Letter pubs.acs.org/NanoLett
Twin Boundary-Assisted Lithium Ion Transport Anmin Nie,†,⊥ Li-Yong Gan,‡ Yingchun Cheng,‡ Qianqian Li,§ Yifei Yuan,† Farzad Mashayek,¶ Hongtao Wang,§ Robert Klie,⊥ Udo Schwingenschlogl,‡ and Reza Shahbazian-Yassar*,†,⊥,¶ †
Department of Mechanical Engineering-Engineering Mechanics, Michigan Technological University, 1400 Townsend Dive, Houghton, Michigan 49931, United States ‡ Department of Physical Science and Engineering, King Abdullah University of Science & Technology, Thuwal 23955-6900, Kingdom of Saudi Arabia § Institute of Applied Mechanics, Zhejiang University, Hangzhou 310027, P. R. China ⊥ Department of Physics and ¶Mechanical and Industrial Engineering Department, University of Illinois at Chicago, Chicago, Illinois 60607, United States S Supporting Information *
ABSTRACT: With the increased need for high-rate Li-ion batteries, it has become apparent that new electrode materials with enhanced Li-ion transport should be designed. Interfaces, such as twin boundaries (TBs), offer new opportunities to navigate the ionic transport within nanoscale materials. Here, we demonstrate the effects of TBs on the Li-ion transport properties in single crystalline SnO2 nanowires. It is shown that the TB-assisted lithiation pathways are remarkably different from the previously reported lithiation behavior in SnO2 nanowires without TBs. Our in situ transmission electron microscopy study combined with direct atomic-scale imaging of the initial lithiation stage of the TB-SnO2 nanowires prove that the lithium ions prefer to intercalate in the vicinity of the (101̅) TB, which acts as conduit for lithium-ion diffusion inside the nanowires. The density functional theory modeling shows that it is energetically preferred for lithium ions to accumulate near the TB compared to perfect neighboring lattice area. These findings may lead to the design of new electrode materials that incorporate TBs as efficient lithium pathways, and eventually, the development of next generation rechargeable batteries that surpass the rate performance of the current commercial Li-ion batteries. KEYWORDS: Twin boundary, lithium-ion transport, in situ STEM, atomic scale, tin oxide nanowires
I
pathways for ions because of the enhanced effective ion conductivity and reduction of the effective diffusion length.9,10 Twin boundaries (TBs), which are observed in a broad class of materials, are known to have novel physical and chemical properties11−15 that affect the transport within the host medium. Early studies have shown that diffusion of impurity ions or vacancies can be enhanced within TBs. Typical examples are the transport of sodium and oxygen along TBs in WO3.15 Such enhanced ionic transport is desirable for increasing the rate performance of lithium-ion batteries. Recent first-principles calculations16 for LiCoO2 showed that the
n the past two decades, many experiments and theoretical calculations have shown that ionic transport properties can be substantially different at interfaces in comparison to the bulk areas.1−4 There are several explanations on how interfaces affect the transport behaviors. These include the presence of interfacial strain,5 which may result in the interfacial structure being more open than the bulk structure, which allows atoms to pass through more easily, or the fact that the interstitial atom and vacancy concentration in the interface are usually different from those of the bulk,6 which affects the rate of the diffusion. Another reason can be found in the space-charge zone at the interface,2 which can alter the transport of charged species through the interface. In nanoionics,7,8 it is highly expected that the interfaces and interfacially induced disorders act as fast © XXXX American Chemical Society
Received: October 24, 2014 Revised: December 11, 2014
A
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
cathode voltage can be decreased by 0.2 V in the vicinity of TBs relative to the perfect crystal. Moreover, there was a diffusion constant difference of three orders of magnitude between Li migration along and across the TBs. To date, no experimental work has reported the Li-ion transport within TBs at the atomic scale. Uncovering the atomistic mechanisms of the ion diffusion within TBs can provide new insights into tailoring the ionic transport behavior by microstructure engineering in battery electrode materials. In this work, we have investigated the transport properties of lithium ions in the presence of TBs in SnO2 nanowires, which are widely studied anode materials in Li-ion batteries. By using state-of-the-art aberration corrected scanning transmission electron microscopy (STEM) with the potential to identify atomic distances as small as 0.7 Å17 and to image light elements such as oxygen,18 lithium,19,20 and hydrogen21 in crystal structures, we directly monitored the dynamics of lithiation in TBs. The result indicates that the lithium transport pathway is remarkably different in the presence of a (101̅) TB. Atomicscale imaging and chemical analysis demonstrate that more lithium ions occupy near the TB than that in perfect lattice area at the initial lithiation stage of the TB-SnO2 nanowire. Firstprinciples simulations also prove that lithium ions energetically prefer to accumulate near the TB defect, which may act as conduit for significantly enhanced lithium-ion transport inside the nanowires. Figure 1, panel a shows the morphology and structure of a pristine SnO2 nanowire containing a TB marked by the arrows. The inset selected area electron diffraction (SAED) pattern taken from the [010] direction indicates that the twinning plane and shear direction are (101̅) and [101], respectively. Atomic-resolution high-angle annular dark field (HAADF) imaging is used to confirm the presence of TBs in SnO2 nanowire viewed in the [010] direction (Figure 1b). Such twins typically form during the synthesis and growth of SnO2 nanowires.22 The HAADF contrast exhibits a Z1.7 dependence with respect to the atomic number Z of a given atomic column.23 In the HAADF image of the twin structure, the oxygen atom columns are not visible, and the bright dots are indicative of Sn atom columns. Mirror symmetry is identified in Figure 1, panel b. Figure 1, panel c shows an atomic-scale annular bright field (ABF) image, acquired simultaneously with the HAADF image (Figure 1b). The oxygen columns can be clearly resolved in between the Sn atom rows (overlapped by inset modeled atoms) even though their contrasts are much weaker than those of the tin atom columns. To the authors’ knowledge, this is the first time that the oxygen atomic column in a TB in SnO2 is shown. The atomic structure of {101} twins in rutile structure TiO2 and SnO2 have been obtained by conventional high-resolution transmission electron microscopy (HRTEM).22 However, the positions of the oxygen atoms were not determined, since the metal atom columns dominate the lattice in conventional HRTEM images. In contrast, the ABF image provides a direct view to oxygen atoms in the SnO2 structure. These HAADF and ABF images show that the boundary plane consisting of Sn atoms is shared by two twinned domains, and the Sn sublattice structure is mirror symmetric. However, the oxygen atoms that form octagonal cages around metal atoms do not exhibit mirror symmetric with respect to the boundary plane. Prior calculations using the ionic shell model24 proposed that the atomic positions in one grain are the mirror-symmetric positions of the other but shifted by 1 /2 < 111>. By combining our experimental observations and
Figure 1. Atomic structure of the (101̅) TB in a SnO2 nanowire. (a) TEM image and corresponding electron diffraction pattern of individual TB-SnO2 nanowire. (b) Atomic resolution HAADF image of the (101̅) TB taken along the [010] axis. (c) Atomic resolution ABF image of the (101̅) TB taken along the [010] axis. The inset shows the Sn and O positions. (d) Atomic model of the (101̅) TB viewed along the [010] direction. Yellow frames highlight oxygen octahedrons near TB. Oxygen and tin atoms are shown in red and blue color, respectively. Size differences of the atoms indicate different level of the atoms. (e) Simulated HAADF image based on the atomic model in panel d. (f) Simulated ABF image based on the atomic model in panel d. Unit cell is indicated by black frame.
previous calculations,24 the atomic configuration of the (101̅) twin boundary in the SnO2 nanowire viewed along [010] is shown in Figure 1, panel d. The twin structure is such that the Sn atoms are mirror symmetric, but the oxygen atoms around the Sn atoms have 1/2 < 111> displacements from the mirror symmetric positions. The octagonal oxygen cages around Sn atoms (indicated by yellow lines) clearly show the spatial geometry relationship of the atom positions along the TB. The simulated HAADF and ABF images shown in Figure 1, panels e and f were obtained based on the atomic model in Figure 1, panel d.25 A comparison to the experimental observations in Figure 1, panels b and c indicates good agreement with respect to the atom positions along the twin boundary. Subsequently, the TEM analysis focused on the nanowires that contained TBs in order to reveal the role of twinning on the lithiation behavior of SnO2 nanowires. To start with, it is informative to address the lithiation behavior of the SnO2 nanowire without TB. Movie S1 (Supporting Information) shows the lithiation process of the SnO2 nanowire without TB. Lithiated strips that act as the reaction front traverse the nanowire during lithiation. Figure 2, panel a gives a clear view of a partially lithiated SnO2 nanowire without TB associated with the contrast changes. There is a distortion between the lithiated and unlithiated parts of the nanowire. According to the B
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
shows darker contrast strips due to the higher sensitivity to chemical composition variations. The LAADF image in Figure 2, panel d, which is more sensitive to the strain contrast,26−28 reveals that the strain field is associated with the lithium-ion intercalation in the nanowire. As a comparison, corresponding atomic-scale HAADF and LAADF images of the unlithiated part of the TBs can be found in Figure S1 (Supporting Information). The elemental distributions in the reaction front of the TBSnO2 nanowire were also investigated by electron energy loss spectroscopy (EELS) mapping, as shown in Figure 3. The
Figure 2. Direct observation of lithiation in SnO2 nanowires with and without TB viewed along the [010] direction. (a) TEM image of a partially lithiated single crystal SnO2 nanowire. An electron diffraction pattern taken from the reaction front (black box A) is shown on the left side of panel a. (b) TEM image of a partially lithiated TB-SnO2 nanowire. An electron diffraction pattern taken from the reaction front (black box B) is shown on the left side of panel b. (c) Atomic resolution HAADF image of the lithiation front in the TB-SnO2 nanowire taken along the [010] axis. (d) Atomic resolution LAADF image acquired simultaneously with the HAADF image in panel c.
Figure 3. HAADF of a lithiated area with TB and the corresponding EELS maps of O, Sn, and Li elements taken from the rectangular area marked in the HAADF image. The TB and the dark strip along the [001] direction (marked by a red arrow) show stronger Li signals along the directions marked by red arrows, whereas the Sn signals are slightly weaker.
inset SAED pattern in Figure 2, panel a, the axial orientation of the nanowire is found to be [101]. The lithiated strips are along the [001] direction in the (200) plane that traverses the SnO2 nanowire, which indicates that the lithium ions prefer diffusion along the [001] direction.17 Further details about the structural analysis of pristine SnO2 nanowires during lithiation are discussed in our earlier work.17 Interestingly, the presence of TB in SnO2 nanowires leads to a new lithiation scenario. Movies S2 and S3 (Supporting Information) show the behavior of the reaction front propagating into a TB-SnO2 nanowire. The activity of the TB and the (200) planes can be clearly seen by the change of contrast in the nanowire, which indicates that strain develops in the lattices. This is caused by the mass transport (lithium-ion diffusion). Figure 2, panel b shows a TEM image and SAED pattern of a partially lithiated TB-SnO2 nanowire. A lithiation tail with light contrast is visible in the center of the nanowire (Figure 2b). According to the analysis of the corresponding SAED pattern, the tail appears to propagate along the TB and two symmetrical (200) lattice planes (the [001] direction). Additionally, in comparison to the SAED of the pristine TBSnO2 nanowire in Figure 1, panel a, the SAED of the partially lithiated TB-SnO2 nanowire (Figure 2b) shows splitting and blurred spots due to the lithiation-induced lattice distortion in the nanowire. The atomic-resolution HAADF and low-angle annular dark-field (LAADF) images (Figure 2c,d) of the TBSnO2 nanowire provide a more clear view of the geometry of the lithiation trail in the reaction front. Darker contrast strips along the TB and [001] directions are indicative of lithium-ion intercalation in the nanowire. The HAADF image (Figure 2c)
peaks belonging to Li, Sn, and O are shown in the spectra (see Supporting Information, Figure S2). The TB and dark strip along the [001] direction in the HAADF image show much stronger Li signals, which confirms that more lithium is intercalated in these areas. This result agrees well with the appearance of darker contrast in the HAADF image and demonstrates that the TB provides a comparably fast lithiumion diffusion pathway as the [001] channel in the SnO2 lattice. To gain a better understanding of the role of the TB in the lithiation of the SnO2 nanowire, the initial lithiation stage in the TB-SnO2 nanowire was further investigated. Figure 4, panel a shows an atomic-resolution HAADF image of the (101̅) twin structure viewed from the [131] direction before lithiation. The inset is the corresponding fast Fourier transform (FFT) image. The strain map along the TB normal direction of the HAADF image (Figure 4a) was calculated29 as shown in Figure 4, panel b. A small fraction of strain fluctuation exists in the image because of scanning noise. However, there is no obvious lattice distortion near the TB. Figure 4, panel c shows an atomic-scale HAADF image of the (101̅) twin structure after a partial lithiation with a darker contrast near the TB in comparison with other areas of the image. The growth of the darker contrast is associated with the lithium intercalation into the lattice near the TB. Strain mapping along the TB normal direction of the HAADF image (Figure 3c) gives direct evidence that there is a lattice expansion near the TB (Figure 3d). According to Figure 4, panels c and d, the strain evolves in the lattice near the TB rather than in other areas, which demonstrates that lithium prefers to intercalate near the TB in C
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
Figure 4. Strain analysis of the partially lithiated SnO2-TB nanowire. (a) Atomic-scale HAADF image of the (101̅) TB along the [131] direction before lithiation. Inset is the corresponding FFT pattern. (b) Strain map taken along the (101̅) TB normal direction calculated from the HAADF image in panel a. (c) Atomic-scale HAADF image of the (101̅) TB taken along the [131] direction after partial lithiation. Inset is the corresponding FFT pattern. (d) Strain map along the (101̅) TB normal direction calculated from the HAADF image in panel c.
comparison with other lattice planes such as the regular {211} and {101} planes in the SnO2 crystal structure. To elucidate the effect of the intercalated Li ions on the structure evolution of the SnO2 lattice with TB, the experimentally observed TB structure was modeled with five mirror symmetrical Sn layers in each twinned grain (Figure 5a). To account for the nanowire geometry, a vacuum of 11 Å thickness was introduced to separate the slab, and each surface was terminated by oxygen atoms as shown in Figure 5, panel a. Figure 5, panels b and c show that the TB structure is modified by the incorporation of one and two Li ions, respectively. The simulation suggests that the subsequently intercalated Li ion preferably occupies the off-centered octahedral sites on the same side of the TB. The Sn−Sn bond length asymmetrically changes at the TB induced by the intercalation of Li ions, which is indicated in Figure 5, panel e. The dimensional change across the TB is presented to show the anisotropy of strain along various crystallographic directions as defined in Figure 5, panel d. It can be seen that the Sn−Sn distances along the [100] direction and perpendicular to the TB directions always experience the largest expansion among all the directions. This explains why the TB and (200) planes gain more contrast change in STEM images during the propagation of the lithiation reaction front in the nanowire. The DFT simulation was used to investigate the energy evolution during the mass transport inside the TB. The energy of nine inequivalent initial intercalation sites along the [001] direction was investigated as shown in Figure 6, panel a. The nine sites relax to five inequivalent local minima with respect to the TB (Figure 6b), which are marked as Li1, Li2, Li3, Li4, and Li5, respectively. Figure 6, panel c shows the energy of these five sites when Li atoms stay there. Here, we define Li1 as a reference. Evidently, the energy cost is the lowest when Li
Figure 5. Atomic structural models used to investigate the Li intercalation induced structure changes across a TB at various crystallographic directions. Oxygen and tin atoms are shown in red and blue color, respectively. Lithium atoms are shown in green color. (a) Before Li intercalation, (b) after one atom Li intercalation, and (c) after two Li atoms intercalation. (d) The TB normal direction and various crystallographic directions are shown. (e) Strain associated with Sn−Sn bond change in different crystallographic directions induced by one and two Li intercalations.
intercalates at the site near the TB. Our energy calculations indicate that the TB is a preferred site for the Li intercalation. The relative formation energy of the Sn vacancy at different sites along the [001] direction is also investigated as shown in Figure 6, panel a. According to Figure 6, panel d, it is more favorable to create Sn vacancies at the TB than those sites away from TB, which in turn attract the positively charged Li ion due to Columbic force. This work provides the first atomic-scale observations of lithium-ion interactions with TB defects. The in situ TEM results show that TBs are a preferable pathway for lithium ions during lithiation. The geometrical strains and concentration of lithium atoms were considerably higher at the TB in comparison to the neighboring areas. During lithiation, it is energetically favorable to form Sn vacancies around TB, which thus attracts Li atoms to the location of the TB. This observation paves the roadmap for development of new electrode materials with smart design of TB defects or other heterostructures to facilitate the lithium transport in and out of the electrode materials in shorter time frame. The design of such pathways for lithium transfer can assist the realization of high-rate capability in next generation lithium ion batteries. D
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
Letter
Figure 6. Atomic model of SnO2 (101̅) TB with Li intercalation at nine inequivalent sites along the [001] direction (a) before and (b) after relaxation. The TB is shown by the dashed line. Large blue balls show the first (Sn1), second (Sn2), and third (Sn3) nearest vacancy sites along the [001] direction with respect to the TB. The intercalated Li atoms are shown in green color. (c) Relative energy of Li intercalation along the [001] direction. (d) Relative vacancy formation energy of Sn1, Sn2, and Sn3. The value for Sn1 or Li1 was defined as a reference (0 eV).
facilitate the diffusion of Li+ from the Li metal to SnO2. A bias of −3 V was applied to the nanowire side during lithiation. The lithiation can be stopped at intermediate stages by retracting the lithium source and electrolyte from the nanowire. STEM Imaging Simulation. Multislice image simulations were conducted using the Kirkland code.25,30 For all structures calculated in this paper, we have used supercells with a total sample thickness of 10 nm. The convergence and collection angles for the simulation are in agreement with the experimental values. Geometric Phase Analysis Calculations. The STEM− HAADF images were subjected to strain analysis using the GPA phase plug-in for DigitalMicrograph based on the original algorithm by Hytch.29 A cosine mask with g/4 (g = 101̅) was chosen in Fourier space to calculate the phase map and strain field of the HAADF images. DFT Calculations. The DFT calculations were conducted by the Vienna Ab Initio Simulation Package31 with the frozen-core projector-augmented-wave method.32 The spin-polarized Perdew, Burke, and Ernzerhof generalized gradient approximation33 was employed for the exchange-correlation energy. Specifically, Sn was described by 14 (4d105s25p2) valence electrons. An on-site Coulomb correction of the Sn 5d orbital with U − J = 3.5 eV was adopted following a previous study,34 and a cutoff energy of 500 eV was employed for the plane-wave expansion. The (101̅) TB structure was modeled by a supercell with four atomic layers on each side, where twin plane consists of Sn atoms shared by the two domains. The Sn sublattices are mirror symmetric, while the oxygen atoms around Sn are not (1/2 < 111> displacement from mirror-symmetric positions). The supercell has dimensions of a = 11.36 Å, b = 9.30 Å, c = 20.34 Å (α = β = γ = 90°) with 64 Sn and 128 O atoms. To simulate the structural deformation induced by Li intercalation, the experimentally observed nanowire with TB was modeled by a symmetric slab with five Sn layers on each side of the twin plane with the surface terminated by O atoms. The vacuum thickness along the z-axis is about 11 Å. Gamma-centered kpoint meshes of 4 × 5 × 2 and 4 × 4 × 1 are used to integrate the Brillouin zone without and with TB, respectively.
In summary, we have studied the lithiation behavior of SnO2 nanowires containing (101)̅ TBs. It has been found that the lattice near the (101̅) TB can be easily intercalated by lithium ions in contrast to the neighboring lattices. This is consistent with DFT simulations, which show that the lithium ions prefer to stay near the (101)̅ TB and the formation of Sn vacancies is energetically favorable at the (101̅) TB. Both the experimental and simulated results indicate that the lattice near the TB gains strain due to the favorable intercalation of lithium atoms. This work offers important insights for smart design of high-rate rechargeable batteries by engineering the microstructure of electrode materials with TBs and other heterostructures. Methods. Sample Preparation. The SnO2 nanowires were synthesized via a vapor transport method catalyzed by gold nanoparticles. High-purity Sn powder (Alfa Aldrich, 99.9%) was heated in a tube furnace as a source with its container partially covered by the Au-coated Si wafer. SnO2 nanowires without TB were obtained from an hour growth at 800 °C with 250 sccm Ar flow, and the TB-SnO2 nanowires (Figure S3, Supporting Information) were grown at 900 °C under an ambient flow of mixed gas (Ar:O2, 250:5 sccm). Microscopy. The experiments were carried out inside an aberration-corrected JEOL JEM-ARM200CF STEM equipped with a 200 keV Schottky cold-field emission gun, HAADF, LAADF, and ABF detectors as well a postcolumn Gatan Enfina EELS spectrometer. A 22 mrad probe convergence angle was used for all the images and spectra. The HAADF, LAADF, and ABF images were acquired at 90−370 mrad, 40−160 mrad, and 11−22 mrad, respectively. The EELS spectra were obtained with a 45 mrad collection angle. In Situ Experiments. A nanoscale electrochemical open cell was adopted inside the TEM, which is achieved via a Nanofactory TEM-STM holder. A single SnO2 nanowire is attached by conductive epoxy to a gold wire to be the working electrode, and Li metal is attached to a tungsten tip to be the counter electrode. This step is finished inside a glovebox filled with Ar. Transportation of the sample from the glovebox to the TEM will result in a naturally grown Li2O layer on the surface of the Li metal, which acts as a solid-state electrolyte to E
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
Nano Letters
■
ASSOCIATED CONTENT
■
AUTHOR INFORMATION
Letter
(19) Oshima, Y.; Sawada, H.; Hosokawa, F.; Okunishi, E.; Kaneyama, T.; Kondo, Y.; Niitaka, S.; Takagi, H.; Tanishiro, Y.; Takayanagi, K. J. Electron Microsc. 2010, 59, 457−461. (20) Gu, L.; Zhu, C.; Li, H.; Yu, Y.; Li, C.; Tsukimoto, S.; Maier, J.; Ikuhara, Y. J. Am. Chem. Soc. 2011, 133, 4661−4663. (21) Ishikawa, R.; Okunishi, E.; Sawada, H.; Kondo, Y.; Hosokawa, F.; Abe, E. Nat. Mater. 2011, 10, 278−281. (22) Iwanaga, H.; Egashira, M.; Suzuki, K.; Ichihara, M.; Takeuchi, S. Philos. Mag. A 1988, 58, 683−690. (23) Hartel, P.; Rose, H.; Dinges, C. Ultramicroscopy 1996, 63, 93− 114. (24) Lee, W.-Y.; Bristowe, P.; Gao, Y.; Merkle, K. Philos. Mag. Lett. 1993, 68, 309−314. (25) Kirkland, E. J.; Loane, R. F.; Silcox, J. Ultramicroscopy 1987, 23, 77−96. (26) Yu, Z.; Muller, D. A.; Silcox, J. J. Appl. Phys. 2004, 95, 3362− 3371. (27) Phillips, P. J.; De Graef, M.; Kovarik, L.; Agrawal, A.; Windl, W.; Mills, M. Ultramicroscopy 2012, 116, 47−55. (28) Muller, D. A.; Nakagawa, N.; Ohtomo, A.; Grazul, J. L.; Hwang, H. Y. Nature 2004, 430, 657−661. (29) Hÿtch, M.; Snoeck, E.; Kilaas, R. Ultramicroscopy 1998, 74, 131−146. (30) Hillyard, S.; Silcox, J. Ultramicroscopy 1995, 58, 6−17. (31) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I. J. Phys.: Condens. Matter 2009, 21, 395502. (32) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758. (33) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865. (34) Singh, A. K.; Janotti, A.; Scheffler, M.; Van de Walle, C. G. Phys. Rev. Lett. 2008, 101, 055502.
* Supporting Information S
Additional figures, movies, and experimental details. This material is available free of charge via the Internet at http:// pubs.acs.org. Corresponding Author
*E-mail:
[email protected]. Author Contributions
A.N. carried out the TEM experiments and made the data analysis under the direction of R.S.-Y. and R.F.K. R.S.-Y. initiated the project and created the experimental protocols. L.Y.G., Y.C.C., and U.S. performed DFT calculations. F.M. provided the strain analysis. Q.L. and H.T.W. supplied the samples. Y.F.Y. helped to conduct the EELS analysis. A.N. and R.S.-Y. wrote the manuscript, and all the authors contributed to the discussion and revision of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS R.S.-Y. acknowledges the financial support from the National Science Foundation (Awards No. CMMI-1200383 and DMR1410560) and the American Chemical Society-Petroleum Research Fund (Award No. 51458-ND10). The acquisition of the UIC JEOL JEM-ARM200CF is supported by an MRI-R2 grant from the National Science Foundation (Grant No. DMR0959470). Support from the UIC Research Resources Center is also acknowledged. Theoretical simulations reported in this publication were supported by the King Abdullah University of Science and Technology (KAUST).
■
REFERENCES
(1) Heitjans, P.; Indris, S. J. Phys.: Condens. Matter 2003, 15, R1257. (2) Maier, J. Nat. Mater. 2005, 4, 805−815. (3) Sata, N.; Eberman, K.; Eberl, K.; Maier, J. Nature 2000, 408, 946−949. (4) Tuller, H. L. Solid State Ionics 2000, 131, 143−157. (5) Pratt, A.; Lari, L.; Hovorka, O.; Shah, A.; Woffinden, C.; Tear, S. P.; Binns, C.; Kröger, R. Nat. Mater. 2013, 13, 26−30. (6) Beaulieu, L.; Larcher, D.; Dunlap, R.; Dahn, J. J. Electrochem. Soc. 2000, 147, 3206−3212. (7) Maier, J. Solid State Ionics 2003, 157, 327−334. (8) Jamnik, J.; Maier, J. Phys. Chem. Chem. Phys. 2003, 5, 5215−5220. (9) Balke, N.; Jesse, S.; Morozovska, A.; Eliseev, E.; Chung, D.; Kim, Y.; Adamczyk, L.; Garcia, R.; Dudney, N.; Kalinin, S. Nat. Nanotechnol. 2010, 5, 749−754. (10) Zhu, X.; Ong, C. S.; Xu, X.; Hu, B.; Shang, J.; Yang, H.; Katlakunta, S.; Liu, Y.; Chen, X.; Pan, L. Sci. Rep. 2013, 3, 1084. (11) Lu, K.; Lu, L.; Suresh, S. Science 2009, 324, 349−352. (12) Lu, L.; Shen, Y.; Chen, X.; Qian, L.; Lu, K. Science 2004, 304, 422−426. (13) Shilo, D.; Ravichandran, G.; Bhattacharya, K. Nat. Mater. 2004, 3, 453−457. (14) Calleja, M.; Dove, M. T.; Salje, E. K. J. Phys.: Condens. Matter 2003, 15, 2301. (15) Aird, A.; Salje, E. Eur. Phys. J. B 2000, 15, 205−210. (16) Moriwake, H.; Kuwabara, A.; Fisher, C. A.; Huang, R.; Hitosugi, T.; Ikuhara, Y. H.; Oki, H.; Ikuhara, Y. Adv. Mater. 2013, 25, 618−622. (17) Nie, A.; Gan, L.-Y.; Cheng, Y.; Asayesh-Ardakani, H.; Li, Q.; Dong, C.; Tao, R.; Mashayek, F.; Wang, H.-T.; Schwingenschlögl, U. ACS Nano 2013, 7, 6203−6211. (18) Klie, R. F.; Qiao, Q.; Paulauskas, T.; Ramasse, Q.; Oxley, M. P.; Idrobo, J. Phys. Rev. B 2012, 85, 054106. F
dx.doi.org/10.1021/nl504087z | Nano Lett. XXXX, XXX, XXX−XXX
