Electrochemical study on nano-Sn, Li4.4Sn andAlSi0.1powders used as secondary lithium battery anodesChunsheng Wanga,*, A. John Applebya, Frank E. LittlebaCenter for Electrochemical Systems and Hydrogen Research, Texas Engineering Experiment Station, Texas A and M University,College Station, Texas 77843-3402, USAbCenter for Space Power, Texas Engineering Experiment Station, Texas A and M University, College Station, Texas 77843-3118, USAReceived 16 August 2000; accepted 23 August 2000AbstractIt is believed that particle cracking resulting from phase transformation is responsible for the poor cycle performance of lithium alloyanodes. Pulverization effects may be reduced by using, (i) smaller active particles; (ii) active particle composites with different potentialsfor the onset of lithium alloy formation; and (iii) expanded alloys which have undergone a major increase during initial charging. Threealloys of the above types (nano-Sn, AlSi0.1and Li4.4Sn) were studied by electrochemical impedance spectroscopy (EIS) to determine theirelectrochemical kinetics and intrinsic resistance during initial lithium insertion-extraction. The electrodes were prepared by sandwiching adisk of active powder between two nickel screens, so that the contact resistance may be determined by EIS and from a d.c. voltagedifference across the electrode (trans-electrode voltage). A large increase in contact resistance was found during lithium discharge(extraction) from nano-LixSn and LixAlSi0.1alloys, compared with the small increase during the initial charge. This result suggest that thematrix materials should have a small coef®cient of elasticity to give low stress on expansion of the active alloy, together with a large elasticdeformation to compensate for volume reduction. This is contrary to generally accepted argument that the matrix should have a highductility. EIS results for measurement of intrinsic resistance and reaction kinetics during initial lithium insertion into nano-Sn and AlSi0.1alloys show that both solid electrolyte interphase (SEI) ®lms formed on particle surfaces, together with particle pulverization, areresponsible for the high contact resistance. The electrochemical kinetics of both lithium charge and discharge are controlled by contactresistance at high states of charge. # 2001 Elsevier Science B.V. All rights reserved.Keywords: Lithium alloy anode; Electrochemical impedance spectroscopy; Contact resistance1. IntroductionLithium alloys have been extensively studied as possiblereplacements for carbon anodes in lithium ion batteries. Theelectrochemical capacities of lithium alloys may be verylarge (LiAl and Li4.4Sn: 990 mA h/g, Li4.4Si: 4200 mA h/g),compared with that of carbon (372 mA h/g). However, thelarge volume expansion due to existence of two phasedomains (LiAl: 97% [1], Li4.4Sn: 358% [2], Li4.4Si:323% [1]) results in severe particle cracking with lossof electrical contact, giving irreversible capacity losseswhich prevent the widespread use of such alloys in lithiumbatteries.Many attempts have been made to stabilize the morpho-logy of lithium alloy electrodes by minimizing the mecha-nical stress in the electrode caused by this volumeexpansion. Successful methods have included the following.(i) The use of composites of active and inactive materials,e.g. the so-called tin-based composite oxides (TCOs) [3],where a nano-structured active phase is dispersed either inan inert solid electrolyte or in a soft metal matrix formed onthe initial charge, e.g. the system based on Sn2Fe [4]. (ii) Theuse of intermetallic lithium insertion compounds, e.g. InSb[2] and Cu6Sn5[5], where lithium atoms occupy interstitialsites, giving only a small volume expansion. (iii) The used ofmixed active material composites (active/active composite),e.g. SnSb0.14[6], where stepwise lithium insertion into thedifferent active phases buffers volume expansion. It isbelieved that the major volume expansion occurs duringinitial charging, and subsequent dealloying and alloyingsteps result in only small shrinkage and expansion [7]. Asa result, the compound Li4.4Sn may have a more stablecapacity during prolonged charge±discharge cycling. Initialelectrochemical dealloying, followed by electrochemicalJournal of Power Sources 93 (2001) 174±185*Corresponding author. Tel.: 1-409-845-8281; fax: 1-979-845-9287.E-mail address: [email protected] (C. Wang).0378-7753/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved.PII: S 0378-7753(00)00576-0alloying of Li4.4Sn prepared by melting can give usefulinformation on the phase transformations in the Li±Snelectrode at ambient temperature.In most case, the cracking and pulverization of themetallic host matrices is evaluated by the capacity changeon charge±discharge cycling [6]. However this methodcannot monitor cracking in each successive two-phasedomain, with the successive single phases Sn, Li2Sn5, LiSn,Li7Sn3,Li5Sn2and Li22Sn5. Moreover, the cycle life of themetallic electrode not only depends on cracking, but alsoaffected by other factors. These include the stability of thesolid electrolyte interphase (SEI) ®lm on the electrode sur-face, the occurrence of less rapid electrochemical reactionkinetics, and the amount of active material available forreaction. Development of an in situ method to measure thepulverization of the lithium alloy electrode is important as ameans of selecting suitable lithium alloy anode composi-tions. The conductivity measured using d.c. methods hasbeen successfully used to measure the intrinsic resistance ofgraphite electrodes on insertion and extraction of lithium[8]. However, it is not applicable to lithium alloys becauseapplication of a large trans-electrode voltage affects theelectrode potential, whose value is required to detect phasetransformation.In this work, Sn, Li4.4Sn and AlSi0.1powder disks sand-wiched between two nickel screens were used as workingelectrodes. Pulverization of these alloys on cycling wasmonitored by the trans-electrode voltage difference whenthey were charged or discharged on one side only. Thecontact resistance and electrochemical reaction kineticswere evaluated by electrochemical impedance spectroscopy(EIS).2. Experimental2.1. Electrode preparation and cell descriptionCommercial Sn (100 nm), SnO2and AlSi0.1(8 mm) pow-ders were purchased from Argonide Corporation, AldrichChemical Company, Inc., and Valimet, Inc., respectively.The Li4.4Sn compound was prepared by mixing the appro-priate amount