. Angewandte Communications DOI: 10.1002/anie.201310402

Lithium-Ion Batteries

In Situ Three-Dimensional Synchrotron X-Ray Nanotomography of the (De)lithiation Processes in Tin Anodes** Jiajun Wang, Yu-chen Karen Chen-Wiegart, and Jun Wang* Abstract: The three-dimensional quantitative analysis and nanometer-scale visualization of the microstructural evolutions of a tin electrode in a lithium-ion battery during cycling is described. Newly developed synchrotron X-ray nanotomography provided an invaluable tool. Severe microstructural changes occur during the first delithiation and the subsequent second lithiation, after which the particles reach a structural equilibrium with no further significant morphological changes. This reveals that initial delithiation and subsequent lithiation play a dominant role in the structural instability that yields mechanical degradation. This in situ 3D quantitative analysis and visualization of the microstructural evolution on the nanometer scale by synchrotron X-ray nanotomography should contribute to our understanding of energy materials and improve their synthetic processing.

L

ithium-ion batteries (LIBs) are based on insertion reaction chemistry, which introduces microstructural changes into host materials by lithiation and delithiation processes. The changes are particularly pronounced in some high-capacity anode materials, such as silicon- and tin-based anodes, leading to a large volume change, fracture, and pulverization, thereby reducing the battery capacity and cycle life.[1–3] To address the mechanical degradation, a fundamental understanding of the mechanisms of microstructural changes in the electrode as a function of cycling is required. In recent years, a few in situ studies have provided valuable information on the electrode behavior of LIBs.[4–10] However, most of the reported in situ methods are based on two-dimensional (2D) rather than 3D measurements.[8–11] Recently, X-ray tomography at micronscale resolution was used for tracking the degradation in lithium-ion batteries and provided valuable 3D images.[12] Herein, we present for the first time a non-destructive in situ 3D X-ray nanotomography method 1) to monitor the 3D microstructural changes in the electrode at the nanoscale, 2) to quantitatively analyze the 3D microstructural changes (including volume change, feature size, specific area, and [*] Dr. J. J. Wang,[+] Dr. Y. K. Chen-Wiegart,[+] Dr. J. Wang Photon Sciences Directorate, Brookhaven National Laboratory Building 744, Upton, NY (USA) E-mail: [email protected] [+] These authors contributed equally to this work. [**] We acknowledge Biqiong Wang and Prof. Xueliang Sun (Western University, Canada) for providing the Sn sample for this study. This work was supported by a Laboratory Directed Research and Development (LDRD) project at Brookhaven National Laboratory. Use of the NSLS was supported by the U.S. Department of Energy, Office of Basic Energy Science (DE-AC02-98CH10886). Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201310402.

4460

curvature) from the high-quality data and 3) to correlate the morphological changes with the electrochemical reactions and to evaluate the stress on the electrode particles that is induced by the lithiation and delithiation processes. Transmission X-ray microscopy (TXM), using a zone plate objective lens, has been applied to provide non-destructive full-field nanotomography.[13–16] The lack of strict constraints on the environment makes the technique ideally suited for in situ investigations. A newly developed transmission X-ray microscope, which operates at the X8C beamline of the National Synchrotron Light Source (NSLS) at Brookhaven National Laboratory (BNL), provided a large field of view (40 mm), 30 nm resolution, local tomography, and automated marker-free image acquisition and alignment.[17–20] An electrochemical cell that can fully represent a working battery but that is also compatible with the TXM working distance requirement is key for in situ TXM nanotomography measurements. To develop such a working cell is challenging. The detailed requirements, challenges, designs, and advantages of the electrochemical cell can be found in the Supporting Information. The principle of TXM for in situ 3D battery experiments is shown in Figure S1 in the Supporting Information. A series of 2D images (361 images) from 908 to +908 at each stage of the lithiation–delithiation process were taken to reconstruct the 3D structure (Figure S2). These 2D images clearly show an obvious volume change (expansion, shrinkage, cracks, and pulverization) for the alloying–dealloying reaction during the electrochemical cycling. The mechanical degradation is particularly remarkable for some curved and coarse particle surfaces because of high stress formation and release, as some voids were created there. Some mechanical surface degradation is not visible from other angles of view, indicating that the volume change of the tin particles is anisotropic. In other words, it is of great importance to track morphology changes of battery materials in situ and in 3D. In the electrochemical measurement, to maximize the lithiation–delithiation reaction, a very low current density of 10 mA g 1 was used for the charge–discharge cycling. After deducting the capacities from decomposition of the electrolyte and the contribution of the carbon paper, the first discharge capacity was determined to be 867.2 mA h g 1, which is close to the theoretical capacity of Li4.4Sn (993 mA h g 1).[21–22] The second discharge capacity was only 489 mA h g 1; this decrease is due to the mechanical degradation (Figure S3). After these initial two cycles, the Sn anode showed a stable discharge capacity and reversibility. The 3D morphological evolutions of the Sn particles during the first two electrochemical cycles are shown in Figure 1 a (for a movie of each 3D structure, see Movie S1).

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Angew. Chem. Int. Ed. 2014, 53, 4460 –4464

Angewandte

Chemie

chemical composition, the particles first undergo a nearly complete lithiation reaction, as shown by the homogeneous distribution of the normalized attenuation coefficient within the particles from a high X-ray attenuation coefficient (Sn) to a low one (LixSn). During the first delithiation process, the shrinkage and deformation that are due to the lithium ion extraction result in an inhomogeneous mass density distribution in the particle, leading to the higher attenuation value in the core, which is also revealed by the line profile in Figure 1 d. During the second cycle (iv to v), the selected particle shows negligible morphological changes, but the electrochemical reaction is still reversible, as confirmed by the obvious changes in attenuation and the homogeneous phase distribution. This result is consisFigure 1. Three-dimensional morphology information. a) 3D morphological evolution of Sn particles tent with the electrochemical during the first two lithiation–delithiation cycles (see also Movie S1). b) Pseudo cross-sectional images cycling measurements. of a single Sn particle during the first two cycles. The particle shows severe fracture and pulverization at the initial stage of cycling, but stays mechanically stable afterwards while the electrochemical The correlation of morpholreaction still proceeds reversibly. The color scale of this cross-sectional view corresponds to the ogy and chemical information normalized linear attenuation coefficient, which provides a direct visualization of the chemical clearly indicates that Sn anodes composition. The color of the surface mesh, which was generated from the segmented reconstruction reach a structural equilibrium volume, in Figure 1 a (the 3D view) was set to visually correspond to the color in Figure 1 b. after the mechanical degradation c) Normalized X-ray attenuation coefficient histogram for the lithiation process. d) Normalized X-ray stages during the initial electroattenuation coefficients of selected locations. chemical cycling. Overlaid 3D views for samples that underwent this process are shown in Figure 2. Despite volume expansion and initial cracking, after the first To gain further information, a quantitative analysis of the lithiation, the overall structure of the Sn particles had mainly 3D structure that is based on all of the particles (not on remained intact. However, severe pulverization and fracture individual particles) was performed with statistics (Figure 3; were observed for the first delithiation when lithium-ion see also Table S1). First, curvature analysis was carried out to extraction and volume shrinkage occurred. After the second explore the effect of the geometric characteristics of the lithiation, the particles reach a structural equilibrium, and no electrode particles on the stress that is induced by the significant changes occurred. Fracture and pulverization, on lithiation–delithiation process. It is known that high curvaone hand, favor high electrolyte permeation, which contribtures of the particles lead to high axial stress and shear utes to lithium-ion insertion and associated volume expansion stress.[23, 24] However, most previous relevant studies are based during the second lithiation step. On the other hand, they lead to collapse during the volume expansion that occurs during on simulation and theory and lack direct experimental the second lithiation; therefore, few lithium ions were evidence. Herein, the curvature distribution, which is also extracted during the second delithiation process. The chemknown as the interfacial shape distribution (ISD),[25] was ical composition of each particle could be quantified, as the calculated from the surface of all of the particles to quantify tomography reconstruction recovers the linear attenuation the 3D surface shape of the electrode particles. Representing coefficient, which is directly linked to the composition of the the convex and concave features, the distributions of the two chemical species. The 2D chemical-phase maps at the pseudo principal curvatures of the Sn/LixSn particles and the shape cross sections of selected particles were obtained by analyzing evolution of the particles, which is based on a statistical the attenuation coefficient distribution of the Sn and LixSn evaluation of all particles, for the first two cycles are shown in Figure 3 a. The number of concave features increased signifiphases (Figure 1 b). A decrease of the X-ray attenuation cantly after the first lithiation. We propose that this high stress reveals the lithiation reaction from Sn to LixSn (Figure 1 c). led to fracture and pulverization during the first delithiation. During the first cycle (ii to iii), the cross-section images clearly Additional concave features after the first delithiation further reveal significant morphological deformation and pulverizainduced structural instability during the second lithiation. tion, particularly for the first delithiation. In terms of the Angew. Chem. Int. Ed. 2014, 53, 4460 –4464

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.angewandte.org

4461

. Angewandte Communications

Figure 2. Overlaid 3D views of the sample during the first two cycles. a) The fresh electrode (yellow) and the electrode after the first lithiation (red). b) The fresh electrode (yellow) and the electrode after the first delithiation (green). c) The fresh electrode (yellow) and the electrode after the second lithiation (purple). d) The electrode after the second lithiation (purple) and the electrode after the second delithiation (gray).

After the severe structural changes during the first delithiation and the second lithiation, the shape of the particles changed significantly. Although the concave features display no obvious change, the convex features dropped by 50 %, which helped the particles to reach a dynamic structural stability. The overall surface area increased by a factor of approximately 1.6 after the first lithiation and maintained a similar value during the first delithiation. This trend was repeated in the second lithiation and delithiation processes. The specific area (surface area per unit volume), a typical microstructural parameter that is inversely proportional to the feature size, escalated by 32 % after the first delithiation compared to the first lithiation, and by 33 % after the second lithiation relative to the first delithiation. This indicates that a considerable amount of particles underwent fracture and pulverization after the first delithiation and the second lithiation. The relatively small increase of only approximately 2 % after the second delithiation implies that no significant changes occurred at that point. An overall feature size distribution at different cycling stages is shown in Figure 3 b. No obvious change has taken place between the second lithiation and delithiation, and an averaged feature size of 500 nm seems to be optimal for the stability of the system. The analyzed 3D structural parameters are consistent with the 3D images discussed above. These results suggest that better control of the initial electrochemical cycle to minimize fracture and pulverization is critical for retaining electrode performance in the following cycles. The overall volume expansion (Figure 3 c and Table S1) amounts to 159 % after the first lithiation, which is less than

Figure 3. 3D quantitative analysis. a) Curvature distributions at different lithiation–delithiation stage. The amount of the concave feature increases significantly after the first lithiation, from 3.4 % in the fresh sample to 8.7 % in the sample after the first lithiation. This contributes to a drastic morphological change in the delithiation stage that ensues. b) Feature size distribution of the electrode during electrochemical cycling. c) Quantitative volume change of the electrodes during electrochemical cycling. d) Surface area and specific surface area for all stages of the electrochemical cycling.

4462

www.angewandte.org

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Angew. Chem. Int. Ed. 2014, 53, 4460 –4464

Angewandte

Chemie

Figure 4. Detailed surface morphology evolution and internal microstructural changes of an individual particle. a) Surface morphology evolution of the particle. b) Internal microstructure of the same particle with a cut-away view. c) Cross-sectional image showing the chemical phase distribution (see also Movie S2). d) Normalized attenuation coefficient histogram of the cross section during the first cycle. e) Line scanning profile of the normalized attenuation coefficient for the cross section. f) Statistical information on the size-dependent fracture within Sn particles.

the theoretical value of 359 % for a full lithiation to Li4.4Sn.[21, 22] The low volume expansion can be attributed to a carbon coating that acts as a buffering layer on the Sn particles (see the Supporting Information). Along with volume expansion, the surface area (Figure 3 d) also shows a tendency to increase during each lithiation process, but remains constant during each following delithiation. The surface area during each lithiation/delithiation process is subject to the combined influence of volume change and pulverization. If no fracture occurs, volume expansion will lead to an increase in the surface area, whereas a decrease in the volume causes the surface area to shrink. On the other hand, pulverization and fracture result in an increase in the surface area. As a result, during the delithiation process, these two opposing factors (volume shrinkage and fracture) keep the surface area constant. Therefore, the specific surface area, the surface area of a material per unit of volume, is an appropriate parameter to quantify the fracture degree of a material, as opposed to the absolute change in surface area, which involves mixed effects from volume change and electrode fracture. The rapid increase in specific surface area (Figure 3 d) during the first delithiation and the second lithiation further reveals the dominant roles of these two stages in the fracture of the material. The experimental 3D structural parameters that are listed in Table S1 also provide insight for accurately modeling and simulating the electrochemical reaction mechanism. Angew. Chem. Int. Ed. 2014, 53, 4460 –4464

The considerable evolution of the surface morphology was confirmed by the 3D structure of a selected particle, in particular by a cut-away view and a pseudo cross-sectional view to reveal both surface morphology and changes in the chemical composition (Figure 4 a–c; see also Movie S2). This Sn particle shows a representative morphological change during the first cycle; a clear volume expansion during the lithiation and a severe morphology pulverization during the delithiation process were observed. The cross-sectional images in Figure 4 c show a pronounced attenuation contrast between core and shell, indicating a partial delithiation in this large particle, which was confirmed by the normalized attenuation coefficient (Figure 4 d) and the line profile of the normalized attenuation coefficient for this particle (Figure 4 e). When a Sn particle undergoes a lithiation process, the volume expands, which results in a larger area under the curve (fresh: 202.47 mm3 ; 1st lithiation: 315.90 mm3 ; 1st delithiation: 278.77 mm3 ; Figure 4 d). The position of the peak shifts towards a lower attenuation, which indicates a change in the chemical composition from Sn to a LixSn mixture. Similarly, during the delithiation process, the LixSn particles are transformed into Sn, accompanied by a decrease in volume and an increase in the attenuation of X-rays. However, for an incomplete delithiation reaction, the particle forms a core–shell structure instead of being transforming back into a single phase of Sn. Therefore, it exhibits a wider distribution after the incomplete delithiation (blue plot). The

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.angewandte.org

4463

. Angewandte Communications line profile of the normalized attenuation coefficient also reveals the same trend (Figure 4 e). In addition to this selected large particle, a size-dependent mechanical change is demonstrated in this work. Based on statistical analysis, the fracture degrees of individual particles were quantified and plotted against their individual radii to reveal the relationship between fracture and particle size. Here, the degree of fracture was measured by the increase in the specific area in comparison with a particle without fracture that had undergone the same volume change (see the Supporting Information for details). The fracture degree generally increases with an increase in particle size (Figure 4 f). To quantitatively understand the stress and the origin of fracture and pulverization during the delithiation process, we performed a finite element analysis (FEA) on a Li4.4Sn particle that had undergone delithiation and been partially transformed back into Sn (Figures S4 and S5). We calculated the stress field that is built within the particle as a result of the volume difference between Li4.4Sn and Sn when the delithiation front propagates into one third of the particle (radius distance). As shown in Figure S4, the stress field was calculated for three different geometric cases, namely particles that contain 1) a pre-existing feature with highly convex curvature (Figure S4 a), 2) a pre-existing feature with a highly concave curvature (Figure S4 b), or 3) a crack (Figure S4 c). The boundary conditions, materials parameters, and details of the model can be found in the Supporting Information. Convex and concave features both lead to high von Mises stress; the concentrated stress mostly developed on the tip of the convex feature, but a stress that is higher than the yield stress of the material is distributed over nearly the entire interface between the Li4.4Sn and Sn (delithiated Li4.4Sn shell) of the concave feature. This finding corroborates the 3D images, which had revealed that particles undergo more severe fracture and pulverization during the first delithiation. We propose that this observation is due to the fact that the lithiated particles develop remarkable concave features during the first lithiation. The FEA results also explain the significant fracture during the second lithiation, as the particles maintain a similar proportion of concave features after the first lithiation. The FEA results exhibit different effects of crack features. A significant deformation surrounding a crack was observed, possibly because the space around a crack allows for a free volume change during delithiation. This deformation does not preserve the sharp shape of the crack, but rather leads to its transformation into a curved pore. In conclusion, the in situ observation of 3D microstructural evolution at the nanometer scale offers a direct way to look inside the electrochemical reaction of batteries to better understand the mechanism of structural degradation, to guide the engineering and processing of advanced electrode materials, and to produce accurate 3D parameters for theoretical simulations. Herein, we studied a battery with a Sn anode, but the developed electrochemical cell and the X-ray nanotomography can be applied to both cathodic and anodic materials. This method also launches an opportunity to monitor the chemical states and the phase transformations in three dimensions during cycling. Although we discussed

4464

www.angewandte.org

a study of batteries, this method promises wide applications in energy storage, catalysis, materials, and biological science. Received: November 30, 2013 Revised: January 23, 2014 Published online: March 19, 2014

.

Keywords: electrochemistry · in situ imaging · lithium-ion batteries · nanotomography · transmission X-ray microscopy

[1] N. S. Choi, Z. Chen, S. A. Freunberger, X. Ji, Y. K. Sun, K. Amine, G. Yushin, L. F. Nazar, J. Cho, P. G. Bruce, Angew. Chem. 2012, 124, 10134 – 10166; Angew. Chem. Int. Ed. 2012, 51, 9994 – 10024. [2] M. S. Whittingham, Chem. Rev. 2004, 104, 4271 – 4301. [3] D. Wang, J. Yang, X. Li, D. Geng, R. Li, M. Cai, T.-K. Sham, X. Sun, Energy Environ. Sci. 2013, 6, 2900 – 2906. [4] J. Y. Huang, L. Zhong, C. M. Wang, J. P. Wullivan, W. Xu, L. Q. Zhang, S. X. Mao, N. S. Hudak, X. H. Liu, A. Subramanian, H. Fan, L. Qi, A. Kushima, J. Li, Science 2010, 330, 1515 – 1520. [5] R. Bhattacharyya, B. Key, H. Chen, A. S. Best, A. F. Hollenkamp, C. P. Grey, Nat. Mater. 2010, 9, 504 – 510. [6] D. Takamatsu, Y. Koyama, Y. Orikasa, S. Mori, T. Nakatsutsumi, T. Hirano, H. Tanida, H. Arai, Y. Uchimoto, Z. Ogumi, Angew. Chem. 2012, 124, 11765 – 11769; Angew. Chem. Int. Ed. 2012, 51, 11597 – 11601. [7] S. Chandrashekar, N. M. Trease, H. J. Chang, L. S. Du, C. P. Grey, A. Jerschow, Nat. Mater. 2012, 11, 311 – 315. [8] J. Cabana, L. Monconduit, D. Larcher, M. R. Palacn, Adv. Mater. 2010, 22, E170 – E192. [9] S.-C. Chao, Y.-C. Yen, Y.-F. Song, Y.-M. Chen, H.-C. Wu, N.-L. Wu, Electrochem. Commun. 2010, 12, 234 – 237. [10] S.-C. Chao, Y.-F. Song, C.-C. Wang, H.-S. Sheu, H.-C. Wu, N.-L. Wu, J. Phys. Chem. C 2011, 115, 22040 – 22047. [11] J. Nelson, S. Misra, Y. Yang, A. Jackson, Y. Liu, H. Wang, H. Dai, J. C. Andrews, Y. Cui, M. F. Toney, J. Am. Chem. Soc. 2012, 134, 6337 – 6343. [12] M. Ebner, F. Marone, M. Stampanoni, V. Wood, Science 2013, 342, 716 – 720. [13] A. Sakdinawat, D. Attwood, Nat. Photonics 2010, 4, 840 – 848. [14] J. Vila-Comamala, M. Wojcik, A. Diaz, M. Guizar-Sicairos, C. M. Kewish, S. Wang, C. David, J. Synchrotron Radiat. 2012, 19, 705 – 709. [15] G. E. Ice, J. D. Budai, J. W. L. Pang, Science 2011, 334, 1234 – 1239. [16] J. L. Shui, J. S. Okasinski, P. Kenesei, H. A. Dobbs, D. Zhao, J. D. Almer, D. J. Liu, Nat. Commun. 2013, 4, 2255. [17] J. J. Wang, Y. K. Chen-Wiegart, J. Wang, Chem. Commun. 2013, 49, 6480 – 6482. [18] J. Wang, Y. K. Chen, Q. Yuan, A. Tkachuk, C. Erdonmez, B. Hornberger, M. Feser, Appl. Phys. Lett. 2012, 100, 143107. [19] Y. K. Chen-Wiegart, J. S. Cronin, Q. Yuan, K. J. Yakal-Kremski, S. A. Barnett, J. Wang, J. Power Sources 2012, 218, 348 – 351. [20] Y. K. Chen-Wiegart, W. M. Harris, J. J. Lombardo, W. Chiu, J. Wang, Appl. Phys. Lett. 2012, 101, 253901. [21] I. A. Courtney, J. R. Dahn, J. Electrochem. Soc. 1997, 144, 2045 – 2052. [22] W.-M. Zhang, J. S. Hu, Y. G. Guo, S. F. Zheng, L. S. Zhong, W. G. Song, L. J. Wan, Adv. Mater. 2008, 20, 1160 – 1165. [23] Y. K. Chen-Wiegart, Z. Liu, K. T. Faber, S. A. Barnett, J. Wang, Electrochem. Commun. 2013, 28, 127 – 130. [24] C. Lim, B. Yan, L. Yin, L. Zhu, Electrochim. Acta 2012, 75, 279 – 287. [25] J. Alkemper, P. W. Voorhees, Acta Mater. 2001, 49, 897 – 902.

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Angew. Chem. Int. Ed. 2014, 53, 4460 –4464

Supporting Information  Wiley-VCH 2014 69451 Weinheim, Germany

In Situ Three-Dimensional Synchrotron X-Ray Nanotomography of the (De)lithiation Processes in Tin Anodes** Jiajun Wang, Yu-chen Karen Chen-Wiegart, and Jun Wang* ange_201310402_sm_miscellaneous_information.pdf ange_201310402_sm_movie_s1.avi ange_201310402_sm_movie_s2.avi

Supporting Information METHODS Growth of carbon coated Sn particles on carbon paper. The carbon coated Sn particles were synthesized by chemical vapor deposition (CVD) method. In a typical procedure, pure Sn powder precursor (99.8% purity) was loaded in a ceramic boat, which was the then placed at the middle of a quartz tube in a horizontal tube furnace. A piece of carbon paper (Toray Carbon Paper TGP-H-030) was placed beside Sn powder, which acted as a substrate for Sn particles deposition. Prior to the growth of Sn particles, argon was introduced into the quartz tube to eliminate the air. The reaction chamber was heated to 800 °C in 15 min under an atmosphere of flowing Ar and 2% ethylene (200 sccm). The furnace was kept at 800 °C for 2 h and then was cooled to room temperature. After the reaction, it was observed that gray darklike products (carbon coated Sn particles) were uniformly deposited on the surface of the carbon paper substrates. Assembly of the special home-designed battery. To develop such a working cell is very challenging because this cell must i) allow a 180-degree rotation without blocking x-ray beam; ii) be micron scale on the study material within the x-ray beam to meet the x-ray field of view (40×40 µm); iii) function normally as a working battery; and, iv) allow electrochemical measurement for correlating the microstructural changes with the electrochemical reaction stages. A widely used coin cell is insufficient because it blocks beam when it is rotated, leading to a very limited angle of rotation and producing an unusable 3D reconstruction. The supporting materials of the cell along the x-ray beam path must be highly transparent to allow good transmission of the x-ray from the studied electrode. In addition, properly sealing such a cell is critical to ensure that the cell can work normally and be stable for repeated cycling. The complexity of developing the cell has hindered the investigation of in situ 3D microstructural evolution using TXM. Our battery is a closed system using lithium metal and 1M LiPF6 in ethylene carbonate/diethyl carbonate (1:1) as counter electrode and electrolyte, respectively. The working electrode is the above Sn/CP (carbon paper), which is cut to trapezoid shape (~ 40 µm short side, 800 µm long side, and 1 cm height) under optical microscopy (LEICA DM4000M) to meet the field of view of our TXM. Based on Sn loading of 0.8 mg/cm2 from inductively coupled plasma atomic emission spectroscopy (ICP-AES) measurement, the working electrode (0.042 cm2) has 0.0336 mg Sn loading. It is noteworthy that no any additional conductive carbon, binder, current collector or supports (such as widely used copper foil in anode electrode) were used in preparation of the working electrode, so the effects of these inactive components on the electrode’s mechanical degradation during cycling can be avoided, allowing more accurately to observe and study the degradation behavior of the active material itself. The three battery components (Sn, electrolyte, Li metal) were fabricated in a quartz capillary (1 mm diameter) in an Arfilled glove box, and sealed with epoxy. The two gold rods (0.5 mm diameter) connected with the two electrodes, making a circuit with the external potentiostat. Similar to coin cells, our battery is closed system where electrode materials were immersed completely (not just tip contact in in situ TEM experiments) in conventional liquid electrolytes to allow reactions in all dimensions. Our micro-battery also allows performing various electrochemical characterizations and analysis such as cyclic voltagram and discharge/charge profile because the potentiostat is sensitive enough for µA scale current density. In addition, our robust battery can undergo considerable stability, and allow long time electrochemical cycling (over weeks) and x-ray characterization.

In situ Nano-tomgorpahy with Transsmission X-ray Microscopy. The assembled cell was then imaged using transmission x-ray microscopy at beamline X8C, National Synchrotron Light Source (Brookhaven National Laboratory, BNL). This newly developed transmission x-ray microscope (TXM) at BNL provides a markerless, automated tomography. Because of the unique capability of the local tomography character offered in the experiment, the 3D morphology of the volume of interest within the cell could be well reconstructed even though the particles were enclosed in the in situ electrochemical cell. For each electrochemical lithiation-delithiation stage, a nano-tomography dataset was collected with 8 keV x-rays, using 361 projections over an angular range of 180ºwith a field of view of 40 × 40 µm2 (with a 2k × 2k CCD camera binning 2 × 2 camera pixels into one output pixel). The pixel size was 38.9 nm. The in situ electrochemical measurements were performed on a versatile multichannel potentiostat (VMP3). Charge-discharge characteristics were therefore galvanostatically carried out at a potential range between 0.005 to 1.2 V (vs. Li/Li+) at room temperature.[2,3] At each lithiated (0.005V) and delithiated stage (1.2 V), a series of 2D images were recorded to reconstruct 3D morphology.

Calculation of Reversible Capacity and Determination of Li-ion Lithiation. We used 1.2 V as the upper cutoff potential.[2,3] Considering Sn is difficult to be fully lithiated to Li4.4Sn at high rates, we used an ultra-low current density of 10 mA/g to maximum the lithiation. The first galvanostatic cycle between 0.005 to 1.2 V was shown in Fig. S2. It was found that the irreversible capacity is 233.2 mAh/g (solid electrolyte interphase membrane, electrolyte’s decomposition itself and Sn catalyzing decomposing) at the first cycle. To remove the capability contribution of carbon paper, a pure carbon paper electrode without Sn particles was tested with the same condition (54 mAh/g, not shown here). After removing this irreversibile capability related to electrolyte’s decomposition (233.3 mAh/g) and carbon paper effect (54 mAh/g), the reversible charging capacity of Sn at the first lithiation is around 867.2 mAh/g, which is close to the theoretical capacity of Li4.4Sn (993.4 mAh/g).[2,3] 3D Morphological Analysis Method. A standard Filtered Back-projection Reconstruction algorithm was again used to reconstruct the 3D images. The reconstructed volumes were cylinders with 40 µm in both diameter and height. The volumes from different electrochemical cycle states were then registered using commercial software (Avizo, VSG, version 7). A median filter with a kernel size of 3×3×3 voxels was then applied to the original image for noise reduction. The Sn/LixSn and exterior regions (electrolyte and carbon fiber) were labeled via simple threshold segmentation. The histogram of the reconstruction images consist of two distinctive peaks for these two phases and therefore the threshold value can be chosen as the minimum value between the two peaks. A smoothed surface mesh of the Sn/LixSn particles was then generated from the segmented images also using Avizo with a constraint that preserves the particle volumes within the surface meshes. Various 3D parameters were then calculated from the segmented structure and surface meshes: particle feature size distribution, volume change, surface area, specific area, and curvature analysis. The volume change was calculated by voxel counting. The surface area was measured from the surface mesh. Specific area is defined as surface area per unit volume. It was calculated from dividing the surface area of the entire sample by the volume of the entire sample. It is a direct indication of the size change of the sample. A smaller particle of the same shape has larger specific area than a larger particle with the same shape. Therefore, the decrease of specific area indicates the morphological change such as fracture, cracking and pulverization which all lead to increase the surface area while the total volume remains constant. The reciprocal of the specific area was then a common parameter used to characterize the average feature size. The feature size distribution was calculated using customized written software (MatLab, R2011b, MathWorks) with the algorithms described elsewhere by Holzer et al. [4] The principal curvature calculations were carried out using commercial package (Avizo, v.7, VSG). The interfacial shape distribution (ISD) was then plotted using customized written software (MatLab, R2011b, MathWorks) with method developed by Voorhees et al. [5] In the ISD calculation, as the surface meshing in Avizo

results in a triangular mesh with tiles of various areas, an area weighting procedure is used when generating the probability map.

Analysis of Fracture Degree.The specific area (Sv) is a good indicator for the amount of fracture. For the same particle with a fixed volume, when fracture occurs, more surface is exposed, which leads to a higher surface area. Sv takes into account of the volume and can be considered as a ‘normalized’ surface area. For a ball-shape particle with radius r, Considering a volume expansion/shrinkage case for a ball-shape particle where no fracture occurs, the particle. The specific area change in percentage (dS = ΔSv/Sv1) Particle radius change from r1 to r2, where r2 = α*r1, where a is a constant (

)

(

)

Note that dSv (the specific area change in percentage at a no-fracture-case) is a constant, independent of the initial radius of the particle. However, when there is fracture accompanying the volume change, an additional new surface area A is generated: The new specific area

+ A’

As a result, the dSv’ deviates from the dSv where no fracture occurs: dSv’ = dSv + A’’, where A’’ > 0 When a particle fractures more severely, more additional surface area is generated and therefore dSv’ deviates more from the ‘perfect’ dSv, which is a constant. As a result, we can use the difference between the experimental dSv’ and the ‘no-fracture’ dSv to represent the amount of fracture. By our definition: Degree of fracture (%) = A” = (dSv’-dSv)/|dSv| When being lithiated, the particles undergo volume expansion. The experimental data show a volume expansion of 159% during lithiation. Assuming no fracture occurs and approximating the particles as a ball, this leads to αlithiation = 1.167. If no fracture occurs during lithiation, dSvlithiation = -0.429 For lithiation, degree of fracture (%) = (dSv’-(-0.429))/|-0.429| When being delithiated, the particles undergo volume shrinkage. The experimental data show a total volume change of 121% after delithiation, compared with the fresh particles. Again, assuming no fracture occurs and approximating the particles as a ball, this leads to αde-lithiation = 1.0656. If no fracture occurs during delithiation, dSvde-lithiation= - 0.185 For delithiation, degree of fracture (%) = (dSv’-(-0.185))/|-0.185|)

Finite Element Modeling Method. Finite Element Modeling (FEM) was used to simulate the delithiation-induced stresses and due to the volume shrinkage during delithiation of Li4.4Sn. The commercial ABAQUS software was used to generate the mesh and calculate the stress distribution. An illustration of the model and also the boundary conditions are shown in Fig. S4. The particle radius was chosen to be 4.2 µm, which was measured from a representative particle after 1st lithiation. The axial symmetry boundaries were introduced along the horizontal and vertical axes as marked in Fig. S4. Therefore although only a corner of the particle was modeled, it represents the full particle in 3D. In both convex and concave feature simulation, a convex/concave feature with its radius as 1/10 of the particle radius was introduced. The feature radius is 0.42 micron, also marked in Fig. S5 (a-b). The physical size and location of the crack was labeled in Fig. S5c. In all three cases (convex, concave and crack) the shell part of the particle (with distance of 1.4 μm as marked in Fig. S5) was then set to exhibit a volume shrinkage due to delithiation. The materials properties used in this model are listed in Table S2. The shell was assigned with the Sn properties and the core was assigned with the Li4.4Sn, except that the shrinkage happens on the shell part and therefore the shell exhibits the linear change coefficient of -0.119. The stress developed as a result of this volume shrinkage was then calculated.

d

e

CCD

f

Objective Zone Plate Pin hole

Condenser

Sample

X-ray beam

a

c

b

Figure S1. The principle of the transmission x-ray microscope (TXM) for the in situ 3D battery experiment. a, Schematic principle of TXM and in situ 3D cell setup. The tip of the working electrode should be less than 40×40µm to meet the field of view of the TXM.

d

e

f

o

o

Figure S2. Selected 2D TXM projections from -90 to 90 at each stage of the lithiation-delithition process. a, fresh sample, b, after 1st-lithiation, c, after 1st-delithiation, d, after 2nd lithiation and e, after 2nd delithiation. Note the significant changes during the 1st delithiation and 2nd lithiation on the representative particles.

900

1.2

current density: 10 mA/g

E/V vs. Li/Li+

b

0.9 0.6 0.3

Capacity (mAh/g)

800

a

700 600 500 400

0.0 0

200

400

600

800

1000

0

1200

2

4

6

8

10

Cycle number

Capacity (mAh/g)

c 1.2

Potential (V vs Li/Li+)

a

i

1.0

iii

v

0.8 0.6 0.4 0.2

iv

ii

0.0 0

500

1000

1500

2000

2500

3000

3500

Capacity (mAh/g)

Figure S3. The electrochemical pferformance of Sn anode materials. a, the first cycle of v of 0.005-1.2V. b, iv in a voltage range iii density of 10 mA/g i profileii of Sn anode at a current b discharge/charge the reversible discharge performance of the initial ten cycles. It is found that Sn anode shows a stable reversible capacity after the initial two cycles. c. the charge/discharge profile at the first two cycles.

c

i

iii

ii

1 µm

Sn

LixSn

v

iv

Sn

LixSn

1

Sn

0

Normalized attenuation (a.u.)

10 µm

Figure S4. Finite element analysis of a Li4.4Sn particle undergone delithiation when the delihiation front propagates into 1/3 of the particle in its radius distance. The von Mises Stress was shown with three different pre-existing features: (A) high convex curvature feature, (B) a high concave curvature feature, and (C) a crack. Note that the yield stress of the core Li4.4Sn (σy- Li4.4Sn) is 3.6 GPa while the yield stress of the shell Sn (σy-Sn) is 1.26 GPa. Von Mises Stress is a yielding criterion above which materials yield. All axial and shear stresses are taken into account in this one single stress parameter.

Axial symmetry

b

a

c

r = 0.42 micron Li4.4Sn  Sn Volume shrinkage

Li4.4Sn  Sn Volume shrinkage

r = 0.42 micron

Li4.4Sn  Sn

10°

0.9 μm

45° Li4.4Sn

Li4.4Sn

Li4.4Sn

Axial symmetry 1.4 μm fix

4.2 μm

Figure S5. Particle radius is 4.2 μm and de-lithiation of the particle with distance of 1.4 μm. a, with convex feature. b, with concave feature. c, crack existing as a result of lithiation: delithiation front exceeds the tip of the crack. Three boundary conditions were applied: one fix point and two axial symmetry.

Tables

fresh volume of Sn particles (nm3) absolute volume change of Sn (%) (compared to fresh) relative volume change(%) (after each half-cycle) surface area (nm2) specific area (nm-1) Sy specific area change(%) (compared to fresh) relative specific area change (%) (after each half-cycle) 1/specific area (nm), Sv-1* absolute feature size change (%) (compared to fresh) relative feature size change(%) (after each half-cycle)

4.85E+11

after 1st lithiation 7.72E+11

after 1st delithiation 5.87E+11

after 2nd lithiation 7.05E+11

after 2nddelithiation 6.88E+11

100

159

121

145

142

--

+59

-24

+20

-2

0.98E+09 0.00201

1.62E+09 0.00210

1.63E+09 0.00277

2.61E+09 0.00370

2.59E+09 0.00376

100

104

138

184

187

--

+4

+32

+33

+2

496.9

476.2

360.7

270.3

265.9

100

96

73

54

54

--

-4

-24

-25

-2

Table S1. Quantitative 3D morphological analysis of the electrodes undergone electrochemical cycling.

*specific area is defined as the surface area in per unit volume. While specific area is generally inversely proportional to the size, its reciprocal (1/specific area, Sv-1) is then a suitable parameter used to indicate the representative size in a structure. Note that the precise relationship between the Sv-1 and the feature size in a system depends on sample’s geometry. For instance, if the sample geometry can be represented by ball-shape with a radius r, then the Sv-1is:

Note that there is a geometric factor of 1/3, which is due to the ball-geometry.

Material

Young’s Modulus

Poisson's ratio linear change coefficient

(GPa) Sn

42

0.36

--

Li4.4Sn

120

0.25

- 0.119

Table S2. Materials Properties used in the finite element modeling

Movie S1. An In situ TXM movie showing the 3D morphological evolution of Sn particles during the first two lithiation-delithiation cycles.

Movie S2. An In situ TXM movie showing a detailed surface morphology evolution and internall microstructural change of an individual particle during the 1st lithiation-delithiation process.

References

1. R. Li, X. Sun, X. Zhou, M. Cai, X. Sun, J. Phys. Chem. C 2007, 111, 9130-9135. 2. S. D. Beattie, T. Hatchard, A. Bonakdarpour, K. C. Hewitt, J. R. Dahn, J. Electrochem. Soc. 2003, 150, A701-A705. 3. I. A. Courtney, J. R. Dahn J. Electrochem. Soc. 1997, 144, 2045-2052. 4. B. Munch, L. Holzer, J. Am. Ceram. Soc. 2008, 91(12), 4059-4067. 5. J. Alkemper, P. W. Voorhees, Acta Mater. 2001, 49, 897-902.