LETTERS b e a-Fes7(001)Au(111) mS· From Au@ol154u o Initial (112) Crysta After 5 cycles 20(deg) Figure 4 Structural evolution of Si NWs during lithiation. a, XRD patterns of Si NWs before electrochemical cycling the first charg and after five cycles. b-e, TEM data for Si NWs at different stages of the first charge. b, A single-crystalline, pristine Si Nw before electrochemical cycling. The SAED spots inset and HRTEM lattice fringes(bottom) are from the Si 1/3(224) planes. c, Nw charged to 100 mv showing a Si crystalline core and the beginning of the formation of a Li Si amorphous shell. The HREM (bottom) shows an enlarged view of the region inside the box. d, Dark-field image of a NW charged to 50 mV showing an amorphous Li si wire with crystalline Si grains(bright regions) in the core. The spotty rings in the SAED (inset) are from crystalline Si. The HRTEM (bottom) shows some Si crystal grains embedded in the amorphous wire. e, A NW charged to 10 mv is completely amorphous Lia. Si. The SAED (top) shows diffuse rings characteristic of an amorphous material showed diffraction peaks associated with Si, a-FeSi,, Au(the Si Nw (li si alloy) were also observed. Li ions must diffuse radially into catalyst)and stainless steel(SS). The ae-FeSi2 forms at the interface the NW from the electrolyte, resulting in the core-shell phase between the SS and the Si wires during the high temperature(530 distribution. The reason for phase distribution along the length is C)NW growth process. The a-FeSi, was not found to appreciably not yet understood. At 50 mv, the Si Nw became mostly react with Li during electrochemical cycling, although a small amorphous with some crystalline Si regions embedded inside the amount of reaction has been reported2. After Si NWs were core, as seen from the dark-field image and HRTEM(Fig. 4d) charged to 150 mV, the higher angle Si peaks disappeared. Only AED showed spotty rings representative of a polycrystalline the Si(111)peak was still visible, but its intensity was greatly and diffuse rings for the amorphous phase. At 10 mV decreased. This is consistent with the disappearance of the 4e), all of the Si had changed to amorphous Li,Si,as nitial crystalline Si and the start of the formation of amorphous indicated by the amorphous rings in the SAED. These TEM Li, Si. The four broad peaks that appeared in the lower angles observations were consistent with the XRD results(Fig. 4a)and re due to the formation of LiisAug(see Supplementary voltage charging curves( Fig 2b) Information, Fig. S5). At 100 mV, the pure Au peaks disappeared, indicating that the Au had completely reacted with METHODS Li. The Si(111) peak was very weak at 100 mV, and disappeared completely at 50 mV. It appears that Si NWs Si NWs were synthesized using the VLS process on stainless steel substrates using remain amorphous after the first charge, consistent with the Au catalyst. The electrochemical properties were evaluated under an argon non-flat voltage charging/discharging curve in Fig. 2b. This atmosphere by both cyclic voltammetry and galvanostatic cycling in a three contrasts, however, with other studies on Si electrodes 25, 26, which electrode configuration, with the Si NWs on the stainless steel substrate as the have reported the formation of crystalline, Li, z Si at potentials working electrode and Li foil as both reference and counter-electrodes. No less than 30-60 mv. In situ XRD studies have determined binders or conducting carbon were used. The charge capacity referred to here is he total charge inserted into the Si Nw, per mass unit, during Li insertion, that this crystalline phase only forms at <50 mv for films whereas the discharge capacity is the total charge removed during Liextractiot hicker than 2 um(ref. 27). We did not observe this to be For electrical characterization, single Si Nw devices were contacted with metal the case in our Si NWs, most likely because of their shape electrodes by electron-beam lithography or focused-ion beam deposition.For nd small dimensions more detailed descriptions of NW synthesis, TEM and XRD characterization, The local structural features of Si NWs during the first Li electrochemical testing, and device fabrication, see the Supplementary insertion were studied with transmission electron microscopy Information. (TEM)and selected area electron diffraction(SAED). The as- grown Si NWs were found to be single-crystalline. Figure 4b Received 23 July 2007; accepted 14 November 2007; published 16 December 2007. shows an example of a typical Si Nw with a(112) growth direction". Figure 4c shows a Si Nw with a(112)growth direction that was charged to 100 mV. In this case there were two pistoia G. Lithium Batteries Scerce arf phases present, as expected from the voltage profile. Both A, Lesh, G. C. Huggins, R. A. All-solid lithium electrodes with mixed-conductor rystalline and amorphous phases were clearly seen. The distribution of the two phases was observed both across the Shodai, T, Okada, S, Tobishima, S. Yamaki. I Study of Li,_M N(M: Co, Ni or Cu)system for use the length. The SAED showed the spot pattern for the crystalline as negative-electrode materia s, Dupont. L Tarascon, L-M. 86-88, 785-789(196) -a ide diameter (a crystalline core and an amorphous shell) and along 4 poirot, P Laruelle, S asavajjula, U, Wang. C. Appleby. A. I. Nano-and bulk phase(Si), but weak diffuse rings from the amorphous pha for Sources 163,1003-1039(2007) @2008 Nature Publishing Group© 2008 Nature Publishing Group showed diffraction peaks associated with Si, a-FeSi2, Au (the Si NW catalyst) and stainless steel (SS). The a-FeSi2 forms at the interface between the SS and the Si wires during the high temperature (530 8C) NW growth process. The a-FeSi2 was not found to appreciably react with Li during electrochemical cycling, although a small amount of reaction has been reported24. After Si NWs were charged to 150 mV, the higher angle Si peaks disappeared. Only the Si(111) peak was still visible, but its intensity was greatly decreased. This is consistent with the disappearance of the initial crystalline Si and the start of the formation of amorphous LixSi. The four broad peaks that appeared in the lower angles are due to the formation of Li15Au4 (see Supplementary Information, Fig. S5). At 100 mV, the pure Au peaks disappeared, indicating that the Au had completely reacted with Li. The Si(111) peak was very weak at 100 mV, and disappeared completely at 50 mV. It appears that Si NWs remain amorphous after the first charge, consistent with the non-flat voltage charging/discharging curve in Fig. 2b. This contrasts, however, with other studies on Si electrodes25,26, which have reported the formation of crystalline, Li3.75Si at potentials less than 30 –60 mV. In situ XRD studies have determined that this crystalline phase only forms at ,50 mV for films thicker than 2 mm (ref. 27). We did not observe this to be the case in our Si NWs, most likely because of their shape and small dimensions. The local structural features of Si NWs during the first Li insertion were studied with transmission electron microscopy (TEM) and selected area electron diffraction (SAED). The as￾grown Si NWs were found to be single-crystalline. Figure 4b shows an example of a typical Si NW with a k112l growth direction29. Figure 4c shows a Si NW with a k112l growth direction that was charged to 100 mV. In this case there were two phases present, as expected from the voltage profile. Both crystalline and amorphous phases were clearly seen. The distribution of the two phases was observed both across the diameter (a crystalline core and an amorphous shell) and along the length. The SAED showed the spot pattern for the crystalline phase (Si), but weak diffuse rings from the amorphous phase (LixSi alloy) were also observed. Li ions must diffuse radially into the NW from the electrolyte, resulting in the core– shell phase distribution. The reason for phase distribution along the length is not yet understood. At 50 mV, the Si NW became mostly amorphous with some crystalline Si regions embedded inside the core, as seen from the dark-field image and HRTEM (Fig. 4d). The SAED showed spotty rings representative of a polycrystalline sample and diffuse rings for the amorphous phase. At 10 mV (Fig. 4e), all of the Si had changed to amorphous Li4.4Si, as indicated by the amorphous rings in the SAED. These TEM observations were consistent with the XRD results (Fig. 4a) and voltage charging curves (Fig. 2b). METHODS Si NWs were synthesized using the VLS process on stainless steel substrates using Au catalyst. The electrochemical properties were evaluated under an argon atmosphere by both cyclic voltammetry and galvanostatic cycling in a three￾electrode configuration, with the Si NWs on the stainless steel substrate as the working electrode and Li foil as both reference and counter-electrodes. No binders or conducting carbon were used. The charge capacity referred to here is the total charge inserted into the Si NW, per mass unit, during Li insertion, whereas the discharge capacity is the total charge removed during Li extraction. For electrical characterization, single Si NW devices were contacted with metal electrodes by electron-beam lithography or focused-ion beam deposition. For more detailed descriptions of NW synthesis, TEM and XRD characterization, electrochemical testing, and device fabrication, see the Supplementary Information. Received 23 July 2007; accepted 14 November 2007; published 16 December 2007. References 1. Nazri, G.-A. & Pistoia, G. Lithium Batteries: Science and Technology (Kluwer Academic/Plenum, Boston, 2004). 2. Boukamp, B. A., Lesh, G. C. & Huggins, R. A. All-solid lithium electrodes with mixed-conductor matrix. J. Electrochem. Soc. 128, 725–729 (1981). 3. Shodai, T., Okada, S., Tobishima, S. & Yamaki, J. Study of Li3 – xMxN (M:Co, Ni or Cu) system for use as anode material in lithium rechargeable cells. Solid State Ionics 86–88, 785 –789 (1996). 4. Poizot, P., Laruelle, S., Grugeon, S., Dupont, L. & Tarascon, J.-M. Nano-sized transition-metal oxides as negative-electrode materials for lithium-ion batteries. Nature 407, 496–499 (2000). 5. Kasavajjula, U., Wang, C. & Appleby, A. J. Nano- and bulk-silicon-based insertion anodes for lithium-ion secondary cells. J. Power Sources 163, 1003 –1039 (2007). 50 nm 100 nm 2 nm 50 nm 2 nm 40 nm 3.3 Å Crystal Amorphous Crystal Amorphous 3.3 Å Crystal Amorphous Crystal Amorphous 5 nm Amorphous (111) (220) Crystal Intensity (a.u.) 2θ (deg) 20 60 80 100 40 α-FeSi2 (001) Si (111) Au(111) From SS ( Li15Au4) From α-FeSi2 From Si Initial 150 mV 100 mV 50 mV 10 mV After 5 cycles From Au ·112Ò Figure 4 Structural evolution of Si NWs during lithiation. a, XRD patterns of Si NWs before electrochemical cycling (initial), at different potentials during the first charge, and after five cycles. b–e, TEM data for Si NWs at different stages of the first charge. b, A single-crystalline, pristine Si NW before electrochemical cycling. The SAED spots (inset) and HRTEM lattice fringes (bottom) are from the Si 1/3(224) planes. c, NW charged to 100 mV showing a Si crystalline core and the beginning of the formation of a LixSi amorphous shell. The HREM (bottom) shows an enlarged view of the region inside the box. d, Dark-field image of a NW charged to 50 mV showing an amorphous LixSi wire with crystalline Si grains (bright regions) in the core. The spotty rings in the SAED (inset) are from crystalline Si. The HRTEM (bottom) shows some Si crystal grains embedded in the amorphous wire. e, A NW charged to 10 mV is completely amorphous Li4.4Si. The SAED (top) shows diffuse rings characteristic of an amorphous material. LETTERS 34 nature nanotechnology | VOL 3 | JANUARY 2008 | www.nature.com/naturenanotechnology