6
Solid State Ionics 58 (1992) 333-338 North-Holland SOLID STATE lowlcs Study on lithium intercalation into MO& Nobuyuki Imanishi, Mie Toyoda, Yasuo Takeda and Osamu Yamamoto Department of Chemistry, Faculty ofEngineering, Mie University, 1515 Kamihama-cho, Tsu, Mie 514, Japan Received 10 June 1992; accepted for publication 1 July 1992 The electrochemical lithium intercalation behavior into 2H-MoS, in aprotic electrolytes was studied. The discharge product consisted of the large expanded LiMo& and lT-LiMo& The solvent molecule was found to co-intercalate with lithium into the interlayer space during the first discharge process. The reversibility of the lithium intercalation depended on the kind of solvent. Propylene carbonate in the interlayer space was not extracted after the charge process, while dimethoxyethane was removed with lithium ions. The co-intercalation of the solvents with lithium was not observed for the chemically prepared IT-MO& 1. Introduction MoS, has been studied as a cathode material for high energy secondary lithium cells with aprotic or- ganic electrolytes [ l-5 1. The structurally modified Li,Mo& has been known to discharge at about 2 V showing a good cycling behavior [ 6 1. This Li/MoS system was commercialized by Moli Energy Limited and the “AA” size cell showed higher performance than Ni-Cd cell under some conditions, such as C/ 5 rate cycling [ 7,8]. Thus, this cell has been ex- pected to be applied as a power source for portable devices. A problem of this cathode material is to in- duce the decomposition of the electrolytes, which re- sults in the decrease of cycle life [ 91. However, the decomposition mechanism has not been clarified yet. The lithium intercalation process into MoS2 is an interesting phenomenon from the view point of fun- damental research, because the intercalation induces the change of electronic and magnetic properties, and structural modification of the host lattice [ 10-l 31. As lithium was intercalated into 2H-MO&, the phase transition to 1 T structure was found in the range from x=0.2 to x= 1.0 in Li,MoS, [ 61. In this study, we have investigated the charge-discharge characteris- tics of the Li/2H-MoS, cell and the structural change of 2H-MO& by lithium intercalation. Especially, the effect of electrolyte solvent has been examined. 2. Experimental The MO& powder was provided from Sumitomo Metal Mining Co., Ltd., whose average particle size was 10 urn in diameter. The powder was dried at 150’ C for several days before use. Chemically lithiated MoS2 was prepared by the re- action of MO!& with n-butyllithium. Five grams of MO& was mixed and stirred with 40 m$! of 1.6 M n- butyllithium in hexane at room temperature for three days. The reaction products were washed with n-hex- ane, and then dried under vacuum at room temper- ature. The lithium content in the reaction products was analyzed by atomic absorption method to be de- termined as Li1.r2MoSz. Charge-discharge tests were carried out by using the coin-type cell (CR2025) as shown in fig. 1. One hundred mg of the MoS2 powder was pressed into a tablet in 8 mm diameter under 10 MPa. The lithium Lithium sheets I Separators & electrolyte Mesh ~ /--I-/--- Anode can / G Cathode 1 Cathode can Fig. 1. The experimental coin-type cell for electrochemical measurements. Elsevier Science Publishers B.V.

Study on lithium intercalation into MoS2

Embed Size (px)

Citation preview

Solid State Ionics 58 (1992) 333-338

North-Holland

SOLID STATE lowlcs

Study on lithium intercalation into MO&

Nobuyuki Imanishi, Mie Toyoda, Yasuo Takeda and Osamu Yamamoto Department of Chemistry, Faculty ofEngineering, Mie University, 1515 Kamihama-cho, Tsu, Mie 514, Japan

Received 10 June 1992; accepted for publication 1 July 1992

The electrochemical lithium intercalation behavior into 2H-MoS, in aprotic electrolytes was studied. The discharge product

consisted of the large expanded LiMo& and lT-LiMo& The solvent molecule was found to co-intercalate with lithium into the

interlayer space during the first discharge process. The reversibility of the lithium intercalation depended on the kind of solvent.

Propylene carbonate in the interlayer space was not extracted after the charge process, while dimethoxyethane was removed with

lithium ions. The co-intercalation of the solvents with lithium was not observed for the chemically prepared IT-MO&

1. Introduction

MoS, has been studied as a cathode material for high energy secondary lithium cells with aprotic or- ganic electrolytes [ l-5 1. The structurally modified Li,Mo& has been known to discharge at about 2 V showing a good cycling behavior [ 6 1. This Li/MoS system was commercialized by Moli Energy Limited and the “AA” size cell showed higher performance than Ni-Cd cell under some conditions, such as C/ 5 rate cycling [ 7,8]. Thus, this cell has been ex- pected to be applied as a power source for portable devices. A problem of this cathode material is to in- duce the decomposition of the electrolytes, which re- sults in the decrease of cycle life [ 91. However, the decomposition mechanism has not been clarified yet.

The lithium intercalation process into MoS2 is an interesting phenomenon from the view point of fun- damental research, because the intercalation induces the change of electronic and magnetic properties, and structural modification of the host lattice [ 10-l 31. As lithium was intercalated into 2H-MO&, the phase transition to 1 T structure was found in the range from x=0.2 to x= 1.0 in Li,MoS, [ 61. In this study, we have investigated the charge-discharge characteris- tics of the Li/2H-MoS, cell and the structural change of 2H-MO& by lithium intercalation. Especially, the effect of electrolyte solvent has been examined.

2. Experimental

The MO& powder was provided from Sumitomo Metal Mining Co., Ltd., whose average particle size was 10 urn in diameter. The powder was dried at

150’ C for several days before use. Chemically lithiated MoS2 was prepared by the re-

action of MO!& with n-butyllithium. Five grams of MO& was mixed and stirred with 40 m$! of 1.6 M n- butyllithium in hexane at room temperature for three days. The reaction products were washed with n-hex- ane, and then dried under vacuum at room temper- ature. The lithium content in the reaction products was analyzed by atomic absorption method to be de- termined as Li1.r2MoSz.

Charge-discharge tests were carried out by using the coin-type cell (CR2025) as shown in fig. 1. One hundred mg of the MoS2 powder was pressed into a tablet in 8 mm diameter under 10 MPa. The lithium

Lithium sheets I Separators & electrolyte Mesh ~

/--I-/--- Anode can

/

G

Cathode 1 Cathode can

Fig. 1. The experimental coin-type cell for electrochemical

measurements.

Elsevier Science Publishers B.V.

334 N. Imanishi et al. /Lithium intercalation into MoS,

sheet with 0.34 mm in thickness was used as the an-

ode. The electrodes were separated by a microporous polypropylene sheet. The electrolytes used in the present study were 1 M solution of LiC104 in pro- pylene carbonate (PC ) /dimethoxyethane (DME ) [ 1: 1 by volume] and 1 M solution of LiC104 in DME, which were supplied from Mitsubishi Petro- chemical Co., Ltd. The test cells were cycled galva- nostatically between 0.9 and 3.5 V at a constant cur- rent density of 200 pA cmp2. The lattice parameters were measured by X-ray diffractometer (Rigaku Ro- taflex RU-200B) with Cu KCI radiation. The XRD

measurements for lithium intercalated products were done by covering the sample with aluminium thin film with 7 pm in thickness in order to avoid the

contact with moisture. The silicon powder was used as the inner standard material. The construction of the cell, preparation of the sample for XRD mea- surements and all other procedures were performed in an Ar atmosphere dry box.

3. Results and discussion

The first five cycles of the Li/2H-MoS2 cell using PC-DME and DME as the electrolytes are shown in fig. 2. The characteristic features of the first dis- charge of these cells are the initial large voltage drops and following voltage plateau at 1.3-l. 1 V. Fig. 2a shows that the cell voltages are still as high as 1.1 V after 170 mA hgg ’ discharge, which corresponds to Li,.,Mo!$. However, the first charge capacity was extremely low and the cycle efficiency calculated by the ratio of charge-discharge capacity of the first cycle was only 13%. The charge and discharge capacities of the 5th cycle diminished markedly compared to the first charge-discharge one. Such a low charge ca- pacity suggests that some of irreversible reaction oc- curs during the voltage plateau of the discharge process.

On the other hand, the Li (DME ( 2H-MoS2 cell showed rather improved cycling behaviors than in the Li 1 PC-DME I2H-MoS, cell as shown in fig. 2b. The first discharge profile showed almost the same as that of the PC-DME cell. However, the first charge process and following charge and discharge cycles showed much higher capacities. The cycle efficiency

Li I PC-DME l2H-MoS,

-v-- I m-~--r

o I.. ~~~ 1 1 50 100 150

specific capacity / mAhg-’

Li 1 DME I 2H-MO&

:-:s:,-

5th ’

I 1 1 I 0 50 100 150

specific capacity / mAhg-’

Fig. 2. The first five charge-discharge cycles of the Li/Mo& cell:

(a) 1M LiClOJPC-DME; (b) 1M LiClO,/DME.

was about 53% in the first cycle and showed no de- crease in the following cycles,

The structural change in 2H-MoS2 during the first

charge-discharge cycle was analyzed by XRD mea- surements. Fig. 3 shows the change of XRD patterns of the cathode material for the various steps of dis- charge and charge. These patterns showed several new peaks besides those of original MoS2. If the peak at the lowest angle, 5.1’ (d= 14.3 1 A) in fig. 3a, is as- sumed as 00 1 reflection of MoS2, the peaks indicated by circles at 10.2”(8.665 A), 15.3”(5.786 A) and 20.5’ (4.329 A) can be indexed as 002,003 and 004 reflections, respectively. From these results it is sug- gested that the lithiated MoS2 has the expanded in- terlayer space after the lithium intercalation. In ad- dition to the expanded phase, the third phase could be also observed as shown in fig. 3. It would be clear from a peak at 13.9” and much broader peaks at around 43 and 54’) which are indicated by triangles in the figure, appeared in the patterns of the deep discharged sample to 168 mA hg- I. As shown later, these peaks are considered to belong to the 1T phase.

The large expansion of the interlayer is often in- duced by co-intercalation of a solvent. The solvent co-intercalation is confirmed by the fact that the dis- charged products in PC-DME and DME show the

N. Imanishi et al. /Lithium intercalation into MoS,

a) PC-DME b) DME

__. L_d’ I_ ._ , _ . .- ._ , I I 1

3 20 40 60 2 H /degree(CuK a: )

I

SI 002

103 SI

_ __. -.__ ,__s . _ _ .- \_’ 3 20 40 60

2 0 /degree(CuK LY )

Fig. 3. X-ray diffraction patterns of MoSz during the first cycle: (0 ) expanded MoS2 phase, ( A ) IT phase. (a) PC-DME; after (a) 0,

(b) 17, (c) 84, (d) 168 mA hg-’ discharge, and (e) 25 mA hg-’ charge after 168 mA hg-’ discharge. (b) DME; after (a) 0, (b) 15,

(c) 75, (d) 150 mA hg-’ discharge, and (e) 75 mA hg-’ charge after 150 mA hg-’ discharge.

different XRD patterns in each other as shown in fig. 3. In the case of PC-DME, more expanded interlayer space was observed. This is also convinced visibly as shown in fig. 4. Photographs (a) and (b) show the MO& tablets discharged to 17 and 170 mA hg- ‘, re- spectively. In this experiment, the cathode tablet was

prepared only with MO& powder, without using any carbon material as a conducting material and PTFE as binding agent. Apparently, the volume of cathode tablet was drastically increased after being dis- charged to 170 mA hg- ‘. The discharged product was heated at 300°C in vacua for one day, whose XRD

Fig. 4. The photographs of the MO& cathode tablet (a) after 17 mA hg-’ and (b) 170 mA hg-’ discharge.

336 N. Imanishi et al. /Lithium intercalation into MoS,

‘I Si 7, Al

/ 1

/ ”

Fig. 5. X-ray diffraction patterns of (a) MO& after the first dis- charged to 170 mA hg-’ and (b) heat-treated discharged sample at 300°C in vacua: (0 ) expanded phase, ( LI ) IT phase.

pattern is shown in fig. 5. The peaks indicated as the expanded layers disappeared perfectly. This means that the intercalated solvent molecule was extracted

away and the interlayer distance returned to that of original 2H-MO&.

The extraction of the solvent molecule was also

observed in the XRD patterns of the MO& cathode being charged as shown in fig. 3. The new peaks ap- peared in the discharge process completely dimin- ished after being charged to 75 mA hg-’ in the DME cell as the solvent. On the other hand, the new peaks

remained at the same positions in the cathode of the cell with PC-DME. It is considered that the DME can easily co-intercalate into and deintercalate from the interlayer space with lithium ion, while the PC molecules are strongly adsorbed onto the MO& layer. Thus, the low cyclability in the PC-DME electrolyte may be due to the prevention of deintercalation pro-

cess by the PC molecule. The discharged product contained the lT-phase in

addition to the expanded phase. And it has been re- ported in some papers that only lT-MO& was ob- tained after lithium intercalation into 2H-MO& [ 6,12,13 1. In order to understand the role of a sol- vent in the lithium intercalation process, it is im- portant to know the phase transformation mecha- nism and electrochemical behavior of lT-MO&. Fig. 6 shows the (1120) projections of the stacking se-

f

S MO

0 s Li -

2H ( P63 /mmc 1 lT(P3ml)

Fig. 6. ( 1120) sections of the structure for 2H- and IT-MoS,.

quences of 2H- and IT-MO&. The transformation

from the 2H phase to the 1T phase is equivalent to a sliding of one S layer in a S-MO-S slab accom- panied with the sliding of S layer in an adjacent slab to opposite direction at the same distance. The tri- gonal prismatic coordination occurs when 4d levels of MO atom split into three groups which are lower stabilized d,2 orbitals and two higher e’(dxy, dx2 -y2) and e”(dxz, dyz) groups. When intercala- tion has occurred, newly added electron must be at the higher energy levels. This leads to the change in the coordination of MO atom from trigonal pris- matic to octahedral [ 10 1.

1 T-MoS2 was prepared by the electrochemical de- lithiation of Li ,.12MoS2 prepared by the reaction of MoS, with n-butyllithium. The XRD pattern of the prepared sample is shown in fig. 7a, which is quite different from that of the 2H-type structure. As no X-ray diffraction profile of IT-MO& has been re- ported in the JCPDS cards, it was simulated by using the program RIETAN [ 141. The conditions of the simulation are as follows; the space group is P3ml and equivalent positions for MO are x= 0, y= 0, Z= 0 and for S were x=0.66, y=O.33, z=O.25 and for Li werex=O, y=O, z=O.S, where the lattice parameters were assumed to be same as 2H-MoS2 and no pre- ferred orientation of crystal was considered. The simulated patterns are shown in fig. 7b. The result is in good agreement with the patterns of the pre- pared sample. The main reason for the differences between observed and calculated data is that the ap- plied lattice parameters (a and c) for calculations might differ slightly from ones of the prepared sam- ple. From these considerations, the prepared sample was concluded to be lT-MoS.

Charge-discharge characteristics of 1 T-MO& were

N. Imanishi et al. /Lithium intercalation into MO& 337

Al Al ’ I

3 20 60

20 / degree (CuKa)

20 / degree (CuKcc)

Fig. 7. X-ray diffraction patterns of (a) a prepared lT-MO& and

(b) simulated patterns by using a computer.

Li I PC-DME I IT-LiMoS, 4, I I

I I I I 50 100 150

specific capacity / rnAhg-’

Fig. 8. The first five charge-discharge cycles of a Li/ IT-MoS,

cell. 1M LiCIO,/PC-DME.

measured at 200 yA cm-* in PC-DME. As shown in fig. 8, the cycling efficiency was over 80% at the first cycle, but the capacity of each process was only 50% of the theoretical one. It is worth while to note that the cycling behaviors are much similar to the second and following cycles of the Li/2H-MoS2 cell in DME solvent. This suggests that the phase transformation from 2H to 1T phase occurs during the first dis- charge process. XRD patterns of IT-MoS2 dis- charged in PC-DME and DME solvents are shown in fig. 9. No extra peak due to the solvent interca- lation is observed for both electrolytes. It is sug-

1 , I

20 40 60

2 Q /degree(CuK a )

Fig. 9. X-ray diffraction patterns of lT-MO& discharged in PC-

DME and DME: (a) before discharge, (b) in PC-DME, (c) in

DME.

gested that electrolyte solvent does not intercalate into the interlayer space of lT-MoS2. Moreover, the large structural change such as the phase transfor- mation also does not occur. We conclude from these results that the electrolyte solvent can co-intercalate only at the period of sliding process of atomic layers in 2H-MoS2, that is, 2H to 1T phase transition.

4. Conclusion

Phase transformation from the 2H to the 1 T phase occurs at 1.1 V and the solvent molecule co-inter- calate with lithium. The reversibility of the solvent co-intercalation depends on its character. PC re- mains in interlayer space even after charge, but DME goes out with lithium. It is possible that reduction of

PC molecule occurs in interlayer space, because PC is comparatively easily reducible. The solvent mol- ecule does not intercalate into the interlayer space of 1 T-MO&.

References

[ 1 ] J.J. Aubom, Y.L. Barberio, K.J. Hanson, D.M. Schleich and

M.J. Martin, J. Electrochem. Sot. 134 (1987) 580. [2] A.J. Jacobson, R.R. Chianelli and M.S. Whittingham, J.

Electrochem. Sot. 126 (1979) 2277.

338 N. Imanishi et al. /Lithium intercalation into MoS2

[3] MS. Whittingham and R. Gamble, Mat. Res. Bull. 10

( 1975) 363.

[4 F.C. Laman, J.A.R. Stiles, R.J. Shank and K. Brand& J.

Power Sources 14 (1985) 201.

[ 51 D. Ilic, K. Wiesener and W. Schneider, J. Power Sources 14

(1985) 223.

[9] M.W. Juzkowandl.D.Gay, J. Phys. Chem. 95 (1991) 9911.

[lo] E. Sandre, R. Brec and J. Rouxel, J. Solid State Chem. 88

(1990) 269.

[ 11 ] P.J. Mulhern, Can. J. Phys. 67 (1989) 1049.

[ 121 C. Julien, T. Sekine and M. Balkanski, Solid State Ionics 48

(1991) 225. [6]M.A.PyandR.R.Haering,Can. J.Phys.61 (1983) 76. [ 131 K. Chrissalis, M. Zamani, K. Kambas, J. Stoemenos and [ 71 F.C. Laman, M.W. Matsen and J.A.R. Stiles, J. Electrochem. N.A. Economou, Mater. Sci. Eng. B3 (1989) 145.

Sot. 133 (1986) 2441. [ 141 F. Izumi. J. Miner. Sot. Japan 17 (1985) 37.

[8] J.A.R. Stiles, New Mater. New Proc. 3 (1985) 89.