In Situ Formed Shields Enabling Li2CO3-Free ... - ACS Publications

ACS2GO © 2019. ← → → ←. loading. To add this web app to the home screen open the browser option menu and tap on Add to homescreen...
0 downloads 0 Views 4MB Size
Download PDF
Research Article Cite This: ACS Appl. Mater. Interfaces 2019, 11, 898−905

www.acsami.org

In Situ Formed Shields Enabling Li2CO3‑Free Solid Electrolytes: A New Route to Uncover the Intrinsic Lithiophilicity of Garnet Electrolytes for Dendrite-Free Li-Metal Batteries Jian-Fang Wu,† Bo-Wei Pu,‡ Da Wang,‡ Si-Qi Shi,*,‡ Ning Zhao,§ Xiangxin Guo,§ and Xin Guo*,†

Downloaded via UNIV OF LOUISIANA AT LAFAYETTE on January 10, 2019 at 14:03:56 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Laboratory of Solid State Ionics, School of Materials Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, P. R. China ‡ School of Materials Science and Engineering, Shanghai University, Shanghai 200444, P. R. China § College of Physics, Qingdao University, Qingdao 266071, P. R. China S Supporting Information *

ABSTRACT: Introduction of inorganic solid electrolytes is believed to be an ultimate strategy to dismiss dendritic Li in high-energy Li-metal batteries (LMBs), and garnet-type Li7La3Zr2O12 (LLZO) electrolytes are impressive candidates. However, the current density for stable Li plating/stripping in LLZO is still quite limited. Here, we create in situ formed Li-deficient shields by the high-temperature calcination at 900 °C. By this novel process, the formation of Li2CO3 on LLZO is restrained, and then we successfully obtain Li2CO3-free LLZO after removing the Li-deficient compounds. Without any surface modification, Li2CO3-free LLZO shows an intrinsic “lithiophilicity” characteristic. The contact angles of metallic Li on LLZO garnets are assessed by the first-principle calculation to confirm the lithiophilicity characteristic of LLZO electrolytes. The wetting of metallic Li on the Li2CO3-free LLZO surface leads to a continuous and tight Li/LLZO interface, resulting in an ultralow interfacial resistance of 49 Ω cm2 and a homogeneous current distribution in the charge/discharge processes of LMBs. Consequently, the current density for the stable Li plating/stripping in LLZO increases to 900 μA cm−2 at 60 °C, one of the highest current density for LMBs based on garnet-type LLZO electrolytes. Our findings not only offer insight into the lithiophilicity characteristics of LLZO electrolytes to suppress dendritic Li at high current densities but also expand the avenue toward high-performance, safe, and long-life energystorage systems. KEYWORDS: solid electrolytes, dendritic Li, interfaces, Li7La3Zr2O12, Li-metal battery in LMBs,1,3,4 including employing polymer electrolytes,5,6 ionic liquids,7 concentrated electrolytes8 or additives,9,10 artificial SEI layers,11 mechanical protective layers,12,13 nanostructure design,14 homogeneous Li-flux through selective deposition,15 and others.16 In principle, inorganic solid electrolytes with high shear moduli (>1.8 shear moduli of Li) are expected to suppress the formation of dendritic Li,17 but many of them are hindered in LMBs due to the low lithium-ion conductivity or the deteriorated stabilities. Fortunately, cubic garnet Li7La3Zr2O12 (LLZO) and its variants show the highest lithium-ion conductivity of ∼10−3 S cm−1 in oxide-based solid electrolytes and extremely good stability against metallic Li,18−20 thus having potential applications in LMBs. Normally, solid-solid contact dominates the LLZO and metallic Li interface, and Li2CO3 on the surface hinders the wetting of metallic Li on the

1. INTRODUCTION Li-metal batteries (LMBs) are considered as the “Holy Grail” of energy-storage systems, and vast improvement in energy density could be expected due to the extremely high theoretical specific capacity (3860 mA h g−1) and the lowest negative electrochemical potential (−3.04 V versus the standard hydrogen electrode) of metallic Li.1 However, LMBs suffer from a poor cycle life and unsatisfactory Coulombic efficiency and encounter severe safety issues due to uncontrollable formation and growth of Li dendrites.2 Metallic Li is reactive with most organic electrolytes, causing the formation of a solid electrolyte interphase (SEI) layer on the Li surface; the brittle SEI layer cannot tolerate the volume expansion of metallic Li during plating/stripping, thus it continuously breaks and recovers; the current distribution on metallic Li becomes uneven, and Li dendrites nucleate and grow at the sites with locally high current density.3 In batteries with high current densities proportioning to its energy densities, the formation of Li dendrites can be speed up. Extensive studies have been conducted to prevent Li dendrites © 2018 American Chemical Society

Received: October 21, 2018 Accepted: December 5, 2018 Published: December 5, 2018 898

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces

of 98.8%, was obtained by hot-pressing in Ar at 1100 °C for an hour in a graphite die under 10 MPa pressure. The samples were afterward calcined at 900 °C in air for 24 h. Finally, the samples were polished in a glovebox filled with pure Ar using SiC abrasive papers. Other garnet-type electrolytes, such as Li6.4Al0.2La3Zr2O12 (Al-LLZO) and Li6.4La3Zr1.4Ta0.6O12 (Ta-LLZO), were prepared similarly. 2.2. Wetting of Molten Li on the LLZO Electrolyte. Metallic Li was melted in a stainless-steel container on a hot plate at 350 °C for 10 min in a glovebox filled with pure Ar. A drop of molten Li was poured on one surface of the heated LLZO pellet and then allowed few minutes to spread out on the surface. 2.3. Battery Assembly. Through the method in the wetting experiment, symmetric Li/Ga-LLZO/Li cells were assembled. All-solid-state Li/LiFePO4 batteries were assembled as follows: First, poly(vinylidene fluoride) (PVDF) and Li(CF3SO2)2N (LiTFSI) were dissolved in N-methyl-2 pyrrolidon (NMP) by magnetic stirring; carbon-coated-LiFePO4 and Ketjenblack were mixed by grinding in an agate mortar for 1 h and the mixture transferred to the above NMP solution and stirred for another 24 h. The mass ratio of LiFePO4, Ketjenblack, and PVDF is 75:15:10, and the mass of LiTFSI is 80% of that of LiFePO4. The obtained slurry was coated on the side of the electrolyte pellet without metallic Li. After being pressed with a stainless-steel plate by a pressure of 10 N, the solid electrolyte with electrode coating was dried in vacuum at 80 °C for 12 h to remove NMP and trace moisture. The loading masses of the electrodes were about 2 mg cm−2. Solid Li/LiFePO4 batteries were assembled as follows: Cathode pastes with 70% LiFePO4, 20% Ketjenblack, and 10% PVDF were prepared on stainless-steel plates. Then, the paste was pressed on the side of the electrolyte pellet without metallic Li to get a battery and 50 μL Li+-containing ionic liquid (1 mmol LiTFSI in 1 mL [EMIM][TFSI]) was added to enhance the Li-ion conduction between the cathode and the solid electrolyte. The LLZO pellets for battery tests had a diameter of ∼15 mm and a thickness of ∼0.8 mm. All the procedures were conducted in an Ar-filled glovebox with O2 < 0.1 ppm and H2O < 0.1 ppm. 2.4. Material Characterizations. Scanning electron microscopy (SEM, Sirion 200) images of the cross section of the LLZO ceramics and the Li/LLZO interface were recorded. Crystalline phases of the ceramics were investigated through X-ray diffraction (XRD, XRD7000S). Raman spectroscopy (LabRAM HR800) was conducted on the surface of the LLZO samples, which were transferred in a sealed glass bottle in the glovebox. X-ray photoelectron spectroscopy (XPS, AXIS-ULTRA DLD-600W) was conducted on the LLZO samples to check their surface chemical compositions. The 7Li NMR spectra were gathered on a Bruker AVANCE III 400 NMR spectrometer, and the chemical shifts of 7Li were calibrated by using a 1 M LiCl solution. Alternating current (AC) impedance measurements were conducted in the temperature range of 25−60 °C using a Solartron 1260 impedance and a gain-phase analyzer in the frequency range of (1−5) × 106 Hz at an amplitude of 50 mV. Li-ion blocking Ag electrodes were used. The samples were heated at the desired temperatures for an hour before each measurement. The dimensions of the sample for the AC impedance test were 10 × 10 × 1 mm3. Galvanostatic charge and discharge performances of Li/Ga-LLZO/ Li cells and LiFePO4/Ga-LLZO/Li batteries were investigated using a LANHE CT2001A charge/discharge system (Wuhan LAND Electronics Co.). The cutoff voltages of the LiFePO4/Ga-LLZO/Li battery were set at 4.0 and 2.8 V. Before the charge/discharge, the allsolid-state battery was heated at 60 °C for 24 h to reach a thermal equilibrium. 2.5. Theoretical Calculations. First-principle calculations of the surface energy and the interfacial energy were carried out based on the projector-augmented wave method of the density functional theory and conducted with the Vienna Ab initio Simulation Package (VASP).35−37 The exchange−correlation part of the density function was treated using the generalized gradient Perdew−Burke−Ernzerhof approximation.38 The plane wave cutoff energy and the Monkhorst− Pack k-point mesh was set to be 600 eV and 3 × 3 × 1, respectively, for the Li(001)/garnet(001) interfaces. The self-consistency con-

LLZO surface, resulting in insufficient contact between LLZO and metallic Li, and an inhomogeneous current distribution during Li plating/stripping happens, therefore, Li dendrites nucleate at the interface. Solid electrolytes are polycrystalline materials, including grain boundaries, pores, and cracks, along which dendritic Li can grow;21−25 as a result, the threshold current density for stable Li plating/stripping is limited. Surface modification, enhancing interfacial contact between LLZO and metallic Li, was employed to suppress dendritic Li formation and growth.20,22,26−33 With a ∼5 nm thick Al2O3 deposited on the surface of LLZO by atom layer deposition, the current density increased to 200 μA cm−2 at room temperature. 20 Notably, all the chosen materials for modifications can easily alloy with metallic Li, which makes metallic Li wet LLZO, i.e., the contact angle θ < 90°, thus leading to a homogeneous current distribution and a high threshold current density. These works come from the fact that metallic Li cannot wet the Li2CO3-coated LLZO surface. If metallic Li can wet the Li2CO3-free LLZO surface, a conformal interface between them, a homogeneous current distribution, and a high threshold current density can be expected. However, it is very difficult to obtain Li2CO3-free LLZO solid electrolytes. Sakamoto et al.34 tried to remove Li2CO3 on the LLZO surface by polishing and calcination at 400−500 °C, but they observed contact angles >95°; therefore, metallic Li did not wet the LLZO electrolyte, which is due to the residual Li2CO3 and LiOH on the LLZO surface. In a very recent work of Goodenough,33 the introduction of carbon, which reacts with Li2CO3 on the LLZO surface in Ar at 700 °C, eliminated Li2CO3; as a result, metallic Li wetted LLZO and a current density of 400 μA cm−2 was achieved at 65 °C. However, it is noted that carbon can alloy with metallic Li easily. Therefore, the intrinsic surface property, i.e., wettability of Li2CO3-free LLZO to metallic Li, is still a pending question. In this work, we tuned the surface chemistry of solid electrolytes and demonstrated their intrinsic “lithiophilicity”. We created Li-deficient compounds on LLZO by the hightemperature calcination at 900 °C, which inhibited the formation of Li2CO3 on the LLZO surface; after removing the Li-deficient compounds, Li2CO3-free LLZO was obtained, providing us a perfect platform to answer the fundamental question: Whether LLZO garnets are intrinsically lithiophilic? Our results show that the contact angles between Li2CO3-free garnets and metallic Li are below 90 °C and vary with different dopants, which are further verified by the first-principle calculations. The method to realize the wetting of metallic Li on LLZO in this work is more feasible and cost-effective than the surface modification. The wetting of metallic Li on the surface of the LLZO electrolyte produces continuous and tight contact between them and a low interfacial resistance of 49 Ω cm2, which facilitates the suppression of dendritic Li in the LLZO electrolyte, and the current density for stable Li plating/ stripping increases to 900 μA cm−2 at 60 °C. In addition to the LLZO materials, this work is also beneficial to other oxidebased ion conductors because it is easy for Li+ ions to leak out from the highly conductive electrolytes to form lithiophobic Li2CO3.

2. EXPERIMENTAL SECTION 2.1. Preparation of Solid Electrolytes. Li6.25Ga0.25La3Zr2O12 (Ga-LLZO) powder was prepared via a solid-state reaction by mixing Li2CO3, Ga2O3, La2O3, and ZrO2 in stoichiometric ratio but with 10 mol % Li2CO3 in excess. Ga-LLZO electrolyte, with a relative density 899

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces

Figure 1. (a) SEM image of Ga-LLZO ceramic. Schematics of interactions between (b) unprotected LLZO surface and humid air and (c) surface protected by the Li-deficient compound and humid air. (d) XRD pattern, (e) AC impedance spectra, and (f) Arrhenius plot of Ga-LLZO ceramic.

Figure 2. (a) XPS results of O 1s (up) and C 1s (bottom) and (b) Raman spectrum of the calcined and polished Ga-LLZO electrolyte. vergence criterion for the energy was set to be 10−4 eV, and the geometry relaxation was converged when all the forces were less than 0.05 eV Å−1. The Li/garnet interface models were constructed by Li(001) slab and garnet(001) slab, which are the low-energy surfaces. The surface energy of Li(001) containing six atomic layers was calculated to be 0.447 J m−2 with its structure built on the basis of ICSD-642110, and the garnet structures were constructed based on the neutron diffraction results reported in literature.39 Structural details after optimization are listed in Table S1. The Li(001) surface was modeled as a (3 × 3) slab, whose in-plane lattice translation vectors are 13.065 Å, approaching that of garnet slabs (∼13 Å). The coherent interface approximation was applied, in which the soft Li(001) (3 × 3) surface slab (108 Li) is strained to match the dimensions of garnet(001) (1 × 1) surface slabs (8Li7−xM0.125La3Zr2−yO12, M = Ga, Al, x = 0.375, y = 0; M = Ta, x = 0.125, y = 0.125). A vacuum region of 15 Å was included in the interface supercell. As proposed by Sharafi et al.,34 the back half atoms farthest away from the interface of each slab were fixed at their bulk-like positions, and thus only one interface was contained in each supercell. Interface configurations are shown in Figures 3 and S3.

The calculation of the contact angle between Li(001) and LLZO variant (001) is based on the Young−Dupré equation40

Wad = σLi(1 + cos θ)

(1)

where Wad is the interfacial work of adhesion, σLi is the surface energy of Li, and θ is the contact angle. Wad can be evaluated according to41 Wad =

Einterface − E Li ‐ slab − Egarnet ‐ slab S

(2)

where Einterface, ELi‑slab, and Egarnet‑slab refer to the total energies of the Li(001)/garnet(001) interface supercell, isolated Li(001), and garnet(001) surface slabs, respectively, and S refers to the interfacial area.

3. RESULTS AND DISCUSSION 3.1. Li2CO3-Free LLZO Solid Electrolytes. Highly densified and conductive Ga-LLZO samples with coarse grains were prepared by hot pressing. The microstructure, crystalline phase, and conductivity of the Ga-LLZO electrolyte were characterized, and the results are shown in Figures 1 and S1. As 900

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces shown in Figure 1a, the electrolyte is well sintered without any visible pores, and a high density of 98.8% is demonstrated by the Archimedes method. Because the grain size of the sample is tens of micrometers, the number of the grain boundaries in the direction perpendicular to the surface is low. Li2CO3 easily forms on the LLZO surface in humid air (Figure 1b);33,42 to avoid the formation of Li2CO3 on the LLZO surface, we calcined the LLZO electrolytes at a high temperature of 900 °C. After the high-temperature calcination at 900 °C in air for 24 h, Li-deficient lanthanum−zirconium compounds were formed in situ on the Ga-LLZO surface (Figure S1) due to the decomposition of LLZO; the Lideficient compounds protect the inner cubic garnet from Li2CO3 (Figures 1c and S1). After removing the Li-deficient compounds, cubic garnet without any impurity phase was obtained (Figure 1d). The conductivity of the electrolyte is 1.24 mS cm−1 at 25 °C and increases to 2.71 mS cm−1 when the temperature increases to 60 °C, as calculated from the AC impedance spectra (Figure 1e), and an activation energy of 0.2 eV is determined from the Arrhenius plot (Figure 1f). According to our previous works,43,44 the Ga-LLZO solid electrolyte is a pure ionic conductor with a lithium-ion transference number approximating to 1. To confirm that there is no Li2CO3 on the Ga-LLZO surface, XPS and Raman analyses were conducted. Figure 2a shows the XPS results of O 1s and C 1s on the Ga-LLZO surface. The characteristic peak for Li2CO3 at around 290 eV does not show up in the C 1s spectrum, and the peaks in the spectrum can be indexed to the absorbed carbon on the LLZO surface.34,45 In the Raman spectrum (Figure 2b), all the peaks can be indexed to cubic garnet LLZO. Therefore, there is no Li2CO3 on the polished Ga-LLZO, and Li2CO3-free LLZO is a perfect platform to demonstrate the intrinsic lithiophilicity characteristic of LLZO electrolytes. 3.2. Intrinsic Lithiophilicity Characteristic of LLZO Electrolytes. The wetting process of molten metallic Li on the Li2CO3-free Ga-LLZO surface is shown in Figure 3a. Initially, the contact angle (θ) of molten metallic Li on the Ga-LLZO surface is about 56°. After being heated at 350 °C for 2 min, the molten metallic Li spreads out and the contact angle gets smaller and approaches ∼0°. According to the SEM images of the cross-section of the Li/LLZO interface in Figure 3b, metallic Li is conformally coated on Ga-LLZO and a continuous and tight interfacial contact is observed, demonstrating the lithiophilicity of the Ga-LLZO electrolyte. The same phenomenon is also observed for other LLZO variants, such as Li6.4Al0.2La3Zr2O12 (Al-LLZO) and Li6.4La3Zr1.4Ta0.6O12 (Ta-LLZO), as shown in Figure S2, and their initial contact angles are about 59.5 and 65°, respectively. In previous works,20,22,27−29,32,33 most researchers claimed a lithiophobicity characteristic of the LLZO electrolytes and the lithiophilicity can only be realized with the help of surface modifications, but they attributed the lithiophobicity characteristic to Li2CO3 on the LLZO surface. In this work, we demonstrate that the lithiophilicity characteristic is intrinsic to LLZO. To further verify the lithiophilicity of the LLZO garnets, we conducted first-principle calculation to obtain the contact angles of metallic Li on the LLZO garnets, and the results are presented in Figures 3c,d, S3, and Table S2. The contact angles for Li/LLZO, Li 6.625 Ga 0.125 La 3 Zr 2 O 12 (Li/Ga-LLZO), Li6.625Al0.125La3Zr2O12 (Li/Al-LLZO), and

Figure 3. (a) Wetting of molten Li on Ga-LLZO surface. (b) SEM image of the interface between metallic Li and Ga-LLZO. Interfacial structure, calculated works of adhesion (Wad), and contact angles (θ) between metallic Li(001) and (c) LLZO(001) and (d) GaLLZO(001).

Li6.875La3Zr1.875Ta0.125O12 (Li/Ta-LLZO) interfaces are about 73, 57, 53, and 68°, respectively, which are smaller than 90°, manifesting that metallic Li wetting the Li2CO3-free LLZO electrolytes is a general phenomenon, in spite of different compositions. We can also regulate the surface chemistry of garnets by doping. Moreover, the calculated contact angles are quite consistent with the experimental observations in Figures 3a and S2. 3.3. Performances of Symmetric Li/LLZO/Li Cells. The lithiophilicity characteristic of the LLZO electrolyte makes a continuous and tight contact between metallic Li and the electrolyte, which facilitates the lithium ion transport across the interface, leading to a low interfacial resistance. The resistance of the symmetric Li/LLZO/Li cell was investigated by the AC impedance spectrometer at 25 °C, and the result is shown in Figure 4a. The Li/LLZO interfacial resistance is calculated by subtracting the electrolyte resistance from the total cell resistance, dividing by 2, and then normalizing to the sample surface area. The total resistance of the symmetric cell is 233 Ω cm2. The resistance of the electrolyte is calculated from the lithium-ion conductivity and the dimension of the sample and then the Li/LLZO interfacial resistance is 901

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces

Figure 4. (a) AC impedance spectrum of the Li/Ga-LLZO/Li symmetric cell and (b) comparison of the Li/LLZO interfacial resistances.20,22,27,30,32,33,46−49 (c) The voltage profiles and (d) AC impedance spectra of the Li/Ga-LLZO/Li symmetric cell during the galvanostatic cycling at room temperature. (e) The voltage profiles of Li/Ga-LLZO/Li symmetric cell with a current density of 900 μA cm−2 at 60 °C. (f) Static 7Li NMR spectra of Ga-LLZO solid electrolyte before and after cycling test.

cm−2 in both directions for 200 h. The corresponding voltages are stable under a constant current but increase with increasing current density, which indicates no formation of dendritic Li in the solid electrolyte. During the cycling process, the cell resistance (Figure 4d) decreases after 55 cycles and then remains stable, which is related to the surface activation process.27 The Li/Ga-LLZO interfacial resistance after the surface activation process decreases to 35 Ω cm2. Impressively, when the working temperature goes up to 60 °C, the current density for the stable Li plating/stripping in Ga-LLZO increases to 900 μA cm−2 (Figure 4e), which is one of the highest values for ceramic solid electrolytes. To verify no formation of dendritic Li in the LLZO sample after cycling, the sample was carefully polished to dislodge the Li electrodes and then sealed in a quartz tube for static 7Li NMR test. In Figure 4f, only a narrow signal at around 0 ppm corresponding to lithium ions in Ga-LLZO is recorded in each spectrum. The narrow character without any satellite transition agrees with the high mobility of lithium ions in the Ga-LLZO electrolyte. Notably, the typical resonance peak for metallic Li, in the range of 200−300 ppm, does not appear in the samples

determined to be 49 Ω cm2. Many researchers measured the interfacial resistances between metallic Li and LLZO electrolytes,20,22,27−30,32,33,46−49 but their reported values are far larger than 100 Ω cm2 (Figure 4b). The existence of Li2CO3 on the LLZO surface leads to a poor interfacial contact and a high interfacial resistance.46,50 Han et al. decreased the Li/LLZO interfacial resistance to a very low level of 1 Ω cm2, with a 5 nm thick Al2O3 layer deposited on the LLZO electrolyte.20 With amorphous Si,22 ZnO,27 Ge,29 and Al,28 one could decrease the Li/LLZO interfacial resistance to ∼100 Ω cm2 as well (Figure 4b). With the aid of carbon, the Li/LLZO interfacial resistance was reduced to 28 and 105 Ω cm2 by Goodenough33 and Yang,32 respectively. However, it has to be pointed out that we achieved a low interfacial resistance of 49 Ω cm2 without any surface modification. Wetting Ga-LLZO by metallic Li leads to a low interfacial resistance as well as a homogeneous current distribution under galvanostatic cycling; thus, the capability of the Ga-LLZO electrolyte to suppress the formation of dendritic Li is enhanced. Figure 4c exhibits cycling at room temperature and under different current densities of 100, 200, and 300 μA 902

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces

Figure 5. (a, c) Plots of cell potential versus specific capacity, (b, d) charge capacity, discharge capacity, and Coulombic efficiency of the (a, b) allsolid-state, and (c, d) solid batteries during galvanostatic charge/discharge cycling. The insets show LEDs lit by the batteries.

LMBs, two kinds of batteries with metallic Li anode and LiFePO4 composite cathode were assembled and tested. Allsolid-state and solid LMBs underwent long charge/discharge cycling testing between 2.8 and 4.2 V, with results given in Figure 5. The all-solid-state battery delivers a specific capacity of 148 mA h g−1 (Figure 5a,b) in the first cycle at a charge/ discharge rate of C/10 at 60 °C, which reaches 87% of the theoretical capacity of LiFePO4, and the battery maintains a specific capacity of 116 mA h g−1 after 40 cycles. The capacity decrease can be attributed to the unstable contact between the cathode and the electrolyte after several cycles.51 Notably, in the first charge/discharge process, the Coulombic efficiency is 94%, which increases to above 99.5% in the following cycles, which is a very high value for LMBs.1 With the aid of Licontaining ionic liquid, the recession of the capacity (Figure 5c,d) is much slower at room temperature. The charge/ discharge behavior and the high Coulombic efficiency demonstrate that no dendritic Li penetrate the Ga-LLZO electrolyte. The high current density for stable Li plating/stripping in LLZO electrolytes is one of the critical factors determining the performances of high-energy LMBs. Except for the Li/LiFePO4 system, the LLZO electrolytes show potential applications in burgeoning Li−air and Li−sulfur batteries due to the ability to tolerate high current density.1 For example, when the weight of active material in the cathode of a Li−sulfur battery reaches 5 mg cm−2, a current density of ∼840 μA cm−2 is needed at a charge/discharge rate of 0.1C, which is still within the tolerance of the Ga-LLZO electrolyte.

before and after cycling, which is a direct evidence of the absence of dendritic Li in the Ga-LLZO electrolyte. Therefore, the Ga-LLZO electrolyte can suppress the formation of dendritic Li at a current density of up to 900 μA cm−2. According to Chiang et al.,25 there are flaws in the insufficiently contacting Li/LLZO interface, and once an uneven current is formed, metallic Li nucleates fill in the flaws and drive crack propagation at these sites; then, the cracks expand and penetrate the electrolyte in the following charge/ discharge processes. Continuous and tight interface between LLZO and metallic Li, originating from the lithiophilicity of LLZO, can dismiss these flaws effectively and produce a homogeneous current distribution, and the metallic nuclei are suppressed at the interface. Consequently, the capability of LLZO electrolytes to suppress dendritic Li is greatly enhanced, and the current density of stable Li plating/strapping increases. Lithium ions in LLZO or other highly conductive oxidebased solid electrolytes can easily react with H2O and CO2 in moist atmosphere to form Li2CO3,33,42 which seriously deteriorates the wettability of metallic Li on the surface.20 For example, in the presence of Li2CO3, the contact angle gets to be about 127°, as shown in Figure S2c. As a result, the contact at the Li/LLZO interface is insufficient, the interfacial resistance becomes high, and the current density for the stable Li plating/stripping is quite limited. Surface modifications by means of materials that can alloy with metallic Li were used to improve the performances of LLZO electrolytes.20,22,26−30 However, Li-deficient shields can be formed in situ at high temperatures due to the decomposition of solid electrolytes, which restrains the formation of Li2CO3 on the surface, then metallic Li wets the electrolyte well and the capability of the electrolyte to suppress lithium dendrites is enhanced. Such a strategy is feasible and cost-effective and makes the solid electrolytes very suitable for practical application in LMBs. 3.4. Performances of Solid LiFePO4/LLZO/Li Batteries. To demonstrate the application of the Ga-LLZO electrolyte in

4. CONCLUSIONS We obtain Li2CO3-free LLZO surfaces using a novel method and demonstrated the intrinsic lithiophilicity characteristic of LLZO electrolytes. After the high-temperature calcination at 900 °C, Li-deficient compounds are in situ formed on the LLZO surface, which shields LLZO from the formation of 903

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces

(4) Cheng, X. B.; Zhang, R.; Zhao, C. Z.; Zhang, Q. Toward Safe Lithium Metal Anode in Rechargeable Batteries: A review. Chem. Rev. 2017, 117, 10403−10473. (5) Khurana, R.; Schaefer, J. L.; Archer, L. A.; Coates, G. W. Suppression of Lithium Dendrite Growth Using Cross-Linked Polyethylene/Poly(Ethylene Oxide) Electrolytes: A New Approach for Practical Lithium-Metal Polymer Batteries. J. Am. Chem. Soc. 2014, 136, 7395−7402. (6) Li, N. W.; Shi, Y.; Yin, Y. X.; Zeng, X. X.; Li, J. Y.; Li, C. J.; Wan, L. J.; Wen, R.; Guo, Y. G. Smart Solid Electrolyte Interphase Layer for Long Life Lithium Metal Anodes. Angew. Chem., Int. Ed. 2018, 57, 1− 6. (7) Li, N. W.; Yin, Y. X.; Li, J. Y.; Zhang, C. H.; Guo, Y. G. Passivation of Lithium Metal Anode via Hybrid Ionic Liquid Electrolyte toward Stable Li Plating/Stripping. Adv. Sci. 2017, 4, No. 1600400. (8) Qian, J.; Henderson, W. A.; Xu, W.; Bhattacharya, P.; Engelhard, M.; Borodin, O.; Zhang, J. G. High Rate and Stable Cycling of Lithium Metal Anode. Nat. Commun. 2015, 6, No. 6362. (9) Ding, F.; Xu, W.; Graff, G. L.; Zhang, J.; Sushko, M. L.; Chen, X.; Shao, Y.; Engelhard, M. H.; Nie, Z.; Xiao, J.; Liu, X.; Sushko, P. V.; Liu, J.; Zhang, J. G. Dendrite-Free Lithium Deposition via SelfHealing Electrostatic Shield Mechanism. J. Am. Chem. Soc. 2013, 135, 4450−4456. (10) Zheng, J.; Engelhard, M. H.; Mei, D.; Jiao, S.; Polzin, B. J.; Zhang, J.-G.; Xu, W. Electrolyte Additive Enabled Fast Charging and Stable Cycling Lithium Metal Batteries. Nat. Energy 2017, 2, No. 17012. (11) Yan, K.; Lee, H. W.; Gao, T.; Zheng, G.; Yao, H.; Wang, H.; Lu, Z.; Zhou, Y.; Liang, Z.; Liu, Z.; Chu, S.; Cui, Y. Ultrathin TwoDimensional Atomic Crystals as Stable Interfacial Layer for Improvement of Lithium Metal Anode. Nano Lett. 2014, 14, 6016− 6022. (12) Park, K.; Goodenough, J. B. Dendrite-Suppressed Lithium Plating from a Liquid Electrolyte via Wetting of Li3N. Adv. Energy Mater. 2017, 7, No. 1700732. (13) Liu, K.; Pei, A.; Lee, H. R.; Kong, B.; Liu, N.; Lin, D.; Liu, Y.; Liu, C.; Hsu, P. C.; Bao, Z.; Cui, Y. Lithium Metal Anodes with an Adaptive “Solid-Liquid” Interfacial Protective Layer. J. Am. Chem. Soc. 2017, 139, 4815−4820. (14) Lin, D.; Liu, Y.; Liang, Z.; Lee, H. W.; Sun, J.; Wang, H.; Yan, K.; Xie, J.; Cui, Y. Layered Reduced Graphene Oxide with Nanoscale Interlayer Gaps as a Stable Host for Lithium Metal Anodes. Nat. Nanotechnol. 2016, 11, 626−632. (15) Yan, K.; Lu, Z.; Lee, H.-W.; Xiong, F.; Hsu, P.-C.; Li, Y.; Zhao, J.; Chu, S.; Cui, Y. Selective Deposition and Stable Encapsulation of Lithium through Heterogeneous Seeded Growth. Nat. Energy 2016, 1, No. 16010. (16) Zhang, R.; Chen, X. R.; Chen, X.; Cheng, X. B.; Zhang, X. Q.; Yan, C.; Zhang, Q. Lithiophilic Sites in Doped Graphene Guide Uniform Lithium Nucleation for Dendrite-Free Lithium Metal Anodes. Angew. Chem., Int. Ed. 2017, 56, 7764−7768. (17) Tikekar, M. D.; Choudhury, S.; Tu, Z.; Archer, L. A. Design Principles for Electrolytes and Interfaces for Stable Lithium-Metal Batteries. Nat. Energy 2016, 1, No. 16114. (18) Murugan, R.; Thangadurai, V.; Weppner, W. Fast Lithium Ion Conduction in Garnet-Type Li7La3Zr2O12. Angew. Chem., Int. Ed. 2007, 46, 7778−7781. (19) Du, F.; Zhao, N.; Li, Y.; Chen, C.; Liu, Z.; Guo, X. All Solid State Lithium Batteries Based on Lamellar Garnet-Type Ceramic Electrolytes. J. Power Sources 2015, 300, 24−28. (20) Han, X.; Gong, Y.; Fu, 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. 2017, 16, 572−579. (21) Ren, Y.; Shen, Y.; Lin, Y.; Nan, C.-W. Direct Observation of Lithium Dendrites Inside Garnet-Type Lithium-Ion Solid Electrolyte. Electrochem. Commun. 2015, 57, 27−30.

Li2CO3, and Li2CO3-free LLZO is obtained after removing Lideficient compounds. The Li2CO3-free LLZO clearly shows an intrinsic lithiophilicity characteristic, without any surface modification. This is also confirmed by the first-principle calculations, in which all the calculated contact angles between metallic Li and Li2CO3-free garnets are below 90°, though varying with different dopants. The lithiophilicity characteristic of LLZO electrolytes leads to a continuous and tight interfacial contact between metallic Li and the LLZO electrolytes, which facilitates the lithium-ion transport across the interface. As a result, a relatively low interfacial resistance of 49 Ω cm−2, as well as a homogeneous current distribution in the charge/ discharge processes of LMBs is achieved, increasing the current density for the stable Li plating/stripping in LLZO to a high level of 900 μA cm−2 at 60 °C. Owing to the ability to suppress dendritic Li, garnet electrolytes potentially enable high-energy LMBs. In addition to the particular LLZO materials, this work is also beneficial to other oxide-based ion conductors for highperformance, safe, and long-life energy-storage systems.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.8b18356.

■

XRD patterns of the calculated LLZO; XPS plots of samples: freshly calcined, exposed in air for 7 days, and polished and exposed in air for 7 days; wetting of molten Li on different garnet electrolytes: Al-LLZO, Ta-LLZO, and polished-Ga-LLZO exposed in air for 7 days; interfacial structure, calculated works of adhesion (Wad), and contact angles (θ) between metallic Li(001) and AlLLZO(001) and Ta-LLZO(001); lattice parameters of cubic lithium, LLZO, and its variants obtained from firstprinciple calculation; dimension parameters, interfacial works of adhesion, and contact angles of Li(001)/ garnet(001) interface (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (S.-Q.S.). *E-mail: [email protected]. Tel/Fax: +86-27-87559804 (X.G.). ORCID

Xin Guo: 0000-0003-1546-8119 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China (Grant Nos 51672096, 51622207, U1630134, and 51802187).

■

REFERENCES

(1) Yang, C.; Fu, K.; Zhang, Y.; Hitz, E.; Hu, L. Protected LithiumMetal Anodes in Batteries: From Liquid to Solid. Adv. Mater. 2017, 29, No. 1701169. (2) Tarascon, J. M.; Armand, M. Issues and Challenges Facing Rechargeable Lithium Batteries. Nature 2001, 414, 359−367. (3) Lin, D.; Liu, Y.; Cui, Y. Reviving the Lithium Metal Anode for High-Energy Batteries. Nat. Nanotechnol. 2017, 12, 194−206. 904

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905

Research Article

ACS Applied Materials & Interfaces (22) Luo, W.; Gong, Y.; Zhu, Y.; Fu, K. K.; Dai, J.; Lacey, S. D.; Wang, C.; Liu, B.; Han, X.; Mo, Y.; Wachsman, E. D.; Hu, L. Transition from Super-Lithiophobicity to Super-Lithiophilicity of Garnet Solid-State Electrolyte. J. Am. Chem. Soc. 2016, 138, 12258− 12262. (23) Cheng, E. J.; Sharafi, A.; Sakamoto, J. Intergranular Li Metal Propagation through Polycrystalline Li6.25Al0.25La3Zr2O12 Ceramic Electrolyte. Electrochim. Acta 2017, 223, 85−91. (24) Aguesse, F.; Manalastas, W.; Buannic, L.; Lopez Del Amo, J. M.; Singh, G.; Llordes, A.; Kilner, J. A. Investigating the Dendritic Growth during Full Cell Cycling of Garnet Electrolyte in Direct Contact with Li Metal. ACS Appl. Mater. Interfaces 2017, 9, 3808− 3816. (25) Porz, L.; Swamy, T.; Sheldon, B. W.; Rettenwander, D.; Frömling, T.; Thaman, H. L.; Berendts, S.; Uecker, R.; Carter, W. C.; Chiang, Y.-M. Mechanism of Lithium Metal Penetration through Inorganic Solid Electrolytes. Adv. Energy Mater. 2017, No. 1701003. (26) Tsai, C. L.; Roddatis, V.; Vinod Chandran, C.; Ma, Q.; Uhlenbruck, S.; Bram, M.; Heitjans, P.; Guillon, O. Li7La3Zr2O12 Interface Modification for Li-Dendrite Prevention. ACS Appl. Mater. Interfaces 2016, 8, 10617−10626. (27) Wang, C.; Gong, Y.; Liu, B.; Fu, K.; Yao, Y.; Hitz, E.; Li, Y.; Dai, J.; Xu, S.; Luo, W.; Wachsman, E. D.; Hu, L. Conformal, Nanoscale ZnO Surface Modification of Garnet-Based Solid-State Electrolyte for Lithium Metal Anodes. Nano Lett. 2017, 17, 565−571. (28) Fu, K. K.; Gong, Y.; Liu, B.; Zhu, Y.; Xu, S.; Yao, Y.; Luo, W.; Wang, C.; Lacey, S. D.; Dai, J.; Chen, Y.; Mo, Y.; Wachsman, E.; Hu, L. Toward Garnet Electrolyte−Based Li Metal Batteries: An Ultrathin, Highly Effective, Artificial Solid-State Electrolyte/Metallic Li Interface. Sci. Adv. 2017, 3, No. e1601659. (29) 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, No. 1606042. (30) Fu, K. K.; Gong, Y.; Fu, Z.; Xie, H.; Yao, Y.; Liu, B.; Carter, M.; Wachsman, E.; Hu, L. Transient Behavior of the Metal Interface in Lithium Metal-Garnet Batteries. Angew. Chem., Int. Ed. 2017, 56, 14942−14947. (31) Li, Y.; Xu, B.; Xu, H.; Duan, H.; Lu, X.; Xin, S.; Zhou, W.; Xue, L.; Fu, G.; Manthiram, A.; Goodenough, J. B. Hybrid Polymer/Garnet Electrolyte with a Small Interfacial Resistance for Lithium-Ion Batteries. Angew. Chem., Int. Ed. 2017, 56, 753−756. (32) Shao, Y.; Wang, H.; Gong, Z.; Wang, D.; Zheng, B.; Zhu, J.; Lu, Y.; Hu, Y.-S.; Guo, X.; Li, H.; Huang, X.; Yang, Y.; Nan, C.-W.; Chen, L. Drawing a soft interface: An Effective Interfacial Modification Strategy for Garnet-Type Solid-State Li Batteries. ACS Energy Lett. 2018, 3, 1212−1218. (33) Li, Y.; Chen, X.; Dolocan, A.; Cui, Z.; Xin, S.; Xue, L.; Xu, H.; Park, K.; Goodenough, J. B. Garnet Electrolyte with an Ultra-Low Interfacial Resistance for Li-Metal Batteries. J. Am. Chem. Soc. 2018, 140, 6448−6455. (34) Sharafi, A.; Kazyak, E.; Davis, A. L.; Yu, S.; Thompson, T.; Siegel, D. J.; Dasgupta, N. P.; Sakamoto, J. Surface Chemistry Mechanism of Ultra-Low Interfacial Resistance in the Solid-State Electrolyte Li7La3Zr2O12. Chem. Mater. 2017, 29, 7961−7968. (35) Blöchl, P. E. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, No. 17953. (36) Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, No. 11169. (37) Shi, S.; Gao, J.; Liu, Y.; Zhao, Y.; Wu, Q.; Ju, W.; Ouyang, C.; Xiao, R. Multi-Scale Computation Methods: Their Applications in Lithium-Ion Battery Research and Development. Chin. Phys. B 2016, 25, No. 018212. (38) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, No. 3865.

(39) Xie, H.; Alonso, J. A.; Li, Y.; Fernández-Díaz, M. T.; Goodenough, J. B. Lithium Distribution in Aluminum-Free Cubic Li7La3Zr2O12. Chem. Mater. 2011, 23, 3587−3589. (40) Young, T. An Essay on the Cohesion of Fluids. Philos. Trans. R. Soc. London 1805, 95, 65−87. (41) Liu, Y.; Ning, X.-S. Influence of α-Al2O3 (0001) Surface Reconstruction on Wettability of Al/Al2O3 Interface: A First-Principle Study. Comput. Mater. Sci. 2014, 85, 193−199. (42) Han, F.; Yue, J.; Chen, C.; Zhao, N.; Fan, X.; Ma, Z.; Gao, T.; Wang, F.; Guo, X.; Wang, C. Interphase Engineering Enabled AllCeramic Lithium Battery. Joule 2018, 2, 497−508. (43) Wu, J. F.; Chen, E. Y.; Yu, Y.; Liu, L.; Wu, Y.; Pang, W. K.; Peterson, V. K.; Guo, X. Gallium-Doped Li7La3Zr2O12 Garnet-Type Electrolytes with High Lithium-Ion Conductivity. ACS Appl. Mater. Interfaces 2017, 9, 1542−1552. (44) Wu, J. F.; Pang, W. K.; Peterson, V. K.; Wei, L.; Guo, X. Garnet-Type Fast Li-Ion Conductors with High Ionic Conductivities for All-Solid-State Batteries. ACS Appl. Mater. Interfaces 2017, 9, 12461−12468. (45) Chen, Y.-T.; Jena, A.; Pang, W. K.; Peterson, V. K.; Sheu, H.-S.; Chang, H.; Liu, R.-S. Voltammetric Enhancement of Li-Ion Conduction in Al-Doped Li7−xLa3Zr2O12 Solid Electrolyte. J. Phys. Chem. C 2017, 121, 15565−15573. (46) Cheng, L.; Crumlin, E. J.; Chen, W.; Qiao, R.; Hou, H.; Franz Lux, S.; Zorba, V.; Russo, R.; Kostecki, R.; Liu, Z.; Persson, K.; Yang, W.; Cabana, J.; Richardson, T.; Chen, G.; Doeff, M. The Origin of High Electrolyte-Electrode Interfacial Resistances in Lithium Cells Containing Garnet Type Solid Electrolytes. Phys. Chem. Chem. Phys. 2014, 16, 18294−18300. (47) Buschmann, H.; Dolle, J.; Berendts, S.; Kuhn, A.; Bottke, P.; Wilkening, M.; Heitjans, P.; Senyshyn, A.; Ehrenberg, H.; Lotnyk, A.; Duppel, V.; Kienle, L.; Janek, J. Structure and Dynamics of the Fast Lithium Ion Conductor “Li7La3Zr2O12”. Phys. Chem. Chem. Phys. 2011, 13, 19378−19392. (48) Liu, T.; Ren, Y.; Shen, Y.; Zhao, S.-X.; Lin, Y.; Nan, C.-W. Achieving High Capacity in Bulk-Type Solid-State Lithium Ion Battery Based on Li6.75La3Zr1.75Ta0.25O12 Electrolyte: Interfacial Resistance. J. Power Sources 2016, 324, 349−357. (49) Sharafi, A.; Meyer, H. M.; Nanda, J.; Wolfenstine, J.; Sakamoto, J. Characterizing the Li−Li7La3Zr2O12 Interface Stability and Kinetics as a Function of Temperature and Current Density. J. Power Sources 2016, 302, 135−139. (50) Cheng, L.; Park, J. S.; Hou, H.; Zorba, V.; Chen, G.; Richardson, T.; Cabana, J.; Russo, R.; Doeff, M. Effect of Microstructure and Surface Impurity Segregation on the Electrical and Electrochemical Properties of dDense Al-Substituted Li7La3Zr2O12. J. Mater. Chem. A 2014, 2, 172−181. (51) Zhang, Z.; Zhang, Q.; Shi, J.; Chu, Y. S.; Yu, X.; Xu, K.; Ge, M.; Yan, H.; Li, W.; Gu, L.; Hu, Y.-S.; Li, H.; Yang, X.-Q.; Chen, L.; Huang, X. A Self-Forming Composite Electrolyte for Solid-State Sodium Battery with Ultralong Cycle Life. Adv. Energy Mater. 2017, 7, No. 1601196.

905

DOI: 10.1021/acsami.8b18356 ACS Appl. Mater. Interfaces 2019, 11, 898−905