- Email: [email protected]
Lithium isotope effects upon electrochemical lithium insertion to host material from ionic liquid medium
Progress in Nuclear Energy 53 (2011) 999e1004
Contents lists available at ScienceDirect
Progress in Nuclear Energy journal homepage: www.elsevier.com/locate/pnucene
Lithium isotope effects upon electrochemical lithium insertion to host material from ionic liquid medium Shun Saito, Yuta Takami, Masahiro Yoshizawa-Fujita, Satoshi Yanase, Takao Oi* Faculty of Science and Technology, Sophia University, 7-1 Kioicho, Chiyodaku, Tokyo 102-8554, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 October 2010 Received in revised form 7 February 2011 Accepted 29 April 2011
Lithium was electrochemically inserted from a Liþ ion containing ionic liquid into graphite or tin to observe lithium isotope effects that accompanied the insertion. While no preferential uptake of the lithium isotopes was detected with graphite, the lighter isotope, 6Li, was preferentially fractionated into tin with the single-stage lithium isotope separation factors, S, ranging from 1.004 to 1.008 at 25 C. It was speculated that a Liþ ion was inserted into graphite together with an anionic component of the ionic liquid and, upon the reduction of the Liþ ion to a lithium atom, the anion was released from graphite, while a Liþ ion alone was inserted into tin. Molecular orbital calculations supported this speculation in a qualitative fashion. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Lithium isotopes Isotope effects Ionic liquid Graphite Separation factor Reduced partition function ratio
1. Introduction Lithium has been and still is a major target of isotope separation study. This is in part because of the importance of the isolated and enriched isotopes of lithium in nuclear science and industry. The heavier isotope of lithium, 7Li, is used as pH controller of coolants in nuclear fission reactors. A large demand for the lighter isotope, 6Li, is expected in DT fusion power reactors in the future where lithium compounds rich in 6Li will be required for the tritium breeder blanket. The only method that was applied to a large-scale lithium isotope separation is the amalgam method (Palko et al., 1976). In this method, lithium is distributed between the amalgam phase and the aqueous or organic electrolyte solution phase and lithium isotope separation is practiced based on the lithium isotope exchange reaction between the two phases, 7
Li(Hg) þ 6Liþ(solution) ¼ 6Li(Hg) þ 7Liþ(solution),
(1)
where ALi(Hg) and ALiþ(solution) denote the lithium isotope A in the oxidation state 0 surrounded by mercury atoms in the amalgam phase and that isotope in the oxidation state þ1 surrounded by
* Corresponding author. Tel.: þ81 3 3238 3359; fax: þ81 3 3238 3361. E-mail address: [email protected] (T. Oi). 0149-1970/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.pnucene.2011.04.017
solvent molecules in the electrolyte phase, respectively. The reported value of the 7Li-to-6Li single-stage separation factor, S, defined as, S ¼ (7Li/6Li)solution/(7Li/6Li)Hg,
(2)
where (7Li/6Li)B denotes the 7Li-to-6Li isotopic ratio in phase B, is 1.049e1.062 for the water solvent (Fujie et al., 1986), meaning that 6 Li is preferentially fractionated in the amalgam phase. Although the value of 1.049e1.062 is large and attractive, the use of toxic mercury brings about biological and environmental problems and makes the amalgam method difficult to be applied to large-scale lithium isotope enrichment in the future. As a possible alternative of mercury, we investigated tin (Yanase et al., 2000), gallium (Zenzai et al., 2008, 2010), zinc (Mouri et al., 2008), graphite (Yanase et al., 2003) and metal oxides and sulfides (Mouri et al., 2007; Asano et al., 2008). We found that the metals, metal sulfide and graphite preferentially take up 6Li, while metal oxides showed little preference or were slightly 7Li-specific. Any host materials for lithium insertion so far investigated, however, did not show lithium isotope effects that surpassed those on mercury. The electrolyte solution so far investigated was in most cases a mixed solution of ethylene carbonate (EC) and methylethyl carbonate (MEC) containing LiClO4. Larger lithium isotope effects may be obtained with solvents other than carbonates. Ionic liquids (ILs) may be candidates of solvents for lithium isotope separation
1000
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
processes based on redox reactions of lithium. They are being investigated as solvents in secondary lithium ion batteries with higher performance than those of concurrent organic solvents (Sato et al., 2004). We attempted to use an IL as an alternative of the EC/ MEC mixture solution. In this paper, we report lithium isotope effects accompanying electrochemical lithium insertion into graphite or tin from an IL containing lithium ions and, in addition, results of molecular orbital (MO) calculations conducted to help the elucidation of the lithium isotope effects observed in the experiments. 2. Experimental 2.1. Electrodes Two kinds of material were used as cathode (the host of the lithium insertion). One was commercially available graphite electrodes for the use as cathodes of secondary lithium ion batteries (graphite powder daubed on a copper foil; capacity, 1.6 mAh/cm2; density 1.1 g/cm3), purchased from Hohsen Corp. Another one was tin wires with a purity of 99.9% and the diameter of 0.5 mm purchased from Kojundo Chemical Laboratory Co. Ltd. Lithium foils with a purity of 99.8% and about 0.3 mm thick purchased from Honjo Metals Co. Ltd. were used as anodes and reference electrodes. 2.2. Electrolyte solution The electrolyte solution used was an IL, N,N-diethyl-Nmethoxyethyl ammonium bis(trifluoromethanesulfonyl)imide {[(C2H5)2(CH3)(CH3OC2H2)N]þ[(CF3SO2)2N]} (DEME-TFSI) containing lithium bis(trifluoromethanesulfonyl)imide (LieTFSI). In electrolytic experiments with graphite electrodes, EC was added to the electrolyte solution. DEME-TFSI, LieTFSI and EC were purchased from Kanto Chemical Co. Inc., Morita Chemical Industries Co. Ltd. and Wako Pure Chemical Industries Ltd., respectively. The electrolyte solution was prepared under a dry Ar atmosphere. 2.3. Lithium insertion The experimental apparatus used was basically the same as the one used in our previous studies (Yanase et al., 2000; Mouri et al., 2008; Yanase et al., 2003; Mouri et al., 2007; Asano et al., 2008) and schematically drawn in Fig. 1. It was composed of a power supply (a Hokuto Denko Corporation HJ-201B battery charge/discharge unit), a cylinder-shaped three-electrode electrochemical glass cell (electrolytic cell) of about 12 cm in height and 3 cm in diameter and a data acquisition unit consisting of an A/D converter and a personal computer. The volume of the electrolyte solution placed in the cell was about 10 cm3. The size of the graphite electrode was 1 cm 1 cm, and the weight of the tin electrode was about 0.1 g. The electrolytic cell was build up in a dry argon atmosphere as shown in Fig. 1. The lithium insertion was performed in the constant currentconstant voltage mode, which was designed to avoid the possible deposit of lithium on the cathode surface. That is, the electrolysis (lithium insertion) was at first carried out in a constant current mode (1 mA for graphite electrodes and 0.2 mA for tin electrodes). As the electrolysis proceeded, the electric potential of the cathode against the reference electrode (cathode potential), in general decreased and reached the pre-set value (0.02 V for graphite electrodes and 0.2 V for tin electrodes), the value chosen based on our previous similar studies in which the deposit of lithium on surfaces of cathode materials was not observed at this cathode potential. The electrolytic mode was then automatically changed to the constant voltage mode. The electrolysis continued until the integrated quantity of electricity reached the predetermined value
Fig. 1. Schematic drawing of experimental apparatus. 1, Personal computer; 2, A/D converter; 3, Power supply; 4, reference electrode; 5, cathode (graphite or tin); 6, anode; 7, electrolytic cell; 8, electrolyte solution; 9, stirrer tip.
and was discontinued manually. During the electrolysis, the temperature of the electrolytic cell was kept at 25 C and the electrolyte solution was stirred with a magnetic stirrer.
2.4. Analyses and measurements Lithium inserted in graphite electrodes was extracted with a dilute (ca. 0.1 M) hydrochloric acid. The procedure of lithium extraction from the lithium inserted tin cathodes has been described elsewhere (Yanase et al., 2000). Lithium extraction from the electrolyte solution was carried out as follows: To ca. 10 cm3 of the electrolyte solution after the electrolysis was added about the same volume of pure water. The two layers were stirred furiously with a magnetic stirrer till they became a sort of an opaque mixture. This mixture was then centrifuged at 2000 rpm for 10 min to separate it into the aqueous and IL phases again. Lithium in the electrolyte solution was extracted into the aqueous phase in this procedure. The amount of lithium inserted into the cathodes was determined by measuring the lithium concentration in a part of the aqueous solution containing the extracted lithium by flame photometry with a Thermo Electron Corp. SOLAAR M mkII atomic absorption spectrometer. The chemical form of lithium in another part of the aqueous solution was converted to lithium iodide through cation and anion exchanges and addition of HI. The lithium iodide thus obtained was subjected to mass analysis for the 7Li/6Li isotopic ratio determination. The 7Li/6Li ratios of the mass samples were determined by the surface ionization technique with a Finnigan MAT261 mass spectrometer. The procedure of the lithium isotopic measurements was described elsewhere (Oi et al., 1991). The 7Li-to-6Li single-stage separation factor, S, was determined by Eq. (3) S ¼ (7Li/6Li)solutions/(7Li/6Li)host
(3)
where (7Li/6Li)solution and (7Li/6Li)host are the 7Li-to-6Li isotopic ratios of the electrolyte solution and the cathode, respectively. By definition, S is larger than unity when 6Li is preferentially taken up by the host material. The equilibrium constant, K, of the isotope exchange reaction that is the basis of the deviation of the S value from unity may be expressed as Li(host) þ 6Liþ(solution) ¼ 6Li(host) þ 7Liþ(solution)
7
(4)
where ALi(host) and ALiþ(sol) denote the lithium isotope A in cathodes (graphite or tin) and the electrolyte solution, respectively.
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
The powder X-ray diffraction (XRD) patterns were recorded using a Rigaku RINT2100PC X-ray diffractometer with the CuKa radiation (l ¼ 1.54056 Å) in the 2q range of 5e80 at room temperature. 2.5. Molecular orbital (MO) calculations MO calculations were performed to discuss the lithium isotope effects observed in the present study. As a model of lithium species in the electrolyte solutions, we considered Li(TFSI)2, a lithium ion surrounded by two anionic components of the IL used as the solvent. A lithium atom in graphite was modeled as a lithium atom sandwiched by two polycyclic aromatic hydrocarbon (pah) molecules of the same kind. The LieSn clusters, LieSn, Li2Sn and Li2Sn2, were employed as models of a lithium atom inserted in tin, i. e., LieSn alloys. For comparison, we also considered Li(EC)4þ as a model of lithium species in the EC/MEC mixture solvent. We have already shown in our previous paper (Yanase and Oi, 2002) that EC molecules are preferentially solvated to a Liþ ion in EC/MEC mixture solutions. The structures of those model species were first optimized and frequencies were calculated at respective optimized structures. Reduced partition function ratios (rpfrs; (s/s’)f) were calculated using the calculated frequencies: f ðs=s0 Þf ¼ Pi ¼ 1
ui expðui =2Þ=f1 expðui Þg o .n u0i =2 1 exp u0i
(5)
u0i exp
where ui ¼ hcui/(kT) and u’i ¼ hcu’i/(kT), f is the degree of freedom of molecular vibration, h is the Planck’s constant, c is the velocity of light, ui and u’i are the wave numbers of the ith molecular vibration of the heavier and the lighter isotopic species, respectively, k is the Boltzmann constant, and T is the absolute temperature (Bigeleisen and Mayer, 1947). The K value of Eq. (4) can be estimated using the values of the rpfrs calculated at the optimized structures:
K ¼ ðs=s0 ÞfLiþðSolutionÞ =ðs=s0 ÞfLiðHostÞ
(6)
where (s/s’)f Liþ(Solution) and (s/s’)f Li(Host) represent the rpfr value of the lithium ion in the electrolyte solution and the lithium atom in the host material, respectively. We made no frequency correction since we were mainly interested in relative magnitudes of the rpfr values. The theoretical level was B3LYP for all the calculations, and the basis set used was 6e311 þ G(d) for Li(TFSI)2, 6e311G(d) for a lithium atom sandwiched by two pah molecules and Li(EC)4þ and 3e21G** for the LieSn clusters. The calculations were made using Gaussian 98 and 03 program packages (Frisch et al., 2002). The Gauss View program (Gaussian Inc.) and the Free Wheel program (Butch Software Studio) were used for the graphics. No symmetry consideration was made in the geometry optimization calculations: For each of the structures considered, bond lengths, bond angles and dihedral angles were varied independently to achieve the geometry optimization. Only the mono
1001
isotope substitutions were considered for all the possible combinations of isotopic species. 3. Results and discussion The experimental conditions and results are summarized in Table 1. The results of the MO calculations were summarized in Table 2, and some selected optimized structures are depicted in Fig. 2. A couple of explanations may be necessary concerning those structures. The TFSI anion is bonded to the Liþ ion through two sulfonyl oxygen atoms, ones on each of the two sulfonyl groups (Fig. 2a). It thus works as a bidentate ligand, which leads to the solvation number 4 of the Liþ ion in Li(TFSI)2. In Li(EC)4þ, each EC molecule is bonded to the Liþ ion through its carbonyl oxygen atom, and the Liþ ion is tetrahedrally coordinated with the solvation number of 4 (Fig. 2b). In Li(pyrene)2, the two pyrene molecules are nearly parallel, and the Li atom is located equidistantly from the two molecular planes (Fig. 2c). 3.1. Ionic liquid-graphite system For the IL-graphite system, the electrolysis was continued for 8510e24700 s and the amount of lithium inserted was 0.043e0.057 mmol. The changes of the cathode potential and electric current with the passage of the insertion time of Run IL1 are drawn in Fig. 3. The cathode potential gradually decreased from its original value with increasing insertion time, reached the pre-set value of 0.02 V within 2000 s and hereafter it was kept constant at that voltage. The electric current was kept constant at its original value during the period in which the cathode potential was changing. Upon the cathode potential reaching its pre-set value, the current started decreasing and its final value was 0.15 mA. In this system, the addition of EC to the electrolyte solution was crucially important; without the addition of EC, no appreciable Li insertion was observed. EC was required to form solid electrolyte interface (SEI) on graphite edges at the early stage of the insertion. It is generally accepted that stable surface films, often called SEIs, are formed on graphite electrodes upon the first charging and thereby the carbon surface is passivated (Yanase et al., 2003; Yan et al., 2008). The SEI has lithium ion conductivity but does not show electronic conductivity (Peled, 1979). It, once formed, suppresses further solvent decomposition, but through it lithium ions can be inserted (intercalated) within graphite. The value of S ranged from 0.998 to 1.001, meaning that no appreciable lithium isotope effects were observed in this system within experimental uncertainties. This is quite contrastive to the EC/MEC-graphite system (Yanase et al., 2003) in which lithium ions in the mixture of EC and MEC were electrochemically inserted (intercalated) into graphite to form LiC6 lithium-graphite intercalation compound (Li-GIC) and, accompanying the Li-GIC formation, 6 Li was preferentially fractionated into the graphite phase with the S value ranging from 1.007 to 1.025. Fig. 4 shows the XRD pattern of a lithium inserted graphite electrode (Experimental conditions: temperature, 25 C; period of electrolysis, 13900 s; the amount of Li
Table 1 Summary of experimental conditions and results. Run No
ILL IL2 IL3 Sn1 Sn2
Host of Li insertion
Graphite Graphite Graphite Tin Tin
Electrolyte solution Li concn. (M)
Amount of EC added (g)
0.990 0.995 0.937 1.036 0.970
3.210 0.719 0.155 0 0
Temperature : 25 C, volume of electrolyte solution : 10 cm3, solvent: DEME-TFSI.
Period of electrolysis (sec)
Amount of Li inserted (m mol)
S
14100 24700 8510 73500 65500
0.053 0.057 0.043 0.014 0.012
0.998 1.001 0.999 1.008 1.004
1002
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004 Table 2 Summary of MO calculations on reduced partition function ratios of Li species at 25 C. Li species
rpfr
LiTFSI Li(TFSI)2 Li(EC)4þ Li(benzene)2 Li(phenanthrene)2 Li(pyrene)2 Li(benzoperylene)2 Li-Sn Li2Sn Li2Sn2
1.0775 1.0833 1.0849 1.0367 1.0365 1.0336 1.0208 1.0176 1.0198 1.0220
inserted (estimated value), 0.046 mmol) in the present study. The peak of LiC6 is clearly observed at 2q ¼ 24.03 , which indicates that lithium was certainly intercalated as a neutral atom between graphene layers to form the Li-GIC as in the case of the EC/MECgraphite system. The peak at 2q ¼ 21.34 is ascribable to Li2CO3, which was supposed to have been formed by reactions of lithium with the moisture and carbon dioxide in the air. MO calculations indicate that the rpfr of the lithium ion in the DEME-TFSI solvent is nearly equivalent to that in the EC/MEC solvent as shown in Table 2, which indicates that the S value in the present IL-graphite system should be nearly equivalent to that in the EC/MEC-graphite system. The discrepancy between the theoretical estimation that S value should be nearly equivalent in the IL-graphite system and in the EC/ MEC-graphite system and the experimental results that no appreciable lithium isotope effects were observed in the IL-graphite system can be understood if we assume that the lithium ion is inserted into graphite as LiTFSI, i. e., solvated by a TFSI anion rather
than in totally de-solvated form, and after the insertion the TFSI anion is released from graphite, leaving the neutral lithium atom between graphene layers. In this case, the lithium isotope exchange reaction may be expressed as Li(TFSI)2 þ 7LiTFSI(graphite) ¼ 7Li(TFSI)2 þ 6LiTFSI(graphite) (7)
6
where LiTFSI(graphite) denotes LiTFSI in graphite. Assuming that the rpfr value of LiTFSI(graphite) is equal to that of LiTFSI, the equilibrium constant of this isotope exchange reaction is calculated to be 1.005 at 25 C. Interaction of the LiTFSI molecule with graphene layers may increase the value of rpfr, which would make the equilibrium constant value even smaller. Anyhow, although the value of the equilibrium constant of Eq. (7) may be slightly larger than unity, meaning that 6Li is slightly preferentially fractionated into the graphite phase, the magnitude of the deviation of the S value from unity may be too small to be detected experimentally beyond experimental uncertainties. The calculated distance between the two pah molecules of Li(pah)2, the models of a lithium atom in graphite, is 3.74e4.82 Ǻ, while the calculated size of the LiTFSI molecule (the shorter diameter of LiTFSI regarded as a spheroid) is 4.2e4.5 Ǻ. The calculations thus show that the interlayer distance of graphere layers can accommodate LiTFSI molecules. Although there is no experimentally explicit evidence that a lithium ion was inserted between graphene layers accompanied by a TFSI anion in the present experiments, the following may serve as a supporting evidence for this. We noticed that, after the Li insertion, the Li inserted graphite was much easier to strip off from the copper base of the electrode in the IL-graphite system than in the EC/MEC-graphite system. This may be interpreted as indicating that the adhesion of the graphite to the copper base was heavily
Fig. 2. Optimized structures of a) Li(TFSI)2, b) Li(EC)4þ, c) Li(pyrene)2, and d) Li2Sn2. No significance is attached to the relative sizes of the spheres representing the different kinds of atoms.
4 3 2
Cathode potential 1 0 0
1003
2 Current / mA Cathode potential / V
Current / mA Cathode potential / V
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
Current 5000 10000 Insertion time / sec
15000
Fig. 3. Changes of the cathode potential and electric current with the passage of the insertion time of Run IL1. Solid line, cathode potential; broken line, electric current.
weakened by the deformation of the graphite caused by the instantaneous insertion of LiTFSI and the simultaneously-occurring expansion of spacing between two adjacent graphene layers.
3.2. Ionic liquid-tin system Two experiments of the IL-tin system were carried out in which the insertion of the solvated lithium ion was highly unthinkable and the lithium ion was expected to be inserted in the totally desolvated form. The insertion time was 65500 and 73500 s and the amount of inserted lithium was 0.012 and 0.014 mmol. The changes of the cathode potential and electric current with the passage of the insertion time of Run Sn2 are drawn in Fig. 5. They are quite different from those of the IL-graphite system in Fig. 3. The cathode potential immediately dropped to the pre-set value at the very beginning of the electrolysis and was kept at that value thereafter. The period of the constant electric current of 0.2 mA was practically none. The electric current was very low at first, then gradually increased, and, after showing a maximum value, gradually decreased. Lithium ions are reducedly inserted into the surface of the tin wire to form LieSn alloys and inserted lithium atoms diffuse to the inner (core) part of the wire (Yanase et al., 2000). The formation of the LieSn alloys increases the volume of the surface part of the wire. Due to this swelling, cracks are formed along grain boundaries of the surface area of the tin wire, and this cracking increases the surface area of the tin wire, making lithium insertion easier, which leads to the increase in the electric current. Diffusion of neutral lithium atoms towards the core part of the wire must be very slow, which accounts for the gradual decrease in the electric current in the latter part of the electrolysis. The value of S of two Runs was 1.008 and 1.004. The preferential enrichment of 6Li in the tin phase is thus clearly observed as in the case of the EC/MEC-tin system (Yanase et al., 2000). This is consistent with the results of the MO calculations in a qualitative fashion that the rpfr of the lithium ion in the electrolyte solution is larger than that of the lithium atom in tin as seen in Table 2, although the quantitative agreement between the experiment and the calculations is poor. For a better agreement of the two, the frequency correction for the rpfr calculations and better models of a lithium atom in LieSn alloys are certainly required.
Fig. 4. XRD pattern of a lithium inserted graphite electrode. -, copper: C, graphite; ,, Li2CO3; B, LiC6.
1
Current Cathode potential
0 0
20000 40000 60000 Insertion time / sec
Fig. 5. Changes of the cathode potential and electric current with the passage of the insertion time of Run Sn2. Solid line, cathode potential; broken line, electric current.
4. Conclusions To summarize, we make the following statements: While no lithium isotope effects were observed when lithium ions in the DEME-TFSI ionic liquid solvent are inserted into graphite, the preferential uptake of the lighter isotope, 6Li, by tin from the same electrolyte was clearly observed. It was speculated that, when a lithium ion was inserted into graphite, it is accompanied by a TFSI anion, while it was inserted into tin in the totally de-solvated form. The TFSI anion seems to be bonded to the lithium ion too strongly. An ionic liquid whose anionic component interacts with the lithium ion more weakly than TFSI may be preferable as the solvent of the electrolyte solution for the lithium isotope separation process based on the redox reaction of lithium. Acknowledgements Professor Y. Fujii, Tokyo Institute of Technology (Titech) kindly offered the use of the mass spectrometer. We acknowledge Dr. M. Nomura, Titech, for his assistance in mass spectrometric measurements of lithium isotopic ratios. References Asano, K., Yanase, S., Oi, T., 2008. Lithium isotope effect accompanying electrochemical insertion of lithium into Tin(IV) sulfide. J. Nucl. Sci. Technol. Suppl. 5, 24e29. Bigeleisen, J., Mayer, M.G., 1947. Calculation of equilibrium constants for isotopic exchange reactions. J. Chem. Phys. 15, 261e267. Frisch, M. J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Zakrzewski, V.G., Montgomery, J.A., Stratmann, R.E., Jr., Burant, J.C., Dapprich, S., Millam, J.M., Daniels, A.D., Kudin, K.N., Strain, M.C., Farkas, O., Tomasi, J., Barone, V., Cossi, M., Cammi, R., Mennucci, B., Pomelli, C., Adamo, C., Clifford, S., Ochterski, J., Petersson, G.A., Ayala, P.Y., Cui, Q., Morokuma, K., Rega, N., Salvador, P., Dannenberg, J.J., Malick, D.K., Rabuck, A.D., Raghavachari, K., Foresman, J.B., Cioslowski, J., Ortiz, J.V., Baboul, A.G., Stefanov, B.B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Gomperts, R., Martin, R.L., Fox, D.J., Keith, T., Al-Laham, M.A., Peng, C.Y., Nanayakkara, A., Challacombe, M., Gill, P.M.W., Johnson, B., Chen, W., Wong, M.W., Andres, J.L., Gonzalez, C., Head-Gordon, M., Replogle, E.S., and Pople, J.A., Gaussian, Inc., Pittsburgh PA, 2002. Fujie, M., Fujii, Y., Nomura, M., Okamoto, M., 1986. Isotope effects in electrolytic formation of lithium amalgam. J. Nucl. Sci. Technol. 23, 330e337. Mouri, M., Asano, K., Yanase, S., Oi, T., 2007. Lithium isotope effects accompanying electrochemical insertion of lithium into metal oxides. J. Nucl. Sci. Technol. 44, 73e80. Mouri, M., Yanase, S., Oi, T., 2008. Observation of lithium isotope effects accompanying electrochemical insertion of lithium into zinc. J. Nucl. Sci. Technol. 45, 384e389. Oi, T., Kawada, K., Hosoe, M., Kakihana, H., 1991. Fractionation of lithium isotopes in cation-exchange chromatography. Sep. Sci. Technol. 26, 1353e1375. Palko, A.A., Drury, J.S., Begun, G.M., 1976. Lithium isotope separation factors of some two-phase equilibrium systems. J. Chem. Phys. 64, 1828e1837. Peled, E., 1979. The electrochemical behavior of alkali and alkaline earth metals in nonaqueous battery systems - the solid electrolyte interphase model. J. Electrochem. Soc. 126, 2047e2051.
1004
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
Sato, T., Maruo, T., Marukane, S., Takagi, K., 2004. Ionic liquids containing carbonate solvent as electrolytes for lithium ion cells. J. Power Sources 138, 253e261. Yan, J., Xia, B., Su, Y., Zhou, X., Zhang, J., Zhang, X., 2008. Phenomenologically modeling the formation and evolution of the solid electrolyte interface on the graphite electrode for lithium-ion batteries. Electrochim. Acta 53, 7069e7078. Yanase, S., Oi, T., 2002. Solvation of lithium ion in organic electrolyte solutions and its isotopic reduced partition function ratios studied by ab initio molecular orbital method. J. Nucl. Sci. Technol. 39, 1060e1064.
Yanase, S., Oi, T., Hashikawa, S., 2000. Observation of lithium isotope effect accompanying electrochemical insertion of lithium into tin. J. Nucl. Sci. Technol. 37, 919e923. Yanase, S., Hayama, W., Oi, T., 2003. Lithium isotope effect accompanying electrochemical intercalation of lithium into graphite. Z. Naturforsch 58a, 306e312. Zenzai, K., Yanase, S., Zhang, Y.-H., Oi, T., 2008. Lithium isotope effect accompanying electrochemical insertion of lithium into gallium. Prog. Nucl. Energy 50, 494e498. Zenzai, K., Yasui, A., Yanase, S., Oi, T., 2010. Lithium isotope effect accompanying electrochemical insertion of lithium into liquid gallium. Z. Naturforschg A 65a, 461e467.
Contents lists available at ScienceDirect
Progress in Nuclear Energy journal homepage: www.elsevier.com/locate/pnucene
Lithium isotope effects upon electrochemical lithium insertion to host material from ionic liquid medium Shun Saito, Yuta Takami, Masahiro Yoshizawa-Fujita, Satoshi Yanase, Takao Oi* Faculty of Science and Technology, Sophia University, 7-1 Kioicho, Chiyodaku, Tokyo 102-8554, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 October 2010 Received in revised form 7 February 2011 Accepted 29 April 2011
Lithium was electrochemically inserted from a Liþ ion containing ionic liquid into graphite or tin to observe lithium isotope effects that accompanied the insertion. While no preferential uptake of the lithium isotopes was detected with graphite, the lighter isotope, 6Li, was preferentially fractionated into tin with the single-stage lithium isotope separation factors, S, ranging from 1.004 to 1.008 at 25 C. It was speculated that a Liþ ion was inserted into graphite together with an anionic component of the ionic liquid and, upon the reduction of the Liþ ion to a lithium atom, the anion was released from graphite, while a Liþ ion alone was inserted into tin. Molecular orbital calculations supported this speculation in a qualitative fashion. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Lithium isotopes Isotope effects Ionic liquid Graphite Separation factor Reduced partition function ratio
1. Introduction Lithium has been and still is a major target of isotope separation study. This is in part because of the importance of the isolated and enriched isotopes of lithium in nuclear science and industry. The heavier isotope of lithium, 7Li, is used as pH controller of coolants in nuclear fission reactors. A large demand for the lighter isotope, 6Li, is expected in DT fusion power reactors in the future where lithium compounds rich in 6Li will be required for the tritium breeder blanket. The only method that was applied to a large-scale lithium isotope separation is the amalgam method (Palko et al., 1976). In this method, lithium is distributed between the amalgam phase and the aqueous or organic electrolyte solution phase and lithium isotope separation is practiced based on the lithium isotope exchange reaction between the two phases, 7
Li(Hg) þ 6Liþ(solution) ¼ 6Li(Hg) þ 7Liþ(solution),
(1)
where ALi(Hg) and ALiþ(solution) denote the lithium isotope A in the oxidation state 0 surrounded by mercury atoms in the amalgam phase and that isotope in the oxidation state þ1 surrounded by
* Corresponding author. Tel.: þ81 3 3238 3359; fax: þ81 3 3238 3361. E-mail address: [email protected] (T. Oi). 0149-1970/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.pnucene.2011.04.017
solvent molecules in the electrolyte phase, respectively. The reported value of the 7Li-to-6Li single-stage separation factor, S, defined as, S ¼ (7Li/6Li)solution/(7Li/6Li)Hg,
(2)
where (7Li/6Li)B denotes the 7Li-to-6Li isotopic ratio in phase B, is 1.049e1.062 for the water solvent (Fujie et al., 1986), meaning that 6 Li is preferentially fractionated in the amalgam phase. Although the value of 1.049e1.062 is large and attractive, the use of toxic mercury brings about biological and environmental problems and makes the amalgam method difficult to be applied to large-scale lithium isotope enrichment in the future. As a possible alternative of mercury, we investigated tin (Yanase et al., 2000), gallium (Zenzai et al., 2008, 2010), zinc (Mouri et al., 2008), graphite (Yanase et al., 2003) and metal oxides and sulfides (Mouri et al., 2007; Asano et al., 2008). We found that the metals, metal sulfide and graphite preferentially take up 6Li, while metal oxides showed little preference or were slightly 7Li-specific. Any host materials for lithium insertion so far investigated, however, did not show lithium isotope effects that surpassed those on mercury. The electrolyte solution so far investigated was in most cases a mixed solution of ethylene carbonate (EC) and methylethyl carbonate (MEC) containing LiClO4. Larger lithium isotope effects may be obtained with solvents other than carbonates. Ionic liquids (ILs) may be candidates of solvents for lithium isotope separation
1000
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
processes based on redox reactions of lithium. They are being investigated as solvents in secondary lithium ion batteries with higher performance than those of concurrent organic solvents (Sato et al., 2004). We attempted to use an IL as an alternative of the EC/ MEC mixture solution. In this paper, we report lithium isotope effects accompanying electrochemical lithium insertion into graphite or tin from an IL containing lithium ions and, in addition, results of molecular orbital (MO) calculations conducted to help the elucidation of the lithium isotope effects observed in the experiments. 2. Experimental 2.1. Electrodes Two kinds of material were used as cathode (the host of the lithium insertion). One was commercially available graphite electrodes for the use as cathodes of secondary lithium ion batteries (graphite powder daubed on a copper foil; capacity, 1.6 mAh/cm2; density 1.1 g/cm3), purchased from Hohsen Corp. Another one was tin wires with a purity of 99.9% and the diameter of 0.5 mm purchased from Kojundo Chemical Laboratory Co. Ltd. Lithium foils with a purity of 99.8% and about 0.3 mm thick purchased from Honjo Metals Co. Ltd. were used as anodes and reference electrodes. 2.2. Electrolyte solution The electrolyte solution used was an IL, N,N-diethyl-Nmethoxyethyl ammonium bis(trifluoromethanesulfonyl)imide {[(C2H5)2(CH3)(CH3OC2H2)N]þ[(CF3SO2)2N]} (DEME-TFSI) containing lithium bis(trifluoromethanesulfonyl)imide (LieTFSI). In electrolytic experiments with graphite electrodes, EC was added to the electrolyte solution. DEME-TFSI, LieTFSI and EC were purchased from Kanto Chemical Co. Inc., Morita Chemical Industries Co. Ltd. and Wako Pure Chemical Industries Ltd., respectively. The electrolyte solution was prepared under a dry Ar atmosphere. 2.3. Lithium insertion The experimental apparatus used was basically the same as the one used in our previous studies (Yanase et al., 2000; Mouri et al., 2008; Yanase et al., 2003; Mouri et al., 2007; Asano et al., 2008) and schematically drawn in Fig. 1. It was composed of a power supply (a Hokuto Denko Corporation HJ-201B battery charge/discharge unit), a cylinder-shaped three-electrode electrochemical glass cell (electrolytic cell) of about 12 cm in height and 3 cm in diameter and a data acquisition unit consisting of an A/D converter and a personal computer. The volume of the electrolyte solution placed in the cell was about 10 cm3. The size of the graphite electrode was 1 cm 1 cm, and the weight of the tin electrode was about 0.1 g. The electrolytic cell was build up in a dry argon atmosphere as shown in Fig. 1. The lithium insertion was performed in the constant currentconstant voltage mode, which was designed to avoid the possible deposit of lithium on the cathode surface. That is, the electrolysis (lithium insertion) was at first carried out in a constant current mode (1 mA for graphite electrodes and 0.2 mA for tin electrodes). As the electrolysis proceeded, the electric potential of the cathode against the reference electrode (cathode potential), in general decreased and reached the pre-set value (0.02 V for graphite electrodes and 0.2 V for tin electrodes), the value chosen based on our previous similar studies in which the deposit of lithium on surfaces of cathode materials was not observed at this cathode potential. The electrolytic mode was then automatically changed to the constant voltage mode. The electrolysis continued until the integrated quantity of electricity reached the predetermined value
Fig. 1. Schematic drawing of experimental apparatus. 1, Personal computer; 2, A/D converter; 3, Power supply; 4, reference electrode; 5, cathode (graphite or tin); 6, anode; 7, electrolytic cell; 8, electrolyte solution; 9, stirrer tip.
and was discontinued manually. During the electrolysis, the temperature of the electrolytic cell was kept at 25 C and the electrolyte solution was stirred with a magnetic stirrer.
2.4. Analyses and measurements Lithium inserted in graphite electrodes was extracted with a dilute (ca. 0.1 M) hydrochloric acid. The procedure of lithium extraction from the lithium inserted tin cathodes has been described elsewhere (Yanase et al., 2000). Lithium extraction from the electrolyte solution was carried out as follows: To ca. 10 cm3 of the electrolyte solution after the electrolysis was added about the same volume of pure water. The two layers were stirred furiously with a magnetic stirrer till they became a sort of an opaque mixture. This mixture was then centrifuged at 2000 rpm for 10 min to separate it into the aqueous and IL phases again. Lithium in the electrolyte solution was extracted into the aqueous phase in this procedure. The amount of lithium inserted into the cathodes was determined by measuring the lithium concentration in a part of the aqueous solution containing the extracted lithium by flame photometry with a Thermo Electron Corp. SOLAAR M mkII atomic absorption spectrometer. The chemical form of lithium in another part of the aqueous solution was converted to lithium iodide through cation and anion exchanges and addition of HI. The lithium iodide thus obtained was subjected to mass analysis for the 7Li/6Li isotopic ratio determination. The 7Li/6Li ratios of the mass samples were determined by the surface ionization technique with a Finnigan MAT261 mass spectrometer. The procedure of the lithium isotopic measurements was described elsewhere (Oi et al., 1991). The 7Li-to-6Li single-stage separation factor, S, was determined by Eq. (3) S ¼ (7Li/6Li)solutions/(7Li/6Li)host
(3)
where (7Li/6Li)solution and (7Li/6Li)host are the 7Li-to-6Li isotopic ratios of the electrolyte solution and the cathode, respectively. By definition, S is larger than unity when 6Li is preferentially taken up by the host material. The equilibrium constant, K, of the isotope exchange reaction that is the basis of the deviation of the S value from unity may be expressed as Li(host) þ 6Liþ(solution) ¼ 6Li(host) þ 7Liþ(solution)
7
(4)
where ALi(host) and ALiþ(sol) denote the lithium isotope A in cathodes (graphite or tin) and the electrolyte solution, respectively.
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
The powder X-ray diffraction (XRD) patterns were recorded using a Rigaku RINT2100PC X-ray diffractometer with the CuKa radiation (l ¼ 1.54056 Å) in the 2q range of 5e80 at room temperature. 2.5. Molecular orbital (MO) calculations MO calculations were performed to discuss the lithium isotope effects observed in the present study. As a model of lithium species in the electrolyte solutions, we considered Li(TFSI)2, a lithium ion surrounded by two anionic components of the IL used as the solvent. A lithium atom in graphite was modeled as a lithium atom sandwiched by two polycyclic aromatic hydrocarbon (pah) molecules of the same kind. The LieSn clusters, LieSn, Li2Sn and Li2Sn2, were employed as models of a lithium atom inserted in tin, i. e., LieSn alloys. For comparison, we also considered Li(EC)4þ as a model of lithium species in the EC/MEC mixture solvent. We have already shown in our previous paper (Yanase and Oi, 2002) that EC molecules are preferentially solvated to a Liþ ion in EC/MEC mixture solutions. The structures of those model species were first optimized and frequencies were calculated at respective optimized structures. Reduced partition function ratios (rpfrs; (s/s’)f) were calculated using the calculated frequencies: f ðs=s0 Þf ¼ Pi ¼ 1
ui expðui =2Þ=f1 expðui Þg o .n u0i =2 1 exp u0i
(5)
u0i exp
where ui ¼ hcui/(kT) and u’i ¼ hcu’i/(kT), f is the degree of freedom of molecular vibration, h is the Planck’s constant, c is the velocity of light, ui and u’i are the wave numbers of the ith molecular vibration of the heavier and the lighter isotopic species, respectively, k is the Boltzmann constant, and T is the absolute temperature (Bigeleisen and Mayer, 1947). The K value of Eq. (4) can be estimated using the values of the rpfrs calculated at the optimized structures:
K ¼ ðs=s0 ÞfLiþðSolutionÞ =ðs=s0 ÞfLiðHostÞ
(6)
where (s/s’)f Liþ(Solution) and (s/s’)f Li(Host) represent the rpfr value of the lithium ion in the electrolyte solution and the lithium atom in the host material, respectively. We made no frequency correction since we were mainly interested in relative magnitudes of the rpfr values. The theoretical level was B3LYP for all the calculations, and the basis set used was 6e311 þ G(d) for Li(TFSI)2, 6e311G(d) for a lithium atom sandwiched by two pah molecules and Li(EC)4þ and 3e21G** for the LieSn clusters. The calculations were made using Gaussian 98 and 03 program packages (Frisch et al., 2002). The Gauss View program (Gaussian Inc.) and the Free Wheel program (Butch Software Studio) were used for the graphics. No symmetry consideration was made in the geometry optimization calculations: For each of the structures considered, bond lengths, bond angles and dihedral angles were varied independently to achieve the geometry optimization. Only the mono
1001
isotope substitutions were considered for all the possible combinations of isotopic species. 3. Results and discussion The experimental conditions and results are summarized in Table 1. The results of the MO calculations were summarized in Table 2, and some selected optimized structures are depicted in Fig. 2. A couple of explanations may be necessary concerning those structures. The TFSI anion is bonded to the Liþ ion through two sulfonyl oxygen atoms, ones on each of the two sulfonyl groups (Fig. 2a). It thus works as a bidentate ligand, which leads to the solvation number 4 of the Liþ ion in Li(TFSI)2. In Li(EC)4þ, each EC molecule is bonded to the Liþ ion through its carbonyl oxygen atom, and the Liþ ion is tetrahedrally coordinated with the solvation number of 4 (Fig. 2b). In Li(pyrene)2, the two pyrene molecules are nearly parallel, and the Li atom is located equidistantly from the two molecular planes (Fig. 2c). 3.1. Ionic liquid-graphite system For the IL-graphite system, the electrolysis was continued for 8510e24700 s and the amount of lithium inserted was 0.043e0.057 mmol. The changes of the cathode potential and electric current with the passage of the insertion time of Run IL1 are drawn in Fig. 3. The cathode potential gradually decreased from its original value with increasing insertion time, reached the pre-set value of 0.02 V within 2000 s and hereafter it was kept constant at that voltage. The electric current was kept constant at its original value during the period in which the cathode potential was changing. Upon the cathode potential reaching its pre-set value, the current started decreasing and its final value was 0.15 mA. In this system, the addition of EC to the electrolyte solution was crucially important; without the addition of EC, no appreciable Li insertion was observed. EC was required to form solid electrolyte interface (SEI) on graphite edges at the early stage of the insertion. It is generally accepted that stable surface films, often called SEIs, are formed on graphite electrodes upon the first charging and thereby the carbon surface is passivated (Yanase et al., 2003; Yan et al., 2008). The SEI has lithium ion conductivity but does not show electronic conductivity (Peled, 1979). It, once formed, suppresses further solvent decomposition, but through it lithium ions can be inserted (intercalated) within graphite. The value of S ranged from 0.998 to 1.001, meaning that no appreciable lithium isotope effects were observed in this system within experimental uncertainties. This is quite contrastive to the EC/MEC-graphite system (Yanase et al., 2003) in which lithium ions in the mixture of EC and MEC were electrochemically inserted (intercalated) into graphite to form LiC6 lithium-graphite intercalation compound (Li-GIC) and, accompanying the Li-GIC formation, 6 Li was preferentially fractionated into the graphite phase with the S value ranging from 1.007 to 1.025. Fig. 4 shows the XRD pattern of a lithium inserted graphite electrode (Experimental conditions: temperature, 25 C; period of electrolysis, 13900 s; the amount of Li
Table 1 Summary of experimental conditions and results. Run No
ILL IL2 IL3 Sn1 Sn2
Host of Li insertion
Graphite Graphite Graphite Tin Tin
Electrolyte solution Li concn. (M)
Amount of EC added (g)
0.990 0.995 0.937 1.036 0.970
3.210 0.719 0.155 0 0
Temperature : 25 C, volume of electrolyte solution : 10 cm3, solvent: DEME-TFSI.
Period of electrolysis (sec)
Amount of Li inserted (m mol)
S
14100 24700 8510 73500 65500
0.053 0.057 0.043 0.014 0.012
0.998 1.001 0.999 1.008 1.004
1002
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004 Table 2 Summary of MO calculations on reduced partition function ratios of Li species at 25 C. Li species
rpfr
LiTFSI Li(TFSI)2 Li(EC)4þ Li(benzene)2 Li(phenanthrene)2 Li(pyrene)2 Li(benzoperylene)2 Li-Sn Li2Sn Li2Sn2
1.0775 1.0833 1.0849 1.0367 1.0365 1.0336 1.0208 1.0176 1.0198 1.0220
inserted (estimated value), 0.046 mmol) in the present study. The peak of LiC6 is clearly observed at 2q ¼ 24.03 , which indicates that lithium was certainly intercalated as a neutral atom between graphene layers to form the Li-GIC as in the case of the EC/MECgraphite system. The peak at 2q ¼ 21.34 is ascribable to Li2CO3, which was supposed to have been formed by reactions of lithium with the moisture and carbon dioxide in the air. MO calculations indicate that the rpfr of the lithium ion in the DEME-TFSI solvent is nearly equivalent to that in the EC/MEC solvent as shown in Table 2, which indicates that the S value in the present IL-graphite system should be nearly equivalent to that in the EC/MEC-graphite system. The discrepancy between the theoretical estimation that S value should be nearly equivalent in the IL-graphite system and in the EC/ MEC-graphite system and the experimental results that no appreciable lithium isotope effects were observed in the IL-graphite system can be understood if we assume that the lithium ion is inserted into graphite as LiTFSI, i. e., solvated by a TFSI anion rather
than in totally de-solvated form, and after the insertion the TFSI anion is released from graphite, leaving the neutral lithium atom between graphene layers. In this case, the lithium isotope exchange reaction may be expressed as Li(TFSI)2 þ 7LiTFSI(graphite) ¼ 7Li(TFSI)2 þ 6LiTFSI(graphite) (7)
6
where LiTFSI(graphite) denotes LiTFSI in graphite. Assuming that the rpfr value of LiTFSI(graphite) is equal to that of LiTFSI, the equilibrium constant of this isotope exchange reaction is calculated to be 1.005 at 25 C. Interaction of the LiTFSI molecule with graphene layers may increase the value of rpfr, which would make the equilibrium constant value even smaller. Anyhow, although the value of the equilibrium constant of Eq. (7) may be slightly larger than unity, meaning that 6Li is slightly preferentially fractionated into the graphite phase, the magnitude of the deviation of the S value from unity may be too small to be detected experimentally beyond experimental uncertainties. The calculated distance between the two pah molecules of Li(pah)2, the models of a lithium atom in graphite, is 3.74e4.82 Ǻ, while the calculated size of the LiTFSI molecule (the shorter diameter of LiTFSI regarded as a spheroid) is 4.2e4.5 Ǻ. The calculations thus show that the interlayer distance of graphere layers can accommodate LiTFSI molecules. Although there is no experimentally explicit evidence that a lithium ion was inserted between graphene layers accompanied by a TFSI anion in the present experiments, the following may serve as a supporting evidence for this. We noticed that, after the Li insertion, the Li inserted graphite was much easier to strip off from the copper base of the electrode in the IL-graphite system than in the EC/MEC-graphite system. This may be interpreted as indicating that the adhesion of the graphite to the copper base was heavily
Fig. 2. Optimized structures of a) Li(TFSI)2, b) Li(EC)4þ, c) Li(pyrene)2, and d) Li2Sn2. No significance is attached to the relative sizes of the spheres representing the different kinds of atoms.
4 3 2
Cathode potential 1 0 0
1003
2 Current / mA Cathode potential / V
Current / mA Cathode potential / V
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
Current 5000 10000 Insertion time / sec
15000
Fig. 3. Changes of the cathode potential and electric current with the passage of the insertion time of Run IL1. Solid line, cathode potential; broken line, electric current.
weakened by the deformation of the graphite caused by the instantaneous insertion of LiTFSI and the simultaneously-occurring expansion of spacing between two adjacent graphene layers.
3.2. Ionic liquid-tin system Two experiments of the IL-tin system were carried out in which the insertion of the solvated lithium ion was highly unthinkable and the lithium ion was expected to be inserted in the totally desolvated form. The insertion time was 65500 and 73500 s and the amount of inserted lithium was 0.012 and 0.014 mmol. The changes of the cathode potential and electric current with the passage of the insertion time of Run Sn2 are drawn in Fig. 5. They are quite different from those of the IL-graphite system in Fig. 3. The cathode potential immediately dropped to the pre-set value at the very beginning of the electrolysis and was kept at that value thereafter. The period of the constant electric current of 0.2 mA was practically none. The electric current was very low at first, then gradually increased, and, after showing a maximum value, gradually decreased. Lithium ions are reducedly inserted into the surface of the tin wire to form LieSn alloys and inserted lithium atoms diffuse to the inner (core) part of the wire (Yanase et al., 2000). The formation of the LieSn alloys increases the volume of the surface part of the wire. Due to this swelling, cracks are formed along grain boundaries of the surface area of the tin wire, and this cracking increases the surface area of the tin wire, making lithium insertion easier, which leads to the increase in the electric current. Diffusion of neutral lithium atoms towards the core part of the wire must be very slow, which accounts for the gradual decrease in the electric current in the latter part of the electrolysis. The value of S of two Runs was 1.008 and 1.004. The preferential enrichment of 6Li in the tin phase is thus clearly observed as in the case of the EC/MEC-tin system (Yanase et al., 2000). This is consistent with the results of the MO calculations in a qualitative fashion that the rpfr of the lithium ion in the electrolyte solution is larger than that of the lithium atom in tin as seen in Table 2, although the quantitative agreement between the experiment and the calculations is poor. For a better agreement of the two, the frequency correction for the rpfr calculations and better models of a lithium atom in LieSn alloys are certainly required.
Fig. 4. XRD pattern of a lithium inserted graphite electrode. -, copper: C, graphite; ,, Li2CO3; B, LiC6.
1
Current Cathode potential
0 0
20000 40000 60000 Insertion time / sec
Fig. 5. Changes of the cathode potential and electric current with the passage of the insertion time of Run Sn2. Solid line, cathode potential; broken line, electric current.
4. Conclusions To summarize, we make the following statements: While no lithium isotope effects were observed when lithium ions in the DEME-TFSI ionic liquid solvent are inserted into graphite, the preferential uptake of the lighter isotope, 6Li, by tin from the same electrolyte was clearly observed. It was speculated that, when a lithium ion was inserted into graphite, it is accompanied by a TFSI anion, while it was inserted into tin in the totally de-solvated form. The TFSI anion seems to be bonded to the lithium ion too strongly. An ionic liquid whose anionic component interacts with the lithium ion more weakly than TFSI may be preferable as the solvent of the electrolyte solution for the lithium isotope separation process based on the redox reaction of lithium. Acknowledgements Professor Y. Fujii, Tokyo Institute of Technology (Titech) kindly offered the use of the mass spectrometer. We acknowledge Dr. M. Nomura, Titech, for his assistance in mass spectrometric measurements of lithium isotopic ratios. References Asano, K., Yanase, S., Oi, T., 2008. Lithium isotope effect accompanying electrochemical insertion of lithium into Tin(IV) sulfide. J. Nucl. Sci. Technol. Suppl. 5, 24e29. Bigeleisen, J., Mayer, M.G., 1947. Calculation of equilibrium constants for isotopic exchange reactions. J. Chem. Phys. 15, 261e267. Frisch, M. J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Zakrzewski, V.G., Montgomery, J.A., Stratmann, R.E., Jr., Burant, J.C., Dapprich, S., Millam, J.M., Daniels, A.D., Kudin, K.N., Strain, M.C., Farkas, O., Tomasi, J., Barone, V., Cossi, M., Cammi, R., Mennucci, B., Pomelli, C., Adamo, C., Clifford, S., Ochterski, J., Petersson, G.A., Ayala, P.Y., Cui, Q., Morokuma, K., Rega, N., Salvador, P., Dannenberg, J.J., Malick, D.K., Rabuck, A.D., Raghavachari, K., Foresman, J.B., Cioslowski, J., Ortiz, J.V., Baboul, A.G., Stefanov, B.B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Gomperts, R., Martin, R.L., Fox, D.J., Keith, T., Al-Laham, M.A., Peng, C.Y., Nanayakkara, A., Challacombe, M., Gill, P.M.W., Johnson, B., Chen, W., Wong, M.W., Andres, J.L., Gonzalez, C., Head-Gordon, M., Replogle, E.S., and Pople, J.A., Gaussian, Inc., Pittsburgh PA, 2002. Fujie, M., Fujii, Y., Nomura, M., Okamoto, M., 1986. Isotope effects in electrolytic formation of lithium amalgam. J. Nucl. Sci. Technol. 23, 330e337. Mouri, M., Asano, K., Yanase, S., Oi, T., 2007. Lithium isotope effects accompanying electrochemical insertion of lithium into metal oxides. J. Nucl. Sci. Technol. 44, 73e80. Mouri, M., Yanase, S., Oi, T., 2008. Observation of lithium isotope effects accompanying electrochemical insertion of lithium into zinc. J. Nucl. Sci. Technol. 45, 384e389. Oi, T., Kawada, K., Hosoe, M., Kakihana, H., 1991. Fractionation of lithium isotopes in cation-exchange chromatography. Sep. Sci. Technol. 26, 1353e1375. Palko, A.A., Drury, J.S., Begun, G.M., 1976. Lithium isotope separation factors of some two-phase equilibrium systems. J. Chem. Phys. 64, 1828e1837. Peled, E., 1979. The electrochemical behavior of alkali and alkaline earth metals in nonaqueous battery systems - the solid electrolyte interphase model. J. Electrochem. Soc. 126, 2047e2051.
1004
S. Saito et al. / Progress in Nuclear Energy 53 (2011) 999e1004
Sato, T., Maruo, T., Marukane, S., Takagi, K., 2004. Ionic liquids containing carbonate solvent as electrolytes for lithium ion cells. J. Power Sources 138, 253e261. Yan, J., Xia, B., Su, Y., Zhou, X., Zhang, J., Zhang, X., 2008. Phenomenologically modeling the formation and evolution of the solid electrolyte interface on the graphite electrode for lithium-ion batteries. Electrochim. Acta 53, 7069e7078. Yanase, S., Oi, T., 2002. Solvation of lithium ion in organic electrolyte solutions and its isotopic reduced partition function ratios studied by ab initio molecular orbital method. J. Nucl. Sci. Technol. 39, 1060e1064.
Yanase, S., Oi, T., Hashikawa, S., 2000. Observation of lithium isotope effect accompanying electrochemical insertion of lithium into tin. J. Nucl. Sci. Technol. 37, 919e923. Yanase, S., Hayama, W., Oi, T., 2003. Lithium isotope effect accompanying electrochemical intercalation of lithium into graphite. Z. Naturforsch 58a, 306e312. Zenzai, K., Yanase, S., Zhang, Y.-H., Oi, T., 2008. Lithium isotope effect accompanying electrochemical insertion of lithium into gallium. Prog. Nucl. Energy 50, 494e498. Zenzai, K., Yasui, A., Yanase, S., Oi, T., 2010. Lithium isotope effect accompanying electrochemical insertion of lithium into liquid gallium. Z. Naturforschg A 65a, 461e467.