Metal fluoride passivation coatings prepared by atomic layer deposition for Li-ion batteries

Abstract

The fabrication of robust interfaces between transition metal oxides and non-aqueous electrolytes is one of the great challenges of lithium ion batteries. Atomic layer deposition (ALD) of aluminum tungsten fluoride (AlW.sub.xF.sub.y) improves the electrochemical stability of LiCoO.sub.2. AlW.sub.xF.sub.y thin films were deposited by combining trimethylaluminum and tungsten hexafluoride. in-situ quartz crystal microbalance and transmission electron microscopy studies show that the films grow in a layer-by-layer fashion and are amorphous nature. Ultrathin AlW.sub.xF.sub.y coatings (<10 Å) on LiCoO.sub.2 significantly enhance stability relative to bare LiCoO.sub.2 when cycled to 4.4 V. The coated LiCoO.sub.2 exhibited superior rate capability (up to 400 mA/g) and discharge capacities at a current of 400 mA/g were 51% and 92% of the first cycle capacities for the bare and AlW.sub.xF.sub.y coated materials. These results open new possibilities for designing ultrathin and electrochemically robust coatings of metal fluorides via ALD to enhance the stability of Li-ion electrodes.

Claims

1. A battery comprising: LiCoO.sub.2 cathode core; a metal fluoride passivation coating conformally coated on the cathode core; the metal fluoride passivation coating consisting of AlW.sub.xF.sub.yC.sub.z excluding oxygen, the metal fluoride coating comprising W particles uniformly dispersed in an ALF.sub.3 matrix; and wherein x>0, y>0, and z> or equal to 0.

2. The battery of claim 1, wherein the metal fluoride passivation coating has a thickness of less than 1 nm.

3. The battery of claim 1, wherein the AlF.sub.3 matrix is amorphous.

4. The battery of claim 3, wherein the W particles have a size of ˜1 nm.

5. The battery of claim 1, wherein z>0.

6. The battery of claim 5, wherein the metal fluoride passivation coating also comprises particles of WC.sub.z disposed in the matrix of AlF.sub.3.

7. The battery of claim 1, wherein the battery retains 99% of its initial capacity after 50 charge-discharge cycles.

8. The battery of claim 1, wherein the LiCoO.sub.2 cathode core comprises LiCoO.sub.2 powder, carbon and polyvinylidenefluoride in N-methyl pyrrolidone.

9. The battery of claim 1, further comprising a current collector, the LiCoO.sub.2 cathode core disposed on the current collector.

10. The battery of claim 1, further comprising an electrolyte, the electrolyte comprising a LiPF.sub.6 solution in ethylene carbonate and ethylmethyl carbonate.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The foregoing and other objects, aspects, features, and advantages of the disclosure will become more apparent and better understood by referring to the following description taken in conjunction with the accompanying drawings, in which:

(2) FIGS. 1A-B illustrate (FIG. 1A) in-situ QCM mass-uptake during two alternating cycles of WF.sub.6 and TMA and (FIG. 1B) HRTEM image of the films deposited using alternating exposures to TMA-WF.sub.6 on an ALD Al.sub.2O.sub.3-coated Si wafer (inset: NBD of the AlW.sub.xF.sub.y film with beam size of ˜50 nm).

(3) FIGS. 2 A-B illustrates (FIG. 2A) First-cycle voltage profiles of bare (black) and AlW.sub.xF.sub.y-coated (red) LiCoO.sub.2 between 4.4-2.5 V at 20 mA/g; the inset shows first- and second-charge voltage profiles of the coated LiCoO.sub.2 and (FIG. 2B) Discharge capacities of bare (black) and coated (red) Li+/LiCoO.sub.2 cells at a current rate of 20 mA/g between 4.4-2.5 V; the inset shows Coulombic efficiencies of the samples.

(4) FIGS. 3A-C show Voltage profiles of (FIG. 3A) bare and (FIG. 3B) AlW.sub.xF.sub.y coated LiCoO.sub.2 on cycles 5, 25, and 50 and corresponding dQ/dV plots of bare and coated LiCoO.sub.2 (inset), as well as (FIG. 3C) Al2p XPS spectra of AlW.sub.xF.sub.y-coated LiCoO.sub.2 before and after 50th cycle.

(5) FIG. 4 illustrates rate capabilities of bare (black) and AlW.sub.xF.sub.y-coated (red) LiCoO.sub.2 electrodes.

(6) FIG. 5 shows XPS spectrum of tungsten in the ALD AlW.sub.xF.sub.y film deposited on silicon

(7) FIG. 6 shows XPS survey spectrum of ALD AlW.sub.xF.sub.y film deposited on LiCoO.sub.2 laminate.

(8) FIG. 7 illustrates the self-limiting growth of TMA-WF.sub.6 ALD.

(9) FIG. 8 illustrates characterization of an ALD coated AlW.sub.xF.sub.y layer.

(10) FIG. 9 illustrates the interactions of the layers in one embodiment demonstrating the resistance to diffusion of lithium.

(11) FIG. 10 is a graph illustrating the sequence of cycles in one embodiment where a 3:1 cycle sequence is applied.

(12) FIG. 11 illustrates a graph of the number of supercycles for various ratios demonstrating a similar thickness with different compositions resultant from the differing ratios.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(13) Atomic layer deposition (ALD) is a modified form of chemical vapor deposition (CVD) that uses the self-limiting nature of specific precursors to deposit films in a layer-by-layer fashion. ALD is particularly well-suited for coating electrode surfaces in that a conformal coating can be applied with precise control of thickness and composition. Previous studies on ALD-coated electrodes have primarily focused on metal oxides such as Al.sub.2O.sub.3, TiO.sub.2, LiAlO.sub.2, and LiTaO.sub.3 because the ALD chemistry of these oxides is well known. Metal fluoride growth by ALD is complex and challenging, mainly due to the lack of suitable fluorine precursors. For example, HF, a highly aggressive chemical etching agent, has been used to deposit CaF.sub.2, ZnF.sub.2, and SrF.sub.2. More recently, alternative ALD chemistries have been developed such as MgF.sub.2 and LiF ALD using either TaF.sub.5 or TiF.sub.4 as the fluorine precursor for optical applications. However, the substrate temperatures in these cases were 300-400° C.; high enough to degrade battery electrode laminates containing polymeric binders. Another potential limitation of AlF.sub.3 for Li-Ion batteries is that it is a wide-bandgap dielectric and hence electrically insulating. Although still promising as a coating, methods to enhance the material's conductivity while maintaining its superb resistance to chemical attack could be advantageous.

(14) U.S. Pat. No. 8,921,799 and pending application published as U.S. Pat. App. Pub. No. 2012/0187305 describe a general method and materials from the method relating to atomic layer deposition of a composite coating. Described herein is a method for and composition and product having ultrathin, amorphous, composite aluminum-tungsten-fluoride (AlW.sub.xF.sub.y) films on, in one embodiment, LiCoO.sub.2 electrodes via ALD using, in one embodiment, trimethyaluminum (TMA) and tungsten hexafluoride (WF.sub.6), at 50° C. to 300° C. preferably at 200° C. Such films are created with metal fluoride and do not utilize a metal oxide. There is no oxidant step involved in TMA-WF.sub.6 process; TMA reduces tungsten terminated surface in various embodiments, AlW.sub.xF.sub.yC.sub.z is formed where x and y are any non-zero positive number and wherein z can be zero (no carbide), or any positive number. These films are highly conducting, but incorporate AlF.sub.3 in the composite. Although LiCoO.sub.2 is the most commercially successful cathode material for Li-ion batteries, the practical use of LiCoO.sub.2 is limited, in part, to surface reactions involving cobalt dissolution, electrolyte oxidation, as well as structural transformations occurring at potentials higher than ˜4.3 V (vs. Li/Li+). As such, LiCoO.sub.2 might be considered as a model system for surface studies on Li-ion cathodes. Ultrathin AlW.sub.xF.sub.y coatings (˜1 nm) on LiCoO.sub.2 are shown to significantly increase stability relative to bare LiCoO.sub.2 when cycled up to 4.4 V. These results reveal new possibilities for designing ultrathin and electrochemically robust coatings of metal fluorides via ALD, and potentially other techniques, to enhance the stability of Li-ion electrodes. Using high vapor pressure precursors are beneficial when implemented in role-to-role ALD or spatial ALD.

(15) ALD of AlW.sub.xF.sub.y was accomplished using alternating exposures of WF.sub.6 and TMA. In-situ quartz crystal microbalance (QCM) measurements recorded during alternating, 1 second WF.sub.6 and TMA exposures at 200° C. showed a staircase pattern comprised of 160 ng/cm.sup.2 steps consistent with layer-by-layer growth (FIG. 1A). Substrate temperature below 200 C is desirable when ALD is applied to electrode laminates composed of active material, carbon, and polymeric binder. Additional QCM measurements verified that each of the precursor exposures was self-limiting under these conditions. FIG. 7 illustrates the self-limiting growth of TMA-WF.sub.6 ALD. Ellipsometric measurements indicated a growth rate of 2.56 Å/cycle, and combined with the QCM data yield a film density of 6.5 g/cm.sup.3. The resulting films appear amorphous by X-ray diffraction are highly conducting; four-point probe measurements give a resistivity of 3.5×10.sup.−2 Ωcm. The growth rate of the film, deposited by alternating cycles of TMA-WF.sub.6, is relatively high compared to that of typical oxides (˜1-1.5 Å/cycle) and is due in part to tungsten and/or tungsten carbide byproducts that have growth rates of 2.5 Å/cycle.

(16) X-ray photoelectron spectroscopy (XPS) survey scans, performed during depth-profiling of these films deposited on silicon, revealed that they are comprised of W, C, Al, and F. Higher resolution XPS analysis demonstrated that the Al and F are bound as AlF.sub.3, and that the W may be present as both metallic W and tungsten carbide (WC.sub.x) (See FIG. 5). For brevity, in this application these films are referred to as AlW.sub.xF.sub.y or AlW.sub.xF.sub.yC.sub.z where z may equal zero to reflect that carbide may or may not be included. High resolution transmission electron microscopy (HRTEM) and nano-beam diffraction (NBD) measurements, taken from cross-sections of an AlW.sub.xF.sub.y film deposited on Al.sub.2O.sub.3-coated silicon (FIG. 1B), confirmed the amorphous nature of the films. These measurements also revealed that the AlW.sub.xF.sub.y exists as a nanocomposite of ˜1 nm particles in a matrix of lower density. It is believed that the darker particles in FIG. 1B are W and WC.sub.x, while the surrounding matrix of lower density material is AlF.sub.3. In one embodiment, particles are uniformly dispersed in the matrix from TEM, but can be controlled by ALD sequence; WF.sub.6-WF.sub.6 pulsing might generate localized nanoparticles. It is believed that TMA pulse on WF.sub.6 terminated surface reduces tungsten to form metallic or carbide form while fluorine create stable bonding with Al.

(17) To evaluate the electrochemical properties of the AlW.sub.xF.sub.y coating, 5 ALD cycles of TMA-N.sub.2-WF.sub.6-N.sub.2 (˜1 nm) were applied on laminates of LiCoO.sub.2. XPS survey scan of coated LiCoO.sub.2 confirmed that the AlW.sub.xF.sub.y films are deposited on the laminates (FIG. 6). FIG. 8 illustrates characterization of a ALD coated ALW.sub.xF.sub.yC.sub.z layer. FIG. 2A shows first-cycle voltage profiles of two Li+/LiCoO.sub.2 cells, with and without AlW.sub.xF.sub.y coatings, between 4.4-2.5 V at 20 mA/g. Uncoated LiCoO.sub.2 exhibited first-cycle capacities of ˜170 mAh/g compared to ˜165 mAh/g of the coated sample. This difference, though fairly negligible, might be attributed to the inactive nature of the coating layer resulting in slightly higher first-cycle impedance. The polarization on charging, however, disappeared after the first cycle indicating possible changes, such as a lithiated phase in the film, allowing diffusion of lithium ions and similar lithium extraction/insertion voltages as the uncoated electrode (inset FIG. 2A). Recent molecular dynamics (MD) calculations on Al.sub.2O.sub.3 coating suggests that the Al.sub.2O.sub.3 coating layer absorbs lithium until it reaches the thermodynamically stable composition during the lithiation process. The characteristic hexagonal to monoclinic transformation of LiCoO.sub.2 occurring at ˜4.1 V is observed for both cells, though slightly less pronounced for the coated electrode. This is an indication that the ALD coating process did not drastically alter the bulk characteristics of the material; similar to previous coating studies on LiCoO.sub.2. The discharge capacity of the uncoated LiCoO.sub.2 continuously decreased while the coated electrode showed excellent capacity retention over the course of cycling between 4.4-2.5 V. The bare LiCoO.sub.2 retained only 85% of its initial capacity, while the 1 nm AlW.sub.xF.sub.y coating enabled a 99% capacity retention after 50 charge-discharge cycles as shown FIG. 2B. The capacity loss of the bare material is attributed to parasitic, interfacial reactions as previously reported. It should be noted that thicker (˜10 nm) AlW.sub.xF.sub.y coatings on LiCoO.sub.2 showed discharge capacities of just ˜10 mAh/g, revealing a limitation in practical thicknesses. There is a trade-off between protection capability and facile diffusion of lithium ion. A 1 nm thick AlW.sub.xF.sub.y allows diffusion of lithium and protect underlying electrodes.

(18) FIGS. 3A and 3B show voltage profiles and corresponding differential capacity (dQ/dV) plots of bare and coated samples on cycles 5, 25, and 50. Voltage profiles of the uncoated LiCoO.sub.2 clearly showed increased polarization on charge and discharge with continued cycling accompanied by a broadening and shifting of the dQ/dV peaks; typical of increasing impedance and structural degradation. On the other hand, voltage profiles and dQ/dV plots of the AlW.sub.xF.sub.y-coated material showed little change upon cycling. This data reveals that the phase transition associated with lithium ordering at ˜4.1 V is a relatively “soft” transition and a significant fraction of the lithium can be repeatedly, and reversibly, extracted without significant damage to the bulk of the structure. Furthermore, it is clear that surface damage, mitigated here by the AlW.sub.xF.sub.y film, is a much greater factor in the degradation of LiCoO.sub.2 under these cycling conditions. XPS studies were employed on AlW.sub.xF.sub.y-coated materials to probe the chemical composition of the surface before/after the electrochemical cycling. FIG. 3C shows Al2p.sub.3/2 spectra of pristine, coated LiCoO.sub.2 and after 50 cycles. The binding energy of the pristine and cycled samples was fitted to 74.9 eV and 75.0 eV, respectively, indicating only marginal change in the binding energy after electrochemical cycling and a stable AlW.sub.xF.sub.y coating under insertion/extraction of lithium. The fitted value is higher than those of Al.sub.2O.sub.3 (74.2 eV) and LiAlO.sub.2 (73.4 eV), and is closer to O—Al—F bonding as reported earlier.

(19) FIG. 4 compares the rate capability up to 400 mA/g for the coated and uncoated LiCoO.sub.2. AlW.sub.xF.sub.y-coated LiCoO.sub.2 exhibited superior rate capability at all current rates compared to the uncoated material, but the most dramatic improvements occurred at the highest rates (200-400 m A/g). Discharge capacities at a current of 400 mA/g were 51% and 92% of the first-cycle capacities for bare and AlW.sub.xF.sub.y materials respectively. It is well known that reactions at electrode/electrolyte interfaces result in surface films that greatly increase cathode electrode impedances, thereby, causing capacity fade and poor power performance. Furthermore, it has been shown, in agreement with our results in FIG. 3B, that mitigating reactions at the surface of LiCoO.sub.2 results in more stable bulk processes. As such, the data herein clearly demonstrate the improved surface stability of AlW.sub.xF.sub.y-coated LiCoO.sub.2, resulting in excellent electrochemical performance.

(20) FIG. 9 illustrates the interactions of the layers in one embodiment demonstrating the resistance to diffusion of lithium. The AlWxFyCz coating resists HF and transition metal dissolution from the cathode.

(21) Various cycles and supercycle arrangements can be utilized. FIG. 10 is a graph illustrating the sequence of cycles in one embodiment where a 3:1 cycle sequence is applied. FIG. 11 illustrates a graph of the number of supercycles for various ratios demonstrating a similar thickness with different compositions resultant from the differing ratios. Generally, the ALD supercycle comprises (a[TMA-H2O].Math.bWF6)×n=total number. In that supercycle, a and b can be, for example, 3:1, 2:1, 1:1, 4:2. The total thickness may be the same for different compositions by varying the a:b ratio. For example, (3:1)×3, (2:1)×4, (1:1)×6, (4:2)×2˜similar thickness with different compositions.

(22) In summary, a novel ALD process allowing the deposition of metal fluoride materials has been developed for surface protection of Li-ion battery electrodes. A ˜1 nm AlW.sub.xF.sub.y film deposited on LiCoO.sub.2 electrodes has been shown to dramatically enhance cycle life and rate capability of Li+/LiCoO.sub.2 cells. These AlW.sub.xF.sub.y thin film composites appear to combine the chemical inertness of AlF.sub.3 with the high electrical conductivity of a metal. This study suggests new opportunities for ALD and the design of advanced surface structures enabling high capacity lithium ion batteries. Moreover, adjusting the composition of the ALD composites may yield further improvements in performance.

(23) Atomic Layer Deposition: Aluminum fluoride-based composite films were deposited via sequential pulsing of trimethylalumium (TMA) (97%, Sigma Aldrich) and tungsten hexafluoride (WF.sub.6) (99.8%, Sigma Aldrich). The deposition was performed at 200° C. in a hot-walled viscous flow ALD reactor. TMA and WF.sub.6 precursors were maintained at room temperature and ultrahigh purity N.sub.2 was used as a carrier gas with 300 sccm. The base pressure of the ALD reaction chamber was maintained at ˜1.0 Torr. TMA and WF.sub.6 were alternatively pulsed into the 15 sccm of N.sub.2 carrier flow with the following time sequence: 1 s WF.sub.6 dose—5 s purge—1 s TMA dose—5 s purge. An in situ QCM study was performed to study the nature of the deposition. The QCM measurements typically used longer N.sub.2 purge times of 10 s to allow the QCM signal to stabilize after each precursor exposure. The thicknesses of AlW.sub.xF.sub.y coating layers were determined by ex situ spectroscopic ellipsometry using a Cauchy model (Alpha-SE, J. A. Woollam Co.). The resistance of the film deposited on fused silica was determined by performing current-voltage (I-V) measurements using a four point probe measurement (Keithley Model 6487 current voltage source). X-ray photoelectron spectroscopy of AlW.sub.xF.sub.y films deposited on silicon was performed by Evans Analytical Group.

(24) Transmission electron microscopy: Microstructure and crystallinity of AlW.sub.xF.sub.y films deposited on silicon wafer were analyzed using TEM. TEM analysis was performed by Evans Analytical Group (Sunnyvale, Calif.). Cross-section TEM samples were prepared using the in-situ focused ion beam (FIB) lift out technique on an FEI Strata Dual Beam FIB/SEM. The samples were capped with a protective layer of carbon prior to FIB milling, and were imaged with a FEI Tecnai TF-20 FEG/TEM operated at 200 kV in bright-field (BF) TEM mode, high-resolution (HR) TEM mode, and nano-beam diffraction (NBD) mode.

(25) Electrochemical measurements: The LiCoO.sub.2 electrodes were prepared by mixing a slurry of commercial LiCoO.sub.2 powder (Sigma-Aldrich), Super-P carbon, and polyvinylidenedifluoride (PVDF) with a mass ratio of 84:8:8 in N-methyl pyrrolidone (NMP). The mixed slurry was caste on an aluminum foil current collector. AlW.sub.xF.sub.y films were deposited via ALD on the finished electrode laminates with thicknesses of ˜1 nm. 2032-type coin cells were assembled in an Ar-filled glovebox (water and oxygen≤1 ppm) with metallic lithium anodes. The electrolyte consisted of a 1.2M LiPF.sub.6 solution in ethylene carbonate and ethylmethyl carbonate (3:7 mixture). Charge-discharge measurements were recorded on a MACCOR potentiostat at room temperature under a current rate of 20 mA/g between 2.5 V and 4.4 V. Rate capability experiments were conducted with a constant charge rate of 20 mA/g and discharge rates of 20, 100, 200, and 400 mA/g.

(26) X-ray photoelectron spectroscopy: The chemical state of aluminum before and after electrochemical cycling was analyzed by XPS (Sigma Probe: Thermo VG Scientific) with monochromatic Al Kα radiation (1486.6 eV). The cycled 2032 coin cells were disassembled and washed with dimethyl carbonate (DMC), and transferred directly to XPS to eliminate air-exposure during sample transport.

(27) The foregoing description of illustrative embodiments has been presented for purposes of illustration and of description. It is not intended to be exhaustive or limiting with respect to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the disclosed embodiments. It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.