Cation Mixing Properties toward Co Diffusion at the LiCoO2 Cathode

Dec 19, 2016 - All-solid-state Li-ion batteries (ASS-LIBs) are expected to be the next-generation battery, however, their large interfacial resistance...
26 downloads 0 Views 3MB Size
Download PDF
Research Article www.acsami.org

Cation Mixing Properties toward Co Diffusion at the LiCoO2 Cathode/ Sulfide Electrolyte Interface in a Solid-State Battery Jun Haruyama,*,† Keitaro Sodeyama,‡,⊥,∥ and Yoshitaka Tateyama*,‡,§,⊥ †

Global Research Center for Environment and Energy Nanoscience (GREEN), ‡Center for Materials Research by Information Integration (CMI2), and §Center for Green Research on Energy and Environmental Materials, National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan ⊥ Elements Strategy Initiative for Catalysts and Batteries, Kyoto University, Goryo-Ohara, Nishikyo-ku, Kyoto 615-8245, Japan ∥ PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 333-0012, Japan S Supporting Information *

ABSTRACT: All-solid-state Li-ion batteries (ASS-LIBs) are expected to be the next-generation battery, however, their large interfacial resistance hinders their widespread application. To understand and resolve the possible causes of this resistance, we examined mutual diffusion properties of the cation elements at LiCoO2 (LCO) cathode/β-Li3PS4 (LPS) solid electrolyte interface as a representative system as well as the effect of a LiNbO3 buffer layer by first-principles calculations. Evaluating energies of exchanging ions between the cathode and the electrolyte, we found that the mixing of Co and P is energetically preferable to the unmixed states at the LCO/LPS interface. We also demonstrated that the interposition of the buffer layer suppresses such mixing because the exchange of Co and Nb is energetically unfavorable. Detailed analyses of the defect levels and the exchange energies by using the individual bulk crystals as well as the interfaces suggest that the lower interfacial states in the energy gap can make a major contribution to the stabilization of the Co ↔ P exchange, although the anion bonding preference of Co and P as well as the electrostatic interactions may have effects as well. Finally, the Co ↔ P exchanges induce interfacial Li sites with low chemical potentials, which enhance the growth of the Li depletion layer. These atomistic understandings can be meaningful for the development of ASS-LIBs with smaller interfacial resistances. KEYWORDS: lithium ionic conductor, solid electrolyte, first-principles calculations, interfacial resistance, mutual diffusion

■

composed of an interposed insulating layer (e.g., Li4Ti5O12,15 LiNbO3:LNO,16 LiTaO3,17,18 and Li2SiO319), cathode (e.g., LiCoO2:LCO), and sulfide electrolyte (e.g., Li4GeS4−Li3PS4: thio-lithium superior ionic conductor (LISICON) and glass ceramic). In particular, LNO-coated LCO particles exhibit very low interfacial resistance.16 However, the underlying mechanism is still unclear, making it difficult to identify the ratedetermining factor of ASS-LIB operation. One possible mechanism for this resistance is the formation of an ionic space−charge layer (SCL, see ref 20 for details) on the electrolyte side.6 In a previous study, we constructed representative oxide cathode/sulfide electrolyte interfaces by first-principles calculations.21 The calculated energies of Livacancy formation at the interfaces suggested Li-depletion-layer growth at the beginning of the charge process. However, to the best of our knowledge, there is no clear evidence of SCL, e.g.

INTRODUCTION

Development of large-scale and stable energy storage is strongly needed throughout the world due to an ever-increasing demand for smart grids and electronic vehicles.1−4 All-solid-state lithium-ion batteries (ASS-LIBs), in which the organic liquid electrolyte typical of LIBs is replaced with an inorganic solid electrolyte, have many advantages: high energy density, safety, shelf life, and long cycle life.5−7 Although the practical applications of ASS-LIBs were hindered by the low ionic conductivity of solid electrolytes, this challenge has been overcome by the development of sulfide-type electrolytes. The ionic conductivities of recently discovered sulfide materials are comparable to those of organic liquid electrolytes (ca. 10−3 S/ cm).8−12 Thus, the rate-determining step of the charge− discharge process become the Li extraction/insertion reaction at the interface between the cathode and sulfide electrolyte.6,13,14 The buffer layer coating has been found to have an important effect on this reaction. Reduced interfacial resistance has been observed in the impedance spectra of composite cells © 2016 American Chemical Society

Received: July 10, 2016 Accepted: December 6, 2016 Published: December 19, 2016 286

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces

conditions and three constructed interfaces are described in Supporting Information, section S1 and Figure S1, respectively. We mention that our simulation interface cell does not treat the cathode voltage directly, and discussion for applied voltage is simply based on Li-vacancy formation energies. A more sophisticated approach such as that used in ref 37 is necessary in future studies. We adopted LPS as the sulfide electrolyte because of its high Li-ion conductivity and simple crystal structure.38 In addition, DFT calculations for the independent surfaces are available.39−43 Because the properties of the three selected components are typical of ASS-LIBs, the three interfaces can include the origin of the resistance observed at the cathode/ sulfide interfaces. Comparing the Li-vacancy formation energies at these interfaces, we found that subsurface Li in the LPS side can begin to transfer at the undervoltage condition only at the LCO/LPS interface, which suggests the Li-depletion-layer growth at the beginning of charging, leading to the interfacial resistance. To discuss the transition metal diffusions at cathode/sulfide interfaces in ASS-LIBs, we focused on the cation exchange properties at the interfaces. We assumed that cation exchange is an elementary step causing the wide range of Co distributions observed in experiments. We then evaluate possibility of the cation diffusion from the exchange energy defined as

the direct observations of Li-ion depletion in working cells. In addition, the size of the SCL suggested by the electric potentials in ASS-LIBs22 is difficult to explain in terms of the Debyescreening length of ion-conductor materials.23 Therefore, further investigations are necessary to determine the contribution of the SCL to the interfacial resistance. A second possible mechanism is microstructure/composition modification induced by continuing charge−discharge cycles; a series of applied bias voltages or Li-ion migrations could induce undesirable chemical reactions or atomic diffusions at interfaces. For example, Sakuda et al. found a wide range of Co distributions from the LCO cathode to glass-ceramic-type sulfide electrolytes.24 Their elemental mapping technique, which combined scanning transmission electron microscopy with energy dispersive X-ray spectroscopy, showed that lots of the Co atoms were distributed in the interfacial region of sulfide electrolyte (Co was even found 50 nm from the interface). Furthermore, a subsequent measurement confirmed that an interposition of the buffer layer reduces the leakage of the Co atoms.25 In the same experiment, the presence of a buffer layer suppressed the increase of the interfacial resistance as the charge−discharge cycles progressed. Woo et al. reported a similar observation at Al2O3 coated LCO/thio-LISICON interfaces.26 Recent thermodynamic analyses of electrochemical stabilities based on first-principles calculations revealed that sulfide type electrolytes are oxidized at around 2 V (vs Li/Li+),27−29 and decomposition reactions at the LCO/sulfide interfaces form Co sulfide phases such as CoS2 and Co3S4.28−30 These experiments and theoretical predictions suggest that the Co atoms can migrate (diffuse) from the cathode to the electrolyte during charge−discharge processes. Repetition of these atomic diffusions can induce detrimental transformations at the LCO/sulfide interfaces, e.g. phase transition or resistive-layer formation. Therefore, suppression of interfacial Co diffusion will reduce the interfacial resistance of ASS-LIBs. This article is organized as follows. First, we evaluate the cation exchange energies and suggest a way of outflow the Co atoms. We then verify the means by which the buffer layer reduces the leakage of the Co diffusion. To understand the origins of these effects, we examine the structural and the electronic properties of the unmixed and mixed interfaces. Sample estimation of Li vacancy formation energies around the interfaces is carried out as well. Based on the results, we clarify microscopic aspects of mutual diffusion at solid/solid interfaces, frequently observed between ion-conducting solid interfaces.31,32

pristine Eex (A i , Bj) = Etot(A i ↔ Bj) − Etot

(1)

Here Epristine is the total energy of the unmixed interface tot evaluated by the DFT+U calculation and Etot(Ai ↔ Bj) represents the total energy of a relaxed interface in which the position of the ith A atom is replaced with the jth B atom from the pristine interface. The value of the exchange energy is a quantitative measure for the exchange of the two selected elements. We chose the elements of Co, Li, P, and Nb to examine the exchange properties at the three interfaces. In calculations of this study, the slabs of these interfaces are assumed to be sharp and the materials are largely intact with small distortions. Many experiments including some of previous calculations suggest the formation of interlayers.24−29 Therefore, this study only corresponds to either the early stage of interfacial reaction, or if the interface is kinetically stabilized.

■

RESULTS First, we examined the exchange energies Eex, mainly focusing on the Co and Li atoms. Figure 1 shows the exchange energies of Co ↔ Li at the LCO/LPS interface. We use the VESTA to illustrate atomic structures throughout this article.44 We selected various Li sites; the top-row Li atoms (Li1, 2, 3) and bottom-row Li (Li4, 5, 6) are 6-fold coordinated (LiS6) and 4fold coordinated (LiS4), respectively. That is, they occupy 4b and 8d sites in the orthorhombic Pnma crystal, respectively.45 The Li1 and Li4 atoms are adsorbed on the edges of the CoO2 layer. In addition, the Co1 and Co2 atoms form CoO4 and CoO6 polyhedrons, respectively. The exchange energies for the sites up to the second-layer Co from the LCO/LPS boundary are displayed in Figure 1b. The exchange energies of distant Co atoms show no significant differences compared with that of Co2. The relaxed interface structures after the exchanges of Co1 ↔ Li2 and Co2 ↔ Li5 are illustrated in Figure S2 in the Supporting Information. From the viewpoint of Co1 atom, the exchange with the neighboring Li (Li1) has rather low exchange

■

METHODS In this study, we examined the LCO and β-Li3PS4 (LPS) solid electrolyte interface with and without interposed an buffer layer (LNO) by density functional theory (DFT) calculations: LCO(110)/LPS(010), LCO(110)/LNO(11̅ 0 ), and LNO(11̅0)/LPS(010). The three interfaces were constructed in our previous study.21 Systematic procedures were employed to find stable interfacial matching of two crystal solids. We used the spin-unpolarized PBE functional33 and DFT corrected Hubbard U terms implemented in QUANTUM ESPRESSO code.34 We checked interface polarization effects using the effective screening medium (ESM) method.35,36 As differences of the total energies between conventional periodic boundary condition (PBC) and nonrepeated slab approach are negligible, we show PBC results only. The detailed computational 287

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces

energies are negative (less than −1 eV); the interface structures after the Co ↔ P exchanges are more stable than the pristine LCO/LPS interface. Except for the Co1 atom, the Co atoms close to the LCO/ LPS boundary show lower exchange energies than those far from the boundary; the LCO/LPS interface has the tendency to convert the Co and P elements. The relative high exchange energies of the Co1 atom result from the change of the relaxed structures after Co ↔ P replacements. We show the relaxed structures of the Co1 ↔ P3 and Co2 ↔ P3 exchanges in Figure S3 in the Supporting Information. In the case of Co1 ↔ P3 case, the replaced P atom captures the O atom from the second-layer CoO6 octahedron, and the interface structure forms PO4 and CoO5 polyhedrons at the edge of the CoO2 layer. In contrast, the interface structure of the Co2 ↔ P3 interface does not show this structural change, the replacement P3 atom forms a PO6 octahedron directly. Thus, these structural differences cause the Co1 ↔ P3 interface to have a higher exchange energy than the Co2 ↔ P3 interface. These exchange energies indicate strong mixing of the Co and P atoms not only near but also far from the boundary. The preference of cobalt (relative to phosphorus) for sulfur could be driving the defect formation. To reveal the effect of the buffer layer on the leakage of Co atoms, we evaluated the Co and P exchange energies in the presence of an LNO buffer layer. Figure 3 shows the exchange

Figure 1. (a) Structure of the LCO(110)/LPS(010) interface and (b) corresponding Co ↔ Li exchange energies. The Li, O, P, S, and Co atoms are depicted as light green, red, purple, yellow, and blue spheres, respectively. Blue, purple, and light green polyhedrons represent CoO6, PS4, and LiS4 complexes, respectively.

energy (ca. 0.2 eV). On the other hand, the other Li sites (Li2− 6) show sufficiently high exchange energies (1−3 eV). Note that the LiS4 sites have slightly higher exchange energies than those of the LiS6 sites. In consequence, the Co atoms can only migrate to the neighboring Li sites, while Co diffusions through the Li sites in the LPS seems unlikely. Next, we examined another possible exchange, i.e. the mixing of Co and P atoms. Figure 2 shows the Co ↔ P exchange energies at the LCO/LPS interface. We selected Co and P atoms as the fourth layer from the LCO/LPS boundary. Co1 forms a CoO4 tetrahedron while Co2, 3, and 4 form CoO6 octahedrons. All of the P atoms are located in PS4 tetrahedrons. Note that the atomic structures of Figures 1 and 2 correspond to different view angles. At a glance, the Co ↔ P exchange

Figure 3. Structures of (a) LCO(110)/LNO(11̅0 ) and (b) LNO(11̅0)/LPS(010). The Nb atoms and NbO6 complexes are depicted as green spheres and green polyhedrons, respectively. Corresponding exchange energies for (c) Co ↔ Li/Nb in the LCO/ LNO interface and (d) P ↔ Li/Nb in the LNO/LPS interface are also shown.

energies of Co ↔ Li/Nb and P ↔ Li/Nb at the LCO/LNO and LNO/LPS interfaces, respectively. We selected the Co and P atom closest to the boundaries. The Co and P atoms form CoO6 and PS4 polyhedrons, respectively. The Nb and Li atoms in the LNO slab are calculated to the third layer. It is noted that Nb1 at the LNO/LPS interface forms a NbO4 tetrahedron, whereas the other Nb atoms form NbO6 octahedrons. At the LCO/LNO interface, the Co ↔ Li and Co ↔ Nb exchange energies are positive, being 1−2 and 1 eV, respectively. At the LNO/LPS interface, except for the Nb1

Figure 2. (a) Structure of the LCO(110)/LPS(010) interface and (b) corresponding Co ↔ P exchange energies. 288

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces atom, the P ↔ Li/Nb exchange energies are also positive and higher than 1 eV. The negative exchange energy of P1 ↔ Nb1 (−1.1 eV) indicates that the NbO4 tetrahedrons can be replaced by PO4 tetrahedrons. If LNO contains a large amount of NbO6 octahedrons, the P atom cannot alternate the Nb atom. Although coated buffer layers are regarded to be an amorphouslike phase,17 we expect that there are not too many tetrahedral sites. Consequently, the Co atoms cannot penetrate the LNO layer, which is consistent with the experimental fact that buffer layers suppress the wide range of Co distributions.24−26 We should point out that the defected interface structures considered here are not necessarily the final configurations. Close to the interface (because this reaction is diffusion limited), more defects are likely to occur because of the large driving force for their formation. This tendency of mixing might lead to amorphization of a layer at the interface or nucleation of an interfacial phase in extreme cases (and with sufficiently high temperature). The calculated defected states here can be regarded as the intermediates between the pristine structure and the final state; the negative calculated energy shows that there is a minimal energetic barrier in the beginning of this process.

LCO/LPS interface with and without the Co2 ↔ P3 exchange are shown in Figure 4, where states near the Fermi levels are

DISCUSSION We now discuss the properties of the LCO/LPS interfaces from the viewpoints of bond lengths and electronic states. First, the bond lengths of Co/P−O and P/Co−S before and after the Co2 ↔ P3 exchange are shown in Figure S4 in the Supporting Information. In the pristine LCO/LPS interface, the Co−O and P−S bonds have 1.92 and 2.06 Å (bond lengths in LCO and LPS bulk crystals), respectively. After the Co2 ↔ P3 exchange, the Co/P−O bond lengths decrease to 1.63−1.84 Å and the PO6 octahedron shrinks compared with other CoO6 structures. The exchanged P atom moves from the center of PO6 octahedron and inclines the edge side of the CoO2 layer. On the other hand, the bond lengths of P/Co−S increase to 2.08− 2.15 Å and the CoS4 tetrahedron expands with the Co atom remaining at the center of the CoS4 tetrahedron. These deformations do not seem to mainly contribute to the negative exchange energies for the Co ↔ P. We next consider an exchange of Co ↔ P using individual LCO and LPS bulk crystals to discuss the defects levels clearly (the computational details for the bulk crystals are shown in Section S1 in the Supporting Information). We calculated the electronic structures of LCO and LPS bulk crystals after introducing PCo and CoP defect, respectively. The projected densities of states (PDOSs) of their crystals are shown in Figure S5 in the Supporting Information. In the LCO crystal, the valence and conduction bands consist of Co-3d and O-2p orbitals. There is no defect state in the energy gap; two electrons are introduced by the one PCo defect. It could be interpreted that P has a + 5 charge, namely P is not reduced nor oxidized (and two Co are reduced to +2 in this calculation system). In the LPS crystal, its valence band consists mainly of S 3p and its conduction band (around 3 eV) consists of S 3p and P 3p orbitals, respectively. In addition, there are Co 3d defect states around the energy gap. The Co could be regarded as +5 (not +3) because of tetrahedral CoS4 configuration and S is not oxidized. In the interface system, on the other hand, the electrons introduced by PCo defects can occupy interface states arising from CoO4 tetrahedrons at the interface. The PDOSs of the

Figure 4. PDOSs of the LCO/LPS interface (a) pristine and (b) with Co2 ↔ P3 exchange. The black, green, blue, red, purple, and yellow lines represent the PDOS of the total, Li, Co, O, P, and S atoms, respectively. The dotted lines indicate Fermi levels. The energy origins are set at the top levels of the occupied band of the pristine interface.

■

attributed to pseudotetrahedral CoO4 placed at the edge of the CoO2 layer.46 The levels of such interface states frequently appear in the energy gaps of the bulk crystals (especially such oxide/sulfide interfaces, see ref 21). Here we analyze the exchange energies in detail again by using the LCO and LPS bulk crystals in addition to the interface systems. The total energies of the LCO and LPS bulk crystals with various defects are listed in Table S3 in the Supporting Information. The Co ↔ P exchange energy calculated with the individual LCO and LPS bulk crystals is +1.98 eV, while that for Co2 ↔ P3 in the interface system is −2.18 eV (see Figure 2). Since the difference between the bulk crystal systems and the interface ones is the presence of the interfacial states in the latter, we attribute the lower exchange energy to the electronic occupations of the lower interfacial states. Note that the Co ↔ Nb exchanges do not have this effect significantly, because the Co and Nb at the LCO/LNO interface mostly form the octahedrons, in contrast to the Co insertion into the tetrahedron in the LCO/LPS case. We also examine the charge compensation effect by the Li transfer across the interface. The exchange energy for the Co2 ↔ P3 is −2.18 eV as described above. Meanwhile, the interface system including PCo + 2 VLi (Li vacancy) in the LCO side and CoP + 2 Lii (Li interstitial) in the LPS side (the Co2 ↔ P3 exchange with 2 Li transfer from the LCO side to the LPS) has the energy around −0.9 eV, with respect to the pristine interface system. This indicates that the charge compensation by the Li transfer together with the cation exchange is not the main cause of the negative exchange energies for the Co ↔ P, and the electronic effect induced by the interface formation without the Li transfer has more contribution. Note that the electrostatic attractive interactions between the charged defects can be another explanation for the negative exchange energies. However, the other charged pairs such as the Co ↔ Nb 289

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces

deformations and the subsequent changes of Li diffusion barriers by the interfacial cation exchanges are to be elucidated, which will be in the future studies. The atomistic understanding as gained in this study may lead to the development of ASSLIBs with small interfacial resistance and exhibiting good cyclability.

exchange do not have significant energy gains. Thus, the electronic states confined around the interfaces seem to have more contribution. The interface system with the charge compensation (PCo + 2 VLi in the LCO slab and CoP + 2 Lii in the LPS slab) can be compared with the same defects in the individual bulk crystals, because all the slabs are charge neutral. The energy sum of the PCo + 2 VLi in the LCO bulk and the CoP + 2 Lii in the LPS bulk is +2.84 eV with respect to the pristine bulk reference (see Table S3), which is much larger than around −0.9 eV for the interface system. This difference can be attributed to the interface formation. We also investigate the initial stage of the charging with a bulk model of LCO + 2 VLi as the reference briefly. If the Co ↔ P occurs in this stage (PCo + 2 VLi in the LCO bulk and CoP in the LPS bulk), the exchange energy can be −1.62 eV under the comparison with the bulk crystals. Thus, LixCoO2 (0.5 < x < 1) cathode can easily introduce the PCo defects in the LCO side, which might be one cause of the cathode degradation. Finally, we calculated the Li-vacancy formation energies at the LCO/LPS interface with 6 Co ↔ P exchanges. The number of exchange was set to reproduce the observed Co concentration.24 The computational details, relaxed interface structure, and formation energies are shown in section S1 and Figures S6 and S7 in the Supporting Information. The Livacancy formation energies at the LCO/LPS interface with Co ↔ P exchanges are lower than those of pristine interface because additional electrons from PCo defects in the interface are easily extracted. Consequently, one PCo defect can produce two additional Li vacancies. This can also be framed in the context of thermodynamics of the bulk phases; LiCoS2 has a lower voltage than LiCoO2,47 and so as Co is increasingly surrounded by sulfur the interface takes on more of the character of LiCoS2. As a result, the Li-vacancy formation at the LCO/LPS interface with Co ↔ P exchanges are energetically more preferable, suggesting that the Co ↔ P exchanges induce interfacial Li sites with low chemical potentials. Therefore, the growth of Li depletion layer can be enhanced in the presence of the Co ↔ P exchanges.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications Web site. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b08435. Computational details of the calculations of the interface and crystal systems. Cell parameters in the interface and bulk crystal calculations. Structural information for the pristine interfaces of LCO/LPS, LCO/LNO, and LNO/ LPS. Interface structures of LCO/LPS with Co ↔ Li and Co ↔ P exchanges. Bond lengths of Co/P−O and P/ Co−S before and after the Co2 ↔ P3 exchange at the LCO/LPS interface. PDOSs of LCO and LPS crystals including PCo and CoP defects as well as bulk crystal structures. Total energies of the LCO and LPS bulk crystals including various defects. Relaxed interface structure of LCO/LPS with 6 Co ↔ P exchanges. Formation energies of the Li vacancies at LCO/LPS interface in the case with 6 Co ↔ P exchanges (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (J.H.). *E-mail: [email protected] (Y.T.). ORCID

Yoshitaka Tateyama: 0000-0002-5532-6134 Notes

The authors declare no competing financial interest.

■

■

ACKNOWLEDGMENTS The authors appreciate Dr. Kazunori Takada for his meaningful discussions. The authors also thank Dr. Yoshinori Tanaka for the valuable discussions. This work was supported in part by JSPS and MEXT KAKENHI Grant Numbers JP16K17969, JP15K05138, and JP15H05701. The calculations were carried out on the supercomputers in NIMS, Institute for Solid State Physics, and The University of Tokyo as well as the supercomputers in Kyushu University. This research also used computational resources of the HPCI system through the HPCI System Research Project (Project IDs: hp150055, hp150068, hp160040, hp160080).

CONCLUSIONS In this study, we investigated the cation-mixing properties at oxide cathode/sulfide electrolyte interfaces in ASS-LIBs by comparing the exchange energies using a DFT+U treatment. The results for the representative LCO/LPS interfaces reveal that the Co ↔ P exchange energies are negative, which strongly suggests preferential Co and P mixing at the LCO/LPS interface. The detailed analyses suggest that the lower interfacial states in the energy gap seem to make a major contribution to the stabilization of the Co ↔ P exchange, although the anion bonding preference of Co and P as well as the electrostatic interactions may have effects as well. These results imply that the Co-interdiffusion is likely to occur without requiring the nucleation and growth of an interfacial layer, and can be therefore expected to proceed rapidly. We further verified that an interposition of a LNO buffer layer could suppress this mixing because the exchange of Co and Nb is energetically prohibitive. These exchange preferences based on the calculated energetics are consistent with the elemental analysis measurements and first-principles thermodynamic calculations. Furthermore, the Co ↔ P exchanges induce interfacial Li sites with low chemical potentials. For comprehensive understanding of the origin, the structural

■

REFERENCES

(1) Armand, M.; Tarascon, J.-M. Building Better Batteries. Nature 2008, 451, 652−657. (2) Li, B. H.; Wang, Z.; Chen, L.; Huang, X. Research on Advanced Materials for Li-Ion Batteries. Adv. Mater. 2009, 21, 4593−4607. (3) Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D. Challenges in the Development of Advanced Li-Ion Batteries: A Review. Energy Environ. Sci. 2011, 4, 3243−3262. (4) Goodenough, J. B.; Park, K.-S. The Li-Ion Rechargeable Battery: A Perspective. J. Am. Chem. Soc. 2013, 135, 1167−1176. (5) Dudney, N. J. Thin Film Micro-Batteries. Electrochem. Soc. Interface 2008, 17, 44−48. 290

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces (6) Takada, K. Progress and Prospective of Solid-State Lithium Batteries. Acta Mater. 2013, 61, 759−770. (7) Luntz, A. C.; Voss, J.; Reuter, K. Interfacial Challenges in SolidState Li Ion Batteries. J. Phys. Chem. Lett. 2015, 6, 4599−4604. (8) Kanno, R.; Murayama, M. Lithium Ionic Conductor ThioLISICON. J. Electrochem. Soc. 2001, 148, A742−A746. (9) Mizuno, F.; Hayashi, A.; Tadanaga, K.; Tatsumisago, M. New, Highly Ion-Conductive Crystals Precipitated from Li2S−P2S5 Glasses. Adv. Mater. 2005, 17, 918−921. (10) Kamaya, N.; Homma, K.; Yamakawa, Y.; Hirayama, M.; Kanno, R.; Yonemura, M.; Kamiyama, T.; Kato, Y.; Hama, S.; Kawamoto, K.; Mitsui, A. A Lithium Superionic Conductor. Nat. Mater. 2011, 10, 682−686. (11) Seino, Y.; Ota, T.; Takada, K.; Hayashi, A.; Tatsumisago, M. A Sulfide Lithium Super Ion Conductor is Superior to Liquid Ion Conductors for Use in Rechargeable Batteries. Energy Environ. Sci. 2014, 7, 627−631. (12) Kato, Y.; Hori, S.; Saito, T.; Suzuki, K.; Hirayama, M.; Mitsui, A.; Yonemura, M.; Iba, H.; Kanno, R. High-Power All-Solid-State Batteries Using Sulfide Superionic Conductors. Nat. Energy 2016, 1, 16030. (13) Takada, K.; Ohta, N.; Zhang, L.; Xu, X.; Hang, B. T.; Ohnishi, T.; Osada, M.; Sasaki, T. Interfacial Phenomena in Solid-State Lithium Battery with Sulfide Solid Electrolyte. Solid State Ionics 2012, 225, 594−597. (14) Takada, K. Interfacial Nanoarchitectonics for Solid-State Lithium Batteries. Langmuir 2013, 29, 7538−7541. (15) Ohta, N.; Takada, K.; Zhang, L.; Ma, R.; Osada, M.; Sasaki, T. Enhancement of the High-Rate Capability of Solid-State Lithium Batteries by Nanoscale Interfacial Modification. Adv. Mater. 2006, 18, 2226−2229. (16) Ohta, N.; Takada, K.; Sakaguchi, I.; Zhang, L.; Ma, R.; Fukuda, K.; Osada, M.; Sasaki, T. LiNbO3-Coated LiCoO2 as Cathode Material for All Solid-State Lithium Secondary Batteries. Electrochem. Commun. 2007, 9, 1486−1490. (17) Takada, K.; Ohta, N.; Zhang, L.; Fukuda, K.; Sakaguchi, I.; Ma, R.; Osada, M.; Sasaki, T. Interfacial Modification for High-Power Solid-State Lithium Batteries. Solid State Ionics 2008, 179, 1333−1337. (18) Xu, X.; Takada, K.; Fukuda, K.; Ohnishi, T.; Akatsuka, K.; Osada, M.; Hang, B. T.; Kumagai, K.; Sekiguchi, T.; Sasaki, T. Tantalum Oxide Nanomesh as Self-Standing One Nanometer Thick Electrolyte. Energy Environ. Sci. 2011, 4, 3509−3512. (19) Sakuda, A.; Kitaura, H.; Hayashi, A.; Tadanaga, K.; Tatsumisago, M. Modification of Interface between LiCoO2 Electrode and Li2S− P2S5 Solid Electrolyte Using Li2O−SiO2 Glassy Layers. J. Electrochem. Soc. 2009, 156, A27−A32. (20) Maier, J. Ionic Conduction in Space Charge Regions. Prog. Solid State Chem. 1995, 23, 171−263. (21) Haruyama, J.; Sodeyama, K.; Han, L.; Takada, K.; Tateyama, Y. Space−Charge Layer Effect at Interface between Oxide Cathode and Sulfide Electrolyte in All-Solid-State Lithium-Ion Battery. Chem. Mater. 2014, 26, 4248−4255. (22) Yamamoto, K.; Iriyama, Y.; Asaka, T.; Hirayama, T.; Fujita, H.; Fisher, C. A. J.; Nonaka, K.; Sugita, Y.; Ogumi, Z. Dynamic Visualization of the Electric Potential in an All-Solid-State Rechargeable Lithium Battery. Angew. Chem., Int. Ed. 2010, 49, 4414−4417. (23) Braun, S.; Yada, C.; Latz, A. Thermodynamically Consistent Model for Space-Charge-Layer Formation in a Solid Electrolyte. J. Phys. Chem. C 2015, 119, 22281−22288. (24) Sakuda, A.; Hayashi, A.; Tatsumisago, M. Interfacial Observation between LiCoO2 Electrode and Li2S−P2S5 Solid Electrolytes of All-Solid-State Lithium Secondary Batteries Using Transmission Electron Microscopy. Chem. Mater. 2010, 22, 949−956. (25) Ohtomo, T.; Hayashi, A.; Tatsumisago, M.; Tsuchida, Y.; Hama, S.; Kawamoto, K. All-Solid-State Lithium Secondary Batteries Using 75Li2S·25P2S5 Glass and the 70Li2S·30P2S5 Glass−Ceramic as Solid Electrolytes. J. Power Sources 2013, 233, 231−235. (26) Woo, H. J.; Trevey, J. E.; Cavanagh, A. S.; Choi, Y. S.; Kim, S. C.; George, S. M.; Oh, K. H.; Lee, S.-H. Nanoscale Interface

Modification of LiCoO2 by Al2O3 Atomic Layer Deposition for SolidState Li Batteries. J. Electrochem. Soc. 2012, 159, A1120−A1124. (27) 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, 23685−23693. (28) Zhu, Y.; He, X.; Mo, Y. First Principles Study on Electrochemical and Chemical Stability of Solid Electrolyte−Electrode Interfaces in All-Solid-State Li-Ion Batteries. J. Mater. Chem. A 2016, 4, 3253−3266. (29) Richards, W. D.; Miara, L. J.; Wang, Y.; Kim, J. C.; Ceder, G. Interface Stability in Solid-State Batteries. Chem. Mater. 2016, 28, 266−273. (30) Yokokawa, H. Thermodynamic Stability of Sulfide Electrolyte/ Oxide Electrode Interface in Solid-State Lithium Batteries. Solid State Ionics 2016, 285, 126−135. (31) Brazier, A.; Dupont, L.; Dantras-Laffont, L.; Kuwata, N.; Kawamura, J.; Tarascon, J.-M. First Cross-Section Observation of an All Solid-State Lithium-Ion “Nanobattery” by Transmission Electron Microscopy. Chem. Mater. 2008, 20, 2352−2359. (32) Okumura, T.; Nakatsutsumi, T.; Ina, T.; Orikasa, Y.; Arai, H.; Fukutsuka, T.; Iriyama, Y.; Uruga, T.; Tanida, H.; Uchimoto, Y.; Ogumi, Z. Depth-Resolved X-Ray Absorption Spectroscopic Study on Nanoscale Observation of the Electrode−Solid Electrolyte Interface for All Solid State Lithium Ion Batteries. J. Mater. Chem. 2011, 21, 10051−10060. (33) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (34) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; Dal Corso, A.; de Gironcoli, S.; Fabris, S.; Fratesi, G.; Gebauer, R.; Gerstmann, U.; Gougoussis, C.; Kokalj, A.; Lazzeri, M.; Martin-Samos, L.; Marzari, N.; Mauri, F.; Mazzarello, R.; Paolini, S.; Pasquarello, A.; Paulatto, L.; Sbraccia, C.; Scandolo, S.; Sclauzero, G.; Seitsonen, A. P.; Smogunov, A.; Umari, P.; Wentzcovitch, R. M. QUANTUM ESPRESSO: A Modular and Open-Source Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter 2009, 21, 395502. (35) Otani, M.; Sugino, O. First-Principles Calculations of Charged Surfaces and Interfaces: A Plane-Wave Nonrepeated Slab Approach. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 73, 115407. (36) Hamada, I.; Otani, M.; Sugino, O.; Morikawa, Y. Green’s Function Method for Elimination of the Spurious Multipole Interaction in the Surface/Interface Slab Model. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 165411. (37) Leung, K.; Leenheer, A. How Voltage Drops Are Manifested by Lithium Ion Configurations at Interfaces and in Thin Films on Battery Electrodes. J. Phys. Chem. C 2015, 119, 10234−10246. (38) Liu, Z.; Fu, W.; Payzant, E. A.; Yu, X.; Wu, Z.; Dudney, N. J.; Kiggans, J.; Hong, K.; Rondinone, A. J.; Liang, C. Anomalous High Ionic Conductivity of Nanoporous β-Li3PS4. J. Am. Chem. Soc. 2013, 135, 975−978. (39) Dahéron, L.; Martinez, H.; Dedryvère, R.; Baraille, I.; Ménétrier, M.; Denage, C.; Delmas, C.; Gonbeau, D. Surface Properties of LiCoO2 Investigated by XPS Analyses and Theoretical Calculations. J. Phys. Chem. C 2009, 113, 5843−5852. (40) Kramer, D.; Ceder, G. Tailoring the Morphology of LiCoO2: A First Principles Study. Chem. Mater. 2009, 21, 3799−3809. (41) Lepley, N. D.; Holzwarth, N. A. W.; Du, Y. A. Structures, Li+ Mobilities, and Interfacial Properties of Solid Electrolytes Li3PS4 and Li3PO4 from First Principles. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 88, 104103. (42) Lepley, N. D.; Holzwarth, N. A. W. Modeling Interfaces between Solids: Application to Li Battery Materials. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 92, 214201. (43) Sanna, S.; Schmidt, W. G. Lithium Niobate X-Cut, Y-Cut, and Z-Cut Surfaces from ab initio Theory. Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 81, 214116. 291

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292

Research Article

ACS Applied Materials & Interfaces (44) Momma, K.; Izumi, F. VESTA: A Three-Dimensional Visualization System for Electronic and Structural Analysis. J. Appl. Crystallogr. 2011, 44, 1272−1276. (45) Homma, K.; Yonemura, M.; Kobayashi, T.; Nagao, M.; Hirayama, M.; Kanno, R. Crystal Structure and Phase Transitions of the Lithium Ionic Conductor Li3PS4. Solid State Ionics 2011, 182, 53− 58. (46) Qian, D.; Hinuma, Y.; Chen, H.; Du, L.-S.; Carroll, K. J.; Ceder, G.; Grey, C. P.; Meng, Y. S. Electronic Spin Transition in Nanosize Stoichiometric Lithium Cobalt Oxide. J. Am. Chem. Soc. 2012, 134, 6096−6099. (47) Aydinol, M. K.; Kohan, A. F.; Ceder, G.; Cho, K.; Joannopoulos, J. Ab Initio Study of Lithium Intercalation in Metal Oxides and Metal Dichalcogenides. Phys. Rev. B: Condens. Matter Mater. Phys. 1997, 56, 1354−1365.

292

DOI: 10.1021/acsami.6b08435 ACS Appl. Mater. Interfaces 2017, 9, 286−292