WO2005076389A9 - Self-contained, alloy type, thin film anodes for lithium-ion batteries - Google Patents

Abstract

An electrically conductive substrate is provided upon which one or more layers of active materials, i.e. materials that act in a reversible manner with the charge carriers of an electrolyte, are formed. Prior to formation of these layers, the substrate is patterned so as to contain cavities. The active materials are then formed in one or more layers so that the total cumulative height of the layers is less than the entire depth of each of the cavities. The remaining 'unfilled' depth of the cavities provides suitable room for volumetric expansion and contraction of the active materials, resulting in little or no net volume change in the anode at any point during the cycling of the batteries.

Description

SELF-CONTAINED, ALLOY TYPE, THIN FILM ANODES FOR LITHIUM-ION BATTERIES

CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims priority from co-pending serial no. 60/532,318 entitled Self- contained Alloy Type Thin Film Anodes for Lithium-Ion Batteries filed December 23, 2003, the entirety of which is hereby incorporated by reference.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH [0002] This invention was made with government support under Grant No. NOOO 14-00-0516 awarded by the Office of Naval Research and Grant No. CTS. 0000563 awarded by the National Science Foundation. The government has certain rights in this invention.

BACKGROUND OF THE INVENTION [0003] The emergence of a myriad of lightweight, miniature consumer electronic devices along with promising advances made in hybrid motor vehicles have increased the demand for longer life, higher energy density rechargeable batteries. Since its commercialization in 1991 by the Sony Corporation, the rechargeable lithium-ion (Li-ion) battery has been the most promising in terms of fulfilling the diverse needs ofthe device manufacturers. Although Li- ion batteries are at the forefront of battery technology, there is an ever-growing need to improve the energy density as devices shrink in size to dimensions unimaginable even five years ago.

[0004] Although carbon based materials have been the preferred anode used in a majority of commercial lithium-ion batteries since Sony first marketed the lithium-ion battery, intense research is currently being directed towards identifying higher capacity anode materials for the next-generation lightweight and compact lithium-ion batteries. Original research conducted on bulk alloy type anodes comprising elemental tin or silicon have shown their potential for high capacity, while also identifying the undesirable volume expansion and contraction resulting in crumbling and mechanical degradation. M. Winter and J.O. Besenhard, Electrochim. Acta, 45, 31 (1999). In an effort to reduce the deleterious effects of decrepitation, multi-component tin containing oxides were used which yielded substantial improvements in cyclability. Idota et al., Science, 276, 1395 (1997). However, tin based oxide materials exhibit irreversible capacities as high as ~50% ofthe total capacity. R. A. Huggins, J. Power Sources, 81-82, 13 (1999). To Teduce the irreversible capacity, researchers have used non-oxide based inactive materials to withstand the volumetric stresses ofthe active phases. IS. Kim, G.E. Blomgren and P.N. Xumta, "Si/SiC Nanocomposite Anodes Synthesized using High-energy Mechanical Milling", J. of Power Sources, 130 (2004) 275- 280; LS. Kim and Prashant N. Kumta, "High Capacity Si/C Nanocomposite Anodes for Li- ion Batteries", J. of Power Sources 136 (2004) 145-149; LS. Kim, G.E. Blomgren, and P.N. Kumta, "Nanostructured Si/TiB₂ Composite Anodes for Li-Ion Batteries", Electrochemical and Solid State Letters 6 (8) A157-A161 (2003); I.S. Kim, G.E. Blomgren and P.N. Kumta, "Sn/C Composite Anodes for Li-ion Batteries", Electrochemical and Solid State Letters Vol. 7, (3) A44-A48 (2004); LS. Kim, P.N. Kumta, and G.E. Blomgren, 'Si/TiN Nanocomposites: New Anode Materials for Li-ion Batteries', Electrochemical and Solid-State Letters (2000) 3 (11) 493-496. Results of metallic alloys containing active and inactive phases have also shown promise. M. Winter and J. O. Besenhard, Electrochim. Acta, 45, 31 (1999); R.A. Huggins, J. Power Sources, 81-82, 13 (1999); W.J. Weydanz, M. Wohlfahrt-Mehrens, and R.A. Huggins, J. Power Sources, 81-82, 236 (1999). Other researchers recognized the importance of small grain size along with a ductile inactive phase to stabilize the capacity of alloy type anodes. O. Mao, R.L. Turner, LA. Courtney, B.D. Fredericksen, M.I. Buckett, L.J. Krause and J.R. Dahn, Electrochem. Solid-State lett, 2(1), 3 (1999). Pyrolized polymers were also used to produce nano-dispersed silicon in a carbon matrix yielding reversible specific capacities up to 600 mAh/g. A.M. Wilson, J.N. Reimers, E.W. Fuller, and J.R. Dahn, Solid State Ionics, 74249 (1994); W. Xing, A.M. ^"Wilson, G. Zank, and J.R. Dahn, Solid State Ionics, 93, 239 (1997); A.M. Wilson, W. Xing, G. Zank, B. Yates, and J.R. Dahn, Solid State Ionics, 100, 259 (1997). Recently, Song et al. have reported capacities as high as ~2200 mAh/g in nanocrystalline thin films of Mg₂Si 30 run thick deposited by pulsed laser deposition. S. Song, K.A. Striebel, R.P. Reade, G.A. Roberts, and E.J. Cairns, J. Electrochem. Soc, 150 (1), A121 (2003). Although these systems yield specific capacities greater than graphite (372 mAh/g), there have been no reports to date on the electrochemical response of materials or systems exhibiting near theoretical specific capacity of silicon (-4000 mAh g).

[0005] Lithium metal has been widely used as the anode material in lithium-ion batteries because lithium ion is the lightest cation (other than H⁺) and is a highly mobile species. Lithium metal is also highly desirable because it can deliver a high theoretical energy density (3860 mAh/g) compared to graphite, the conventional anode material, LiCβ (372 mAh/g). Many attempts have been made to fabricate and commercialize rechargeable batteries containing Li metal anodes, including an attempt by Nippon Telephone and Telegraph (NTT) Corporation using a lithium metal anode and amorphous V₂O₅ cathode. J. Yamaki and S. Tobishima in Handbook of Battery Materials, ed. by J. Besenhard, Wiley-VCH, Weinheim, 1999; Y. Sakurai, S. Sugihara, M. Shibata and J. Yamaki, NTT ev., 7, 60 (1995). Unfortunately, cells containing lithium metal anodes have historically had many problematic issues leading to research of other non-lithium metal based anodes. Lithium metal anodes tend to form dendritic structures gradually after a repeated number of charge discharge cycles. The dendrites can cause a short circuit condition between the anode and cathode resulting in failure of the cell. The lithium dendrites can also he broken off during cycling and coated with an insulating solid-electrolyte interfacial (SEI) layer (E. Peled, J. Electrochem. Soc, 126, 2047 (1979)), resulting in "dead", inactive lithium. The "dead" lithium ultimately leads to poor cycling efficiencies. Safety is the other large problem encountered with lithium metal anodes. Heat can be generated in a lithium metal cell when it is abused. Some examples of heat generation occurring by thermal decomposition or reaction are (J. Yamaki and S. Tobishima in Handbook of Battery Materials, supra): 1) by a reaction between an anode and electrolyte, 2) by the thermal decomposition of an organic electrolyte, 3) by a reaction between a cathode and electrolyte, 4) by the thermal decomposition of the cathode, 5) by the thermal decomposition of the anode, 6) by an entropy change in a cathode active material (and an anode active material) and 7) by current passing through a cell with high internal resistance (I²R Joule heating). Although preventive safety steps (pressure vents, thermal fuses, current fuses, etc.) have been adopted by manufacturers (Id), the safety issues surrounding lithium metal anodes in rechargeable ^"batteries have prompted the need for development of alternative anode materials.

[0006] Starting in the early 1980's, metallic lithium was replaced by lithium insertion materials (M. Armand, in Materials for Advanced Batteries, ed. by D. Murphy, J. Broadhead and B. Steele, Plenum Press, New York, p. 145 (1980>; B. Scrosati, J. Electrochem. Soc, 139, 2776 (1992)) having lower redox potentials than the cathode, leading to the rocking chair cell, so called because both electrodes simultaneously undergo intercalation/de-intercalation. Amongst all the lithium insertion materials considered for use as anodes, lithiated carbons have been the most promising and hence most widely studied and used. The carbon based anodes have lower redox potentials than polymers, metal oxides, or chalcogenides. The carbon based anodes also have better dimensional staT->ility than lithium alloys. Furthermore, the carbons are relatively cheap and abundant. M. Winter and J. Besenhard in Handbook of Battery Materials, ed. by J. Besenhard, Wiley- VCH, Weinheim, 1999. [0007] The electrochemical response of carbon based anodes depends on the structure ofthe parent carbonaceous material, the type of electrolyte, and the interactions between the electrolyte and the anode. The quality and number of sites available for reversible lithium intercalation and de-intercalation depend on many factors including the crystallinity, texture, microstructure, and micromorphology of the parent carbonaceous material. J. Besenhard, in Progress in Intercalation Research, ed. by W. Muller-Warmuth and R. Schollhorn, Kluwer, Dordrecht, p. 457 (1994); J. Dahn, A. Sleigh, H. Shi, B. Way, W. ^"Weydanz, J. Reimers, Q. Zhong and U. von Sacken, in Lithium Batteries, New Materials, Developments, and Perspectives, ed. by G. Pistoia, Elsevier, Amsterdam, p. 1 (1994); D. Fauteux and R. Koksbang, J. Appl. Electrochem., 23, 1 (1993); K. Sawai, Y. Iwakoshi and T. Ohzuku, Sol. St. Ionics, 69, 273 (1994); A. Mabuchi, Tanso, 165, 298 (1994); R. Yazami, and M. Munshi, in Handbook of Solid State Batteries and Capacitors, ed. by M. Munshi, World Scientific, Singapore, p. 425 (1995); J. Dahn, A. Sleigh, H. Shi, J. Reimers, Q. Zhong and B. Way, Electrochim. Acta, 28, 1179 (1993); R. Yazami, in Lithium Batteries, New Materials, Developments, and Perspectives, ed. by G. Pistoia, Elsevier, Amsterdam, p. 1 (1994); O. Yamamoto, Y. Takeda, N. Imanishi and R. Kanno, in New Sealed Rechargeable Batteries and Supercapacitors, ed. by B. Barnett, E. Dowgiallo, G. Halpert, Y. M-atsude, Z. Takehara, The Electrochemical Society, Pennington, NJ, PV 93-23, p. 302 (19»93); R. Yazami and D. Guerard, J. Pow. Sources, 43-44, 39 (1993); N. Takami, A. Satoh, WL. Hara and T. Ohsaki, J. Electrochem. Soc, 142, 371 (1995); A. Satoh, N. Takami and T. OTisaki, Sol. St. Ionics, 80, 291 (1995); M. Endo, J. Nakamura, A. Emori, Y. Sasabe, K. Takeixchi and M. Inagaki, Mol Cryst. Liq. Cryst, 245, 171 (1994); R. Yazami, K. Zhaghib and M. Deschamps, Mol. Cryst Liq. Cryst., 245, 165 (1994); T. Zheng, J. Xue and J. Dahn, Chem. Mater., 8, 389 (1996); T. Zheng, Y. Liu, E. Fuller, S. Tseng, U. von Sacken and J. Dahn, J. Electrochem. Soc, 142 2581 (1995); O. Yamamoto, N. Imanishi, Y. Takeda and H. Kashiwagi, J. Pow. Sources, 54 271 (1995); G. Li, R. Xue, L. Chen and Y. Huang, J. Pow. Sources, 54, 271 (1995); J Yamaura, Y. Ohzaki, A. Morita, and A. Ohta, J. Pow. Sources, 43-44, 233 (1993); G. Li, Z Lu, B. Huang, H. Huang, R. Xue and L. Chen, Sol. St. Ionicsm, 81, 15 (1995); N. Takami, A, Sato and T. Ohsaki, in Lithium Batteries, ed. by S. Surampudi and V. Koch, The Electrochemical Society, Pennington, NJ, PV 93-24, p. 44 (1993). Most carbonaceous materials can be classified as graphitic or non-graphitic (disordered). [0008] Graphitic carbons contain sp² -hybridized carbon atoms that are arranged in a planar hexagonal fashion, such that 'graphene' layers are held together by weak van der Waals forces. The term graphite is actually reserved for carbons that have perfect AB (hexagonal graphite, more common) or perfect ABC (rhombohedral graphite, less common) layer stacking. A pure single polytype of graphite is difficult to obtain because the free energies associated with the AB and ABC polytypes are very close in value. The typical process of lithium intercalation into graphitic carbons is called 'staging' (J. Dahn, Phys. Rev. B., 44, 9170, (1991)) in which the ABA layers slide and form an AAA configuratio-n, as shown in Fig.l. Graphitic carbons have a theoretical maximum capacity of 372 mAh/g. However, the first charge cycle typically exceeds 372 mAh/g due to the SEI formation and corrosion-like reactions of Li_xC₆. M. Winter and J. Besenhard in Handbook of Battery Materials, ed. by J. Besenhard, Wiley-VCH, Weinheim, 1999.

[0009] Non-graphitic carbons are also potentially attractive anode materials for two reasons. First, the cross-linking in non-graphitic carbons between graphene layers tends to prevent co- intercalation of solvent molecules. J. Dah, A. Sleigh, H. Shi, B. Way, ^"W. Weydanz, J. Reimers, Q. Zhong and U. von Sacken, in Lithium Batteries, New Materials, Developments, and Perspectives, ed. by G. Pistoia, Elesevier, Amsterdam, p. 1 (1994); R. Fong, U. von Sacken and J. Dahn, J. Electrochem. Soc, 137, 2009 (1990); J. Besenhard, M. Winter, J. Yang and W. Biberarcher, J. Pow. Sources, 54, 228 (1995). Second, in comparison to graphite, non-graphitic carbons can provide more sites for lithium incorporation. As a result, higher capacities than 372 mAh/g are possible in non-graphitic carbons. Despite mild progress on high reversible capacity non-graphitic carbons, significant technical hurdles still exist such as high first cycle irreversible loss and loss of capacity upon further cycling. [0010] While carbon anodes have been used extensively in commercial batteries since the first commercialized lithium-ion battery, lithium alloy type anode materials have also been the target of many research studies. Lithium alloys were first considered as possible anode materials by researchers working on molten salt electrolyte batteries that operated at temperatures in excess of 400°C. R. Huggins in Handbook of Battery Materials, ed. by J. Besenhard, Wiley-VCH, Weinheim, 1999. Two major alloy systems were studied, the Li-Al (N. Yao, L. Heredy and R. Sauders, J. Electrochem. Soc, 118, 1039 (197 1); E. Gay, J. Electrochem. Soc, 123, 1591 (1976)) and Li-Si (S. Lai, J. Electrochem. Soc, 123, 1196 (1976); R. Sharma and R. Seefurth, J. Electrochem. Soc, 123, 1763 (1976); R.. Seefurth and R. Sharma, J. Electrochem. Soc, 124, 1207 (1976)) systems. Since then, lithium has been observed to electrochemically alloy at room temperature with other elements besides Al and Si, such as, Mg, Ca, Ge, Sn, Sb, Bi, As, Ag, Au, Pt, Cd, etc. M. Winter and J. Besenhard, Electrochim. Acta., 45, 31 (1999). All of these elements (M) form alloys with- lithium according to the following reaction: Li_xM => xLi⁺ + xe^~ + M [0011] One ofthe most appealing characteristics of several lithium alloy type anode systems is their high theoretical specific capacity (gravimetric and volumetric) compared to graphite. A second desirable characteristic is their low redox potential (typically less than 0.5 V versus Li/Li⁴), leading to a high open circuit voltage. Tables 1 and 2, shown below, give the specific capacities and plateau potentials of several candidate next-generation active materials for lithium-ion batteries. Table 1. Theoretical specific capacities of promising lithium alloy anodes.

[0012] Lithium alloy anode materials have not been commercialized due to the large scale volume expansion and contraction observed in these zintl phase alloys during cycling which results in crumbling and mechanical degradation. Id. Ln essence, the volume changes create high stresses within the lithium alloy anode material, eventually leading to fracture, pulverization, and subsequent loss of electrical contact between the pulverized particle and the rest of the electrode. After the initial loss of electrical contact, the particle remains virtually inactive for the rest ofthe useful life ofthe battery. On a large scale, the degradation can lead to a swift decline, even after only a few cycles.

Table 2. Plateau potentials of some promising binary lithium alloys (Li_yM) at 400°C.

[0013] Ln an effort to reduce the effects of decrepitation, multi-component tin containing oxides were identified by researchers at Fuji in 1997. These oxide based glasses undergo electrochemical reduction during the insertion of lithium ions resulting in the formation of nanosized elemental tin clusters and a lithiated oxide glass. Subsequent cycling leads to alloying of the reduced tin with lithium while the oxide glass essentially acts as a spectator enduring the large volumetric stresses generated. The tin oxide based nanocomposite glasses resulted in substantial improvements in cyclability. Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, and T. Miyasaka, Science, 276, 1395 (1997). However, the first cycle reduction of the oxide materials to form tin clusters results in irreversible capacities as high as ~50% ofthe total capacity. R.A. Huggins, J. Power Sources, 81-82, 13 (1999). To reduce the irreversible capacity, researchers have attempted to synthesize nanocomposite structures either ex-situ prior to insertion of lithium or in-situ during the insertion of lithium ions similar to the tin oxide glasses. In both cases, the electrochemically active material is contained within an electrochemically inactive but electrically conducting matrix that serves to withstand the large stresses generated, thereby acting as a 'glue' holding the active material. The only difference being that in the former case, the electrochemically inactive phase is thermodynamically and chemically stable with respect to the active species while in the latter, the matrix phase is generated by electrochemical reduction and phase separation. LS. Kim, G.E. Blomgren and P.N. Kumta, "Si/SiC Nanocomposite Anodes Synthesized using High-energy Mechanical Milling", J. of Power Sources, 130 (2004) 275-280; LS. Kim and Prashant N. Kumta, "High Capacity Si/C Nanocomposite Anodes for Li-ion Batteries", J. of Power Sources 136 (2004) 145-149; LS. Kim, G.E. Blomgren, and P.N. Kumta, "Nanostructured Si/TiB₂ Composite Anodes for Li-Ion Batteries", Electrochemical and Solid State Letters 6 (8) A157-A161 (2003); LS. Kim, G.E. Blomgren and P.N. Kumta, "Sn/C Composite Anodes for Li-ion Batteries", Electrochemical and Solid State Letters Vol. 7, (3) A44-A48 (2004); LS. Kim, P.N. Kumta, and G.E. Blomgren, 'Si/TiN Nanocomposites: New Anode Materials for Li-ion Batteries', Electrochemical and Solid-State Letters (2000) 3 (11) 493-496. O. Mao, R. Turner, I. Courteney, B. Fredericksen, M. Buckett, L. Krause and J. Dahn, Electrochem. and Sol.-St. Let., 2,3 (1999); O. Mao and J. Dahn, J. Electrochem. Soc, 146, 405 (1999); O. Mao and J. Dahn, J. Electrochem. Soc, 146, 414 (1999); O. Mao and J. Dahn, J. Electrochem. Soc, 146, 423 (1999); K. Kepler, J. Vaughey and M. Thackeray, Electrochem. and Sol.-St. Let., 2, 307 (1999); J. Vaughey, J. O'Hara and M. Thackeray, Electrochem. and Sol.-St. Let., 3, 13 (2000); L. Fransson, J. Vaughey, K. Edstrom and M. Thackeray, J. Electrochem. Soc, 150, A86 (2003).

[0014] The use of an electrochemically inactive matrix serves to withstand the volumetric strains experienced by the active phases. Results of several metallic alloy systems containing active and inactive phases have also demonstrated promise. M. Winter and J. Besenhard, Electrochim. Acta., 45, 31 (1999); R.A. Huggins, J. Power Sources, 81-82, 13 (1999); J. yang, M. Winter, and J.O. Besenhard, Solid State Ionics, 90, 281 (1996); R. Huggins, Solid State Ionics, 113-115, 57 (1998); W. Weydanz, M. Wohlfahrt-Mehrens and R. Huggins, J. Power Sources, 81-82, 237 (1999). Researchers have recognized the importance of small grain size along with a ductile inactive phase in order to stabilize the capacity of alloy type anodes. O. Mao, R. Turner, I. Courtney, B. Fredericksen, M. Buckett, L. Krause and J. Dahn, Electrochem. Solid-State Lett., 2(1), 3 (1999). Disordered carbon generated by pyrolysis of polymers was also used to produce nano-dispersed silicon in a carbon matrix yielding reversible specific capacities up to 600 mAh/g. A. Wilson, J. Reimers, E. Fuller, and J. Dah, Solid State Ionics, 74, 249 (1994); W. Xing, A. Wilson, G. Zank, and J. Dahn, Solid State Ionics, 93, 239 (1997); A. Wilson, W. Xing, G. Zank, B. Yates, and J. Dahn, Solid State Ionics, 100, 259 (1997). Similar research in our group at Carnegie Mellon University on tin particles dispersed within an amorphous carbon matrix has yielded reversible capacities as high as 500 mAh/g. LS. Kim, G.E. Blomgren and P.N. Kumta, "Sn/C Composite Anodes for Li-ion Batteries", Electrochemical and Solid State Letters Vol. 7, (3) A44-A48 (2004). Research in our group has also demonstrated that amorphous silicon can be contained within nanocrystalline matrices of TiN, TiB₂ and SiC to successfully stabilize reversible capacities up to 400 mAh/g. I. Kim, P. Kumta, and G. Blomgren, Electrochem. Solid-State Lett., 3 (11), 493 (2000); I. Kim, G. Blomgren, and P. Kumta, Electrochem. Solid-State Lett. 6(8) A 157 (2003). Research in bulk lithium alloy anodes has been steadily evolving and appears to be quite promising over that reported in the last 20 years. However, much research still needs to be done to harness the high theoretical capacities of silicon and tin based alloy anodes into realizable practical capacities. [0015] SnO₂ thin film anodes have been synthesized by several methods for use in thin film lithium-ion batteries. Low-pressure chemical vapor deposition has been used to produce thin films of SnO₂ which cycled well out to ~120 cycles. T. Brousse, R. Retoux, U. Herterich, and D.M. Schleich, J. Electrochem. Soc, 145(1), 1 (1998); R. Retoux, T. Brousse, and D.M. Schleich, J. Electrochem. Soc, 146(7), 2472 (1999). The electrochemical properties of sputter deposited SnO₂ anode thin films were investigated by Nam et al. S.C. Nam, Y.S. Soon, WI. Cho, B.W. Cho, H.S. Chun, and K.S. Yun, J. Electrochem. Soc, 148(3), A220 (2001). Electron beam evaporation has also been used to fabricate thin film SnO₂ and Si doped SnO₂ anodes. Y.I. Kim, W.H. Lee, H.S. Moon, K.S. Ji, S. H. Seong, and J.W. Prk, J. Pow. Sources, 101, 253 (2001). An amorphous silicon tin oxynitride (SiTON) thin film anode was also synthesized by rf magnetron sputtering. B. Neudecker, R. Zuhr and J. Bates, J. Pow. Sources, 81-82 27 (1999). The SiTON thin film cycled reasonably well for -12000 cycles, and in a thin film battery configuration withstood a 250°C anneal temperature designed to simulate a solder reflow condition. Spin coating was also used to synthesize SnO₂ thin films that cycled well for 30 cycles, but exhibited an extremely large 1^st cycle irreversible loss of ~70%. J.P. Maranchi, A.F. Hepp and P.N. Kumta, "LiCoO₂ and SnO₂ thin film electrodes for lithium-ion battery applications", Materials Science and Engineering, B (2005) in Press. In fact, all of the oxide based thin film anodes suffer the same problem as their bulk counterparts; they exhibit ~ 50% πreversible loss in the first cycle due to the formation of Li₂O.

[0016] Relatively few researchers have investigated thin or thick films of silicon or silicon- tin. Bourderau et al. had limited success studying CVD deposited thick (1.2 μm) amorphous silicon films on porous nickel substrates. S. Bourderau, T. Brousse, and D. M. Schleich, J. Power Sources, 81-82, 233 (1999). Bourderau et al. perceived a high capacity in the 1^st three cycles of ~1000 mAh/g which faded rapidly to 200 mAh/g by the 20^th cycle. More recently, Lee et al. studied the cycle related stress effects in sputtered silicon thin films. S. Lee, J. Lee, S. Chung, H. Lee, S. Lee, and H. Baik, J. Power Sources, 97-98, 191 (2001). Lee et al. showed that raising the lower limit of the voltage range from 0 V to 0.1 V had a significant positive effect on the cyclability of the a-Si film, and they qualitatively showed a change in film stress during cycling using a laser deflection technique.

[0017] Reversible, large scale volume changes were seen in meso-scale sputtered Si-Sn (unspecified composition in the article) films reported by Beaulieu et al. L.Y. Beaulieu, K. W. Eberman, R.L. Turner, L.J. Krause, and J.R. Dahn, Electrochem. Solid-State Lett. 4(9), A137 (2001). Beaulieu et al. used both in-situ AFM and in-situ optical microscopy to study the reversible moφhological changes in their Si-Sn films. It is important to note that for the optical microscopy experiment, a 7.5 μm Si-Sn film was deposited on a 130 μm disk of copper foil, whereas, in the case of the in-situ afm experiment, a rigid, stainless steel disk coated with a 4 μm thick layer of Cu "to improve the adhesion between the stainless steel substrates and the SiSn films" was used as the substrate. In a later work, research by Beaulieu et al. provided good insight into the nature of the electrochemical reaction of lithium with sputtered amoφhous Si₀.₆₆Sn₀.₃ films. L.Y. Beaulieu, K.C. Hewitt, R.L. Turner, A. Bonakdaφour, A.A. Abdo, L. Christensen, K.W. Eberman, L.J. Krause, and J.R. Dahn, J. Electrochem. Soc, 150 (2), A149 (2003). In that work, an in-situ XRD experiment confirmed that the initially amoφhous Si₀.₆₆Sn₀.₃₄ film remained amoφhous throughout the entire charge/discharge cycle. Beaulieu et al. also used the in-situ AFM technique to study the reaction of Li with patterned and continuous films of materials such as crystalline Sn, crystalline Al, amoφhous Si, and amoφhous Si₀.₆₆Srio.₃₄. L.Beaulieu, S. Beattie, T. Hatchard, and J. Dahn, J. Electrochem. Soc, 150 (4), A419 (2003); L. Beaulieu, T. Hatchard, A. Bonakdaφour, M. Fleischauer, and J. Dahn, J. Electrochem. Soc, 150 (11), A1457 (2003). There was a stark difference in the moφhological behavior between the amoφhous and crystalline materials during cycling. The volume changes were clearly reversible in the amoφhous materials, while the volume changes were inhomogeneous in the crystalline materials due to co-existing phase regions leading to severe moφhological changes. The research presented by Beaulieu et al. Id. has furthered the notion that amoφhous materials at least offer the possibility of reversibly cycling lithium.

[0018] Recently, bulk polycrystalline Si has been shown to become amoφhous upon room temperature electrochemical lithiation. P. Limthongkul, Y. Jang, N. Dudney, and Y. Chiang, Acta Materialia, 51, 1103 (2203); P. Limthongkul, Y. Jang, N. Dudney, and Y. Chiang, Journal of Power Sources, 119-121, 604 (2003). Ln those studies, the researchers have also put forth the approximate Gibbs free energy of formation for the amoφhous LiSi₀. ₇ phase. The Gibbs free energy of formation was calculated using the experimental equilibrium voltage ofthe amoφhous alloy and the Nernst equation.

[0019] Several groups have presented encouraging results on Si films deposited on Cu and Ni substrates at the 11^th International Meeting on Lithium Batteries (IMLB). K. Sayama, H. Yagi, Y. Kato, S. Matsuta, H. Tarui, and S. Fujita i, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 52, Monterey, CA (2002). Available online at http://www.electrochem. org/meetings/satellite/imlb/11/abstracts/piimlbll.htm. T. Yoshida, T. Fujihara, H. Fumimoto, R. Ohshita, M. Kamino, and S. Fujitani, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 48, Monterey CA (2002). Available online at http://www.electrochem. org/meetings/satellite/imlb/11/abstracts/piimlbll.htm. T. Takamura, S. Ohara, J. Suzuki, and K. Sekine, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 257, Monterey, CA (2002). Available online at http://www.electrochem. org/meetmgs/satellite/imlb/11/abstracts/piimlbll.htm. The research performed by Sanyo (K. Sayama, H. Yagi, Y. Kato, S. Matsuta, H. Tarui, and S. Fujitani, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 52, Monterey, CA (2002). Available online at http://www.electrochem. org/meetings/satellite/imlb/11/abstracts/piimlbll.htm) shows that a rough (~ 500 nm roughness) Cu substrate can be used to reversibly cycle 2-10 μm a-Si (plasma CVD and rf magnetron sputtering deposition methods were used) films at high capacity of ~ 3990 mAh/g for at least 10 cycles. The Sanyo researchers have also shown that complete batteries could be assembled with the a-Si electrode and LiCoO₂ cathode that cycle at higher capacities than control cells (using graphite anodes) for at least 50 cycles. T. Yoshida, T. Fujihara, H. Fumimoto, R. Ohshita, M. Kamino, and S. Fujitani, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 48, Monterey CA (2002). Available online at http://www.electrochem. org/meetings/satellite/imlb/11/abstracts/piimlbll.htm. Ln the same work, the researchers have shown that the 5 μm thick films deposited on rough Cu current collectors exhibit excellent rate capability characteristics. Researchers at Petoca Materials Ltd. have also shown excellent cyclability (25 cycles) and high capacity (-3000 mAh/g) for evaporated 40 nm Si films. T. Takamura, S. Ohara, J. Suzuki, and K. Sekine, Abstracts ofthe 11^th International Meeting on Lithium Batteries, 257, Monterey, CA (2002). Available online at http://www.electrochem. org/meetings/satellite/imlb/11/abstracts/piimlbll.htm. The conference abstracts by Sanyo and Petoca are some ofthe few published company research efforts on thin film Si anodes.

[0020] A US patent application publication no. 2002/0048705 Al, Apr. 25, 2002 outlined the idea of using silicon-silver thin film multilayer anodes for lithium-ion battery applications. In this patent application, good reversibility was seen out to 100 cycles. However, it appears that there is a large 1^st cycle ineversible loss (~ 56%). Si-Zr thin film anodes have shown promise at lower Si/Zr compositions (Si₀.₆Zr₀.₄). S. Lee, H. Lee, H. Baik, and S. Lee, Journal of Power Sources, 119-121, 113 (2003). Related research on Si|Ag multilayer thin films and Si-Zr|Ag multilayer thin films also highlighted the fact that Si based systems can show good reversibility for 70 cycles with a reversible capacity of- 17 μAh-cm^"2-μm^"L S. Lee, H. Lee, Y. Park, H. Baik, and S. Lee, Journal of Power Sources, 119-121, 117 (2003). The same group also showed the positive effect of annealing on the electrochemical cyclability of Co-Si alloy thin films. Y. Kim, H. Lee, S. Jang, S. Lim, S. Lee, H. Baik, Y. Yoon, and S. Lee, Electrochimica Acta, 48, 2593 (2003). Fe|Si multilayer thin film anodes have also exhibited excellent reversible capacities up to 50 cycles. J. Kim, H. Lee, K. Lee, S. Lim, and S. Lee, Electrochem, Cornm, 5, 544 (2003). Of all the thin film anode compositions studied by the group in South Korea (S. Lee, H. Lee, H. Baik, and S. Lee, Journal of Power Sources, 119- 121, 113 (2003); S. Lee, H. Lee, Y. Park, H. Baik, and S. Lee, Journal of Power Sources, 119- 121, 117 (2003); Y. Kim, H. Lee, S. Jang, S. Lim, S. Lee, H. Baik, Y. Yoon, and S. Lee, Electrochimica Acta, 48, 2593 (2003); J. Kim, H. Lee, K. Lee, S. Lim, and S. Lee, Electrochem, Comm, 5, 544 (2003)), the Fe|Si system appears to be the most promising. The researchers hint that the electrochemical characteristics of the Fe|Si thin film may be controlled by manipulating interfacial reactions and the extent of chemical reaction between the Fe and Si layers. Lee et al. also studied a full thin film battery comprising a Si-V thin film anode, LiSiPON electrolyte, and LiCoO₂ cathode. S. Lee, H. Baik, and S. Lee, Electrochemistry Communications, 5, 32 (2003). Although the study was not an in-depth investigation, the thin film battery reported by Lee et al. shows the potential of Si based thin films for use as anodes in next generation thin film lithium-ion batteries. [0021] A promising reversible capacity near 1700 mAh/g (50^th cycle) was recently observed by Graetz et al. when they electrochemically cycled a 100 nm evaporated thin film of pure Si on Cu substrate. J. Graetz, C. Ahn, R. Yazami, and B. Fultz, Electrochem. and Sol.-St. Let, 6(9), A194 (2003). Preliminary research performed by Maranchi et al. at Carnegie Mellon University has shown that it is possible to achieve sustained near theoretical (88% of theoretical) reversible capacities in pure sputtered amoφhous Si thin films. J. Maranchi, A. Hepp, and P. Kumta, Electrochem. and Sol.-St. Let, 6(9), A198 (2003), which is hereby incorporated by reference in its entirety. Fig. 2 displays the experimental cycling data for sputtered 250 nm and 1 μm a-Si thin films. Id.

BRIEF SUMMARY OF THE INVENTION [0022] An electrically conductive substrate is provided upon which one or more layers of active materials, i.e. materials that react reversibly with the charge earners in an electrolyte, are formed. Ln the case where lithium ions are the charge carriers, the active materials may be chosen from a group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys. Carbon may be used, as needed, as a buffer layer. Certain materials, such as iron, titanium and chromium, may be used as adhesion layers. For example, a layer of chromium may be used as an adhesive layer between a copper substrate and a layer of carbon. Prior to formation of these layers, the substrate is patterned so as to contain cavities. The active materials, as well as any adhesion and buffer layers, are then formed in one or more layers so that the total cumulative height ofthe layers is no more than, in one embodiment, one-third ofthe total depth of each ofthe cavities. The remaining "unfilled" depth ofthe cavities provides suitable room for volumetric expansion and contraction ofthe active materials, resulting in little or no net volume change in the anode at any point during the cycling ofthe batteries. This anode structure can be fabricated independently ofthe other manufacturing processes used for creation ofthe batteries and can therefore be used as a "drop in" component during the battery manufacturing process.

BRIEF DESCRIPTION OF THE DRAWINGS [0023] For the present invention to be easily understood and readily practiced, the present invention will now be described, for puφoses of illustration and not limitation, in conjunction with the folloΛving figures, wherein:

[0024] FIG. 1A is a representation ofthe structure of fully lithiated graphite (LiC₆) while FIG. IB illustrates the inplane distribution ofthe lithium ions in LiC₆; [0025] FIG. 2 is a comparison of galvanostatic cycling of sputtered 250 nm and 1.0 μm Si films on Cu foil (both at -C/2.5 rate); [0026] FIG. 3 illustrates a portion of a copper substrate;

[0027] FIG. 4 illustrates the substrate of FIG. 3 after a layer of resist has been formed and patterned;

[0028] FIG. 5 illustrates the substrate of FIG. 4 after a hard mask has been formed; [0029] FIG. 6 illustrates the substrate of FIG. 5 after the remainder ofthe resist has been stripped leaving a patterned hard mask; the substrate is then subjected to a directional copper etch;

[0030] FIG. 7A is a top plan view ofthe substrate after the copper etch while FIG. 7B is a close up of trie pattern of cavities formed in the top surface ofthe substrate; [0031] FIG. 8 is a cross-sectional view taken along the lines VIII - VLH in FIG. 7B; [0032] FIG. 9 illustrates the substrate of FIG. 8 after a layer of chromium is formed; [0033] FIG. 10 illustrates the substrate of FIG. 9 after a layer of carbon and a layer of silicon are formed;

[0034] FIG. 11 illustrates the substrate of FIG. 10 after the remainder ofthe hard mask is stripped; [0035] FIG. 12 illustrates a battery constructed with the anode ofthe present disclosure;

[0036] FIG. 13 illustrates electrochemical cycling data of an as-deposited 50 nm C/ 250 nm

Si/50nm C/10 nm Cr/Cu multilayer thin film cycled at 0.4 C (A) specific capacity vs. cycle number and (B) differential capacity;

[0037] FIG. 14 illustrates electrochemical cycling data of an as-deposited 50 nm C/ 250 nm

Si/50nm C/10 nm Cr/Cu multilayer thin film cycled at 0.8 C (A) specific capacity vs. cycle number and (B) differential capacity;

[0038] FIG. 15 illustrates electrochemical cycling data of an as-deposited 250 nm Si/50nm

C/10 nm Cr/Cu multilayer thin film cycled at 0.4 C (A) specific capacity vs. cycle number and (B) differential capacity;

[0039] FIG. 16 illustrates electrochemical cycling data of a 300°C-4h annealed 250 nm

Si/50nm C/10 nm Cr/Cu multilayer thin film cycled at 0.4 C (A) specific capacity vs. cycle number and (B) differential capacity;

[0040] FIG. 17 illustrates electrochemical cycling data of a 500°C-4h annealed 250 nm

Si/50nm C/10 nm Cr/Cu multilayer thin film cycled at 0.4 C (A) specific capacity vs. cycle number and (B) differential capacity; and

[0041] FIG. 18 illustrates SEM images of a 500°C-4h annealed 250 nm Si/50nm C/10 nm

Cr/Cu multilayer thin film cycled at 0.4 C after 75 cycles (A) 2000x and (B) same sample as in (A) with magnification of 5000x.

DETAILED DESCRIPTION OF THE INVENTION [0042] An anode for a lithium ion battery can be constructed as will be described in detail below. Briefly, a substrate of an electrically conductive material, such as polycrystalline copper, has a pattern created thereon. The pattern can be created by standard industrial processes such as photolithography and rapid ion etching. The patterns will generally consist of cavities, which may have cross-sectional shapes such as squares, rectangles, circles, or of other shapes or combinations of shapes. The dimensions of these patterns may conveniently be on the order of from about 1 μm to about 10 mm, although there is conceptually no limit placed upon their sizes. One then partially fills the cavities by depositing thin film layers of one or more active materials chosen from a group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys. Carbon may be used, as needed, as a buffer layer. Certain materials, such as iron, titanium and chromium, may be used as adhesion layers. One method of depositing thin film layers of these materials is by sputter deposition. Other methods could also be used such as, for example, pulsed laser deposition (PLD), chemical vapor deposition (CVD), plasma enhanced CVD, ion assisted sputter deposition etc. One may advantageously deposit multiple layers of these thin films in an alternating fashion.

[0043] In manufacturing the anodes, in one embodiment, one will usually not fill more than one-third (1/3) ofthe total depth ofthe cavities. This allows suitable room for the expected volumetric expansion and contraction ofthe deposited layers during battery cycling. Because the volumetric expansion and contraction ofthe deposited materials will not exceed the height ofthe cavities, there will be no net volume change in the anode/substrate module at any point during cycling ofthe patterned thin film anodes.

[0044] A heat treating step may be performed after the layers are formed, or at intermediate steps before formation of all the layers, as needed for stabilization. For example, a heat treatment after formation of all the layers described above (chromium/carbon/silicon) for a minimum of three hours at 300° C may be performed.

[0045] Anodes may be created in this fashion so that they constitute a "drop-in" component in the fabrication ofthe batteries. The anode of this disclosure therefore greatly simplifies the manufacture of lithium-ion batteries by eliminating the use of conventional graphite/binder/carbon black slurry casting of anodes. The anodes described herein are expected to find use in conventionally fabricated lithium-ion cells using both liquid and polymer gel electrolytes. This invention may reduce the volume ofthe anodes used in lithium-ion batteries while maintaining excellent electrochemical properties such as cycle life, capacity, rate capability, and discharge profile. It may also find applicability in battery types other than lithium-ion batteries.

[0046] Turning to FIG. 3, FIG. 3 illustrates a portion of an exemplary substrate 10 which may be, for example, a polycrystalline copper. The substrate 10 has a layer of resist 12 formed thereon and patterned as shown in FIG. 4. The word "formed" as used herein is intended to be used in its broad sense and to encompass any process step or steps which facilitate the creation ofthe layer. The resist 12 may be any commercially available photo resist, either positive or negative.

[0047] FIG. 5 illustrates the substrate of FIG. 4 after a hard mask has been formed. The hard mask may be, for example, a layer of silicon dioxide 14 formed using any convenient technique. Thereafter, the remaining photo resist 12 is stripped removing the silicon dioxide 14 above it, leaving the pattern of silicon dioxide 14 illustrated in FIG. 6. The silicon dioxide 14, in this example, forms a waffle type pattern as seen more clearly in FIG. 7B. Returning to FIG. 6, a unidirectional etch selective for copper is performed to create wells or cavities 16 as represented by the dotted lines in FIG. 6. The resulting pattern of etched cavities is illustrated more clearly in FIGs. 7A, 7B and 8. FIG. 8 is a cross-sectional view taken along the lines Vm-Viπ in FIG. 7B.

[0048] As seen from the foregoing description, the pattern can be created using standard industrial processes such as photolithography and rapid ion etchings. Although the pattern shown in the illustrated example is substantially square or rectangular, other types of patterns such as circles, other shapes, or combination of shapes can be created. Typical dimensions for the cavities are on the order of lμm to about ten millimeters, although there is conceptually no limit placed upon the size ofthe cavities or the spacing between the cavities. [0049] Once the cavities have been fabricated, one or more layers are then deposited within the wells 16. The layers are constructed of materials chosen from a group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys. The criteria for selecting materials from which active layers are to be formed is that the material must react with lithium, in the case of a lithium battery, in a reversible manner. Other layers may be formed. For example a carbon layer may be used as a buffer layer and layers of chromium, titanium and/or iron may be used as adhesive layers.

[0050] Turning now to FIG. 9, the first layer may be, for example, an adhesive layer 18 of chromium 10 — 50 nm thick. The layer of chromium 18 may be formed in any conventional manner. Thereafter, as shown in FIG. 10, a buffer layer 20 of carbon 10 - 50 nm thick may be formed followed by a 250 nm thick active layer of silicon 22. After each ofthe layers 18, 20 and 22 is formed, the substrate 10 may be cleaned so as to leave only the material formed in the well in place, or cleaning may wait until all ofthe steps are performed. In that case, the remainder ofthe hard mask 14 is stripped off together with any other material formed on the surface thereof. The dimensions given here are exemplary and are directed to the presently prefened embodiment although it is anticipated that other dimensions may provide comparable results as well as other layers of active materials with or without buffer and adhesion layers.

[0051] Although the figures are not to scale, in the disclosed embodiment, one will usually not fill more than one third ofthe total depth ofthe wells 16. Partially filling the wells to any depth less than completely full is intended to allow room for the expected volumetric expansion and contraction ofthe formed layers during battery cycling. Because the volumetric expansion and contraction ofthe formed layers of materials will not exceed the height ofthe wells 16, there will be no net volume changed in the anode/substrate 10 at any point during cycling ofthe patterned thin film anodes. One method of forming the various layers of thin films is by sputter deposition, although other methods may be used. One may advantageously deposit multiple layers from the materials identified above in thin films in an alternating fashion. The particular selection ofthe materials forming the thin film layers is not critical, as long as at least one ofthe selected materials reacts in a reversible manner with the charge carriers in the electrolyte. It is anticipated that the bottom ofthe wells 16 may be roughened to improve adhesion. Ln the case of roughened well bottoms, it is anticipated that if more than one layer is formed, the first layer will be substantially thin such that the pattern ofthe roughened well bottom will be substantially duplicated on the top surface ofthe first layer. Thus, the roughened bottom will "show through" or be "reflected by" the upper surface ofthe first layer thereby facilitating better adhesion with a second layer formed thereon. [0052] FIG. 12 illustrates a battery constructed using the anode ofthe present invention. Although in this example the battery is a lithium battery, the concepts ofthe present invention may be employed in the construction of other types of batteries. Additionally, in other types of batteries, the construction ofthe anode of he present disclosure, namely forming wells or cavities of active material, may be useful in the construction of cathodes in such other batteries.

[0053] Turning to FIG. 12, a battery 28 constructed according to the teachings ofthe present invention is illustrated. The battery 28 has a first terminal 30 connected to an anode 32 constructed according to the teachings ofthe present disclosure. Because anodes constructed according to the present disclosure greatly simplify the manufacturing of lithium-ion batteries by eliminating the use of conventional graphite/binder/carbon black slurry casting of anodes, the anodes ofthe present disclosure may constitute a drop-in component. The battery 28 also comprises a second terminal 36 electrically connected to cathode 38. An electrolyte 42 is positioned between the anode 32 and cathode 38 and may be either a liquid or polymer gel electrolyte. Those of ordinary skill in the art will recognize that because the anode may be formed using a variety of different materials which react reversibly with lithium, the selection ofthe material for cathode 38 and the electrolyte 42 will depend upon the material selected for anode 32. It is anticipated that the selection of an appropriate cathode 38 material and an appropriate electrolyte 42 once the material for the anode 32 has been selected is known in the art need not be repeated here. Finally, the various components comprising the battery 28 are housed within a casing 40. Experimental results

[0054] Initially, the inventors believed that a Cu-carbon-Si-carbon sandwich structure could maintain mechanical stability for increased cycles compared to a simple Cu-Si structure. The starting hypothesis was that the carbon layers could act to reduce any potentially large stresses from developing at the active (Si): inactive (Cu) interface. It was also believed that the carbon would prevent the deleterious formation of any Cu-Si or Li-Cu-Si which could lower the interface fracture energy. In all of the studies, Cr was used as an adhesion promoting layer between the Cu foil and C film.

[0055] To test the aforementioned hypothesis, samples were prepared using r.f. sputtering of type Cu|10 nm Cr|50 nm C|250 nm Si|50 nm C. The electrochemical cycling properties ofthe as-prepared films were examined at two cunent rates 0.4 C and 0.8 C, conesponding to 100 μA/cm² and 200 μA/cm² based on the theoretical capacity of Si. The electrochemical results of the Cu|Cr|C|Si|C samples cycled at rates of 0.4 C and 0.8 C are shown in Figs. 13 and 14, respectively. Fig. 13 A reveals that the C/Si/C sandwich sample starts off with a very high reversible capacity of - 5144 mAh/cc in the 2^nd cycle which fades slowly to - 4450 mAh/cc by the 100^th cycle. Therefore, the capacity fade is 0.14% per cycle. Fig. 13B shows differential capacity plot for the C/Si/C sample cycled at a 0.4C rate. The dQ/dV graph is very similar to those observed before for Cu Si samples with broad peaks indicating the amoφhous nature ofthe reacting constituents and resultant phases. Fig. 14A shows the effect of increasing the cycling rate by a factor of two to 0.8C. The initial reversible capacity was 5088 mAh/cc which faded to 4277 mAh/cc in the 100^th cycle, resulting in a capacity fade of 0.16% per cycle. Therefore, increasing the C-rate by a factor of two led to a slight increase in the capacity fade, but in general did not change the electrochemical characteristics substantially. The differential capacity plot shown in Fig. 14B for the 0.8C sample shows similar reaction peaks to the slower rate sample shown in Fig. 13B.

[0056] Ln an attempt to increase the absolute value of the volumetric capacity, the inventors prepared samples without the 50 nm C overlayer of type Cu|10 nm Cr|50 nm C|250 nm Si. The electrochemical results ofthe C/Si sample tested at a 0.4C rate are shown in Fig. 15. Fig. 15A shows that the initial reversible capacity of 6244 mAh/cc was higher than previously reported for the C/Si/C sample. However, the capacity fade per cycle also increased to 0.21% for the first 100 cycles. The differential capacity graph shown in Fig. 15B also showed consistent electrochemical signatures when compared to the dQ/dV plots shown in Fig. 13B and Fig. 14B. Another strategy to enhance the reversible capacity, while maintaining low fade per cycle was developed. A post-deposition anneal may enhance adhesion between the material layers.

[0057] The same sample tested in Fig. 15, without the C overlayer, was subjected to annealing treatments in Ar atmosphere for 4 hours. Samples were annealed at 300°C and 500°C to test the effect of annealing on the electrochemical properties of the thin film multilayer anodes. The electrochemical results of the Cu|10 nm Cr|50 nm C|250 nm Si sample annealed at 300°C in Ar was tested at a 0.4C rate. Fig. 16A shows that the initial reversible capacity of 5480 mAh/cc was lower than previously reported for the unannealed sample, probably due to the formation of amoφhous SiC which reduces the amount of active Si. However, the capacity fade per cycle was lowered to 0.19% for the first 100 cycles. The differential capacity graph shown in Fig. 16B was consistent with the dQ/dV plots shown in Figs. 13B, 14B and 15B. Encouraged by the slight drop in capacity fade by annealing, the as- deposited Cu|10 nm Cr|50 nm C|250 nm Si sample annealed at 500°C in Ar was tested at a 0.4C rate was also evaluated using constant-cunent cycling. Fig. 17A reveals that the initial reversible capacity was 5379 mAh/cc and the 100^th cycle reversible capacity was 4737 mAh/cc. The capacity fade was excellent at 0.12% per cycle. The higher temperature annealing only slightly dropped the initial reversible capacity, but decreased the capacity fade to a commercially acceptable level. As compared to the original C/Si/C sandwich, the 500 °C annealed C/Si sample had enhanced electrochemical cycling properties and similar electrochemical reaction signatures (as shown in Fig. 17B).

[0058] An initial SEM investigation ofthe 500 °C annealed Cu|10 nm Cr|50 nm C|250 nm Si sample after 75 charge/discharge cycles was conducted. The results at magnifications of 2000x and 5000x are shown in Figs. 18A and 18B, respectively. From the images, one can clearly see that the film did crack substantially. The depth ofthe cracking is not known at this time, but will be the subject of future study. It is of reasonable scientific interest to learn whether the cracks stopped at the C layer, Cr layer, or penetrated completely through the multilayers to the Cu substrate. Furthermore, the cracking pattern is slightly different than those shown previously for unannealed Cu/Si films after cycling at the same rate. The Cu/Si samples showed very straight compressive type failures, whereas the annealed samples exhibit a telephone-cord like compressive failure crack moφhology. The annealing step may influence the state of stress in the film, as well as change the mechanical properties of all the constituent film layers as well as the Cu substrate. The excellent reversibility and high capacity ofthe annealed multilayer film illustrates the promise and potential of such thin film anode systems, justifying substantial future research endeavors. [0059] While the present invention has been described in connection with preferred embodiments thereof, those of ordinary skill in the art will recognize that many modifications and variations are possible. The present invention is intended to be limited only by the following claims and not by the foregoing description which is intended to set forth the presently prefened embodiment.

Claims

What is claimed is: 1. An electrode for a rechargeable battery, comprising: a substrate having a plurality of cavities formed therein; and at least one active layer formed in at least certain of said cavities, said active layer formed of a material that reacts in a reversible manner with a charge carrier of an electrolyte.

2. The electrode of claim 1 wherein said charge carrier in the electrolyte includes lithium ions.

3. The electrode of claim 2 wherein said material is selected from the group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys.

4. The electrode of claim 1 additionally comprising a buffer layer.

5. The electrode of claim 4 wherein said buffer layer includes a layer of carbon.

6. The electrode of claim 1 additionally comprising an adhesion layer.

7. The electrode of claim 6 wherein said adhesion layer is formed of a material selected from the group consisting of chromium, titanium and iron.

8. The electrode of claim 1 wherein said cavities are filled to less than their entire depth.

9. The electrode of claim 1 wherein said certain cavities each contain an active layer of material that reacts in a reversible manner with lithium, a buffer layer, and an adhesive layer.

10. The electrode of claim 9 wherein said substrate includes a copper substrate, said buffer layer includes a layer of carbon, said adhesive layer includes a layer of chromium between said substrate and said buffer layer, and wherein said active layer includes a layer of silicon.

11. The electrode of claim 1 wherein the bottoms of said cavities are rough.

12. The electrode of claim 11 wherein said electrode has at least two layers, a first of said layers being formed on the bottom of said cavities, said first layer being sufficiently thin such that an upper surface of said first layer reflects said rough cavity bottom, and a second layer formed on top of said first layer.

13. A process for manufacturing an electrode for a battery, comprising: forming a plurality of cavities in a substrate; and forming at least one active layer in at least certain of said cavities, said layer formed of a material that reacts in a reversible manner with a charge carrier of an electrolyte.

14. The process of claim 13 wherein said charge carrier in the electrolyte includes lithium ions, and wherein said forming at least one layer includes forming at least one layer of a material selected from the group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys.

15. The process of claim 13 additionally comprising forming a buffer layer.

16. The process of claim 15 wherein said buffer layer is formed of carbon.

17. The process of claim 13 additionally comprising forming an adhesion layer.

18. The process of claim 17 wherein said adhesion layer is formed of a material selected from the group consisting of chromium, titanium and iron.

19. The process of claim 13 wherein said forming at least one layer fills said certain of said cavities to less than their entire depth.

20. The process of claim 13 wherein said forming at least one layer includes forming an active layer of material that reacts in a reversible manner with lithium, a buffer layer, and an adhesive layer.

21. The process of claim 20 wherein said substrate includes a copper substrate and wherein said forming a buffer layer includes forming a layer of carbon, said forming an adhesive layer includes forming a layer of chromium between said substrate and said layer of carbon, and wherein said forming an active layer includes forming a layer of silicon.

22. The process of claim 13 additionally comprising roughening the bottoms of said cavities.

23. The process of claim 22 wherein said forming at least one layer includes forming a first layer on the bottom of said cavities, said first layer being sufficiently thin such that an upper surface of said first layer reflects said rough cavity bottom, and forming a second layer on top of said first layer.

24. A battery, comprising: a first electrode comprised of a substrate having a plurality of cavities formed therein and at least one layer formed in at least certain of said cavities, said layer formed of a material that reacts in a reversible manner with a charge carrier of an electrolyte; a second electrode; and an electrolyte positioned between said first and second electrodes

25. The battery of claim 24 additionally comprising a casing sunounding said first and second electrodes.

26. The battery of claim 24 wherein said charge carcier in the electrolyte includes lithium ions.

27. The battery of claim 26 wherein said material is selected from the group consisting of silicon, tin, germanium, indium, antimony, lead, aluminum, magnesium, calcium, bismuth, arsenic, silver, gold, platinum, cadmium, zinc, phosphorus, sulfur and their alloys.

28. The battery of claim 24 wherein said first electrode additionally comprises a buffer layer.

29. The battery of claim 28 wherein said buffer layer includes a layer of carbon.

30. The battery of claim 24 wherein said first electrode additionally comprises an adhesion layer.

31. The battery of claim 30 wherein said adhesion layer is formed of a material selected from the group consisting of chromium, titanium and iron.

32. The battery of claim 24 wherein said cavities are filled to less than their entire depth.

33. The battery of claim 24 wherein said certain cavities each contain an active layer of material that reacts in a reversible manner with lithium, a buffer layer, and an adhesive layer.

34. The battery of claim 33 wherein said substrate includes a copper substrate, said buffer layer includes a layer of carbon, said adhesive layer includes a layer of chromium between said substrate and said buffer layer, and wherein said active layer includes a layer of silicon.

35. The battery of claim 24 wherein the bottoms of said cavities are rough.

36. The battery of claim 35 wherein said electrode has at least two layers, a first of said layers being formed on the bottom of said cavities, said first layer being sufficiently thin such that an upper surface of said first layer reflects said rough cavity bottom, and a second layer formed on top of^* said first layer.