Mössbauer spectra as a “fingerprint” in tin–lithium compounds: Applications to Li-ion batteries
Graphical abstract
Δ−δ correlation diagram for the different tin sites of the Li–Sn compounds. The symbols denote the different Li–Sn phases and the products obtained at the end of the discharge of η-Cu6Sn5 and SnB0.6P0.4O2.9. The grey and the light-grey areas show Sn-centred polyhedra without and with one Sn first-nearest neighbours, respectively.
Introduction
The increasing demand for electric energy in many applications has lead to intensive research on batteries. Carbonaceous materials are commonly used as negative electrodes in commercial lithium-ion batteries [1], [2], [3]. Carbonaceous materials usually show good cycling performances and little volume change during the lithiation and de-lithiation cycles. However, the maximum theoretical specific capacity is only 372 mA h g−1, corresponding to the formation of the LiC6 graphite intercalation compound. Many intermetallic-based anode materials have been studied to increase the limited capacity and improve the cycling life of graphite [4], [5]. Tin-based lithium storage materials are often referred to as alternative electrodes due to their reasonably low potentials for Li+ insertion and high storage capacities [6], [7], [8], [9]. Poor cyclability due to large volume changes occurring during lithium insertion and extraction is still a major problem, which is responsible for electrode disintegration. Previous papers pointed out that these limitations can be partly both overcome by reducing the metal particle size and using multi-phase systems or compounds [10], [11], [12], [13], [14]. Lithium extensively reacts with tin, up to the maximum stoichiometry of Li22Sn5 [15], corresponding to a specific capacity of 991 mA h g−1 which is considerably higher than that of graphite and coke. However, the Li-insertion mechanism is mainly based on the formation of Li–Sn compounds leading to high volume changes of the particles during cycling. Therefore, reversibility is strongly limited. The volume of the highly lithiated compound, i.e. Li22Sn5, increases by 300% compared to β-Sn, which leads to the formation of microcracks within the electrode.
In order to solve these problems, several authors put forward composite materials in which an electrochemically inactive matrix acts as a buffer to reduce the effects of volume changes. Fuji Photo Film Co. announced a new type of anode materials [13], [16] based on tin-composite oxide (TCO) glasses with significantly higher reversible specific (>600 mA h g−1) and volumetric (>2200 mA h cm−3) capacities. Sony recently released a tin-based amorphous anode material named Nexelion [17] in which the lithium ion storage capacity is 50% higher than that of carbonaceous materials, the overall battery capacity is thus increased by 30%. The mechanisms of the above-mentioned materials are complex. The result of the reactions is the formation of Li–Sn nano-compounds with distinct stoichiometries.
The purpose of this work is to introduce a method to identify the various Li–Sn compounds formed during electrochemical reactions (“fingerprint”). The binary diagram Li–Sn [18] shows the existence of 7 phases: Li2Sn5, LiSn, Li7Sn3 Li5Sn2, Li13Sn5, Li7Sn2 and Li22Sn5 whose structures were characterized by X-ray diffraction (XRD) [19], [20], [21], [22], [23], [24], [25]. The Li-richest phase is known to be Li22Sn5 according to the binary phase diagram, but recent studies revised its crystal structure and reported the Li17Sn4 stoichiometry [26], [27], [28], [29]. These two compounds have close compositions, 4.25 Li/Sn for Li17Sn4 and 4.4Li/Sn for Li22Sn5, and similar Sn local environments. Consequently, it is almost impossible to differentiate the Mössbauer parameters and we will not discuss the correct stoichiometry of the Li-richest phase in this paper. For simplicity, we consider Li22Sn5 as the Li-richest phase in this paper. The Li–Sn phases can be obtained by the following methods: (i) the syntheses by solid state reaction that do not allow us to obtain all Li–Sn pure phases [30] and to control the particles size, (ii) the electrochemical route which yields small quantities of materials [31], and (iii) the microwave-assisted solid-state reaction [29]. Here we used an alternative method, the mechanical alloying [32], which is a powerful technique to synthesize very different materials such as metallic to ionic extended solid solutions, compounds made up of immiscible elements or compounds made up of elements with very different melting point temperatures [33]. Furthermore, mechanosynthesis yields large amounts of rather pure Li–Sn compounds that can be used as references for Mössbauer experiments.
Section snippets
Experimental
The materials were obtained with an SPEX 8000 vibratory mill. The vial was shaken at frequency of 20 Hz in the three orthogonal directions. The impact speed of the balls was several m s−1 and the shock frequencies were several hundred Hz. A pure Sn rod (99.9% Aldrich) and a Li foil (99.9% FMC) were used for the synthesis. The starting materials were put in stoichiometric amounts and were sealed into a stainless steel milling container (25 cm3) with stainless steel balls for 48 h under argon
Results
The XRD patterns are shown in Fig. 1 and show the high crystallinity of our samples. The purity of LiSn, Li13Sn5, Li7Sn2 and Li22Sn5 was confirmed after comparison with the JCPDS database. The XRD pattern of the Sn-richest phase, Li2Sn5, shows the presence of β-Sn. As for the following three phases, Li7Sn3, Li5Sn2 and Li13Sn5, corresponding to small variations of the Li/Sn ratio (2.33; 2.5; 2.6, respectively) it was possible to obtain rather pure phases. Impurities were only detected for Li5Sn2
Discussion
The structural data of the Li–Sn phases [18], [19], [20], [21], [22], [23], [24], [25] are summarized in Table 2, and their structures, as reported by Chouvin et al. [35] can be found in Fig. 3.
Both the Sn-richest phase (Li2Sn5), based on tin stacking, and the Li-richest phase (Li22Sn5), based on lithium stacking, have three-dimensional lattices whereas the other phases have two-dimensional lattices (Fig. 3). The Sn-richest compound, Li2Sn5, is made up of a three-dimensional tin network with
Applications of the Δ−δ correlation diagram
It is possible to determine from Fig. 5 the stoichiometry of the Li–Sn phases formed during the electrochemical process. The exact composition of LixSn, is often difficult to characterize from XRD since the phases obtained by electrochemical reactions with lithium are amorphous and/or nano-crystallized. In this case, the use of 119Sn MS can greatly help in determining the phase composition, for example, Li22Sn5 and Li7Sn2. The spectra of two fully lithiated samples (at the end of the first
Conclusion
Pure Li–Sn crystalline phases were obtained by ball-milling method. Successful syntheses made the characterization of these phases with 119Sn MS possible. The analysis of the isomer shift and quadrupole splitting in terms of crystal structure and ED as defined by Hume–Rothery allowed us to characterize two types of Li–Sn compounds : the Sn-richest compounds (Li2Sn5, LiSn) and the Li-richest compounds (Li7Sn3, Li5Sn2, Li13Sn5, Li7Sn2, Li22Sn5). From these results a Δ−δ correlation diagram was
Acknowledgments
This work was carried out in the framework of ALISTORE, Network of Excellence (Contract no.: SES6-CT-2003-503532). The authors are grateful to the European Community for financial support.
References (43)
- et al.
J. Power Sources
(1997) - et al.
J. Power Sources
(1999) - et al.
J. Power Sources
(2002) - et al.
J. Power Sources
(2003) - et al.
J. Alloys Compds.
(2001) - et al.
Inorg. Chem.
(2003) - et al.
J. Alloys Compds.
(1999) - et al.
Chem. Phys. Lett.
(1999) - et al.
J. Mater. Chem.
(1999) - et al.
J. Electrochem. Soc.
(2002)