Microwave activated lithium intercalation in transition metal sulfides

Materials Research Bulletin, Vol. 32, No. 6, pp. 709-717, 1997 Copyright 0 1997 Elsevier Science Ltd Printed in the USA. All rights resewed 0025-5408/97 $17.00 + .OO

Pergamon

PI1 SOOZS-5408(97)00037-S

MICROWAVE ACTIVATED LITHIUM INTERCALATION IN TRANSITION METAL SULFIDES

E. Benavente and G. Gontilez” Department of Chemistry, Faculty of Sciences, Universidad de Chile, Casilla 653, Santiago de Chile (Refereed) (Received October 10, 1996; Accepted December 18, 1996)

ABSTRACT The reaction rates for the intercalation of lithium in molybdenum and titanium sulfide activated by microwave irradiation at room temperature and atmospheric pressure leading to products of relatively high crystallinity are about two orders of magnitude higher than those by conventional thermal methods. Nevertheless, microwave irradiation of titanium sulfide samples produces appreciable decomposition. A similar effect is observed for the intercalation of some organic and organometallic species in LiMo&. Acceleration observed for microwave assisted lithium intercalation reactions appears to be related with mechanistic changes which facilitate a first stage intercalation. 9 1997 Elsevier Science Lrd KEYWORDS: A. layered compounds, B. intercalation reactions, electrochemical measurements, D. diffusion, D. electrical properties

C.

INTRODUCTION The intercalation of lithium in transition metal dichalcogenides have been the subject of numerous investigations during the last 20 years (1). An important part of the steady interest in these compounds concerns the use of lamellar transition metal sulfides as negative electrodes in lithium-based rechargeable batteries (2). Electrodes based on titanium and molybdenum disulfides have been used in commercially proposed devices (3,4). One of the

*Corresponding author. Mailing address: Prof. Guillermo GonzBlez-Moraga, Departamento de Quimica, Facultad de Ciencias, Universidad de Chile, Casilla 653, Santiago de Chile, Tel. 56-2-678 7255, FAX 56-2-271 3888, E-mail . 709

710

E. BENAVENTE

ef al.

Vol. 32, No. 6

advantages of using transition metal dichalcogenides as electrode materials is the relatively high electronic conductivity of their lithium intercalates. Titanium as well as molybdenum dichalcogenides are intrinsically semiconductors; some titanium derivatives, however, also show semimetallic behavior in the intercalated state (5). Moreover, lithium intercalates are important as synthetic intermediates in the preparation of other intercalation compounds (6). Not only alkali metals may be inserted into the interlaminar spaces of the transition metal dichalcogenides. The intercalation of donor species such as amines, ethers, polymeric as well as organometallic species is well known (7). Attempts to improve the mass transport, a property also required by efficient electrode materials, by co-intercalating polymeric donor species in the interlayer spaces of molybdenum sulfide have been performed recently (8). Nanocomposites produced by the intercalation of lithium and poly(ethylene oxide) or polyacrylonitrile in MO& show lithium diffusion coefficients at room temperature which are more than one magnitude order higher than that of pristine sulfide. One of the major problems of preparative intercalation chemistry is the slow rate of the insertion processes, which often require long reaction times and elevated temperatures. The development of synthetic methods which could accelerate the intercalation softening the reaction conditions in some cases is necessary. Currently, the method most frequently used for intercalating lithium in transition metal dichalcogenides is the reaction of a suspension of the chalcogenide in an inert, with a moderate boiling point solvent with butyl lithium (9). MS + BuLi --+ Li,MSz A near-stoichiometric reaction may be normally achieved by refluxing the suspension of the reactants for a couple of hours. Although the reaction with titanium disulfide is relatively easy, requiring less than an hour to reach intercalation of 1 mol of lithium per mol of sulfide, molybdenum disulfide needs about 24 hours to reach the same intercalation degree. Long reflux times imply not only a waste of time but also an often not desired diminution of the crystallinity of the products; the development of softer intercalation methods is therefore worthwhile, especially for molybdenum sulfide. Ultrasonic irradiation of reaction mixtures strongly increases the rate of intercalation of organic and organometallic compounds into Mo03, ZrS2, and TaS2 (10). However, the effect of ultrasonic irradiation appears to be produced more by alteration of the particle size and damage of the solid surface than by mass transfer improvement. Thus, although the intercalation reaction rates increase significantly by soniiication, the crystallinity of tinal products is much inferior than that of the materials obtained by conventional methods. The topotactic intercalation of donor species in transition metal dichalcogenides by direct methods, as occurs, for instance, in the intercalation of amines in TaSz (1 l), is possible only in very few cases. Microwave treatments offer, in general, an alternative method for activating chemical reactions (12). Industrial use of microwave radiation as an alternative to conventional thermal heating has generated great interest and has been widely applied in the recent years. However, the chemical mechanism of interaction in such processes has not been well understood (13). The question of whether there are microwave specific effects other than dielectric heating is still open. Although some interesting antecedents about the use of microwave for improving intercalation processes are indeed known (14), no information about the application of this method to the intercalation in transition metal chalcogenides is available. In this paper, a study of the intercalation of lithium in both titanium and

Vol. 32, No. 6

METAL SULFIDE INTERCALATION

molybdenum disulfides by microwave-assisted application of the method to the intercalation species is outlined.

711

processes is described. Furthermore, the of some organic and organometallic donor

EXPERIMENTAL Molybdenum sulfide (Fluka, purum, size l-3 pm) was used as received. Butyl freshly prepared according to standard procedures (15). Solvents, n-hexane, were dried and distilled under argon. Polyacrylonitrile (Aldrich), ferrocene styrene (Aldrich) were used as received. Reactions were performed in Schlenk dry argon atmosphere.

lithium was and benzene (Fluka), and flasks under

Lithium Intercalation. Typical procedure: To a suspension of the metal disulfide in n-hexane, one (for titanium) or two (for molybdenum) equivalents of butyl lithium in the same solvent were added at room temperature. One part of the reaction batch was stirred under reflux. The other one was treated in a microwave oven. In both types of experiment, the development of the reaction was monitored by periodically determining the unreacted butyl lithium in 0.2 ml aliquots of the reaction mixtures. This was done with double titration technique (16). The results of some experiments performed by monitoring the amount of lithium intercalated in the solid by atomic absorption spectrophotometry at 670.8 nm with a sensitivity of 0.035 pg/mL do not differ significantly from those obtained by determining BuLi. Furthermore, final intercalation products were always characterized by both X-ray difiaction analysis and lithium content. Organic and Organometallic Species Intercalation. As described in ref. 17, a suspension of one equivalent of the lithiated metal disulfide, Li,Mo& (x 2 1.O), in n-hexane or benzene

OP

0

Dependence treatment.

of lithium

I

5

intercalation

I

10 time (hrs)

I

15

FIG. 1 degree on the reaction

!

20

25

time for conventional

thermal

E. BENAVENTE et al.

712

Vol. 32, No. 6

was treated with two equivalents of the donor following a procedure similar to that outlined above for lithium intercalation. Purity and crystallinity of pristine metal sulfides as well as their intercalation products were estimated by powder X-ray diffraction analysis. Diffractograms were recorded in the range 2O > 28 > 20” on a Siemens D-5000 diffractometer (using Cu Ka radiation and a graphite monochromator, h = 1.5418 A). The stoichiometry of the products was determined by elemental analysis (vide infra). The state of the donor in the intercalation products was corroborated by thermal analysis and, in the case of ferrocene, by cyclic voltammetry. Electric conductivity measurements performed by ac electrochemical impedance analysis (EG&G PAR Model 63 10) were used to exclude the possibility of donor-MoS1 mechanical mixtures. Elemental analysis: LiO $MoSZ(C~HCH=CH z)017, C 5.82, H 0.48, Li 2.55%; LiosMoSz[(CsHs)2Fe]o Is C 5.33, H 0.44, Li 1.96%; and LiosMoS2(PAN), 0, C 18.43, H 2.25, N 7.01, Li 2.1%. Microwave irradiation was performed in a domestic oven (Samsung RE 725-TC) at 2450 MHz and 325 W power. Irradiation was performed stepwise in order to avoid temperature and pressure jumps. Temperature was maintained at about 60°C. Near atmospheric pressure was secured by connecting the reaction vessel to a glove. RESULTS

AND DISCUSSION

Lithium Intercalation. Figures 1 and 2 illustrate the variation of the intercalation degree of lithium in MO& with the reaction time for the conventional thermal treatment and for that using microwave activation, respectively. These patterns permit us to appreciate the efficiency of both methods. For molybdenum sulfide, the reaction rate reached by microwave-assisted intercalation is superior to that reached by the thermal method. On the other hand, the reaction yield, in terms of the excess of the butyl lithium concentration,

0.8 -

0

20

50

85

125

190

220

255

340

time(sec)

Dependence of lithium activation method.

intercalation

FIG. 2 degree on the reaction

time

for the microwave

Vol. 32, No. 6

METAL SULFIDE INTERCALATION

713

which is necessary to reach a given intercalation degree, is more favorable in the thermal method. Thus, for obtaining compounds Li,Mo!& with x in the range 0 < x < 1, the initial stoichiometric ratio in the microwave-assisted method must be 2: 1, twice that required in the thermal procedure. For the latter, with an initial ratio of 2:1 the intercalation can go much further than x = 1, as observed in Figure 1. From the point of view of the quality of the products, the microwave method results are, in general, favorable, the crystallinity of the products prepared by microwave activation being at least as good as that of those prepared by conventional methods. Although lithium intercalation rates higher than x = 1 may be always easily afforded, they were systematically avoided because of secondary reactions of intercalated lithium with the matrix leading to the formation of LiS2 (18). In the case of the titanium disulfide, the reaction time may be significantly reduced by microwave irradiation. However, as observed in the diffractograms reproduced in Figure 3, the product contains some impurities, indicating a decomposition of the sample. The chemical reactivity of TiS2 being much higher than that of MO% can account at least partially for the observed instability of the product under the conditions of the microwave activation experiments. Moreover, it is known that the microwave absorption by metallic materials is very strong, delivering a great quantity of energy in the process. It is therefore probable that lithium-intercalated titanium disulfide, because of its metallic behavior, undergoes specific energy absorptions leading to secondary reactions with partial decomposition of the sample. Intercalation of Organic and Organometallic Species. Direct intercalation of organic and organometallic donors in molybdenum sulfide has been not observed until now. Ion exchange reactions, which, in general, have proven to be an appropriate method for the intercalation of donor species in many layered transition metal derivatives (19), appear to have poor efficiency in the case of molybdenum sulfide. A good procedure for intercalating this matrix has proven to be the activation of the host matrix by an exfoliation process achieved by rapid hydrolysis of the lithiated sulfide (20). Thin layers of the matrix formed in the interphase of immiscible liquid mixtures favor the host-guest interactions leading to intercalation products. Intercalation rates .obtained by this procedure are, however, not

2 theta (degrees) X-ray diffiction

FIG. 3 pattern of LiTi& prepared by microwave activation.

714

E. BENAVENTE

et al

Vol. 32, No. 6

TABLE I Reaction Times and Product Interlayer Distances of Intercalation Compounds Prepared by Conventional and Microwave-Assisted Methods

Host MO& M0S* LiMo& LiMo& LiMoSz LiMoSz LiMoSI LiMoS,

Guest

Method

Li Ii Styrene” Styrene Ferrocene” Ferrocene PAN” PAN

convenkmnl microwave conventional microwave conventional microwave conventional microwave

Time [S] 8.64 3.4 8.64 3.4 8.64 3.4 8.64 3.4

lo4 102 IO4 102 lo4 IO2 lo4 102

Product Interlayer Distances [A] 6.15 6.15 11.31

I 1.27 11.29 11.28

i 1.34 11.43

a C6H5CH=CH2, b (C5H5)*Fe, ’ [-CH$H(CN)-I,,.

always high enough for a monolayer of the guest in the interlayer MO& spaces, the products often being better described as pillared species (20). Better intercalation yields may be obtained by a method developed recently in our laboratory (17), which consists of the prior reduction of the matrix by butyl lithium, followed by the insertion of the donor in a dry nonaqueous medium. This method is especially appropriate for obtaining a new series of polyacrylonitrile(PAN)-based nanocomposites (21). Nevertheless, for the donor intercalation in LiMoSz, relatively long reaction times (about 24 hours) are still necessary. As observed in Table 1, the comparison of the reaction times required for the intercalation of a series of donors by conventional thermal methods and the microwave activation procedure shows that the latter is considerably faster. The ctystallinity of the products is, in general, acceptable, as observed in typical diffractograms illustrated in Figure 4. Elemental analysis, thermal stability, and electrical conductivity determinations confirm that in each case the compounds obtained by both procedures, thermal and microwave activation, are essentially the same and that they actually correspond to nanocomposites in which the donors are located in the van der Waals spaces of the MoS2. The interlaminar distances in the intercalated compounds, mostly about 11.3 A, being comparable to those observed for similar preparations (20), are always large enough for the insertion of the corresponding donors. Nevertheless, besides the size of the donor, there are other subtle effects-mainly associated to the host-guest charge transfer and electrostatic host-host, host-guest, and guest-guest interactions-which also contribute to the expansion of the interlayer spaces. Experiments directed to a better understanding of the factors determining the structural properties of MO& intercalation products are in progress. Intercalation Mechanism. The most relevant feature of the experiments discussed here is certainly the relative high reaction rates obtained by the microwave-assisted intercalation processes. Independent of the initial stoichiometry of the reagents, the reaction is very rapid, indeed about loo-fold faster than that observed in conventional methods. The use of microwave dielectric loss heating for accelerating the intercalation of pyridine and substituted pyridines into a-VO(PO4) .2&O has been reported (14). Spectacular results were observed, namely, reaction rates about two magnitude orders faster than those observed with conventional thermal methods. However, considering the conditions under

Vol. 32, No. 6

METAL SULFIDE INTERCALATION

715

a

b

C

2

L

6

0

2 theta

10 12 (degrees)

FIG. 4 X-ray powder diffraction patterns of (a) LiMo$, Lio,6MoS2[(CsH5)2Fe]olsand (d) Lio,sMoS#AN)I O.

lr,

16

18

20

(b) Li0.sMoSz(CaHsCH=CH& I,, (c)

716

E. BENAVENTE

et al.

Vol. 32, No. 6

which the experiments were carried out, 200°C and pressures of ca. 50 atmospheres, such results could be explained more by those extreme rigorous reaction conditions than by a direct influence of microwave energy absorption on the mass transport in the solid. It is therefore interesting to note that in the present case, the intercalation process was accelerated while maintaining both a relatively low temperature and atmospheric pressure. An explanation for these features may be found in the dielectric nature of microwave heating in which heat is produced directly in the sites of dielectric loss. Dielectric heating could explain to some extent the effect of microwave irradiation on the reaction rates of the system discussed here. On one hand, dielectric heating could lead to a simultaneous reaction in a large number of sites, probably greater than those available for solute solvent interactions. On the other hand, it could lead to an increase of the diffusion rate of intercalated lithium in the solid, which is known to be strongly dependent on the temperature. Although in the cases discussed above, the reaction temperature during microwave irradiation was always below the boiling point of the solvent and there was never over-pressure in the vessel, local heating associated with either endothermic process could explain the observed effects. However, detailed kinetic studies needed to obtain mechanistic information and learn about the existence of other more specific effects are in progress. The shape analysis of the curves in Figures 1 and 2 indicates that the different behavior of the conventional and microwave methods for intercalating lithium in metal transition dichalcogenides may be related to intercalation staging. In the case of thermal activation a relatively quick intercalation is observed below x = 0.75. In the microwave-assisted reaction, however, after a short induction step, the intercalation curve shows a smooth behavior which could be interpreted as a mechanism in which a first-stage intercalation is produced. ACKNOWLEDGMENTS The authors are grateful to Mrs. L. Guajardo for technical collaboration. This work was partially supported by DTI, Universidad de Chile, FONDECYT, Fundaci6n Andes (Grant C12510), and European Union (Contract CII-CT93-0330). REFERENCES 1. M.S. Whittingham and A.J. Jacobson, Intercalation 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Chemistry, Academic Press, New York (1982). C. Julien and G.A. Nazri, Solid State Batteries, Kluwer, Boston, MA (1994). K.M. Abraham, J.S. Foosand S.B. Brummer, US Patent4.489.145 (1988). D. Fouchard and L. Lechner, Electrochim. Actu 38, 1193 (1993). R.H. Friend and A.D Yoffe, A& Physics 36, 1 (1987). D.W. Murphy and G. W. Hull, J. Chem. Phys. 62,973 (1975). M.S. Wittingham, Prog. SolidSt. Chem. 12, 41 (1978). G. Gonzalez, M.A. Santa Ana, E. Benavente, J.P. Donoso, T.J. Bonagamba, N.C. Mello, and H. Panepucci, Solid State Ionics 85,225 (1996). M.D. Dines, Mater. Rex Bull. 10,287 (1987). K. Chatakondu, M.L.H. Grenn, M.E. Thompson and K.S. Suslick, J. Chem. Sot., Chem. Commun 900 ( 1987). F.R. Gamble, J.H. Osiecki, M. Cais, R. Pisharody, F.J. DiSalvo and T.H. Geballe, Science 174, 4931(1971).

Vol. 32, No. 6

METAL SULFIDE INTERCALATION

717

12. R.N. Gedye, F.E. Smith and K.C. Westaway, Can. J. Chem. 66, 17 (1988). 13. D.M.P. Mingos and D.R. Baghurst, Chem. Sot. Rev. 20, 1 (1991). 14. K. Chatakondu, M.L.H. Green, D.M.P. Mingos, and S.M. Reynolds, J. Chem. Sot., Chem. Commun. 1515 (1989). 15. N. Nudelman and A. Vitale, Org. Prep. Proced. 13,44 (1981). 16. G.M. Whitesides, C. Casey, and J.K. Krieger, .I. Am. Chem. Sot. 93, 1379 (1971). 17. E. Benavente and G. Gonzalez, submitted to Bol. Chil. Quim. 18. R. Schoellhom, Solid State Zonics, 39,23 (1990). 19 L.S. Selwyn, W.R. McKinnon, U. von Sacken, and C.A. Jones, Solid State Zonics, 22,337 (1987). 20. W.M.R. Divigalpitiya, R.F. Flindt and S.R. Morrinson, Science 246,369 (1989). 2 1. G. Gonzalez, E. Benavente and M.A. Santa Ana, unpublished data.