Failure Mechanism and Interface Engineering for NASICON

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Energy, Environmental, and Catalysis Applications

Failure Mechanism and Interface Engineering for NASICON Structure All-solid-state Lithium Metal Batteries Linchun He, Qiaomei Sun, Chao Chen, Jin An Sam Oh, Jianguo Sun, Minchan Li, Wenqiang Tu, Heng-Hui Zhou, Kaiyang Zeng, and Li Lu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 22 May 2019 Downloaded from http://pubs.acs.org on May 26, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Failure Mechanism and Interface Engineering for NASICON Structure All-solid-state Lithium Metal Batteries Linchun He1, Qiaomei Sun1, Chao Chen1, 2, Jin An Sam Oh1,4,5, Jianguo Sun1, Minchan Li1†, Wenqiang Tu1, Henghui Zhou3, Kaiyang Zeng1, Li Lu1, 2,* 1

Department of Mechanical Engineering National University of Singapore, Singapore 117575 2 National

University of Singapore Suzhou Research Institute Suzhou 215123, P. R. China

3 College

of Chemistry and Molecular Engineering Peking University, Beijing 100871, China

4

Graduate School for Integrative Sciences and Engineering, National University of Singapore, Singapore 138632

5

Singapore Institute of Manufacturing Technology, A*STAR (Agency for Science, Technology and Research), 2 Fusionopolis Way, Innovis, Singapore 138634 †Deceased *

on 20 April 2019

[email protected]

Abstract All-solid-state lithium metal batteries (ASSLiMBs) have been considered as one of the most promising next-generation high energy storage systems that replace liquid organic electrolytes by solid-state ones (SSE). Among many different types of SSE, NASICON-structured Li1+xAlxGe2x(PO3)4

(LAGP) shows high ionic conductivity, high stability against moisture and wide working

electrochemical windows. However, it is unstable when it is in contact with molten Li, hence largely limiting its applications in ASSLiMB. To solve this issue, we have studied reaction processes and mechanisms between LAGP and molten Li, based on which, a failure mechanism is hence proposed. With better understanding the failure mechanism, a thin thermosetting Li salt polymer- P(AA-co-MA)Li layer is coated on the bare LAGP pellet before contacting with the

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 26

molten Li. To further increase the ionic conductivity of P(AA-co-MA)Li, LiCl is added in P(AAco-MA)Li. A symmetric cell of Li/interface/LAGP/interface/Li that is prepared using molten LiSn alloy, and is galvanically cycled at current densities of 15, 30 and 70 μA cm-2 for 100 cycles showing stable low overpotentials of 0.036, 0.105 and 0.257 V, respectively. These electrochemical results demonstrate that the interface coating of P(AA-co-MA)Li can be an effective method to avoid interfacial reaction between the LAGP electrolyte and molten Li.

Keywords All-solid-state lithium metal battery; NASICON structure; Solid-state electrolyte; Failure mechanism; Interface

1.Introduction Currently Li ion battery is the main energy storage system for portable electronics as well as large energy applications such as electric vehicle, submarine etc. With the increased demons in energy density and safety for storage devices, the current commercial electrodes and format of the batteries are however unable to meet these requirements.

1

Therefore, designing new energy

storage systems with high safety and energy density become urgent needs. All-solid-state lithium metal batteries (ASSLiMB) are promising candidates for next-generation energy storage to replace current commercial Li ion batteries. 2-5 One of the key differences between the current Liion battery and ASSLiMB is use of solid-state electrolytes (SSEs). Among various SSEs ceramic SSEs such as garnet

8-10

and NAICON

11-13

3-4, 6-7,

structure are outstanding SSE due to their

high ionic conductivity (10-4-10-3 S cm-1), wide operation electrochemical windows and relatively high stability in the ambient condition. Since NASICON structured Li1+xAlxGe2-x(PO3)4/ Li1+xAlxTi2-x(PO3)4 (LAGP/LATP) are stable with moisture 14, they have been widely used in the Li air battery 15. However, LAGP/LATP is unable to be directly in contact with Li metal since


Page 3 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ti4+/Ge4+ in the LATP/LAGP will be reduced by Li to Ti2+/Ge2+ or even Ti0/Ge0

16-17.

Furthermore, the relative high electronic conductivity of pure NASICON structure SSE was also not satisfied the requirement of a long cycle cell.18-20 To utilize the high ionic conductivity and moisture stability of NASICON structured LAGP/LATP SSEs and high capacity Li metal as the anode, it is essential to design an interface layer which is stable against LAGP/LATP as well as Li metal. There are many possible candidates, for example, poly(oxyethylene) (PEO)

21

and

poly(ethylene glycol) methyl ether acrylate 22 films. However, these soft polymers may easily be reduced and penetrated by Li dendrite when working at elevated temperature

23-24.

Dendrite

growth and penetration through polymer electrolyte still limit application of polymer electrolytebased cells. The main reason is associated with the low mechanical strength of polymer electrolyte, thus, higher shear strength polymer interface layer was used to decrease or eliminate penetration of dendrite. 25-28 The pure SSEs such as garnet SSE (LLZO) also suffers from dendrite growth or Li deposition during galvanic cycle which may be caused by its voids inside. As it was reported, the dendrite Li grows from one side voids to the other side one until resulting a short circuit during galvanic cycle.

29-31

The shear modulus of grain boundary may be also reduced by

Li during cycle which also results dendrite growth along grain boundary of ceramic SSEs.

32

Therefore, intrducing a high mechanical strength interface layer may also suppress the growth of Li dendrite into the voids and grain boundary of ceramic SSEs. Atomic layer deposition (ALD) of LiPON

33,

Al2O3

34-35,

artificial solid state electrolyte

25

were used as the physical barriers for

directly contact of SSEs with Li. There are two possible ways to integrate Li with SSE and hence with ASSLiMB. The first way is to deposit Li on SSE and then current collector, and the second way is to deposit Li on current collector and then to press the current collector with Li layer onto SSE. The uneven surfaces of Li foil and SSEs will cause high impedance due to voids and improper contact if they are directly pressed together. To minimize voids and interfacial resistance, various liquid state Li alloys such as Li-Sn 36, Li-Ge 37, Li-Nb 38, Li-Au alloys

3, 40-41

39

and other

had been used in the garnet SSEs. To our best knowledge, liquid Li alloy metal is



still not used in the NASICON-structured LAGP. Therefore, deep understanding the reaction process and mechanism of LAGP with molten Li is essential for successful fabrication of ASSLiMB through molten Li and Li alloys. Among various Li compounds, the lowest molten temperature is about 185 ℃, which gives a prerequisite to the interface layer. Thus the interface layer for ASSLiMB must possess the following four essential properties: (i) thermally stable at least at temperature >185 ℃, (ii) electrochemically stable with molten Li metal /Li compound, (iii) mechanically preventable to Li dendrite penetration, and (iv) highly conductive for Li ions.

In this work, we study the failure process and mechanisms of LAGP with molten Li through the characterization of morphology and Li ion distribution, and valence. In addition, we successfully introduce a thermosetting Li salt polymer as interface layer between LAGP and molten Li-Sn alloy to avoid side reaction and eliminate voids. The effect of LiCl on this interface layer is also explored to reduce interface resistance. Through detailed evaluation of the interface by FE-SEM, EIS and galvanic cycle, we ascribe the effect of interface layer as both physical barriers to avoid contact of LAGP with molten Li compound and Li ionic conductor. The incorporation of LiCl further increases ionic conductivity of the interfacial layer.

2. Experimental procedures and characterization technologies 2.1 Preparation of NASICON structure LAGP pellets Li1.5Al0.5Ge1.5(PO4)3 (LAGP) solid-state electrolyte pellets were synthesized by a modified solidstate reaction method. In a typical processing, stoichiometric amounts of Li2CO3 (99%, Sigmaaldrich) (10% excess), Al2O3 (99.98%, Alfa Aesar), GeO2 (99.999%, Alfa Aesar) and NH4H2PO4 (98%, Sigma-aldrich) were mixed at 150 rpm for 4 h in a zirconia jar filled with ethanol. The mixture was dried at 80 ℃ for 10 h followed by calcination at 380 ℃ for 4 h to decompose


Page 4 of 26

Page 5 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ammonia and carbon dioxide. After dry milling at 100 rpm for 30 mins, this power was then melted at 1350 ℃ for 2 h with heating rate of 10 ℃ min-1. The molten liquid immediately poured to a pre-heated stainless steel of 500 ℃ to obtain dense glass pellet. This glass pellet was maintained at 500 ℃ for 2 h to release internal stress. It was drying ball-milled again to obtain small glass particle and further crystallized at 800 ℃ for 8 h with heating rate of 5 ℃ min-1. The obtained LAGP powder was pelletized and sintered at 900 ℃ for 8 h with the same heating rate as before. After sintering, the pellets were polished to 0.30 mm in thickness.

2.2 Preparation of P(AA-co-MA)Li solution Poly(acrylic acid-co-maleic acid) solution (P(AA-co-MA)) (MW. 3 000, 50 wt. % in H2O, Sigma-aldrich) was dissolved into deionized water and magnetically stirred for 12 h to obtain 27.5-40 wt. % P(AA-co-MA) solution. LiOH (98%, Sigma-aldrich) was added into this solution with P(AA-co-MA) mole ratio of 0.2 to 1.4 to adjust the pH value of Poly(acrylic acid-co-maleic acid)lithium (P(AA-co-MA) Li) solution, which was further magnetically stirred for 12 h.

2.3 Preparation of Li/LAGP/Li cell To investigate stability of the LAGP pellet with molten Li or Li-Sn alloy, LAGP pellet was polished, washed and vacuum dried for further experiment. Li or Li-Sn alloy were heated to 225 ℃ until totally molten in the glove box. Since there was a thin impurity such as oxide layer on the surface of molten Li or Li-Sn alloy, the thin layer on top of the molten Li or Li-Sn alloy were removed before a LAGP pellet being immersed into the molten Li or Li-Sn alloy for a given duration. After reaction, the LAGP pellet was then removed from the glove box and washed to remove unreacted Li by ultrasonic vibration in the deionized water. After vacuum dried at 110 ℃



for 12 h, the pellets were ready for SPM and XPS characterizations. For EIS measurement, the Li foils were attached on the two sides of the polished LAGP (0.3 mm) in the swagelok.

2.4 Preparation of Li/interface/LAGP/interface/Li symmetrical cell P(AA-co-MA)Li solution (2.5 wt. %, pH=7) with or without LiCl was spray-coated on the LAGP pellet that was at 150 ℃ for every 3 s and waited 60 s for drying. The coating step was repeated for 8 times to obtain an interface layer with thickness of 1.5 μm. This LAGP pellet was finally polished to maintain surface roughness to about 1 μm. After coating and polishing, the LAGP pellets were vacuum-dried at 110 ℃ for 12 h and then immediately stored in the glove box. Li foil and Sn power were mixed with mass ratio of 1:0.6 and melted at 225 ℃. Both sides of surface modified LAGP were then coated with the Li-Sn alloy.

2.5 Evaluation of electrochemical properties X-ray diffractometry (XRD) of LAGP powder was conducted by a Shimazu XRD-6000 with a Cu Kα radiation from 10° to 80° of 0.5° min-1. Microstructure was characterized using field emission scanning electron microscope (FE-SEM) HITACHI S-4300. Fourier transform infrared spectroscopy (FTIR) was performed by GladiATR from 4000 to 400 cm-1 at a resolution of 2 cm-1 after P(AA-co-MA) and P(AA-co-MA)Li powders being vacuum dried at 110 ℃ for 12 h. Viscosity of P(AA-co-MA)Li aqueous solution was measured by a CAP 2000+ viscometer with rotation speed of 800 rpm immediately after preparation. Thermal behavior of P(AA-co-MA)Li was measured using a thermal gravity analyzer (TA-60WS) and a simutaneoul differential temperature and gravity analyzer (DTG-60H) from room temperature to 800 ℃ at a heating rate of 5 ℃ cm-1. Electrochemical impedance spectroscopy (EIS), cyclic voltammetry (CV) and galvanic cycle of LAGP, Li/LAGP/Li, Li/interface/LAGP/interface/Li and LiCl added cells were


Page 6 of 26

Page 7 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conducted by Solartron analytical 1400 Cell Test system at room temperature (23.8 ℃). For the EIS at various temperatures, an oven (WEISS WKL34) was used for control the temperature after each temperature maintained 30 mins.

2.6 Scanning Probe Microscope (SPM) Measurement Electrochemical Strain Microscopy (ESM) and SPM images were conducted by a commercial Scanning Probe Microscope (SPM) system (MFP-3D, Asylum Research, U.S.A) under ambient atmosphere. A conductive cantilever with silicon tip coated with Pt layer(AC240-PP, OPUS, U.S.A.) with first eigenmode spring constant of 2 N/m were used. All the DART-ESM mappings were performed with drive amplitude of 5 V on a 256×256 pixel grid at a scan rate of 0.6 Hz. Then the ESM parameters were obtained using the simple harmonic oscillator (SHO) fitting method.

3. Results and discussion 3.1 Failure process and mechanism of LAGP with molten Li metal Figure S1 (a) shows XRD spectrum of prepared LAGP pellet without visible impurity, and no ping holes could be observed from the FE-SEM image. Figure S1 (c) displays the Nyquist curve of LAGP pellet which contains a high frequency semicircle (1,000 k to 7.94 k) and a long straight line. The high frequency semicircle comes from bulk and grain boundary resistance of LAGP and the long tail relates with ionic block Au electrode. After being fitted by equivalent circuit, the ionic conductivity of the LAGP pellet was obtained to be 1.82 x 10-4 S cm-1 at room temperature (23.8 ℃). To investigate the stability of Li metal with LAGP pellet, a Li metal was molten at 225 ℃ with Sn powder in a mass ratio of 1:0.6 to increase wetability of Li anode.

36

Figure 1(a) shows

the LAGP pellet surface morphology changes with time when it was in contact with the molten Li-Sn alloy. After the LAGP pellet being contacted with Li-Sn liquid for 5 s, the contacted



surface of LAGP changed to black. Continue contacting with molten Li results in breaking of this pellet. This observation clearly shows that the LAGP pellet was extremely unstable with liquid Li metal. As shown in Figure 1(b) of the cross section of the LAGP pellet, the original grain size of LAGP pellet is about 1 μm. For the partially reacted region, although still can see big particles, no clear sharp grains could be observed with a surface showing “erosion” type of morphology. On the relatively smooth surface, there are a few cracks. For the heavily reacted region, a lot of nanosize flake-like morphology could be observed. The flake-like morphology seems being caused by deep corrosion from the molten Li. To further explore the reaction process and mechanism of LAGP with molten Li, scanning probe microscope (SPM) based technique was employed. Selected regions of SPM measurement were conducted showing in the Figure S2, which includes lightly reacted region (gray) and heavily reacted one (black). As a contrast, a primary region was also chosen to measure the performance of original LAGP. Electrochemical strain microscopy (ESM) of SPM was used to scan morphology change and Li ionic density of original, lightly and heavily reacted regions. The principle of ESM is based on the relationship between the molar volume change and bias-induced ion movement, which could measure Li ion movement and diffusion. 42 As it was seen in the Figure S3, the primary LAGP has relatively large grain size of about 1 μm in consistent with the FE-SEM image of LAGP (Figure 1 (b)). Lightly and heavily reacted regions were composed of smaller “grain size” with 400 nm and 50 nm (Figure 1 (c)), respectively. These topography images clearly show decrease in “grain size” after reaction. Li ion motion resulting deformation of test surface under applied voltage (2 V). Thus, the deformation was used for representing movable Li ion density of these three regions (Figure 1 (d)). The deformation in the primary grain was uniform and relatively low (average 2.5 pm), representing uniform and relatively low density Li ion distribution. For LAGP regions reacted with Li metal, maximum deformation reached to 4.65 pm (lightly reacted region) and 5.90 pm (heavily reacted region). This implied Li ion density increased in these reacted regions. Since Li metal was the only source of Li ion, the deformation increase based on primary LAGP could be used for


Page 8 of 26

Page 9 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analysis reaction process of LAGP and molten Li. It is noted from the observation that in the lightly reacted region, some of the grain boundaries exhibit higher deformation than the internal grains or gaps, implying LAGP and molten Li reaction started from these grain boundaries. Heavily reacted region of LAGP shows even higher deformation than pristine and lightly reacted regions, which implying Li rich component formed after heavily reacted. Resonance frequencies are related to contact stiffness by considering the dynamics of a clamped-spring coupled cantilever of the scanned sample (Figure 1 (e)). In the lightly reacted region, the wide resonance frequency range can be attributed to obvious different stiffness of LAGP and lightly reacted LAGP. Bimodal Dual AC images were obtained to analysis the structure and component of LAGP (Figure 1 (f)). Different phases represent different composition or orientation of scanned region. For primary LAGP, the phase related with topography was uniform which proved that this region is composed of one component and one orientation crystal structure. For lightly reacted region, most relatively flat regions have the same phase which implied they have the same component, but regions in the grain boundary have different phase from “grain” regions. These different colors represented that they were different components. For heavily reacted regions, even one small particle composed of different phases at different sides, which implied composition or orientation for this one particle was different at different position. To further scan small area of the heavily reacted region, many peel off points could be found (Figure 1 (g)), which may be the reason of pellet broken in the Figure 1 (a). Although SPM could directly observe the morphology and component changing for different reaction regions, the valence of elements is still unknown.

X-ray photoelectron spectroscopy (XPS) of primary and reacted of LAGP were measured (Figure 2) to reveal chemical changes of all elements. To eliminate the effect of chosen regions, two black regions of reacted LAGP were scanned (Figure S4). After calibration by C1s of each sample (284.8 eV), XPS results of each element could be compared. The XPS spectrums of the



Page 10 of 26

two different reacted regions were the same. The results obtained from one example were used to compare with primary XPS images of LAGP. As can be seen from Figure 2, O1s (Figure 2 (b, h)) and P2p (Figure 2 (e, k)) almost the same after reaction with the molten Li. For Li1s, the peak intensity increased after reaction with the molten Li (Figure 2 (c, i)). The increase in intensity of Li should be caused from increasing density of Li atoms, which was coincided with the increase Li ionic activation measured by SPM (Figure 1 (d)). Al2p (Figure 2 (d, j)) slightly changed after reaction with LiSn0.6. The position of peaks slightly decreased and the area ratio of Al2p 1/2 and 3/2 was different. The change of Al2p may from different binding of Al component after reaction which may imply formation of new Al component. Ge 3d also shows obvious change after reaction (Figure 2 (f, l)). The splitting of Ge 3d is small (∆=0.58 eV) and thus has symmetric lineshape. For primary LAGP, a sharp peak was observed at 32.76 eV which was corresponding to Ge4+. A peak at 32.62 eV was observed for the reacted LAGP, but with low intensity. In addition, several high intensity peaks (28.45 eV, 25.62 eV and 23.54 eV) were observed for the reacted LAGP. The three peaks can be ascribed to low valence of Ge element Ge0, Li-Ge alloy and O2s 43 (ref. 23.2 eV). A similar peak of Li-Ge alloy (26.2 eV) was also observed 17. Thus, the reaction of LAGP with the molten Li can be expressed:

Ge4+ +2Li=Ge2+ +2Li +

[1]

Ge2+ +2Li=Ge0 +2Li +

[2]

Ge0 +xLi=Li x Ge

[3]

In summary, LAGP is unstable with molten Li. The reaction started from grain boundaries resulted in change of “grain size” from 1 μm to 100 nm. Li rich component formed during reaction. Composition of gains changed with degree of reaction for one refined particle. Since fast reaction alone the grain boundaries, some grains lost their contact with LAGP after heavy reaction. Part of Ge4+ near the grain boundaries and surfaces was reduced to Ge0 and even form


Page 11 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Li-Ge alloy.

3.2 Characterization of interface modified LAGP pellet and molten Li Since the surface of LAGP pellet and Li foil were not absolutely flat, direct pressing Li foil on LAGP pellet may result in gaps leading to increased interface resistance. To effectively increase the contact area, an interface layer of P(AA-co-MA)Li was introduced between the LAGP pellet and Li and at the same time, this interface layer can effectively avoid Li to directly contact LAGP. The schematic diagram of this interface modification was shown in the Figure 3 (a). The decomposition temperature of the P(AA-co-MA)Li is as high as 370 ℃ so that it enables it to work with molten Li coating at a temperature below 370 ℃ (Figure 3 (c)). Figure S5 shows wettability of Li and Li-Sn alloy on P(AA-co-MA)Li coated alumina plate. It is clear that the wettability of pure Li with P(AA-co-MA)Li is poor. With alloying Sn into Li, wettability increased with increasing Sn content. In a mass ration of Li:Sn at 1:0.6, the molten Li-Sn almost completely wets with P(AA-co-MA)Li coated alumina plate. Therefore LiSn0.6 alloy was used as the anode for testing. For interface modified LAGP, a molten Li-Sn liquid was easily coated on its surface as shown in the Figure 3 (b). Unlike bare LAGP without P(AA-co-MA)Li coating, the P(AA-co-MA)Li coated LAGP did not react with molten LiSn0.6. The coated LiSn0.6 shows smooth surface without color change after prolonged contacting. Cross-sectional image of Li/LAGP/Li pellet is shown in Figure S6. It is clear that interfacial contact is poor whereas very good adhesion with the help of P(AA-co-MA)Li interface layer could be observed from Figure 4 (a-d). The P(AA-co-MA)Li interface layer is 1.52 μm. The grain size and morphology of LAGP near the LiSn0.6 were the same as that of the primary LAGP (Figure S1 (b)), which implied that LAGP was not reacted with the molten Li. To increase ionic conductivity, we incorporated addition lithium salt-LiCl, to the interface layer. Three mole ratios of LiCl to P(AA-co-MA)Li, 0.5:1, 1:1 and 3:1 were used, which donated as interface-LiCl-0.5, interface-LiCl-1 and interface-



LiCl-3, respectively. The morphologies of three different interfaces are shown in Figure 4 (b), (c) and (d). The cross-section images of LiCl added interface are the same as the one without LiCl. All interface layers were controlled to about 1.5 μm. Electrochemical impedance spectroscopy (EIS) of attached Li on LAGP and molten LiSn0.6 coated on interface modified LAGP were measured to evaluate the ionic conductivity and interface binding (Figure 5). The thickness of the LAGP pellet was controlled at 0.3 ± 0.02 mm. For symmetrical cell Li/LAGP/Li without interface layer, the Nyquist curve composed of two depressed semicircles. According to the frequency range of LAGP (Figure S1 (c)), the Nyquist plot from frequency range of 1000 k to 7.94 k was associated with bulk and grain boundary resistance of LAGP pellet and interface layer (P(AA-co-MA)Li). The Nyquist plot in the frequency range from 7.94 k to 63.1 k in Fig. 5 (a) was caused from interface resistance of LAGP (modified LAGP) and Li, which could also be found from the Li/LAGP/Li symmetrical cell with and without interface layer. The second depressed semicircle of Li/LAGP/Li cell should from Li ionic diffusion in the interface layer which may from voids in the interface layers. The total resistance especially interface resistance of Li/interface/LAGP/interface/Li decreased a lot compared with Li/LAGP/Li. This decreased resistance could be assigned to larger contact area between LAGP and Li metal which could be observed in the Figure 4. To increase conductivity of the interface layer, LiCl was introduced into P(AA-co-MA)Li. The interface layer resistance of the cell with additional of 0.5 % LiCl (LiCl0.5) was obviously reduced, and further increase in the content of LiCl content to 1% led to further decrease in the resistance of the cell (LiCl-1). It was noted from the Nyquist plots, when LiCl was increased to 3%, no further decrease in resistance of the interface layer indicating saturation of LiCl contribution. Thus, the increase of ionic conductivity by adding LiCl may originate from the increase in concentration of Li ion at the interface layer leading to reduced Li ion hopping distance. Too much LiCl cannot contribute to the ionic conductivity of interface layer.


Page 12 of 26

Page 13 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The ionic conductivity relates the interface modified Li/interface/LAGP/interface/Li symmetrical cell was studied from -25 ℃ to 75 ℃ in order to understand impact of interface responses at a wide temperature range (Figure 5 (d)). The activation energy was thus calculated based on

 t  A exp(- Ea / KbT ) where

[4]

 t is ionic conductivity at absolute temperature T, A is a pre-exponential factor, Ea is

activation energy which was treated as constant in the relatively narrow temperature range, and

K b is Boltzmann constant. Then log( t ) related with 1000/T was obtained as shown below. log( t )  log( A)  Ea log e / (1000  K b )  1000 / T

[5]

After the fitting of the experimental data, the activation energy of this symmetrical cell Li/interface/LAGP/interface/Li was obtained to be 0.37 eV which is slightly higher than that of the

pure

LAGP

(0.30

eV).

Galvanic

cycle

of

the

symmetrical

cells

of

Li/interface/LAGP/interface/Li without addition of LiCl and with LiCl-1 were performed from 15 μA cm-1 to 70 μA cm-1 at room temperature (23.8 ℃) to investigate the stability of interface modified symmetrical cells (Figure 6). For each current density, 100 cycles were conducted. The maximum over potential was low for both cells. The one with LiCl-1 addition showed smaller overpotential compared with pure P(AA-co-MA)Li as interface layer (Figure 6 (c)), which was coincide with the Nyquist plots in the Figure 5 (b) and (c). The overpotentical of Li/interface/LAGP/interface/Li was not linear related with current density which may imply Li ion diffusion was limited under high (70 μA cm-1) current density. As a compare, the overpotentical of LiCl added symmetric cell was linear related with applied current density (Figure 6 (c)).

4. Conclusion The failure mechanism of LAGP with molten Li was systematically investigated. The failure of



Page 14 of 26

LAGP started from grain boundary and expanded to rest area of grains. Therefore reduced “grain size” was observed. The composition of LAGP particles were changed with gradual reaction. Some part of LAGP surface peeled off because of heavy reaction. Li-rich component was formed and Ge4+ reduced to Ge0 by Li or even formed Ge-Li alloy after reaction with molten Li. Based on this, a thermosetting Li salt P(AA-co-MA)Li was explored and coated on the LAGP pellet to act

as

an

interface

layer

for

molten

LiSn0.6

and

LAGP

pellet.

The

obtained

Li/interface/LAGP/interface/Li symmetric cell showed neglectable defects in the interface layer and thus had smaller interface resistance. The impedance was further decreased when LiCl was introduced to P(AA-co-MA)Li. The galvanic cycles at different current density were applied to investigate interface modified symmetric cell with and without LiCl added. During 100 galvanic cycles at each current density, stability and small corresponding overpotentical (0.036, 0.105 and 0.257 V) were obtained for symmetric cell with LiCl-1 added interface layer.

Author information Corresponding Author *E-mail: [email protected] ORCID Linchun He: 0000-0002-4786-9710 Qiaomei Sun: 0000-0003-3209-7428 Chao Chen: 0000-0003-4381-8884 Jin An Sam Oh: 0000-0001-9336-234X Jianguo Sun: 0000-0002-2222-2478 Wenqiang Tu: 0000-0002-0659-4941 Prof. Kaiyang Zeng: 0000-0002-3348-0018


Page 15 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Prof. Henghui Zhou: 0000-0003-0317-1756 Prof. Li Lu: 0000-0002-3794-2793

Acknowledgments This work is supported by National University of Singapore, National University (Suzhou) Research Institute, and the Natural Science Foundation of China (NSFC 51572182).



References 1. Tu, W.; Xia, P.; Zheng, X.; Ye, C.; Xu, M.; Li, W., Insight into the interaction between layered lithium-rich oxide and additive-containing electrolyte. J. Power Sources 2017, 341, 348356. 2. Dai, J.; Yang, C.; Wang, C.; Pastel, G.; Hu, L., Interface engineering for garnet-based solidstate lithium-metal batteries: materials, structures, and characterization. Adv. Mater. 2018, 30 (48), 1802068. 3. Gao, Z.; Sun, H.; Fu, L.; Ye, F.; Zhang, Y.; Luo, W.; Huang, Y., Promises, challenges, and recent progress of inorganic solid-state electrolytes for all-solid-state lithium batteries. Adv. Mater. 2018, 30 (17), 1705702. 4. Fan, L.; Wei, S.; Li, S.; Li, Q.; Lu, Y., Recent progress of the solid-state electrolytes for high-energy metal-based batteries. Adv. Energy Mater. 2018, 8 (11), 1702657. 5. He, L.; Chen, C.; Kotobuki, M.; Zheng, F.; Zhou, H.; Lu, L., A new approach for synthesizing bulk-type all-solid-state lithium-ion batteries. J. Mater. Chem. A 2019, 7 (16), 97489760. 6. Zheng, F.; Kotobuki, M.; Song, S.; Lai, M. O.; Lu, L., Review on solid electrolytes for allsolid-state lithium-ion batteries. J. Power Sources 2018, 389, 198-213. 7. Bachman, J. C.; Muy, S.; Grimaud, A.; Chang, H. H.; Pour, N.; Lux, S. F.; Paschos, O.; Maglia, F.; Lupart, S.; Lamp, P.; Giordano, L.; Shao-Horn, Y., Inorganic solid-state electrolytes for lithium batteries: mechanisms and properties governing ion conduction. Chem. Rev. 2016, 116 (1), 140-162. 8. Xu, B.; Duan, H.; Xia, W.; Guo, Y.; Kang, H.; Li, H.; Liu, H., Multistep sintering to synthesize fast lithium garnets. J. Power Sources 2016, 302, 291-297. 9. Huang, M.; Liu, T.; Deng, Y.; Geng, H.; Shen, Y.; Lin, Y.; Nan, C.-W., Effect of sintering temperature on structure and ionic conductivity of Li7−xLa3Zr2O12−0.5x (x=0.5~0.7) ceramics. Solid State Ionics 2011, 204-205, 41-45. 10. Allen, J. L.; Wolfenstine, J.; Rangasamy, E.; Sakamoto, J., Effect of substitution (Ta, Al, Ga) on the conductivity of Li7La3Zr2O12. J. Power Sources 2012, 206, 315-319. 11. Zhu, Y.; Zhang, Y.; Lu, L., Influence of crystallization temperature on ionic conductivity of lithium aluminum germanium phosphate glass-ceramic. J. Power Sources 2015, 290, 123-129. 12. Cui, Y.; Mahmoud, M. M.; Rohde, M.; Ziebert, C.; Seifert, H. J., Thermal and ionic conductivity studies of lithium aluminum germanium phosphate solid-state electrolyte. Solid State Ionics 2016, 289, 125-132. 13. Breuer, S.; Prutsch, D.; Ma, Q.; Epp, V.; Preishuber-Pflügl, F.; Tietz, F.; Wilkening, M., Separating bulk from grain boundary Li ion conductivity in the sol–gel prepared solid electrolyte Li1.5Al0.5Ti1.5(PO4)3. J. Mater. Chem. A 2015, 3 (42), 21343-21350. 14. He, K.; Zu, C.; Wang, Y.; Han, B.; Yin, X.; Zhao, H.; Liu, Y.; Chen, J., Stability of lithium ion conductor NASICON structure glass ceramic in acid and alkaline aqueous solution. Solid State Ionics 2014, 254, 78-81. 15. Li, F.; Kitaura, H.; Zhou, H., The pursuit of rechargeable solid-state Li–air batteries. Energy & Environmental Science 2013, 6 (8), 2302. 16. Zhu, Y.; He, X.; Mo, Y., Origin of outstanding stability in the lithium solid electrolyte materials: insights from thermodynamic analyses based on first-principles calculations. ACS Appl. Mater. Interfaces 2015, 7 (42), 23685-93. 17. Liu, Y.; Li, C.; Li, B.; Song, H.; Cheng, Z.; Chen, M.; He, P.; Zhou, H., Germanium thin film protected lithium aluminum germanium phosphate for solid-state li batteries. Adv. Energy Mater. 2018, 1702374.


Page 16 of 26

Page 17 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


18. Fan X.; Xiao J.; Han F.; Yue J.; Chen J.; Chen L.; Deng T.; Jiang J.; C., W., Fluorinated solid electrolyte interphase enables highly reversible solid-state Li metal battery. Sci. Adv. 2018, 4, eaau9245. 19. Fan, X.; Chen, L.; Borodin, O.; Ji, X.; Chen, J.; Hou, S.; Deng, T.; Zheng, J.; Yang, C.; Liou, S.-C.; Amine, K.; Xu, K.; Wang, C., Non-flammable electrolyte enables Li-metal batteries with aggressive cathode chemistries. Nat. Nanotechnol. 2018, 13 (8), 715-722. 20. Han, F.; Westover, A. S.; Yue, J.; Fan, X.; Wang, F.; Chi, M.; Leonard, D. N.; Dudney, N. J.; Wang, H.; Wang, C., High electronic conductivity as the origin of lithium dendrite formation within solid electrolytes. Nature Energy 2019, 4 (3), 187-196. 21. Fergus, J. W., Ceramic and polymeric solid electrolytes for lithium-ion batteries. J. Power Sources 2010, 195 (15), 4554-4569. 22. Zhou, W.; Wang, S.; Li, Y.; Xin, S.; Manthiram, A.; Goodenough, J. B., Plating a dendritefree lithium anode with a polymer/ceramic/polymer sandwich electrolyte. J. Am. Chem. Soc. 2016, 138 (30), 9385-8. 23. Maria, C. B.; Marina, M.; Alberto, Z., Investigation on lithium/polymer electrolyte interface for high performance lithium rechargeable batteries. J. Power Sources 1997, 68 (1997), 52-58. 24. Appetecchi, G. B.; Croce, F.; Mastragostino, M.; Scrosati, B.; Soavi, F.; Zanellit, A., Composite polymer electrolytes with improved lithium metal electrode interfacial properties. J. Electrochem. Soc. 1998, 145 (12), 4133-4134. 25. Hou, G.; Ma, X.; Sun, Q.; Ai, Q.; Xu, X.; Chen, L.; Li, D.; Chen, J.; Zhong, H.; Li, Y.; Xu, Z.; Si, P.; Feng, J.; Zhang, L.; Ding, F.; Ci, L., Lithium dendrite suppression and enhanced interfacial compatibility enabled by an ex situ sei on li anode for LAGP-based all-solid-state batteries. ACS Appl. Mater. Interfaces 2018, 10 (22), 18610-18618. 26. Song, S.; Kotobuki, M.; Zheng, F.; Xu, C.; Savilov, S. V.; Hu, N.; Lu, L.; Wang, Y.; Li, W. D. Z., A hybrid polymer/oxide/ionic-liquid solid electrolyte for Na-metal batteries. J. Mater. Chem. A 2017, 5 (14), 6424-6431. 27. Song, S.; Wu, Y.; Tang, W.; Deng, F.; Yao, J.; Liu, Z.; Hu, R.; Alamusi; Wen, Z.; Lu, L.; Hu, N., Composite solid polymer electrolyte with garnet nanosheets in poly(ethylene oxide). ACS Sustain. Chem. Eng. 2019, 7 (7), 7163-7170. 28. Li, D.; Chen, L.; Wang, T.; Fan, L. Z., 3d fiber-network-reinforced bicontinuous composite solid electrolyte for dendrite-free lithium metal batteries. ACS Appl Mater Interfaces 2018, 10 (8), 7069-7078. 29. Wu, B.; Wang, S.; Lochala, J.; Desrochers, D.; Liu, B.; Zhang, W.; Yang, J.; Xiao, J., The role of the solid electrolyte interphase layer in preventing Li dendrite growth in solid-state batteries. Energy & Environmental Science 2018, 11 (7), 1803-1810. 30. Ren Y.; Shen Y.; Lin Y.; Nan C., Direct observation of lithium dendrites inside garnet-type lithium-ion solid electrolyte. Electrochem. Commun. 2015, 57, 27-30. 31. Sang, L.; Bassett, K. L.; Castro, F. C.; Young, M. J.; Chen, L.; Haasch, R. T.; Elam, J. W.; Dravid, V. P.; Nuzzo, R. G.; Gewirth, A. A., Understanding the effect of interlayers at the thiophosphate solid electrolyte/lithium interface for all-solid-state li batteries. Chem. Mater. 2018, 30 (24), 8747-8756. 32. Yu, S.; Siegel, D. J., Grain boundary softening: A potential mechanism for lithium metal penetration through stiff solid electrolytes. ACS Appl. Mater. Interfaces 2018, 10 (44), 3815138158. 33. Conrad, G.; Mathias, F.; Rene, H., Process related effects upon formation of composite electrolyte interfaces: Nitridation and reduction of NASICON-type electrolytes by deposition of LiPON. J. Power Sources 2017, 362 (2017), 299-307.



34. Liu, Y.; Sun, Q.; Zhao, Y.; Wang, B.; Kaghazchi, P.; Adair, K. R.; Li, R.; Zhang, C.; Liu, J.; Kuo, L. Y.; Hu, Y.; Sham, T. K.; Zhang, L.; Yang, R.; Lu, S.; Song, X.; Sun, X., Stabilizing the interface of NASICON solid electrolyte against li metal with atomic layer deposition. ACS Appl. Mater. Interfaces 2018, 10 (37), 31240-31248. 35. Han, X.; Gong, Y.; Fu, K. K.; He, X.; Hitz, G. T.; Dai, J.; Pearse, A.; Liu, B.; Wang, H.; Rubloff, G.; Mo, Y.; Thangadurai, V.; Wachsman, E. D.; Hu, L., Negating interfacial impedance in garnetbased solid-state Li metal batteries. Nat. Mater. 2016, 10 (1038), 1-9. 36. Wang, C.; Xie, H.; Zhang, L.; Gong, Y.; Pastel, G.; Dai, J.; Liu, B.; Wachsman, E. D.; Hu, L., Universal soldering of lithium and sodium alloys on various substrates for batteries. Adv. Energy Mater. 2018, 8 (6), 1701963. 37. Luo, W.; Gong, Y.; Zhu, Y.; Li, Y.; Yao, Y.; Zhang, Y.; Fu, K. K.; Pastel, G.; Lin, C. F.; Mo, Y.; Wachsman, E. D.; Hu, L., Reducing interfacial resistance between garnet-structured solid-state electrolyte and li-metal anode by a germanium layer. Adv. Mater. 2017, 29 (22). 38. Zhao, N.; Fang, R.; He, M.-H.; Chen, C.; Li, Y.-Q.; Bi, Z.-J.; Guo, X.-X., Cycle stability of lithium/garnet/lithium cells with different intermediate layers. Rare Metals 2018, 37 (6), 473479. 39. Alexander, G. V.; Patra, S.; Sobhan Raj, S. V.; Sugumar, M. K.; Ud Din, M. M.; Murugan, R., Electrodes-electrolyte interfacial engineering for realizing room temperature lithium metal battery based on garnet structured solid fast Li+ conductors. J. Power Sources 2018, 396, 764773. 40. Park, C. M.; Kim, J. H.; Kim, H.; Sohn, H. J., Li-alloy based anode materials for Li secondary batteries. Chem. Soc. Rev. 2010, 39 (8), 3115-3141. 41. Wang, J.; Wang, H.; Xie, J.; Yang, A.; Pei, A.; Wu, C.-L.; Shi, F.; Liu, Y.; Lin, D.; Gong, Y.; Cui, Y., Fundamental study on the wetting property of liquid lithium. Energy Storage Mater. 2018, 14, 345-350. 42. Li, T.; Zeng, K., Probing of local multifield coupling phenomena of advanced materials by scanning probe microscopy techniques. Adv. Mater. 2018, 30 (47), 1803064. 43. Kim, S.; Won Hwang, S.; Choi, S.-H.; Elliman, R. G.; Kim, Y.-M.; Kim, Y.-J., Formation characteristics and photoluminescence of Ge nanocrystals in HfO2. J. Appl. Phys. 2009, 105 (10), 106112.


Page 18 of 26

Page 19 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60




Figure 1 Images of a LAGP pellet reacted with molten Li metal: (a) Surface images of LAGP pellet changed with time. (b) FE-SEM cross-section pictures, DART-ESM (c) topography and (d) deformation amplitude, AC-AFM (e) resonance frequency vs counts number and (f) phases of primary and reacted LAGP pellet. (g) LAGP particle surface peel off after heavy reaction.


Page 20 of 26

Page 21 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2 XPS images of primary LAGP (a-f) and reacted LAGP (g-l).



Figure 3 (a) Schematic illustrations of LAGP/Li interface with and without interface layer, (b) surface of LiSn0.6 after molten coating on interface modified LAGP pellet, and (c) TGA-DSC curve of P(AA-coMA)Li from room temperature to 800 ℃.


Page 22 of 26

Page 23 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4 FE-SEM images of (a) Li/LAGP/Li with P(AA-co-MA)Li as interface layer, (b), (c) and (d) P(AA-co-MA)Li added with mole ratio of 0.5, 1 and 3 LiCl as interface layer, respectively. The thickness of interface layer was ~1.5 μm with clear interface boundary.



Figure 5 EIS of (a) Li/LAGP/Li without interface, (b) with P(AA-co-MA)Li as interface layer and (c) LiCl added in the P(AA-co-MA)Li interface layer. (d) Arrhenius plot of P(AA-co-MA)Li/LAGP/P(AA-coMA)Li from -25 to 75 ℃.


Page 24 of 26

Page 25 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6 Galvanic cycle of (a) Li/LAGP/Li with P(AA-co-MA)Li as interface layer, (b) LiCl added P(AAco-MA)Li interface layer and (c) overpotentical vs current density of interface modified symmetrical cell Li/Interface/LAGP/Interface/Li.



Table of contents graphic


Page 26 of 26

Failure Mechanism and Interface Engineering for NASICON

Recommend Documents