff potential , and then repeat the charge and discharge sequence.

[0050] In the first charging cycle 604, the battery is charged to a potential V1 of 4.2 V. The battery is then discharged, and as Figure 6 As shown in FIG. , the corresponding discharge curve does not exhibit features (at 616) indicating a change in the crystal structure (eg, reference 616 above). Figure 5 The difference 516 discussed above). In a second charging cycle 608, the battery is charged to a higher potential V2, such as 4.3V. The battery is discharged again, and this time the battery exhibits a difference 620 in the discharge curve that indicates a change in the crystal structure. In a third charging cycle 612, the battery is charged to a potential V3 between V1 and V2 (e.g., 4.25V), and then the battery is discharged. Figure 6 As seen in FIG. 6 , the discharge during the third charge cycle 612 also exhibits features indicative of a crystal structure change 620. As such, the electrode undergoes a crystal structure change between 4.20V and 4.25V.

[0051] Figure 6 The process illustrated in can be repeated for a desired number of cycles to achieve the desired accuracy. The potential V * (V1 in the illustrated experiment) depicts that no crystal alteration signature was detected in subsequent discharge cycles, whereas for , the crystal change characteristics are detected. The potential V * , extract the relevant internal state (negative electrode potential V neg ) and compare it with the above reference Figure 3-5 Determined half-cell potential Compare with known values ​​of .

[0052] Figure 7 A flow chart 700 is depicted of a model optimization process according to the present disclosure that uses the potential for crystal structure changes to identify a better fit to the model parameters. O With SOC O The inherent thermodynamic relationship between depends on material properties that typically do not change. However, as mentioned above Figure 3 and 4 As illustrated, materials in which the crystal structure changes have different thermodynamic relationships that can change during use of the unit.

[0053] Due to the relationship between OCP and SOC or capacity for materials that exhibit crystal structure changes, detection of crystal structure changes can be used to partially verify the internal state of the battery cell. The verification process includes using the above-mentioned Figure 4 and 5 A similar approach discussed in the previous section is based on an electrochemical cell model of the potential at which silicon undergoes a change in its crystal structure. (block 704). In particular, the battery is charged to a cell potential that is known to cause a change in the crystal structure and a cell potential that is known not to cause a change in the crystal structure. Between the two charging processes, the difference in the delithiation curves (e.g., curves 504b and 508b) (e.g., Figure 5 The difference 516 shown in the figure is used to determine predictable features that may be observed in the discharge curve of the battery after the crystal structure change has occurred compared to the discharge after the crystal structure change has not yet occurred. When using an electrochemical model, the BMS is based on the full cell potential and other parameters such as V neg Alternatively, in other embodiments, the BMS may operate under an electrochemical model based only on the full cell potential or only on the internal state.

[0054] Thus, based on the electrochemical battery model determined in block 704, a feature based on the internal state of the battery can be predicted (block 708). For example, the BMS can be configured to determine that a feature indicating a change in the crystal structure of the battery electrode exists when the potential changes by approximately 5 mV from a curve for which no crystal structure change has occurred at a specific point on the curve. In another embodiment, the BMS determines the feature when the potential changes by approximately 10 mV from a curve for which no crystal structure change has occurred at a specific point on the curve. In various embodiments, the threshold potential difference can be any value between approximately 5 mV and approximately 10 mV, depending on the capacity ratio between graphite and Si or SiO. In other embodiments, the BMS may be programmed with another desired threshold potential difference indicating a crystal structure change feature based on the specific chemistry of the electrode and the battery.

[0055] In further embodiments, the BMS identifies features indicating changes in the crystal structure by, for example, identifying rapid changes in the potential-based estimated SOC of the silicon oxide. In one particular embodiment, the threshold for such SOC-based feature identification would be an approximate or exact 5% deviation from the capacity-based SOC change.

[0056] In addition, the battery cells can be used in the same manner as above. Figure 6 The described manner is similarly cycled to various charge cutoff potentials (block 712). The curves are compared with each other to determine the charge cutoff potential V * ——It is the highest potential at which no crystal change occurs, and the negative electrode potential V neg - which is the lowest potential at which a change in the crystal structure is detected (block 716). The determined charge cutoff potential can be compared to the potential predicted from the electrochemical model Comparison to verify the predicted internal state. Based on the lithiation potential determined from the electrochemical cell model (Block 704) and V identified from the V, I and T data neg Based on the difference between the charging conditions (block 712 ), the model parameters may be updated to better predict the internal state of the battery cell based on the different charging processes (block 720 ), and the updated parameters may be used in subsequent iterations of method 700 .

[0057] In one embodiment, the detection of crystal structure changes can be directly incorporated into the formulation of the optimization problem to fit the model parameters. Typically, the following optimization problem is proposed:

[0058]

[0059] in J is the cost function, x and z refers to the differential and algebraic state of the model, f, g, hrefers to the function that forms the structure of the electrochemical model, i is the number of experiments, It refers to i The initial conditions of the experiment are is the result of the optimization problem, and the set Upper and lower limits are defined for parameter variations. The function is derived from a physics-based Li-ion battery model consisting of coupled partial differential equations using model reduction techniques. f, g, h .

[0060] For simplicity, experimental subscripts are ignored in the following sections. i The model output is And the experimental measurement value is The structure of the cost function is typically However, by using experiments that identify the potentials at which changes in the crystal structure occur, the cost function can be modified to include more information about the internal state of the system. For example, in one such modification, , where if for a given cycle a structural change is detected then , and if no structural changes are detected then .function is an indicator function for crystal structure changes that maps to 0 or 1 depending on the model state, and is the weight. For the function An example structure would be: If at some time during charging ,but , and 0 otherwise, where It refers to the vector z Therefore, the BMS is configured to select or merge appropriate models based on the detection of crystal structure changes.

[0061] In some embodiments, BMS operating parameters are updated based on detected crystal structure changes while the battery is in service (otherwise referred to as online). In particular, in various embodiments, crystal structure changes are detected to improve the accuracy of voltage prediction and SOC estimation, improve voltage prediction and power prediction accuracy, adapt battery limitations to battery life and control battery aging behavior, and / or optimize fast charging algorithms for Li-ion batteries.

[0062] Figure 8 A flow chart example of a process 800 for improving the accuracy of model voltage prediction and SOC estimation based on the detection of crystal structure changes is depicted. During operation of the battery, the electrochemical cell model is used to generate a voltage prediction and SOC estimation signal using, for example, the above reference. Figure 5and block 704 to estimate the battery's state of charge (block 804). When using an electrochemical model, the BMS can estimate the battery's state of charge based on the full cell potential and internal states (such as V neg , which is equivalent to the half-cell potential of the anode). In other embodiments, the BMS may operate based on the full cell potential or the internal state of the battery. From the experimental data, certain features of the discharge curve are determined (block 808) related to whether crystal structure changes have occurred during the charging process for a battery having a negative electrode comprising Si, SiO, Si alloys, or other materials that can undergo crystal structure changes.

[0063] The observed changes in the characteristics can be attributed to changes in the OCP of the battery cell during delithiation of the negative electrode. Therefore, the internal state associated with the structural change is used to determine which delithiation boundary curve of the negative electrode to use for the model for subsequent model predictions (block 812). In particular, when the electrode of the battery has undergone a crystal structure change, the state of charge curve and power prediction are determined from the curve based on the determined crystal structure change, and when the electrode has not undergone a crystal structure change, the state of charge and power predictions are based on the curve for no crystal structure change.

[0064] Advantageously, the disclosed BMS can thus provide more accurate model-based voltage and state of charge predictions than in conventional BMS systems. Since the algorithms used for power prediction in a BMS rely on models for accurate voltage predictions, improving the quality of the voltage prediction model by detecting and accounting for crystal structure changes results in improved power prediction capabilities for the BMS.

[0065] Furthermore, since the algorithm for accurate SOC estimation relies on voltage error feedback between model predicted voltage 820 and measured cell voltage 824 (block 816 ), improving the model quality by detecting structural changes in the negative electrode while the battery is online results in a more accurate estimate of the battery SOC.

[0066] The BMS is configured to use the internal state of the battery (such as V neg ) to charge the battery to maintain the current and current integral (i.e., capacity) within a specified range, thereby also maintaining the potential within a desired range. Since the disclosed BMS operation provides a more accurate SOC determination, the charging current is more accurately controlled to maintain the potential within the desired range.

[0067] In another embodiment, the BMS is configured to optimize the operating regime of the battery in order to reduce battery aging over the life of the battery. The battery is operated by the BMS to gradually reduce the maximum charge cutoff potential or the applied constant voltage over the life of the battery. For batteries that exhibit crystal structure changes, detection of the crystal structure changes can be used to adapt boundary conditions of the operating regime. In a specific embodiment, detection of the crystal structure changes is coupled to a lithium plating regime close to the battery (which occurs at V neg Since lithium plating behavior is a well-known aging mechanism and potential safety issue, by detecting crystal structure changes, the maximum operating voltage of the battery can be adapted to the battery age to maintain V neg >0 to reduce battery deterioration due to lithium plating.

[0068] Detecting crystal structure changes also helps in the adaptation of the empirical fast charge algorithm in the absence of an electrochemical model. Adaptation is made on a cycle-by-cycle basis after detection of crystal structure changes on discharge. In the next charge cycle, this information is used by the BMS to adapt the parameters of the fast charge algorithm, resulting in an iterative learning control process. In particular, knowledge of crystal structure changes during discharge can be used to adapt the maximum voltage or cut-off criteria of the battery during fast charge to achieve fast charging while minimizing battery aging.

[0069] For example, based on a derived and validated relationship between OCP and SOC, lithiation of the battery can be limited to potential values ​​directly adjacent to, but not exceeding, a determined threshold potential at which the electrode undergoes a crystal structure change. In particular, in one embodiment, the battery is lithiated to a full cell potential that is directly above or directly below (e.g., within 2 mV) the threshold potential at which silicon or silicon-based materials have been determined to undergo a crystal structure change. In various embodiments, lithiation during battery operation can be limited to a charge target half-cell potential that is, for example, approximately 1 mV, approximately 2 mV, approximately 5 mV, or approximately 10 mV less than or greater than the crystal structure change threshold potential. Advantageously, in such operation, a single battery or battery type can be tested using both detailed electrochemical models and simpler models, and battery operating parameters can be optimized more accurately than a BMS in which a crystal structure change is assumed to avoid a crystal structure change, or, if desired, ensure that a crystal structure change occurs.

[0070] In another embodiment, the BMS strategy includes using the verified internal state to control the negative electrode potential to a charge target potential in a region between 70 mV and 55 mV. In such an embodiment, the BMS strategy is used in combination with a detailed electrochemical model.

[0071] In a further embodiment, the BMS strategy includes controlling the negative electrode potential to a charge target potential in a region below 55 mV. Controlling the negative electrode potential in a region below 55 mV can enable additional capacity to be obtained for the battery that would otherwise be unavailable if the battery were operated to a conventional 55 mV potential level, which is only an estimate of the potential at which the crystal structure change occurs rather than an exact determination thereof.

[0072] It will be appreciated that variations of the above-described and other features and functions or alternatives thereof may be desirably combined into many other different systems, applications or methods. Various currently unforeseen or unconceived alternatives, modifications, variations or improvements that are also intended to be covered by the foregoing disclosure may be subsequently made by those skilled in the art.

Claims

1. A battery comprising: Electrodes that exhibit a change in crystal structure upon lithiation beyond a threshold potential; A battery management system comprising a controller configured, while the battery is in service: determining a threshold potential based on a crystal structure change by performing charge and discharge cycles to a plurality of different cutoff potentials and determining the threshold potential based on corresponding discharge curves from the plurality of charge and discharge cycles, comprising: charging the battery to a first potential that exceeds the threshold potential, discharging the battery from the first potential, and storing the first discharge curve in a memory, charging the battery to a second potential that does not exceed the threshold potential, discharging the battery from the second potential, and storing the second discharge curve in a memory, and identifying at least one feature that is present in the first discharge curve and is absent in the second discharge curve; determining a battery operating parameter based on the determined threshold potential; and The battery is operated based on the determined battery operating parameters.

2. The battery of claim 1 , wherein determining the threshold potential based on discharge curves from the plurality of charge and discharge cycles comprises selecting as the threshold potential a lowest cutoff potential among a plurality of cutoff potentials where the corresponding discharge curves do not include the at least one feature.

3. The battery of claim 1, wherein the threshold potential is determined within 2 mV accuracy.

4. The battery of claim 1, wherein the determination of the battery operating parameters comprises selecting a state of charge curve based on the determined threshold potential and a charge cutoff potential from a most recent charge.

5. The battery of claim 1, wherein the determination of the battery operating parameters comprises adapting boundary conditions of a charging process based on the determined threshold potential.

6. The battery of claim 5, wherein the determination of the battery operating parameters comprises selecting a charging target potential that is within 2 mV of the threshold potential.

7. A method of operating a battery using a battery management system, the method comprising: While the battery is in service: determining a threshold potential based on a crystal structure change, wherein an electrode of a battery exhibits the crystal structure change when the electrode is lithiated beyond the threshold potential, by performing charge and discharge cycles to a plurality of different cutoff potentials and determining the threshold potential based on corresponding discharge curves from the plurality of charge and discharge cycles, comprising: charging the battery to a first potential that exceeds the threshold potential, discharging the battery from the first potential, and storing the first discharge curve in a memory, charging the battery to a second potential that does not exceed the threshold potential, discharging the battery from the second potential, and storing the second discharge curve in a memory, and identifying at least one feature that is present in the first discharge curve and is absent in the second discharge curve; determining a battery operating parameter based on the determined threshold potential; and The battery is operated based on the determined battery operating parameters.

8. The method of claim 7, wherein determining the threshold potential based on discharge curves from the plurality of charge and discharge cycles comprises selecting as the threshold potential a lowest cutoff potential among a plurality of cutoff potentials where the corresponding discharge curve does not include the at least one feature.

9. The method of claim 7, wherein the threshold potential is determined within 2 mV accuracy.

10. The method of claim 7, wherein the determining of the battery operating parameter comprises selecting a state of charge curve based on the determined threshold potential and a charge cutoff potential from a most recent charge.

11. The method of claim 7, wherein the determining of the battery operating parameters comprises adapting boundary conditions of the charging process based on the determined threshold potential.

12. The method of claim 11, wherein the determining of the battery operating parameters comprises selecting a charging target potential that is within 2 mV of a threshold potential.

Images