WO2025199502A1 - Fluorine-doped materials for functional coating on lithium-based batteries - Google Patents

Abstract

Improved barriers for protecting cathodes and/or anodes in Li batteries are provided. We have found that F-doping of ALD-grown lithium metal oxides results in a barrier having a highly desirable combination of properties. In particular, these barriers have mixed conductivity where both electronic conductivity and ionic conductivity are important. Furthermore, these barriers also have the desirable mechanical properties of uniform and controllable thickness, and highly conformal deposition.

Description

Fluorine-doped materials for functional coating on lithium- based batteries

FIELD OF THE INVENTION

This invention relates to lithium batteries .

BACKGROUND

With the purpose of enabling widespread energy utili zation, lithium batteries , which are renowned for their high energy density, have emerged as crucial energy storage systems that overcome the challenges of limited energy resources and spatial constraints .

However, lithium batteries often encounter several issues that can result in performance degradation and even failure . These include undesirable reactions of the electrodes with the electrolyte forming transport blocking layers of the so-called SEI ( solid-electrolyte interphase ) at the anode and CEI ( cathode-electrolyte interphase ) at the cathode , dissolution of transition metal ions , oxygen evolution that poses fire and safety risks , and structural deterioration and pulveri zation due to cyclic volume changes by the repeated insertion and extraction of lithium-ions during battery cycling . Although solid-state electrolytes have gained attention for their potential in addressing safety concerns , they can encounter similar obstacles including low ionic conductivity .

To surmount these challenges , the application of functional coatings becomes imperative as they of fer essential electrochemical , chemical , and structural stability to active materials . However, these coatings may also increase resistive losses . The functional coating material deposited onto the active electrode surfaces of lithium batteries can impede calendar aging as well as chemical , electrochemical , and structural degradation during battery cycling such as SEI /CEI formation and dendrite formation .

However, such coating materials can also have undesirable properties such as parasitic voltage drop and poor lithium-ion transport . Thus it would be an advance in the art to provide improved properties for electrode barrier layers for in Li batteries .

SUMMARY

Among a range of coating techniques , such as pulsed laser deposition, sol-gel coating, and radio frequency magnetron sputtering, atomic layer deposition (ALD) stands out for its exceptional uni formity and conformality . ALD of fers precise control over coating thickness at the atomic scale with desired stoichiometric composition .

This work provides an ALD-grown composition ( i . e . , Fluorine-doped ALD-grown ternary lithium oxide ) for a functional coating material that provides notable mixed electronic and ionic conduction, which distinguishes itsel f from similarly reported coatings that are predominantly made of electronically insulating oxide materials .

This work can be applied across various sectors of the battery industry where there is a requirement for protecting the battery electrodes from parasitic reactions and degradation ef fects by providing a thin and uni form physical barrier layer, coupled with advantageous characteristics of fered by mixed electronic and ionic conduction . Additionally, this work is applicable to the applications that prioriti ze battery electrode safety and stability . Signi ficant advantages provided by this work include a mixed conducting barrier coating layer with outstanding mixed electronic and ionic conductivity in contrast to other ALD-deposited oxide coating materials that are mostly insulating including alumina, zirconia and silica . This approach enables functional coatings that facilitate facile transport of Li+ ions while mitigating potential battery safety and stability issues .

BRIEF DESCRIPTION OF THE DRAWINGS

FIGs . 1A-C show several barrier configurations .

FIG . ID shows an exemplary embodiment of the invention .

FIGs . 2A-B show the calculated ef fect of F-doping on the density of states .

FIGs . 3A-B are TEM images of an ALD-grown F-doped barrier .

FIGs . 4A-B show the ef fect of F-doping on electrical conductivity .

FIG . 5A shows there is an optimal F-doping level in this example .

FIG . 5B is an Arrhenius plot comparing LMO with an F- doped LVO layer on it to bare LMO .

FIG . 6 shows calculated energy barriers for Li ion transport for undoped LVO and two di f ferent doping levels of LVO . DETAILED DESCRIPTION

Section A describes general principles relating to embodiments of the invention . Section B describes several relevant experiments in detail .

A) General principles

An exemplary embodiment of the invention is a lithium battery including : an anode , a cathode , an electrolyte , and a barrier disposed on at least one of the cathode and the anode . The barrier includes an ALD ( atomic layer deposition) -grown ternary lithium oxide , where the ALD-grown ternary lithium oxide is Fluorine-doped during the ALD growth .

The barrier preferably permits transport of Li ions , e . g . , by having an ionic conductivity in a range from 10“⁶ S/cm to 10² S/cm .

The barrier preferably has a mixed electrical conductivity provided by both electrons and ions , e . g . , by having an electronic conductivity in a range from 10“⁸ S/cm to 10² S/cm (more preferably larger than 10“⁵ S/cm) , preferably in combination with having an ionic conductivity in a range from 10“⁶ S/cm to 10² S/cm .

Suitable compositions for the ALD-grown ternary lithium oxide include Li-M-0 compositions where M is Ti , Nb, V, W, Mn, Mo , Ta, La, Co , or Ni . More generally, cathode or anode materials that exhibit one- , two- , or three-dimensional Li ion transport through their crystal lattice may be suitable barrier materials . Suitable barrier materials may have an Li-M-0 composition where M is a transition metal possessing more than one oxidation state . Such materials may generate Li ion vacancies in their crystal structure when doped with Fluorine that facilitate facile incorporation of Li-ions from ( and/or, to ) the electrolyte .

In general , barriers as considered herein can protect the cathode and/or anode by various mechanisms , including acting as a physical barrier layer and acting as an arti ficial solid interface . Practice of the invention does not depend critically on such mechanisms .

The barrier can be disposed on the cathode and/or disposed on the anode .

The fluorine atomic fraction of the ALD-grown ternary lithium oxide is preferably between 0 . 5at% and 6at% and more preferably between lat% and 3at% . The ALD-grown ternary lithium oxide can be deposited by repetitions of an ALD cycle including a lithium fluoride subcycle sandwiched between lithium vanadate subcycles to provide the Fluorine doping .

The thickness of the barrier is preferably in a range from 3nm to 20nm .

Suitable cathode material compositions include , but are not limited to : LUhnCh ( Spinel ) , LiFePCt ( Olivine ) , LiCo02 ( Layered) , and LiNi_xMn_yCo_z02 ( Layered) . The barrier layer for the cathodes avoids the formation of the blocking cathode electrolyte interface ( CEI ) and replaces it with an an electronically and ionically conducting barrier layer that is transparent to both Li+-ion and electron transport with minimum voltage loss , suppresses parasitic side reactions with the electrolytes , and blocks the dissolution of transition metal from the cathode and avoids oxygen evolution .

Suitable anode material compositions include , but are not limited to : Silicon, graphite , Li4Ti50i2 , and Lithium metal . The main purposes of the barrier coating for anodes are to avoid the formation of the blocking solid electrolyte interface ( SEI ) and instead provide the anode with an electronically and ionically conducting barrier layer that is transparent to both Li+-ion and electron transport with minimum voltage loss , and also to avoid dendrite formation at the anode .

Suitable electrolyte compositions include , but are not limited to : LiPON and Li7La3Zr20i2 . The electrolyte is preferably a liquid or solid electronic insulator having an ionic conductivity in a range from 10“⁸ S/cm 10^-1 S/cm .

B ) Detailed examples

This section describes some examples of the preceding ideas .

Bl ) Introduction

There is great need to develop ef fective barrier coatings to mitigate and minimi ze degradation of electrode materials in Li-ion batteries ( LIBs ) . The barrier coatings hold the potential to improve electrochemical , chemical , and structural stability of the active materials of the battery components . More speci fically, the coating can act as a physical barrier, ef fectively separating the electrodes from the electrolyte to avoid parasitic reactions and thereby stabili zing the electrode surface . Intense research activity reflected in numerous publications and reports attest to the high interest to find the most ef fective protective coating material and how best to fabricate it . This coating needs to be stable and uni formly and conformally cover the electrode surface down to the particulate level and also must be suf ficiently conductive and thin to minimi ze voltage loss due to the added ohmic resistance of the coating layer . In other words , these coatings may pose an extra transport barrier for lithium-ions and electrons and may be considerably more resistive than the active electrode materials . Therefore , it is critical to ensure facile electronic as well as lithium-ion conduction within the barrier coating layer, since this directly influences the cell properties and performance , including its capacity and calendar li fe .

The focus of this work is to provide a general guideline for uni form mixed conducting coating by using ALD ( atomic layer deposition) as a representative coating methodology . Among a wide range of coating techniques for LIBs , including various wet chemistry techniques , pulsed laser deposition, and radio frequency magnetron sputtering, ALD not only allows targeted implementation of extrinsic doping strategies but also provides superior uni form and conformal coverage particularly in contrast to most conventional coating approaches which may often leave unprotected regions exposed to the same degradation mechanisms at play, possibly resulting in severe capacity fade .

We highlight that functional coatings should be mixed conducting . In other words , they should not only facilitate heterogenous charge trans fer but also provide favorable pathways for electronic and lithium-ion conduction . Lithium-ions in the electrolyte require not only a lithium- ion vacancy but also electrons for incorporation into the coating layer while maintaining charge neutrality across the interface . Absence of these defects in insulating oxide coatings may also disrupt charge balance during charging/discharging, potentially resulting in lithium-ion accumulation at the interface between the coating and the electrolyte . Such charge separation and polari zation can adversely af fect electrochemical performance . To maintain charge neutrality and to achieve ef fective ambipolar di f fusion across the barrier coating layer, the electron trans fer must be balanced with the transport of lithium-ions from the electrolyte . The signi ficance of mixed conduction becomes more pronounced when uni form and conformal coatings are applied via ALD, which may have been underestimated in the case of non-uni form or incomplete coverage . For the coating layers to ful fill their intended functionality, it is essential that the coating layer is suf ficiently thin to avoid additional resistive losses , but thick enough to avoid pinholes and cover the electrode surface uni formly and conformally .

However, most of the coating techniques are prone to non-uni form or incomplete surface coverage and often with exposed regions as well as insuf ficient control of thickness , uni formity and conformality of the protective coating . This is schematically illustrated in FIGs . 1A-C, which compare the behavior of discontinuous and continuous insulating versus mixed conducting barrier coatings on the envisioned incorporation mechanism of lithium-ions into the underlying cathode material .

More speci fically, FIGs . 1A-C show a schematic comparison of plausible pathways for lithium-ion incorporation from the liquid electrolyte 104 into cathode barrier layers and their associated voltage profiles . FIG . 1A shows that for discontinuous insulating coatings 106 , lithium-ions in the electrolyte 104 have ready access to the bare surface of cathode 102 for facile incorporation, but only through grain boundaries in parts of the cathode covered by the barrier coating . FIG . IB shows that for continuous insulating coatings 108 , lithium-ions in the electrolyte 104 can incorporate into the underlying LMO ( lithium manganese oxide ) cathode 102 only via cracks and grain boundaries in the barrier layer 108 . The insulating coating layer blocks bulk incorporation of lithium-ions from the electrolyte directly into the crystal lattice sites of the barrier layer . FIG . 1C shows that for a mixed electronically and ionically conducting coating 110 , such as F-doped LVO as presented in this work, lithium-ion incorporation from the electrolyte is facilitated by lithium-ion vacancies formed as a result of F-doping .

FIG . ID schematically shows a battery incorporating a barrier 110 as on FIG . 1C . Here 102 is the cathode , 104 is the electrolyte , 110 is the barrier, 112 is the anode and 114 is an optional second barrier on the anode .

The schematic voltage profiles expected for various cathode coating methodologies are also shown on FIGs . 1A-C . The discontinuous coating depicted in FIG . 1A provides only limited protection by blocking part of the active cathode surface from parasitic reactions . The electronically insulating and uni formly covering coating layer in FIG . IB is expected to result in a signi ficant voltage loss that naturally leads to a reduced potential for the battery . By contrast , however, there is hardly any voltage loss - except for minimal resistive loss - across the mixed conducting barrier coating due to the high electronic conductivity of mixed conducting barrier layer, as illustrated in FIG . 1C .

Many wide band gap oxides and several lithium ternary oxides have been reported in the literature as barrier coatings including ZrCt , AI2O3, ZnO, and LiaPCt . Although electronically insulating univalent oxides provide ideal chemical barriers to reduce the propensity of deleterious side reactions , they also provide a physical barrier for lithium-ion incorporation at the oxide coating/electrolyte interface ( FIGs . 1A-B ) . These highly insulating oxide coatings with univalent cations do not allow lithium-ions from the electrolyte or the cathode to readily incorporate into the oxide crystal lattice due to lack of charge compensation mechanism . The incorporation reaction can be represented by

Without the presence of electronic defects for charge compensation in the coating material , this reaction would have to overcome a large energy barrier to proceed . As a result , it is likely that lithium-ions in the electrolyte access the underlying cathode material through micro-cracks and grain boundaries in the coating layer . As an example of univalent oxides , there is a literature report of deposition of AI2O3 by ALD on LiNio.5Mn1.5O4 particles , which success fully contributed to capacity retention, but increasing the coating thickness resulted in capacity drop due to the hindered transport of both electrons and lithium-ions . The incorporation of Li in the coating materials improves ionic conductivity compared to these oxides . However, due to their wide band gaps in general , lithium ternary oxides may not have suf ficient electronic conductivity and generally exhibit lower ionic conductivities than active cathode materials . For example , another literature report demonstrated that a lithium aluminate ( LiAlCt ) coating layer shows an ionic conductivity of 5 . 6X 10^-8 S-cm^-1 at room temperature and success fully protected LiNio.5Mn1.5O4 from possible side reactions . A lithium phosphate ( Li3PO4 ) coating was also reported to show an ionic conductivity of 4 . 3X 10^-6 S-cm^-1 at room temperature .

In contrast to these highly insulating oxides , this work instead employs mixed ionically and electronically conducting (MIEC ) doped oxides as ef fective cathode barrier coatings via ALD . A systematic fluorine ( F) -doping strategy is introduced in the anion sublattice of the oxide structure to generate lithium vacancies • Anionic doping can create bonding environments with di f ferent charge densities and of fer variations in local spatial distributions , thus af fecting the activation barrier for lithium-ion transport . In particular, F-doping is known to contribute to the formation of a rigid structure and to decrease lattice constants , thereby impacting the capacity for lithium storage . The higher concentration of vacancies due to F- doping not only lowers the activation barrier for the lithium-ion incorporation but also increases lithium-ion conductivity . It is also likely that doping of F as an ef fective donor can partially lead to electronic transitions in the oxidation state of multi-valent cations for charge compensation, which improves electronic conductivity of the parent oxide material as a result .

This coating strategy distinguishes itsel f from the conventional insulating oxide coatings in several important aspects . First , this study utili zes lithium vanadate ( LVO) , especially LiVOa as the ionically conducting host structure for F-doping for enhanced mixed conductivity . LiVOa is a wide band gap semiconducting material with a band gap usually around 3 . 1 eV, and hence is not quite suitable for electrode materials per se , unless it is extrinsically doped to improve both ionic and electronic transport properties .

Second, this study employs a doping strategy to generate lithium-ion vacancies and facilitate both lithium-ion transport and improved electronic conductivity . The coating material incorporates F-doping by lithium fluoride ( LiF) ALD sub-cycles on the oxygen sublattice , (F_o° ) with an ef fective charge of + 1 . Charge neutrality of the coating material is maintained by the generation of . In Kroger-Vink notation, this defect reaction can be expressed as ,

The increased concentration of lithium-ion vacancies induced by F-doping can improve charge transport across the protective coating . Simultaneously, it facilitates incorporation of lithium-ions from the electrolyte ( or the cathode ) into the protective layer, as denoted by the defect reaction below .

Lastly, an optimal F-doping concentration in the LVO host structure is thoroughly studied by adj usting the ratio of ALD sub-cycles of LVO and LiF . Experimental studies corroborated with theoretical calculations demonstrate that the ideal concentration of F-doping exists and that the optimally doped LVO shows the highest mixed conductivity .

This work reports the important findings of assessing the viability and advantages of mixed electronically and ionically conducting coatings doped with fluorine for lithium-ion battery cathode protection .

B2 ) Results and discussion

FIG . 2A shows the calculated PDOS (proj ected density of states ) of the optimi zed LiVOa 1 X 1 X3 supercell structure from density functional theory calculations . FIG . 2B shows the calculated PDOS of the optimi zed LiVOa-yF_y 1 X 1 X3 supercell structure from density functional theory calculations . The advantage of F-doping in the host lithium ternary oxide, for example, in LVO, can be first foreseen by DFT calculations, revealing a significant reduction in the band gap. First, the LiVOa and LiVOa-yF_y supercells were constructed and optimized. The substitution of an 0 site with a single F atom represents 1.7 at . % of F-doping in LVO host structure. In the monoclinic LiVOa structure, the LiOe octahedra shares the corner with the VO4 tetrahedra. The substitution with a single F atom disrupts the cornersharing bonding environment, shortening the bond length with o

Li from 2.05 to 1.94 A while elongating that with V from o

1.64 to 1.98 A. This bond restructuring can cause electron localization. This change in the bond length can affect the electron density distribution, resulting in the ionic bond characteristic between Li and F with localized electrons. These localized electrons are expected to be electrostatically favorable for lithium-ion transport in the host structure, which enhances lithium-ion conductivity as a result. The LiVOa 1X1X3 supercell structure exhibits a direct band gap of 3.1 eV, aligning with the reported value in the literature. The band structure of the LiVOa-yF_y 1X1X3 supercell structure shows a significant reduction in the gap between conduction band maximum (CBM) and valence band maximum (VBM) from 3.1 eV to 2.5 eV where the Fermi level locates near the CBM.

Along with the enhanced electronic conduction from the band structure analysis, the PDOS analysis reveals that the substitution of an oxygen atom with an F atom results in the perturbation of the electronic states of the neighboring vanadium site. FIGs. 2A and 2B show the PDOS of the pristine LiVOa and LiVOa-yFy structures, respectively. The PDOS of the LiVOa-yFy structure is more prominent at the Fermi level and exhibits a metallic property compared to the pristine LiVOa structure . This suggests a considerable increase in electronic conductivity owing to the additional electronic states near the Fermi level . These new electronic states in proximity to the Fermi level in the LiVOa-yFy structure originate from the hybridi zation between vanadium d and oxygen p orbitals , as seen in more detailed simulations , with the F electronic states mostly lying below the Fermi level . More speci fically, the e and t2 states of the VO4 tetrahedra in the LiVOa-yFy structure are split into these new electronic states in the conduction band . The e and t2 symmetry of VO4 is disturbed due to the F substitution compared to the pristine LiVOa . The split of the e states with the d_x2__y2 and d_z2 orbitals and the split of the t2 states with the d_xv, d_V7, and d_X7 orbitals contribute to the new energy states at the vicinity of the Fermi level .

The doping of a single F atom additionally lowers the formation energy of a lithium vacancy . The defect formation energy as a function of chemical potential of Li was studied within the electrochemical stability regime of well-known cathode materials . The formation energy of lithium vacancy is lowered by 2 eV due to F-doping of 1 . 7 at . % in the LiVOa supercell structure . This lowered formation energy of lithium vacancy is strongly correlated with initiating vacancy-mediated transport .

B2a ) Structural and compositional analysis

In the previous section, the band structure and PDOS analysis suggest the importance of F-doping on the enhancement of charge trans fer, which also inspires an incorporation of F-doping in coating materials .

To fabricate LVO-F films with varying F concentrations , di f ferent ALD sub-cycle ratios ( a : b ) of LiVO_x and LiF were employed. XPS (X-ray photoelectron spectroscopy) was performed to quantify the surface-specific at . % of the pristine LVO and LVO-F films. The XPS FIs spectrums for LVO- F films with different a:b ratios indicate that F at . % was systematically controlled in LVO host structure. The FIs peaks are located at 685 eV and are identified as Li-F bonding. For the pristine LVO film, no distinct XPS FIs peak is observed. The peak area of FIs spectrum is proportional to the number of LiF sub-cycles, which suggests that the higher frequencies of LiF sub-cycles increase the F-doping concentration. The F at . % is in the range of 0.7% to 3.8% with various ALD sub-cycle ratios, a:b. LVO-F-616 film exhibits the minimal F concentration of 0.7%, followed by LVO-F-313 film with a higher F concentration of 1.9% which corresponds to the doping concentration in the band structure and PDOS analysis. In addition to the FIs spectrums, an increase in the relative ratio of at . % of Li to V was observed, which is attributed to the higher counts of LiF sub-cycles while the total number of LiVO_x sub-cycles remains unchanged. The V2pa/2 spectra of LVO-F-313 film can be deconvoluted into two distinct valence states, V⁵⁺ and V⁴⁺, with binding energies of 517.3 eV and 516.1 eV, respectively. The presence of the V⁴⁺ valence state can be ascribed to the monoatomic Ar sputtering to remove carbon contamination of the film surface.

FIG. 3A shows a transmission electron microscopy (TEM) image of LVO-F-313-coated LiMn2O4 (LMO) particle. FIG. 3B is a high-magnification TEM image of LVO-F-313-coated LMO particle .

In FIGs. 3A-B, the HRTEM image further shows the conformal and uniform LVO-F-313 coating with 1.9% of F onto the LMO, as a representative of LVO-F coatings. An aberration-corrected scanning TEM (STEM) image with high- angle annular dark-field (HAADF) detector further confirms the conformal LVO-F-313 coating. This also clarifies the interface between the coating layer and the electrode particle, along with the energy-dispersive spectroscopy

(EDS) mapping of V, 0, F, and Mn. Since LVO-F-313 is three- dimensionally coated onto the particle and additionally onto the TEM grid, low signals of V, 0, and F outside the particle can be observed.

B2b) Electrical properties

FIG. 4A shows measured room-temperature electrical conductivity of pristine LVO and LVO-F-313 film on the hPSi platform. FIG. 4B is an Arrhenius plot for pristine LVO and LVO-F-313 film on the hPSi platform.

The electrical properties of LVO-F coating layer have been investigated using the hPSi platform. In parallel with the DFT calculations in the previous section with 1.7 at . % of F-doping, electrical conductivities of the pristine LVO and LVO-F-313 films were compared. A considerable difference of approximately three orders of magnitude was observed in the electrical conductivities between the two compositions at room temperature, as illustrated in FIG. 4A. The electrical conductivity of LVO and LVO-F-313 films at room temperature is 2.6X10^-9 S-cm^-1 and 1.2X10^-5 S-cm^-1, respectively. FIG. 4B illustrates the activation energy for charge transport in the pristine LVO and LVO-F-313 film. The activation energy extracted from equation 4 for the pristine LVO was 0.69 eV, which was in good agreement with the previously reported value of 0.7 eV. However, the activation energy of LVO-F-313 film was only 0.11 eV, which shows a significant decrease due to the advantageous F-doping. We note that the charge transfer resistance reflects both electronic and ionic resistance. To support the concept of mixed conduction, we performed DC conductivity measurements at room temperature. We found a mixed conductive behavior of LVO-F film with an electronic conductivity of approximately 10“⁶ S-cm^-1.

FIG. 5A is an area specific resistance plot with different F-doping concentration on the LMO (LiM^Ct) platform. FIG. 5B is an Arrhenius plot for LVO-F-313 film on the LMO and the bare LMO.

The effect of mixed-conducting LVO-F coating can be simulated by using LMO as an exemplary cathode material. We hypothesize the existence of an optimal F-doping concentration for the highest electrical conductivity. FIG. 5A presents ASR values of LVO and LVO-F films onto the sputtered LMO surface, which was obtained by using the charge transfer resistance from EIS at room temperature. Starting from the pristine LVO film with ASR value of 7.5 Q-cm², ASR decreases with the incorporation of 0.7 at . % of F in the LVO structure. Notably, ASR continues to decrease and reaches its minimum at a F-doping level of 1.9 at.%, resulting in an ASR value of less than 0.4 Q-cm². However, beyond this optimal F concentration, ASR increases with higher F-doping concentration. When the F concentration exceeds 3%, ASR of LVO-F films remains even higher than that of the pristine LVO film, which can be attributed to the increased defect association that slows down the transport, or to the formation of excessive Li-F bonds with electronically insulating characteristics.

The lowest ASR value from LVO-F-313 film signifies enhanced electrical conductivity, encompassing both electronic and lithium-ion conductivity. This distinguishes itself from other reported ALD oxide films. The existence of this optimal F concentration highlights the importance of confining the doping concentration within speci fic range to obtain the highest mixed electronic and ionic conductivity . Within LVO-F-313 structure with the optimal doping concentration, charge transport might remain unimpeded since an electron locali zation, which is induced by the excess electron from F-doping, can play a key role in the af finity to lithium-ion . In contrast , at the doping concentration beyond the optimal point , electrons would be excessively locali zed around the V-0 bonding, thereby possibly forming electron polarons and lower the electrical conductivity .

To delineate the benefits of the optimal LVO-F coating onto the LMO platform, the activation energy for charge transport is compared between bare LMO and LVO-F-313-coated LMO, as illustrated in FIG . 5B . The electrical conductivity of bare LMO at room temperature is 4 X 10^-8 S/cm and the activation energy of bare LMO resonates with the values reported elsewhere . The activation energy for charge transport decreased by approximately 0 . 1 eV . We highlight that the electrical conductivity reduced by more than three orders of magnitude . Combining the intrinsic electrical properties and the application of the optimal LVO-F coating on the LIB cathode material , we demonstrate that not only the optimal LVO-F coating exhibits enhanced mixed conductivity it also lowers the charge trans fer resistance of LMO and facilitates lithium-ion conduction .

B2c ) Lithium-ion transport properties

To investigate how F-doping can af fect lithium-ion transport within LVO host structure at the atomistic scale , NEB calculations were conducted . The substitution of a single 0 atom with F corresponds to the doping concentration of approximately 1.7 at.%, which is equivalent to the F concentration in LVO-F-313 film.

FIG. 6 shows energy profiles of lithium-ion transport in the Lii-xVOa, Lii-_xV03-yF_y, and Lii-_xVO3-2yF2_Y supercell structures from density functional theory calculations.

For these calculations, several lithium-ion transport trajectories for the Lii-_xV03 and Lii-_xV03-_ayF_ay (a=l,2) structures were considered. The trajectory closest to the doped F atom is selected for lithium-ion to transport in the Lii-_xV03 and Lii-_xV03-_ayF_ay supercell structures. The transport trajectory in the Lii-_xV03-_ayF_ay structures can be slightly deviated from the one in the Lii-_xV03 structure due to the relatively strong coulombic interaction between Li and F atoms .

The energy profiles of lithium-ion transport in Lii-_xV03 and two different Lii-_xV03-a_yF_ay structures are illustrated in FIG. 6. A noticeable trend in the transport energy barrier signifies the importance of the optimal F concentration to achieve the lowest transport barrier. Within the Lii-_xV03 structure, it has been noted that the transport barrier height exceeds 0.64 eV, which is well aligned with the experimental activation energy of 0.69 eV. The Lii-_xVO3-_yF_y structure exhibits the lowest magnitude of transport barrier of 0.13 eV that aligns with the experimental activation energy of 0.11 eV from LVO-F-313 film. The significant difference in the activation energy can be ascribed to the affinity to lithium-ion due to localized electrons from F- doping .

The minor disparity between the experimental and theoretical activation energy can be due to the difference in the crystallinity, i.e., between the experimentally tested amorphous LVO-F-313 film and the crystalline Lii-_xVO3- _yF_y supercell used in the NEB calculations. Compared to the Lii-xVOa-yFy structure , the Lii-_xVO3-2yF2_Y structure shows higher barrier height of 0 . 25 eV . This implies that excessive F- doping can lead to sluggish lithium-ion transport .

CI-NEB calculations support that the optimal concentration of F is close to 1 . 7 % , which agrees with the experimental results . However, these LiVOs-related structures in the computational study may not perfectly mirror the structures of LVO-F-aba films .

B3 ) Conclusion

In summary, this work success fully demonstrates that controlled extrinsic doping by fluorine leads to enhanced electronic and ionic conductivity in F-doped LVO coating material prepared by ALD . LVO-F functional coating layer can be expected not only to protect the cathode/electrolyte interface , but also to improve charge transport , which ef fectively addresses poor electrical conductivity of oxide coating layers . It has been proposed that LVO-F-313 film with the optimal F-doping concentration exhibited the lowest ASR value of 0 . 4 Q, ■ cm² and the activation energy of 0 . 11 eV, positioning itsel f , to the best of our knowledge , as one of the best electronically and ionically conductive ALD coating layer so far . DFT calculations provide further support for enhanced electronic and ionic conductivity observed in the LVO-F-313 film by the decrease in the band gap and electron locali zation responsible for enhancing lithium-ion conductivity . Additionally, NEB calculations reveal a signi ficant decrease in the activation energy for lithium-ion transport in the presence of 1 . 7 at . % F, aligning with the experimental findings . We envision that mixed-conducting coatings will become even more essential in next-generation batteries where achieving uni form coatings is of growing significance. The uniform LVO-F barrier layer coupled with enhanced mixed electronic and ionic conduction can open new avenues for improving battery performance and durability without sacrificing charge transport across the interface with the cathode materials.

B4) Methods

B4a) Deposition of LVO and F-doped LVO (LVO-F)

All LVO and LVO-F films were prepared in an Oxford FlexAL PEALD (Plasma-enhanced ALD) system. One full ALD cycle for LVO-F deposition is composed of the LiF ALD subcycle sandwiched by LiVO_x ALD sub-cycles. Here, the LiVO_x ALD sub-cycles consist of LiO_x ALD sub-cycle and V0_x ALD sub-cycles. Lithium tert-butoxide (LiO^tBu, Strem Chemicals, 98%) and vanadium (V) oxytriisopropoxide (VTIP, Strem Chemicals, 98%) bubblers were used as precursors in the LiVO_x ALD process and they were kept at 150°C and 40°C, respectively. The precursor dose time of LiO^tBu and VTIP was fixed at 740 and 540 ms. 10 s of oxygen plasma was utilized as a co-reactant for the LiVO_x ALD sub-cycles at a flow rate of 30 standard cubic centimeter per minute (SCCM) . The plasma was generated by an inductively coupled plasma (TCP) source operating at a power of 300W using oxygen (99.993%) . For the LiF ALD sub-cycles, LiO^tBu and titanium fluoride (TiF4, Strem Chemicals, 98%) were selected as a precursor and a reactant, respectively. The 100 ms dose of TiF4 was preceded by the dose of LiO^tBu. This HF4 reactant canister was held at 90°C and the substrate temperature for all depositions was fixed at 200°C. Between every dose of precursors and reactants, the manifold and the ALD chamber were purged with Ar (99.999%) for 20 seconds to avoid any cross-contamination from precursors during the deposition. We note that F contents in the films did not originate from the contamination from the ALD chamber and were only controlled by the number of ALD sub-cycle .

These LiVO_x ALD sub-cycles can be repeated for the number of ' a' times while the LiF ALD sub-cycle can be replicated for 'b' times . The value of a and b can vary from 1 to 6 . For instance , LVO-F-313 film represents that a single LiF ALD sub-cycle is sandwiched by the LiVO_x blocks repeated for three times . The number of ALD sub-cycles for the pristine LVO was deliberately set to 18 to achieve a comparable film thickness with other LVO-F films .

B4b ) Materials characteri zation

For all characteri zation, all LVO and LVO-F films were deposited on p-type boron-doped Si ( 100 ) wafers . The composition of the films was determined by x-ray photoelectron spectroscopy (XPS ) using the PHI Versaprobe 4 at Stanford nano shared facility ( SNSF) with Al Ka radiation, employing PHI instrument-speci fic relativity sensitivity factors . All films were sputtered with monoatomic Ar to remove surface carbon contamination . In addition, all XPS spectrums were calibrated using the adventitious carbon peak at 248 . 8 eV .

The thickness of the films was measured by the J . A. Woollam M2000 spectroscopic ellipsometer at Stanford shared facility ( SNF) with the incidence angle of 65 , 70 , 75°, using the General Oscillator Layer for oxides with the Tauc- Lorentz functions provided by J . A. Woollam . The relative thicknesses of the Si and its native oxide layer were incorporated in this model as 1 mm and 16 nm .

The structure of LVO-F film was investigated by grazing incidence x-ray di f fraction ( GIXRD) , using the PANalytical X'Pert2 X-ray diffraction at SNSF with Cu Ka radiation. A glass substrate was used to hold the thin film sample and the incidence angle was set to 6.5°.

Transmission electron microscope (TEM) images were obtained using the Thermo Fisher Spectra double-corrected TEM at SNSF with an accelerating voltage of 300 kV. For TEM analysis, the commercial LMO powder (MTI corporation) was loaded onto the TEM grid and LVO-F film was directly deposited onto the LMO powder.

B4c) Electrical measurements

Electrochemical impedance spectroscopy (EIS) and linear sweep voltammetry (LSV) were performed using the Solartron ModuLab XM ECS system. Two platforms were utilized for EIS: highly B-doped p-type silicon platform and LMO platform. To illustrate intrinsic properties of the films without an electrode layer, highly B-doped p-type silicon with an aluminum back-side coating (hPSi platform) was employed as the bottom electrode in temperature-variant EIS. The charge transfer resistance was used in the Arrhenius plot, which was obtained by subtracting the ohmic resistance from the total resistance and corresponds to the diameter of the semicircles in the Nyquist plot. The hPSi platform was also used for LSV with a scan rate of lOOmV-s^-1.

The LMO platform is selected to explore the charge transfer characteristics of the films with an electrode layer as a proof of concept. It consists of an LMO layer deposited onto platinum (Pt) top and bottom electrode. The Pt electrodes and the LMO layer were deposited using RE (Radio Frequency) sputtering. For ASR calculations, the charge transfer resistance of the films was selected. The electrical conductivity (cr) of all films including bare LMO was determined by equation 3:

L (4)

° ~ RA

L is the thickness of the film determined by the ellipsometry, R is the charge transfer resistance, and A is the geometric area of the top Pt electrode.

The activation energy was extracted from the slope in the Arrhenius plot, using the Arrhenius equation: where cr₀ is a pre-exponential term, E_a corresponds to an activation energy, k_B is the Boltzmann constant, and T is the absolute temperature.

All electrical measurements were conducted on at least

4 different regions of the sample for repeatability. The geometric area of the top electrode masks was 700 x 700

B4d) Density functional theory (DFT) calculations

All DFT calculations were performed using QuantumATK atomistic simulation package. All electronic calculations were completed using the generalized gradient approximation (GGA) - Perdew-Burke-Ernzerhof (PBE) functional with the grid-based pro j ector-augmented-wave (GPAW) pseudopotentials. The calculations employed plane wave basis sets with a cutoff energy of 680 eV and the broadening of the Fermi- Dirac distribution at 300K. The self-consistent field (SCF) cycles were iterated until a total energy difference of 10^-5 eV was achieved to calculate the solutions for Kohn- Sham equation. All crystal structures were optimized using limited Broyden-Fletcher-Goldf arb-Shanno algorithms (LBFGS) integrated in the QuantumATK package.

The monoclinic lithium vanadate (LiVOa) structure, possessing the closest stoichiometry to the ALD-deposited LVO film, was selected as the basis for the calculations. This monoclinic LiVOa structure was based on the Materials Project (MP-19373) . There exists two Wyckoff Li sites and three Wyckoff oxygen sites in this LiVOa cell. The material structures in this work were based on the supercell approach. The atomic coordinates of this unit cell were first optimized with the periodic boundary conditions prior to the construction of the supercell under an atomic force o tolerance of below 0.01 eV- A^-1, using the Monkhorst-Pack approach with a k-point grid of 5X5X5. The 1X1X3 supercell structures were then constructed with periodic boundary conditions. To generate the F-doped LVO structure (LiVOa- yFy) , a single F atom substituted one of the three oxygen sites in the optimized LiVOa structure, which presents approximately 1.7 at . % of doping considering total of 60 atoms in the supercell structures. The LiVOa-yF_y supercell structures with different F-doping sites were considered and the most stable LiVOa-yF_y supercell was selected for further calculations in this work. The atomic coordinates of this LiVOa-yFy supercell structure was also optimized under a o force tolerance of 0.015 eV- A^-1. The reciprocal momentum space was discretized using a Monkhorst-Pack Grid 6x6x4 k- point grid shifted to F. A Gaussian smearing of 0.1 eV was applied to the projected density of states (PDOS) , band structure, and partial electron density calculations.

The structures with single lithium vacancy were generated using the optimized LiVOa or LiVOa-yFy structures and further relaxed using the identical force threshold and k-point grid. These structures with single lithium vacancy are denoted as Lii-xVOa and Lii-_xV03-yF_y, respectively. Various lithium vacancy sites in LiVOa and LiVOa-yFy supercells were considered. The formation energy of lithium vacancy was calculated from the following equation: where E(V_Li) is the total energy of the LiVOa or LiVOa-yFy supercell with lithium vacancy, E(pristine) is the total energy of the LiVOa or LiVOa-yFy supercell without lithium vacancy, p_Lt is the chemical potential of lithium, and n_Li is the number of lithium vacancy.

B4e) Nudged elastic band (NEB) calculations

To elucidate the effect of F-doping on lithium-ion transport in LVO host structure, climbing-image NEB (CI-NEB) was implemented for Lii-_xVOa-y and Lii-_xVO3-ayF_ay supercells using the image dependent pair potential. Considering the size of the supercell structures, each specific value of 'a' corresponds to the F at . % of 0, 1.7, and 3.4% correspondingly. Among possible NEB pathways, the trajectory from second to first lithium vacancy site was selected to allow a lithium-ion to hop to another Li Wyckoff site, as described elsewhere. The structures with lithium vacancy o that satisfy a force tolerance of 0.07 eV- A^-1 were selected as the start and end points for NEB tra ectories. The maximum distance for lithium-ion to diffuse in each image o was fixed at 0.5 A. Subsequently, the NEB trajectory was optimized using the energy and force tolerance of 10“⁵ eV o and 0.07 eV- A^-1, respectively.

Claims

1. A lithium battery comprising: an anode; a cathode; an electrolyte; and a barrier disposed on at least one of the cathode and the anode; wherein the barrier comprises an ALD (atomic layer deposition) -grown ternary lithium oxide; wherein the ALD-grown ternary lithium oxide is Fluorine-doped during ALD growth.

2. The lithium battery of claim 1, wherein the barrier has a mixed electrical conductivity provided by both electrons and ions .

3. The lithium battery of claim 1, wherein the barrier permits transport of Li ions.

4. The lithium battery of claim 3, wherein the barrier has an ionic conductivity in a range from 10“⁶ S/cm to 10² S/cm.

5. The lithium battery of claim 3, wherein the barrier has an electronic conductivity in a range from 10“⁸ S/cm to

10² S/cm.

6. The lithium battery of claim 1, wherein the ALD-grown ternary lithium oxide comprises an Li-M-0 composition wherein M is selected from the group consisting of Ti, Nb, V, W, Mn, Mo, Ta, La, Co, and Ni .

7. The lithium battery of claim 1, wherein the barrier is disposed on the cathode and disposed on the anode.

8. The lithium battery of claim 1, wherein a fluorine atomic fraction of the ALD-grown ternary lithium oxide is between 0.5at% and 6at%.

9. The lithium battery of claim 8, wherein the fluorine atomic fraction of the ALD-grown ternary lithium oxide is between lat% and 3at%.

10. The lithium battery of claim 1, wherein the ALD-grown ternary lithium oxide is deposited by repetitions of an ALD cycle including a lithium fluoride subcycle sandwiched between lithium vanadate subcycles.

11. The lithium battery of claim 1, wherein a thickness of the barrier is in a range from 3nm to 20nm.

12. The lithium battery of claim 1, wherein the cathode includes one or more materials selected from the group consisting of: LiMn2O4 (Spinel) , LiFePCt (Olivine) , LiCo02 (Layered) , and LiNi_xMn_yCo_z02 (Layered) .

13. The lithium battery of claim 1, wherein the anode includes one or more materials selected from the group consisting of: Silicon, graphite, Li4TisOi2, and Lithium metal .

14. The lithium battery of claim 1, wherein the electrolyte includes one or more materials selected from the group consisting of: LiPON and Li7La3Zr20i2.

15. The lithium battery of claim 1, wherein the electrolyte is a liquid or solid electronic insulator and has an ionic conductivity in a range from 10“⁸ S/cm 10^-1 S/cm.