The subject matter disclosed herein relates generally to battery technology and more specifically to the suppression of dendrite formation and ion depletion in rechargeable batteries.
A battery may convert, through oxidation and reduction, chemical energy into electrical energy, and vice versa. For example, during the discharge of the battery, atoms at an anode (e.g., negative electrode) of the battery may oxidize to form cations (e.g., positively charged ions) and free electrons. The free electrons may migrate from the anode to a cathode (e.g., positive electrode) of the battery, thereby generating an electric current through an external circuit that includes an electric load of the battery. Moreover, the cations may also travel to the cathode through an electrolyte interposed between the anode and the cathode. Meanwhile, to charge the battery, an electric current may be applied to the battery to cause the atoms at the cathode to oxidize and form both cations and free electrons. The free electrons may return to the anode through the external circuit while the cations may travel through the electrolyte in order to return to the anode.
Articles of manufacture and methods associated with batteries resistant to dendrite formation and ion depletion are provided. In one aspect, there is provided a battery that includes: a first electrode; a second electrode; an electrolyte interposed between the first electrode and the second electrode; and at least one acoustic device configured to generate acoustic streaming during a charging and/or a discharging of the battery, the charging of the battery triggering cations from the first electrode to travel through the electrolyte and deposit on the second electrode, the discharging of the battery triggering cations from the second electrode to travel through the electrolyte and deposit on the first electrode, the acoustic streaming driving a mixing and/or a turbulent flow of the electrolyte, the mixing and/or the turbulent flow of the electrolyte increasing a charge rate and/or a discharge rate of the battery by at least increasing a diffusion rate of cations and/or anions, and the mixing and/or the turbulent flow further preventing a formation of dendrites on the first electrode and/or the second electrode by at least homogenizing a distribution of the cations and/or anions in the electrolyte.
In some variations, one or more features disclosed herein including the following features can optionally be included in any feasible combination. The homogenization may prevent the formation of dendrites by at least decreasing a concentration gradient of the cations and/or anions in the electrolyte.
In some variations, the homogenization may prevent the formation of dendrites by at least increasing a uniformity of the distribution of the cations and anions in the electrolyte.
In some variations, the homogenization may prevent the formation of dendrites by at least increasing a uniformity of the deposit of cations on the first electrode and/or the second electrode.
In some variations, the mixing flow of the electrolyte may further maximize a transport of cations and/or anions to replace the cations and/or anions depleted from the electrolyte during the charging and/or the discharging of the battery.
In some variations, the electrolyte may comprise a liquid electrolyte including one or more of a water, a carbonate-based electrolyte, an ester-based electrolyte, an ether-based electrolyte, an ionic liquid, a nitrile based electrolyte, a phosphate based electrolyte, a sulfur-based electrolyte, and a sulfone-based electrolyte.
In some variations, the electrolyte may comprise a polymer-based electrolyte, an organic electrolyte, a solid electrolyte, a non-aqueous organic solvent electrolyte, and a gas electrolyte.
In some variations, the first electrode may be an anode of the battery.
In some variations, the anode of the battery may be formed from a metal including at least one of a lithium (Li), potassium (K), magnesium (Mg), copper (Cu), zinc (Zn), sodium (Na), and lead (Pb)).
In some variations, the anode of the battery may be formed from an intercalated material including at least one of a graphite, graphene, and/or titanium dioxide (TiO2)).
In some variations, the anode of the battery may be formed from an alloy including at least one of a silicon (Si), aluminum (Al), and tin (Sn).
In some variations, the anode of the battery may be formed from a conversion material including a copper peroxide (CuO2).
In some variations, the second electrode may be a cathode of the battery.
In some variations, the cathode of the battery may be an intercalation type electrode including at least one of a lithium-intercalated carbon electrode, a lithium-intercalated silicone electrode, a vanadium oxide electrode, a lithium excess electrode, a graphite electrode, and a graphene electrode.
In some variations, the cathode of the battery may be an alloy type electrode including a tin (Sn).
In some variations, the cathode of the battery may be an air electrode including at least one of an oxygen (O) and air.
In some variations, the at least one acoustic device may be a transducer deposited on a substrate. The transducer may be configured to respond to an electrical input signal by at least applying tension and compression within and/or upon the substrate. The substrate may respond to the tension and the compression by at least oscillating to generate a plurality of acoustic waves.
In some variations, the plurality of acoustic waves may include surface acoustic waves, Lamb waves, flexural waves, thickness mode vibrations, mixed-mode waves, longitudinal waves, shear mode vibrations, and/or bulk wave vibrations.
In some variations, the at least one acoustic device may include one or more pairs of interdigital transducers, a layer of conductive material, and/or one or more contact pins.
In some variations, the substrate may be formed from at least a piezoelectric material.
In some variations, the piezoelectric material may include lithium niobate (LiNbO3), lithium titanate (Li2TiO3), barium titanate (BaTiO3), lead zirconate titanate (Pb(ZrxTi1-x)O3 wherein (0≤x≤1)), quartz, aluminum nitride (AlN), langasite, lead magnesium niobate-lead titanate (PMN-PT), lead-free potassium sodium niobate (K0.5Na0.5NbO3 or KNN), a doped derivative of lead-free potassium sodium niobate, and/or polyvinylidene fluoride (PVDF).
In some variations, the at least one acoustic device may be configured to generate a plurality of acoustic waves having a frequency corresponding to an attenuation length of the plurality of acoustic waves. The attenuation length may correspond to a first length of the first electrode, a second length of the second electrode, and/or a distance between the first electrode and the second electrode.
In some variations, the at least one acoustic device may be integrated inside a case of the battery and/or integrated on the case of the battery.
In some variations, the battery may be a coin cell, a pouch cell, or a cylindrical cell.
In some variations, the battery may be coupled with a circuit configured to drive the at least one acoustic device. The circuit may include an integrated battery charging circuit and an automatic resonance search function.
In some variations, a method may include: receiving a feedback signal responsive to one or more acoustic waves, the one or more acoustic waves generated by the at least one acoustic device comprising the battery, and the feedback signal corresponding to at least a partial reflection of the one or more acoustic waves formed by one or more components on an interior of the battery; determining, based at least on the feedback signal, a morphology of the interior of the battery; and controlling, based at least on the morphology of the interior of the battery, an operation of the battery.
In some variations, the controlling of the operation of the battery may include terminating the operation of the battery in response to the feedback signal indicating a presence of dendrites and/or an air bubble on a surface of the first electrode and/or the second electrode.
In some variations, the controlling of the operation of the battery may include terminating the operation of the battery in response to the feedback signal indicating a presence of detached dendrites, a breakage in a solid electrolyte interface layer, and/or a formation of a protective polymer layer on the at least one acoustic device.
In some variations, the operation of the battery may be terminated by electrically decoupling the battery from an electric load of the battery and/or from another battery in a same battery array.
The details of one or more variations of the subject matter described herein are set forth in the accompanying drawings and the description below. Other features and advantages of the subject matter described herein will be apparent from the description and drawings, and from the claims. While certain features of the currently disclosed subject matter are described for illustrative purposes in relation to rechargeable batteries, it should be readily understood that such features are not intended to be limiting. The claims that follow this disclosure are intended to define the scope of the protected subject matter.
The accompanying drawings, which are incorporated in and constitute a part of this specification, show certain aspects of the subject matter disclosed herein and, together with the description, help explain some of the principles associated with the subject matter disclosed herein. In the drawings,
When practical, similar reference numbers denote similar structures, features, or elements.
The charging of a battery may cause the formation of dendrites. For example, charging a lithium (Li) metal battery may cause the formation of lithium dendrites at the anode of the battery as lithium ions returning to the anode from the cathode form irregular, mossy deposits on the anode. The formation of dendrites may gradually reduce the battery's discharge capacity. Furthermore, the dendrites forming on the anode may eventually puncture the separator to come in contact with the cathode and cause an internal short within the battery. Susceptibility to dendrite formation may therefore diminish the safety, rechargeability, capacity, and lifespan of conventional lithium metal batteries. The risk of dendrites forming in lithium metal batteries may be especially high at high current densities, which renders lithium metal batteries unsuitable for applications requiring a high charging rate.
In some example embodiments, a lithium metal battery may include an integrated surface acoustic wave (SAW) device, which may operate during the charging of the lithium metal battery to suppress the formation of lithium dendrites in the lithium metal battery. The surface acoustic wave device may generate acoustic streaming, which may drive rapid submicron boundary layer mixing flow of the electrolyte adjacent to the anode of the lithium metal battery. This surface acoustic wave driven mixing flow may increase the uniformity of the lithium deposit on the anode of the lithium metal battery including by decreasing the lithium concentration gradient that is present during the charging of the lithium metal battery, even when the lithium metal battery is subject to rapid charging. Notably, this surface acoustic wave driven mixing flow may suppress the formation of lithium dendrites even when the chemical composition of the lithium metal battery, such as the inclusion of a carbonate-based electrolyte (e.g., ethylene carbonate (EC) and diethyl carbonate (DEC) and/or the like), renders the lithium metal battery especially susceptible to dendrite formation. Moreover, the surface acoustic wave device may operate to suppress dendrite formation with minimal power consumption (e.g., approximately 10 mWh/cm2), especially relative to the power that is consumed to charge the lithium metal battery.
In some example embodiments, the acoustic streaming generate by a surface acoustic wave device may suppress the formation of lithium dendrites in a lithium metal battery even when the chemical composition of the lithium metal battery, such as the inclusion of a carbonate-based electrolyte (e.g., EC/DEC and/or the like), renders the lithium metal battery especially susceptible to dendrite formation.
Referring to
The difference in the thickness of the electrodes of the baseline battery charged without surface acoustic waves and the battery charged with surface acoustic wave may be even more pronounced at a higher current density (e.g., 6 mAcm−2). Whereas the deposition thickness is slightly increased to 6 μm for the battery having the integrated surface acoustic wave device, the deposition thickness of the baseline cell increased dramatically to 27 μm. This significant change in the thickness of the baseline battery may be an indication of dendrite formation and loose lithium deposition. When viewed from the top, the lithium dendrites may appear thinner and more porous when the baseline battery is subjected to a higher current density. Contrastingly, the battery having the integrated surface acoustic wave device may exhibit a more homogeneous morphology including the presence of lithium chunks indicative of the formation of a homogeneous and stable solid-electrolyte interface (SEI).
The cycleability of the battery having the integrated surface acoustic wave device may be examined at different cycle rates with a carbonate-based electrolyte (e.g., 1 M LiPF6 in EC/DEC). While the battery having the integrated surface acoustic wave device may exhibit an average of 91.5% Coulombic efficiency at 1 mAcm−2, the baseline battery may exhibit an 88% Coulombic efficiency. When cycled at a current density of 2 mAcm−2, the battery having the integrated surface acoustic wave device may retain an 89% Coulombic efficiency while the baseline battery may exhibit an 87% Coloumbic efficiency after the first two cycles. Moreover, the baseline cell may begin to exhibit an unstable electrochemistry profile at the third cycle at the current density of 2 mAcm−2. Contrastingly, the battery having the integrated surface acoustic wave device may maintain optimal cycling performance throughout including by continuing to exhibit a stable electrochemistry profile. For example, the battery having the integrated surface acoustic wave device may maintain a >80% Coloumbic efficiency throughout the cycle period even at high charge rates whereas the Coloumbic efficiency of the baseline battery may degrade even at relatively low charge rates.
As shown in
For example, the lithium iron phosphate battery having the integrated surface acoustic wave device may deliver 130 mAh/g at 1 mAcm−2 current density while the baseline battery may deliver 120 mAcm−2 at 1 mAcm−2 current density. Moreover, the decrease in discharge capacity may be more precipitous for the baseline battery when the induced current density is increased. For instance, the baseline battery delivered 8.3% discharge capacity when the current density is increased from 1 mAcm−2 to 6 mAcm−2. Contrastingly, the battery having the integrated surface acoustic wave device delivered 42% discharge capacity when the current density increased from 1 mAcm−2 to 6 mAcm−2.
Referring again to
Referring again to
The difference in discharge capacity and the retention of discharge capacity may be observed in the voltage profile of the baseline battery shown in
When the porosity of the lithium deposits is quantified, the lithium electrode from the baseline battery may exhibit a porosity of 0.541 whereas the porosity of the lithium electrode in the battery having the integrated surface acoustic wave device is significantly lower at 0.0367. The difference in the porosity and morphology of the lithium deposits may also be observed in the cross-sectional views shown in
The performance of a lithium metal battery may be contingent upon its diffusion properties, which directly affect the charge and discharge rate, capacity, and cycling stability of the lithium metal battery. In most batteries, the fluid velocity in the electrolyte, u, may be negligible. As such, the lithium ions (Li+) that are depleted from the electrolyte into the anode due to the ionic migration that occurs charging may be replaced through diffusion. However, in a lithium metal battery that is being subject to rapid charging, diffusion may be too slow to overcome electrolyte ion depletion. As such, the charge rate of the lithium metal battery may be maximized by recirculating the electrolyte to improve ion transport. For example, electrolyte recirculation may be achieved by introducing surface acoustic wave driven streaming, which may increase the fluid velocity u of the electrolyte, for example, zero to approximately 1 m/s. Nevertheless, in some example embodiments, the surface acoustic wave device may be configured to generate surface acoustic waves that maximizes ion transport while suppressing the formation of lithium dendrites.
Conventional models for dendrite formation in electrochemical cells typically cast dendrite formation as a spatially one-dimensional diffusion problem, conserving the number of ions in the electrolyte subject to a predefined electrical current through the cell. The current may be a function of the electrical potential difference between the electrodes. Contrastingly, according to some example embodiments, the flow of electrolyte, especially impinging flows, may inhibit the early growth of small dendrites. Hence, the convective and diffusive transport of ions in the electrochemical cell may be modeled transverse as well as parallel to an electrode. The cell may be assumed to be near the limiting current density and that slight morphological imperfections along the electrode form “hotspots” that locally enhance the rate by which metal ions adsorb onto the electrode and allow for the initial growth of dendrites. Moreover, acoustically-driven flow in the cell may be assumed to affect the distribution of ions along the electrode in the vicinity of these hotspots.
The attenuation length of the sound wave in the electrolyte after its generation from leakage from the surface acoustic wave device may be 4π2f2/c3sound)×(4μ/3p)−1≈1 cm in the electrolyte solution, where f csound, μ, and p denote the frequency, the speed of sound, the viscosity, and the density of the electrolyte solution, 1.22 g/cm3, respectively. The acoustic waves may propagate in the fluid electrolyte over a length scale roughly corresponding to the size of the battery electrodes, a consequence of choosing the 100 MHz operating frequency for the surface acoustic wave device knowing the size of the prototype battery. The acoustic streaming may be most akin to Eckart streaming, due to the lateral confinement and presence of acoustic attenuation through the bulk of the fluid. The experimental flow field may include many vortical cells of characteristic length and velocity δ and uc, respectively. Moreover, the characteristic streaming velocity may be assumed to be uc≈5 mm/s based on the experimental data, and the thickness of each electrolyte chamber in the battery, i.e., L=50 μm, as a characteristic length. With a 1M LiPF6 in EC:DEC electrolyte, the Reynolds number may be Re=pucL/μ≈0.2-2, indicating laminar, almost viscous, flow as one might expect from the dimensions of the structure.
However, taking the diffusion coefficient of the ions to be of the order of magnitude of 10−9 m2/s may indicate strong ion convection and potentially an ion transport boundary layer of a thickness of l≈0.1-1 μm. This conclusion may follow from the requirement that the leading order convective and diffusive components in the transport equations must become comparable in magnitude within the boundary layer, which is satisfied by requiring that the corresponding Peclet number in the boundary layer is ucl/D≈1.
The analysis may be simplified by assuming a simple shear flow of the characteristic velocity uc. The small thickness of the boundary layer in comparison to the gap between the electrodes and the lack of excess pressure therein supports, at least locally, the assumption for simple shear flow.
The steady mass transport of ions, assuming the electrical field in the battery is effectively screened by the high electrolyte concentration, is governed by Equation (1) below.
u·|∇c=D∇
2
c. (1)
wherein c, u, D may denote ion concentration, velocity field, and the constant ion diffusion coefficient, respectively.
[84] The problem may be simplified by further assuming a 2D problem, in which the x coordinate is along the flow in the boundary layer and the y coordinate traverses the electrodes, which are assumed to be flat and parallel (prior to the physical growth of dendrites). As Equations (2) and (3) below show, the problem may be solved subject to the mass conservation of metal ions in the electrolyte and a harmonic variation in ion concentration along the surface of the lithium electrode, which is associated with local ion depletion areas in the vicinity of hotspots for the growth of dendrites.
wherein A may denote the area between the electrodes along the x and y coordinates in a 2D view of the system, cbulk is the concentration of lithium ions in the electrolyte, ∈ is a small perturbation parameter of the excess ion depletion near hotspots with compare to the level of ion depletion away from hotspots, and k is a perturbation wavenumber of ion depletion, which physically may be taken to account for the density of the hotspots along the Li electrode with a corresponding wavelength of 2π/k that is associated with the characteristic separation between hotspots. The surface of the lithium electrode is given at y=0.
In these expressions, localized minima along the lithium electrodes are permitted, where the ion concentration fully vanishes and hence supports the hotspots. The velocity field in the boundary layer is taken to be u=βyex and v=Oey, where u and v are the components of the velocity field along the ex and ey unit vector directions associated with the x and y coordinates, respectively, and β≈uc/δ is the shear rate along the y coordinate, where δ is a characteristic length of the flow in the boundary layer. The solution of this problem subject to δ=0 (no flow) and δ>0 (simple shear flow in the boundary layer) is provided in the supporting information.
In the absence of flow, the diffusion-limited flux of ions to the electrode, −i, may be given by Equation (4) below.
wherein the negative sign in front of I appears because the flux of ions to the electrode is along the −y axis direction. The flux of ions is locally enhanced near the hotspots, suggesting the initial growth of dendrites in this case may be inevitable.
The presence of flow near the lithium electrode may enhance the advection of lithium ions to the electrode in a manner proportional to Pe1/3, wherein Pe≡ucl/D is the Peclet number. In addition, the flow may further enhance the local transport of lithium ions to the hotspots in a manner proportional to Pe1/3. This result may be consistent with the observation that the enhanced convection of ions along the electrode to the hotspots decreases variations in ion concentration that would otherwise arise. The overall rate of lithium ion adsorption onto the electrode may be given by Equation (5) below.
where the assumption may be that ∈≈Pe−2/3 (albeit similar result appears requiring that 1>>∈>>Pe−2/3) and the function Γ( ) is the Euler Gamma function, in which Γ(⅓)≈2.68 and Γ(⅙)≈5.57.
The first term on the right may indicate the spatially monotonic convective contribution of ion flux to a flat homogeneous electrode and the second term indicates the correction to the spatially non-monotonic convective contribution to the ion flux due to the presence of the hotspots. The third term given simply as O(∈) is an additional convective contribution to the ion flux, which is spatially monotonic and may be obtained numerically. The first and third terms may be products of similarity analysis and hence are mathematically singular at the origins, x=0, and hence the expression for the current in Equation (5) may still be physically valid far from the origin.
The mechanism by which flow inhibits the growth of dendrites may be counterintuitive. The flow enhances the flux of lithium ions (Li30 ) to the electrode and particularly to the hotspots where dendrites may grow, as given independently by the first and second terms on the right side of Equation (5), respectively. The ion flux is spatially perturbed by ion depletion next to hotspots for the growth of dendrites, which is given in the second term in the equation. However, the leading order convection term, which decays like x−1/3 along the electrode, eliminates localized ion flux maxima and hence is the key to the inhibition of dendrites' growth. The combined contribution of both terms eliminates localized ion transport maxima to the electrode and hence eliminates spatially localized growth spots—dendrites—on the electrode.
However, this suppression of dendrite growth may only be over a finite length of the electrode from x=0, where the shear flow (or alternatively the electrode) commences, to x<xcrit. As x increases, the second of the two terms in Equation (5) may become dominant and the hotspots at x≥xcrit will begin to allow dendrite growth. To determine this critical length, we require the slope of ion flux to not change sign with respect to x along the electrode, such that d (−i)/dx<0, thus avoiding localized ion flux maxima along the electrode. Substituting Equation (5) into the non-equality, replacing the spatial derivative of the term sin(kx)−√3 cos(kx) ny its numerical upper bound, 2, and ignoring the second order (O(∈)) spatially monotonic contributions to ion flux along the electrode surface, thus comparing between the contribution of the leading order spatially monotonic ion flux and the leading order (harmonic) contribution to the ion flux from the presence of dendrites, gives the expression below.
wherein α≡31/3(1−∈)/Γ(⅓) and β≡√π(3/2)1/3/Γ(⅙).
The correction to the ion flux due to the presence of hotspots in Equation (5) and in the corresponding estimate of the dendrite free length of the electrode, xcrit, are qualitative results. Their quantitative magnitude may be given from the requirement that the contribution of ion depletion (next to hotspots) to the ion flux appears in the first correction (of the order of ∈≈Pe−2/3) to the leading order (O(1)) convective result. Hence, xcrit indicates that the excitation of flow near the electrode inhibits the growth of dendrites but to a limited electrode length, which is dependent on the properties of the electrode. In particular, xcrit may increase when reducing the density of hotspots and their intensity, that is, reducing the excess of ion depletion next to the hotspots. Alternatively, it is clear that increasing flow intensity further increases xcrit. The curious result here is that this length is independent of the specifics of the flow, but only if the Peclet number is significantly greater than one. Here, the means to ensure the Peclet number is sufficiently large may be acoustic streaming.
Accordingly, in some example embodiments, the frequency of the surface acoustic wave device may be selected to ensure the length scale of attenuation of the acoustic wave matches the distance along the interelectrode gap (e.g., the length of the electrodes, the distance between the electrodes, and/or the like) the flow needs to be driven. The integration of small, high-frequency ultrasound generators to drive electrolyte flow within the inter-electrode gaps may give rise to ion flux distributions that render potential locations of dendrite growth stable within a specific distance from the ultrasound source. The distance may be independent of the details of the flow as long as the Peclet number is sufficiently large. This configuration may be feasible with the acoustic streaming induced by a surface acoustic wave device, even with rapid charge rates and the choice of electrode materials that would normally be considered unrealistic. The lithium copper battery, as an example, may be capable of cycling at a current density of 6 mAcm−2 while maintain a Coulombic efficiency above 80% throughout. Similarly, the lithium iron phosphate (LiFePO4) configuration may be capable of delivering a 95 mAh/g discharge capacity after 100 cycles at 2C charge and discharge rates.
As noted, in some example embodiments, a battery may be fabricated to include an integrated surface acoustic wave device. For example, to fabricate the lithium copper battery described with respect to
A surface acoustic wave device may be fabricated through a lift-off lithography process to deposit, for example, twenty-eight pairs of un-weighted gold chromium (Au/Cr) fingers to form an optimal interdigital transducer (IDT) onto a 500 μm thick 127.68° Y-rotated, X-propagating cut lithium niobate substrate (LiNbO3 (LN), Roditi). The surface acoustic wave device may be coated with parylene C using chemical vapor deposition to prevent the reactions with the electrolyte present in the battery. The baseline battery as well as the battery including the integrated surface acoustic wave device may be assembled inside an argon-filled glovebox, where moisture level and oxygen level are maintained at <1 ppm. The housing for the batteries may include nut, back ferrule, front ferrule, and body for sealing the electrolyte and electrode from exposure to air. Moreover, the current collectors used for the batteries may be formed from stainless steel rods.
To further illustrate,
In some example embodiments, the surface acoustic wave device 810 may be configured to generate surface acoustic waves. However, it should be appreciated that the surface acoustic wave device 810 may also generate other types of acoustic waves including, for example, Lamb waves, flexural waves, thickness mode vibrations, mixed-mode waves, longitudinal waves, shear mode vibrations, and/or bulk wave vibrations. The surface acoustic wave device 810 may include a transducer deposited on a substrate. The transducer may be configured to respond to an electrical input signal by at least applying tension and compression within and/or upon the substrate. The substrate may respond to the tension and the compression by at least oscillating to generate the plurality of surface acoustic waves. The transducer may include one or more pairs of interdigital transducers, a layer of conductive material, and/or one or more contact pins. The substrate may be formed from a piezoelectric material including, for example, lithium niobate (LiNbO3), lithium titanate (Li2TiO3), barium titanate (BaTiO3), lead zirconate titanate (Pb(ZrxTi1-x)O3 wherein (0≤x≤1)), quartz, aluminum nitride (AlN), polyvinylidene fluoride (PVDF), and/or the like.
In some example embodiments, the surface acoustic wave device, for example, the surface acoustic wave device 810, may be integrated inside or outside of the case of a battery. When the surface acoustic wave device is integrated outside of the case of a battery, one or more coupling agents may be used to couple surface acoustic waves into the battery. It should be appreciated that the surface acoustic wave device may be integrated into various different types of battery cells in a variety of different manner. For example, for a pouch cell, the surface acoustic wave device may be attached onto any surface of the pouch cell. For a cylindrical cell, the surface acoustic wave device may be positioned from the bottom and/or top flat surfaces, or along the edges of the cylinder rolls. For a coin cell, the surface acoustic device may be positioned onto the flat surfaces or the edge of the round shape of the coin cell.
In some example embodiments, the morphology of the interior of the battery having the integrated surface acoustic wave device may be determined based at least on a feedback signal formed by a reflection of one or more surface acoustic waves being reflected off the surface of the electrodes of the battery. For example, the surface acoustic wave device may generate one or more surface acoustic waves while the battery is being charged and/or discharged. These surface acoustic waves may propagate, through an electrolyte filling the interior of the battery, toward the one or more electrodes of the battery before being reflected off of the surface of the one or more electrodes. The surface acoustic wave device may be further configured to detect the feedback signals formed by the reflection of these acoustic waves off the surface of the one or more electrodes.
The surface acoustic wave device may exhibit piezoelectric properties. For example, the surface acoustic wave device may include a transducer (e.g., one or more pairs of metallic interdigital transducers, a layer of conductive material, contact pins, and/or the like) deposited on a substrate formed from a piezoelectric material. As such, the surface acoustic wave device may generate the plurality of acoustic waves by at least converting an electrical signal into mechanical energy embodied by the acoustic waves. Furthermore, the surface acoustic wave device may detect the feedback signals by at least converting the mechanical energy of the feedback signals into an electrical signal. However, it should be appreciated that instead of and/or in addition to the surface acoustic wave device, a different detector may be used to detect the feedback signals.
In some example embodiments, the battery having the integrated surface acoustic wave device may be coupled with a controller configured to: determine, based at least on the feedback signal, a morphology of an interior of the battery; and control, based at least on the morphology, an operation of the battery. The controller may be configured to terminate the operation of the battery in response to the feedback signal indicating an adverse morphology including, for example, the presence of dendrites and/or air bubbles on the surface of the first electrode and/or the second electrode. In response to detecting the presence of adverse morphology, the controller may terminate the operation of the battery by at least electrically decoupling the battery from an electric load of the battery and/or another battery in a same battery array.
In some example embodiments, stimulus generation may be accomplished by a class of semiconductor circuits known as “phase locked loops” (PLL), or“frequency synthesizers”. This low-cost solution uses a reference crystal oscillator to produce a highly accurate and stable tone. The frequency is programmable over a specified range with very fine (<0.01 MHz) resolution. However, unlike the benchtop RF signal generators or arbitrary waveform generators (AWG) it replaces, the output amplitude of a phase locked loop is usually fixed. Moreover, phase locked loops may be unable to produce the output power required to drive an acoustic surface wave device, thus requiring an amplification block.
In some example embodiments, a chain of amplifiers may be used to couple the output of the phase locked loop to the input of the surface acoustic wave device, achieving increasingly higher voltage swings (with higher supplies or power consumption) as needed. Furthermore, duty cycle control may be added using the enable signals of clock buffers, attenuators (using dedicated chips or a simple resistor voltage divider) may be used to fine-tune the signal swing, and a power amplifier with a push-pull output stage may be employed to efficiently deliver high current (power) to the surface acoustic wave device. The surface acoustic wave device itself may be modelled as a load of low impedance at the resonance frequency.
In some example embodiments, the power management unit (PMU) may generate, from a single battery or a wall outlet, all of the voltage supplies (such as3.3V, 5V, 24V etc.) required by the various semiconductor chips on the printed circuit board. These circuits are commonly known as “DC-DC converters”. “Boost converters” may be used to step-up voltages from input to output while “low dropout” (LDO) regulators may be used to step-down voltages. If higher efficiency is required, a step-down function may also be achieved using “buck converters”. This unit may replace benchtop power supplies.
In some example embodiments, the micro-controller unit (MCU),such as an Arduino Nano, may serve as the interface between the electronic driver system and the end users. Through general-purpose I2C IO expanders, the microcontroller may translate user inputs and send low-level digital signals to control all components on the printed circuit board (PCB). The microcontroller may be connected through a USB connection to a laptop for maximum programming and testing flexibility. It may also be pre-programmed with a few options (e.g., power on/off, frequency up/down, and/or the like) selected by push-buttons. Accordingly, the resulting surface acoustic wave battery system may be turned into a completely self-contained and user-friendly device.
While the electronics described above may be sufficient to drive the surface acoustic wave device, additional value-added features are still possible. For example, in some example embodiments, the electrical driver system 2000 may include thermistors to monitor temperature on certain sections of the board. Digitized and read by the microcontroller, the measured data may be used to monitor operating conditions or within a feedback loop, for example, to automatically shut down when a given component overheats. The electrical driver system 2000 may also incorporate current sensors on the surface acoustic wave device itself to automatically detect the optimal resonance frequency to combat inevitable device-to-device variations and to account for variations in boundary conditions, particularly when liquid ay be present on the surface of the surface acoustic wave device. These factors may often shift the resonance frequency by 100 kHz or more, which may be enough to significantly reduce the performance of an acoustic transducer with a high Q factor.
For example, the phase locked loop frequency range may be swept by the microcontroller and the output current to the surface acoustic wave device may be measured, digitized, and recorded for each stimulus frequency. A range may be specified in the algorithm to minimize time needed to perform the sweep as well as to allow for the selection of higher harmonics, which can be useful in transducers. The voltage amplitude, V, at the final stage of the signal chain, the driver amplifier, may be constant by virtue of its resistor feedback architecture. As such, the higher the output current amplitude, I, the higher the power, P, delivered to the surface acoustic wave device (e.g., P=VI). The frequency at which the measured current amplitude is maximized may thus correspond to the resonance frequency of the transducer.
In some example embodiments, two-dimensional computations may be performed to support the analysis of various battery cells, in particular to determine the changes in the concentration gradient in a lithium metal battery with and without acoustic streaming as shown in
wherein Ni may denote the flux of charged species in the electrolyte and can be expressed as Ni=−Di∇Ci−ziUmFci∇+V□Ciu, Ci may denote the concentration of ions i, zi may denote the charge transfer number, Di may denote the diffusion coefficient, Um may denote mobility, F is the Faraday constant, V is the battery potential, and u is the velocity vector.
For the lithium metal battery having the integrated surface acoustic wave deivice, the simulation is more complex, necessitating the sequential use of the pressure acoustic, creeping flow, and electrochemistry modules for frequency and time-domain computations. The volume-force terms (Fi) may be obtained first from the attenuating acoustic wave propagating through the electrolyte via the pressure acoustic module, where
and ∂i<L> refers to the gradient of the potential energy of the wave in a linear medium.
The wave attenuation in COMSOL may be modeled with respect to the wave's power (P) as
where u0 is the particle displacement, α is the attenuation coefficient, and f is the operating frequency of the surface acoustic wave device.
The volume forces, Fi, found from this calculation may be used in the creeping flow module, represented by a time-average derived expression from mass and momentum conservation to the second order,
−∂i
providing the acoustic streaming-driven flow field for the electrolyte. This flow field is then used in the electrochemistry module to determine the ion concentration gradient in the electrolyte. The analysis may be useful for a qualitative assessment of the observed phenomena better explored by experiment and theory due to the computational cost of such multiphysics high-frequency phenomena.
In some example embodiments, to prevent corrosion from the electrolyte present in the lithium metal battery cell, the surface acoustic wave device may be protected using a thin, electrochemically compatible, durable, and acoustically-compatible material.
Table 1 below depicts the effects of the parylene film on the performance of the surface acoustic wave device. As shown, the effect of a 200 nm parylene coating may be weak, with a 2% decrease in the displacement, velocity, and acceleration. The parylene film is therefore able to protect the surface acoustic wave device in a harsh environment while imposing a negligible effect (e.g., <1%) on the performance of the surface acoustic wave device.
In some example embodiments, to overcome the difficulties associated with observing electrolyte acoustic streaming flow induced by surface acoustic waves, a “dummy” battery assembly made of transparent acrylic plates with water couple with polystyrene particles to emulate the conditions of the actual battery in an observable fashion for a set of simple experiments devised to partially validate the COMSOL computations and the analysis results—in particular, the induced fluid flow—may be employed.
Because acoustic streaming is predicated upon the existence of viscosity and compressibility in fluid flow, the typical assumptions of incompressible Stokesian flow at small scales or batteries may be inappropriate. Instead, the full Navier-Stokes representation in conservation of momentum is used. Through knowledge of the amplitude distribution of the surface acoustic wave source in the representative setup using laser Doppler vibrometry, a velocity boundary condition at the electrolyte boundary adjacent the surface acoustic wave device may be defined.
Within the fluid domain, the convection-diffusion equation with the lithium ion (Li+) species present in the electrolyte under the action of insertion upon the anode and extraction from the cathode according to the configuration dimensions of the prototype battery and the charge rates of 6 mAcm−2 (equivalent to 6 C) may be included. As shown in
Referring again to
Referring again to Equations (1)-(3), the problem associated with those equations may be rendered dimensionless, and hence simplified, by using the transformations
Doing so may give rise to Equations (10)-(12) below.
which introduces two small parameters, e.g., 1/Pe<<1 (Pe=ucδ/D>>1) in Equation (10) and ∈<<1 in Equation (12). Assuming a simple shear flow in the vicinity of the lithium electrode, u=y and v=0.
Equations (10)-(12) may support a transport boundary layer of ions and hence are associated with a singular asymptotic expansion of the concentration c in 1/Pe. There is therefore an outer concentration field far from the lithium electrode, described by C, and an inner (boundary layer) concentration filed near the electrode, described by c. In order to solve the inner (boundary layer) problem, the coordinate y may be rescaled in the form y=YPe−n, so that the leading order diffusive term satisfies convection. Both concentration fields must satisfy that limy→∞c. The leading order concentration field may be expanded in powers of E according to the series expansion C=C0+∈C1+ . . . and c=∈C1+ . . . .
To leading order, the problem associated with Equations (10)-(12) in the outer field may satisfy the system of equations.
which gives the trivial solution C0=1. In the inner (boundary layer) field, where the transformation y+YPe−n is used, the problem takes the leading order form,
Y∂xc0=∂YYC0.
where n=⅓, so that the leading order diffusive terms is satisfied by convection. The corresponding boundary conditions at the surface of the electrode and far away from the boundary layer (where the inner solution is matched to the outer solution) are then,
c0=0 at Y=0, and c0=1 at Y→∞.
respectively. An analytical similarity to this problem is obtained by using the transformation
The boundary layer problem then translates to
This system of equations is satisfied by
where Γ( ) is the Euler gamma function and Γ(⅓)≈2.68.
Taking the y derivative of the leading order concentration near the surface of the lithium electrode at Y=ζ=0, gives,
Hence, the dimensional flux of ions to the electrode is,
where the negative sign infers that the flux is to the electrode. Thus, it is clear that the current generally increases when the Peclet number (the convective flow) increases and when the characteristic length scale of the flow decreases (shear rate increases) while the surface of the electrode is flat and homogeneous. Moreover, the current decreases downstream since the convection of ions reduce the variations in ion concentration along this direction.
Since C0 is a constant, the next order problem set forth in Equations (10)-(12) in the outer field may satisfy the system of equations
which gives the trivial solution C1=0.
The next order problem Equations (10)-(12) in the inner field is
Y∂xc1=∂YYc0. (16)
c1=1+cos(kx) at Y=0. (17)
c1=0 at Y→∞. (18)
where the transformation y=YPe−1/3 may be used again and further requires that ε≈Pe−2/3 in order to include the perturbation of the ion concentration in Equation (17). This problem may be written as a superposition of three subproblems, where c1=c1,1+c1,2+c1,3. Solving the problem for c1,1, which is given by omitting the forcing term ∂xxc0 from Equation (16) and replacing Equation (17) by c1,1=1 at Y=0, one obtains that
Hence, the corresponding dimenstional flux of ions is,
One can further write the problem for c1,2 by omitting the forcing term ∂xxc0 from Equation (16) and replacing Equation (17) by c1,2=cos kx at Y=0. The problem is written as,
Y∂x
1,2=eikx at Y=0. (22)
1,2=0 at Y→∞. 923)
using the complex variable c1,2 whose real component is c1,2.
Using the transformation c1,2=f(Y)eikx in (21)-923) gives the alternate system of equations,
which is satisfied by the complex solution
f=32/3Γ(2/3)Ai((ik)1/3Y). (24)
where Ai is the Airy function of the first kind. The Airy function decays in the limit Y→∞ subject to the argument (ik)1/3.
The real component of the Y derivative of c1,2 is given by,
Hence, the corresponding dimensional flux of ions is,
Finally, one can write the problem for c1,2 using Equation (16) and replacing Equation (17) by c1,3=0 at Y=0. The problem for c1,3 gives a spatially monotonic solution and requires a numerical solution; however, this solution does not contribute to the leading order solution for the dendrite free length of the electrode. Hence, we refer to the solution of this problem in terms of its order of magnitude of O(ε) in the followings.
The total ion flux to the Li electrode is given by i=i0+ε(i1,1+i1,2+i1,3), which translates to,
where again ε≈Pe−2/3. A similar problem and solution may appear in the case where the value of r is arbitrary while satisfying 1>>ε>>Pe−2/3, with the exception that the forcing term ∂xxc0 does not exist in Equation (16) and hence the result given in Equation (27) does not contain the third term on the right hand side of the equation, given as O(ε).
The subject matter described herein can be embodied in systems, apparatus, methods, and/or articles depending on the desired configuration. The implementations set forth in the foregoing description do not represent all implementations consistent with the subject matter described herein. Instead, they are merely some examples consistent with aspects related to the described subject matter. Although a few variations have been described in detail above, other modifications or additions are possible. In particular, further features and/or variations can be provided in addition to those set forth herein. For example, the implementations described above can be directed to various combinations and subcombinations of the disclosed features and/or combinations and subcombinations of several further features disclosed above. In addition, the logic flows depicted in the accompanying figures and/or described herein do not necessarily require the particular order shown, or sequential order, to achieve desirable results. Other implementations may be within the scope of the following claims.
This application claims priority to U.S. Provisional Patent Application No. 62/882,450, filed on Aug. 2, 2019 and entitled “CHEMISTRY-AGNOSTIC PREVENTION OF ION DEPLETION AND DENDRITE FORMATION IN A LIQUID ELECTROLYTE,” and U.S. Provisional Application No. 62/968,556, filed on Jan. 31, 2020 and entitled “CHEMISTRY-AGNOSTIC PREVENTION OF ION DEPLETION AND DENDRITE FORMATION IN A LIQUID ELECTROLYTE,” the disclosures of which are incorporated herein by reference in their entirety.
This invention was made with government support under Grant Number EE008363 awarded by the Department of Energy. The government has certain rights in the invention.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/US20/44685 | 8/1/2020 | WO |
Number | Date | Country | |
---|---|---|---|
62882450 | Aug 2019 | US | |
62968556 | Jan 2020 | US |