- Email: [email protected]
Novel 2 V rocking-chair lithium battery based on nano-crystalline titanium dioxide
" '
'''
JOURHALO[
POWER ELSEVIER
Journal of P o w e r Sources 68 ( 1997 ) 7 2 0 - 7 2 2
SOURCES
Novel 2 V rocking-chair lithium battery based on nano-crystalline titanium dioxide I. Exnar a, L. K a v a n
b,
S.Y. H u a n g c, M. Gr~itzel ~
Renata AG, 4452 ltmgen, Switzerland h j. Heyrovsky Institute ofPhy~swal Chemtstr3, 18 223 Prague, Czech Repubhc Swtss Federal lnsttute oJ Technology, 1015 Lausamle, Swttzerland Accepted 9 September 1996
Abstract A novel type of 2 V rockmg-chair lithium batteries with nano-crystallme TiO: (anatase) as the negative and LiNio 5CoosOz as the positive electrode with LiN(CF~SO2)z + ethylene carbonate/dlmethoxyethane electrolyte has been studied. The closed R921 button cells showed excellent cycling performance and a charge density of about 50 mAh/g. The advantages of the specific porous morphology of nano-crystalline insertion materials for Li insertion and release are discussed. © 1997 Elsevier Science S.A. Kevwords. Titanium dioxide; Insertion material: R o c k i n g - c h a i r batteries, Lithium battertes
1. Introduction While most of the attention of both the industry and research groups in the field of lithium batteries was focused in recent years on 4 V carbon rocking-chair systems, 2 V batteries seem to be more practical than their 4 V counterparts for application in micro-electronics and/or consumer devices with photovoltaic recharging, e.g. solar watches. This for the following two reasons: (i) lower voltage allows higher integration of microelectronic circuits, and (ii) the photovoltatc recharging system offers only a limited voltage. This paper refers to a study on a new type of 2 V rockingchair lithium battery with nano-crystalline TiO2 ( anatase ) as the negative electrode and LiNiosCo0502 as the positive electrode.
2. Experimental Nano-crystalline titanium dioxide powder (Bayer PKP 09040, anatase, purity > 99%) was used as received. The starting powder was characterized by X-ray diffraction (XRD) by using an X-ray diffractometer Siemens DS-500 with Cu Kc~ radiation. The morphology of the samples was analysed by scanning electron microscope (SEM) using a Hitachi S-900 instrument. The BET surface area was 154 m2/g ( Kr-isotherms, Accusorb 2100 E, Micrometrics ). 0 3 7 8 - 7 7 5 3 / 9 7 / $ 1 7 . 0 0 cO 1997 Elsevier Science S.A All nght~ reserved PHS0378-7753 q 9 6 1 0 2 5 8 1-5
LiCoO2 and LiNio 5Coo ~O2 (Cyprus Foote Mineral) had both a particle size of about 400 nm and a surface area of 4 m2/g. Electrodes were prepared in the form of pressed pellets with 10% graphite (KS-10, Lonza) as the conducting agent and 2% polyvinylidene fluoride (PVDF) (Solef, Solvay) as the binder. Ethylene carbonate (EC) and dimethoxyethane (DME) were from Tomiama. A 1:1 mixture (by mass) was dried over a 4 A molecular sieve ( Union Carbide). LiN (CF3SO2) 2 (Fluorad HQ 115 from 3 M) was dried at 200 °C and at 10 -3 Pa. The prepared 1 M solution L i N ( C F ~ S O z ) z + E C / D M E contained about 10 ppm of water. The preparation of the electrolyte and the assembling of the experimental button cells were performed in an argon-filled glove box ( 1-5 ppm H20, 1 ppm 02). Electrochemical experiments employed a convential threeelectrode cell interfaced to a potentiostatic/galvanostatic setup (PAR EG&G 273 A). The reference and auxiliary electrodes were prepared from pure lithium metal. The closed R921 button cells were further tested at ambient temperature using a Maccor 2000 series battery test equipment. 3. Results and discussion Fig. 1 shows the XRD pattern of the nano-crystalline TiO2 (anatase) powder. The primary particle size, d, was calculated from the X-ray line width using the Scherrer formula
I. Exnar et al. / Journal o / P ( m er Sour¢'e~s 68 ( 1997l 720 722
721
~oo[
c
2.8
i~ .I'' I'~
'1
200 F~
L 20
Nanocr sfaLt ne
Y
]0
40
50
TiO
=r t
~
60
I h ,c,,u: ' - . ~
2.4
--- 033rnMcm ~ (S,nA/cm 2 0 90rnA/cm 2
2.2
\
1.8
',~. . . . . . . .
70
40
A
35 B
~" ~.
20{degreesl CuKc~ F~g. 1. X R D spectra o f TiO~ ( a n a t a s e B a y e r P K P 0 9 0 4 0 )
",
1.6
2". ", LJ,T O 25
"l
00
0.2
04
0.6
0.8
1.0
x=l A/M IM=T[, Co~ Fig. 3 G a l v a n o s t a H c c h a r g e / d i s c h a r g e o f the T I O 2 a n d L i C o O 2 e l e c t r o d e s m cell w d h a h t h m m c o u n t e r e l e c t r o d e
r.
LiCF:~SO3. LiAsF6 and LiPF, by cyclic voltametry [3] showed that 1 M L i N ( C F 3 S O 2 ) 2 + E C / D M E exhibited the best insertions capacities, x, and insertion/extraction reversibilities [ 6]. The solutions in E C / D M C and E C / D E C showed better stability against anodic breakdown at potentials greater than 4 V, but the performance with the nano-crystalline TiO2 electrode was lower which is presumably due to lower conductivity of these solutions. Supposing that the Li +-insertion capacity is limited by the solid-state diffusion of Li +/TiO2 as demonstrated in Fig. 4 we should find in nanosized anatase material the effective diffusion time, t.
i( •
18""
,~a6
t=R2/rrD~0.53 s -
Jl
1 5 0 nm
(3)
k
Fig. 2. S E M i m a g e s o f the n a n o - c r y s t a l l l n e h i m o f B a y e r P K P 0 9 0 4 0 T I O :
]LITHIUM I N T E R C A L A T I O N D Y N A M I C S
d = 0.9cff/31/2cos 0
a) C o m p a c t L a y e r
( 1)
where o~is the X-ray wavelength./3~/2 the corrected width of the main diffraction peak at hall-height and 0 the diffraction angle. The d value of TiO2 Bayer PKP 09040 was l 0 n m [ 1 ]. Fig. 2 shows the SEM images of the sud'ace ofTiOe (anatase ). The particles are roughly spherical, the surface is highly textured with open pores and channels. The particle size is about 10-15 nm, m agreement with the XRD results.
(
7Ll * %
TIO
2 anatase
~Lll I I II III :I :I : I: II ~I II II I II II I I: I]l illlllll
. . . . . .--"'~, . lat~ '+
i I ] J ] I I I I'[ l l l l i l l l l l l ] l l i l l l l [ l l l l l l
I::F11_.~ ~ I-- IU~J.lll ~
M e a n intercalation time: ( t ) = L 2 = 5000s 2D w h e r e L lfl~m, O~10-1Ocm2/s
b) Nanocrystalline Layer ~ . I÷
4. Electrochemistry: Li + insertion
~ : 7Nm
In the Li+-contaming electrolyte solutions and at potentials below the flatband, the insertion of Li + into anatase takes place [ 1,2,4.5]. TiO2 + x( Li + + e I ~ Li,TiOz
(2 )
Fig. 3 displays the galvanostatic charge/discharge characteristics of the TiO, and LiCoO2 electrodes versus lithium metal in I M LiN(CF3SO2) 2 + E C / D M E electrolyte solution. Screening of many combinations of lithium salts and aprotic solvents such as propylene carbonate (PC), DME, EC, dimethyl carbonate ( DMC ), diethyl carbonate I DEC ),
,
M e a n intercalation time: (t) = R 2 = 1.56x103s ~D where
R=7nm, D~10 -10 cm2/s
N a n o c r y s t a l l i n e e n h a n c e m e n t factor =
Ftg. 4. D f f f u m o n m o d e l for c o n v e n t m n a l battery
5000 _ = 3.2x106 1.56x10 3
and nano crystalline Ll+-[on
722
1. Exnar et al. / Journal of Power Sources 68 (1997) 720-722
Volts 30
tical TiO/LiConsNiosO2 button cells we found that the optimal anatase particle size is about 10-15 nm. The galvanostatic charge/discharge behaviour of the TiOz/LiNin 5Coo 502 battery is shown in Fig. 5. The completed crimp-sealed coin-type cell (diameter: 9.5 mm, height: 2.0 ram) delivered a reversible capacity of about 50 m A h / g (based on the mass of active electrode materials at _+0.33 m A / c m 2 in the 1-2.2 V region ) tbr over 300 cycles as shown in Fig. 6.
+
2,0 1.5 1.0 0.S
i corlsf
0
-- 2'7
0
i 5&
l 81
=
0,66 m A l c m 2
i 108
135
Time (hl
Fig. 5. Oalvanostatic charge/dmcharge of the TiOz/L1Ni. ~Co. ~O: button cell. 50
~. ......
i .......
i .......
~o
, . . . . . . .
, . . . . . . .
i
. . . . . . .
i . . . . . . .
i . . . . . . .
--
3o m (11
-~
@ 0,33male@ (~) o,66mAlcd
~ I0
~.
@ ill
....
i ....
/+0
iiiiii1~,1=1
80
.......
i .......
I .......
120 160 200 Cycle number
0,99
mA/cm
i .......
240
2 r .......
280
5. Conclusions The investigated nano-crystalline TiO2 (anatase) and LiNin 5Coo 502 system is an example of promising rechargeable 2 V rocking-chair batteries for low drain devices such as watches, computer memory backup, etc. A standard size R921 coin cell supplied typically 3-5 mAh at _+0.1-0.3 mA. The obtained charge density is about 50 mAh/g. The advantage of the nano-crystalline porous morphology of insertion materials is demonstrated in enhancement of the reversibility and insertion ratios.
320
Fig 6 Cychng perlormance of the TiO2/LiNk~ ~O2 cell. The cell was periodically discharged ( 1.0-2.2 V ) at constant current densmes of 0.33, 0.66 and 0.99 mA/cm-'.
where D is diffusion coefficient of Li + in anatase 1,81 × 10-13 c m 2 / s and R is the particle radius = 7 rim. Nevertheless, the real insertion time of Li + is in order of magnitude higher than predicted by Eq. (3). We have recently suggested that the solid-state diffusion of Li + is not the rate-determining process and that the Li + insertion is controlled by the slow formation of an accumulation layer due to ohmic drops and electron trapping at the surface. Screening of many TiO2-types with different morphology by galvanostatic chronopotentiometry showed that at relatively high charging rate there is some enhancement of the insertion ratio, x, and the reversibility ( Eq. 1) with decreasing particle size of the insertion material. Kavan and co-workers [3,7] observed this eftect in detail by electrochemical mesurements on thin TiO2 films supported on SnOz-coated glass. In prac-
Acknowledgements This work was partially supported by the Swiss Commission for the encouragement of scientific research.
References [ 1 ] S.Y Huang, L Kavan, A. Kay, I. Exnar and M. Gratzel, Active and Pmswe Elec. Comp., 18 ( 19951 23-30. 12] S.Y. Huang, L. Kavan, I Exnar and M. Gratzel. J. Ele~trochem. Soc.. 142 (1995) L142-144. [ 3 ] L. Kavan, M Gr~ttzel, J. Rathousky and A. Zukal, J. Electrochem. Soc, 143 (1996) 394 [4] T. Ohzuku and T Hwm, Electrochim. Acta. 27 (1982) 1263. [5] T. Ohzuku and T Kodama, J. Power Sources. 14 11985 ) 153. [ 6 ] K. Kanamura, K. Yuasa and Z. Takehara, J Power Sources, 20 ( 1987 ) 127. [7 ] L. Kavan, K. Kratochvilovfi and M. Gratze[. J Electroanal Chem., 394 { 1995) 93
'''
JOURHALO[
POWER ELSEVIER
Journal of P o w e r Sources 68 ( 1997 ) 7 2 0 - 7 2 2
SOURCES
Novel 2 V rocking-chair lithium battery based on nano-crystalline titanium dioxide I. Exnar a, L. K a v a n
b,
S.Y. H u a n g c, M. Gr~itzel ~
Renata AG, 4452 ltmgen, Switzerland h j. Heyrovsky Institute ofPhy~swal Chemtstr3, 18 223 Prague, Czech Repubhc Swtss Federal lnsttute oJ Technology, 1015 Lausamle, Swttzerland Accepted 9 September 1996
Abstract A novel type of 2 V rockmg-chair lithium batteries with nano-crystallme TiO: (anatase) as the negative and LiNio 5CoosOz as the positive electrode with LiN(CF~SO2)z + ethylene carbonate/dlmethoxyethane electrolyte has been studied. The closed R921 button cells showed excellent cycling performance and a charge density of about 50 mAh/g. The advantages of the specific porous morphology of nano-crystalline insertion materials for Li insertion and release are discussed. © 1997 Elsevier Science S.A. Kevwords. Titanium dioxide; Insertion material: R o c k i n g - c h a i r batteries, Lithium battertes
1. Introduction While most of the attention of both the industry and research groups in the field of lithium batteries was focused in recent years on 4 V carbon rocking-chair systems, 2 V batteries seem to be more practical than their 4 V counterparts for application in micro-electronics and/or consumer devices with photovoltaic recharging, e.g. solar watches. This for the following two reasons: (i) lower voltage allows higher integration of microelectronic circuits, and (ii) the photovoltatc recharging system offers only a limited voltage. This paper refers to a study on a new type of 2 V rockingchair lithium battery with nano-crystalline TiO2 ( anatase ) as the negative electrode and LiNiosCo0502 as the positive electrode.
2. Experimental Nano-crystalline titanium dioxide powder (Bayer PKP 09040, anatase, purity > 99%) was used as received. The starting powder was characterized by X-ray diffraction (XRD) by using an X-ray diffractometer Siemens DS-500 with Cu Kc~ radiation. The morphology of the samples was analysed by scanning electron microscope (SEM) using a Hitachi S-900 instrument. The BET surface area was 154 m2/g ( Kr-isotherms, Accusorb 2100 E, Micrometrics ). 0 3 7 8 - 7 7 5 3 / 9 7 / $ 1 7 . 0 0 cO 1997 Elsevier Science S.A All nght~ reserved PHS0378-7753 q 9 6 1 0 2 5 8 1-5
LiCoO2 and LiNio 5Coo ~O2 (Cyprus Foote Mineral) had both a particle size of about 400 nm and a surface area of 4 m2/g. Electrodes were prepared in the form of pressed pellets with 10% graphite (KS-10, Lonza) as the conducting agent and 2% polyvinylidene fluoride (PVDF) (Solef, Solvay) as the binder. Ethylene carbonate (EC) and dimethoxyethane (DME) were from Tomiama. A 1:1 mixture (by mass) was dried over a 4 A molecular sieve ( Union Carbide). LiN (CF3SO2) 2 (Fluorad HQ 115 from 3 M) was dried at 200 °C and at 10 -3 Pa. The prepared 1 M solution L i N ( C F ~ S O z ) z + E C / D M E contained about 10 ppm of water. The preparation of the electrolyte and the assembling of the experimental button cells were performed in an argon-filled glove box ( 1-5 ppm H20, 1 ppm 02). Electrochemical experiments employed a convential threeelectrode cell interfaced to a potentiostatic/galvanostatic setup (PAR EG&G 273 A). The reference and auxiliary electrodes were prepared from pure lithium metal. The closed R921 button cells were further tested at ambient temperature using a Maccor 2000 series battery test equipment. 3. Results and discussion Fig. 1 shows the XRD pattern of the nano-crystalline TiO2 (anatase) powder. The primary particle size, d, was calculated from the X-ray line width using the Scherrer formula
I. Exnar et al. / Journal o / P ( m er Sour¢'e~s 68 ( 1997l 720 722
721
~oo[
c
2.8
i~ .I'' I'~
'1
200 F~
L 20
Nanocr sfaLt ne
Y
]0
40
50
TiO
=r t
~
60
I h ,c,,u: ' - . ~
2.4
--- 033rnMcm ~ (S,nA/cm 2 0 90rnA/cm 2
2.2
\
1.8
',~. . . . . . . .
70
40
A
35 B
~" ~.
20{degreesl CuKc~ F~g. 1. X R D spectra o f TiO~ ( a n a t a s e B a y e r P K P 0 9 0 4 0 )
",
1.6
2". ", LJ,T O 25
"l
00
0.2
04
0.6
0.8
1.0
x=l A/M IM=T[, Co~ Fig. 3 G a l v a n o s t a H c c h a r g e / d i s c h a r g e o f the T I O 2 a n d L i C o O 2 e l e c t r o d e s m cell w d h a h t h m m c o u n t e r e l e c t r o d e
r.
LiCF:~SO3. LiAsF6 and LiPF, by cyclic voltametry [3] showed that 1 M L i N ( C F 3 S O 2 ) 2 + E C / D M E exhibited the best insertions capacities, x, and insertion/extraction reversibilities [ 6]. The solutions in E C / D M C and E C / D E C showed better stability against anodic breakdown at potentials greater than 4 V, but the performance with the nano-crystalline TiO2 electrode was lower which is presumably due to lower conductivity of these solutions. Supposing that the Li +-insertion capacity is limited by the solid-state diffusion of Li +/TiO2 as demonstrated in Fig. 4 we should find in nanosized anatase material the effective diffusion time, t.
i( •
18""
,~a6
t=R2/rrD~0.53 s -
Jl
1 5 0 nm
(3)
k
Fig. 2. S E M i m a g e s o f the n a n o - c r y s t a l l l n e h i m o f B a y e r P K P 0 9 0 4 0 T I O :
]LITHIUM I N T E R C A L A T I O N D Y N A M I C S
d = 0.9cff/31/2cos 0
a) C o m p a c t L a y e r
( 1)
where o~is the X-ray wavelength./3~/2 the corrected width of the main diffraction peak at hall-height and 0 the diffraction angle. The d value of TiO2 Bayer PKP 09040 was l 0 n m [ 1 ]. Fig. 2 shows the SEM images of the sud'ace ofTiOe (anatase ). The particles are roughly spherical, the surface is highly textured with open pores and channels. The particle size is about 10-15 nm, m agreement with the XRD results.
(
7Ll * %
TIO
2 anatase
~Lll I I II III :I :I : I: II ~I II II I II II I I: I]l illlllll
. . . . . .--"'~, . lat~ '+
i I ] J ] I I I I'[ l l l l i l l l l l l ] l l i l l l l [ l l l l l l
I::F11_.~ ~ I-- IU~J.lll ~
M e a n intercalation time: ( t ) = L 2 = 5000s 2D w h e r e L lfl~m, O~10-1Ocm2/s
b) Nanocrystalline Layer ~ . I÷
4. Electrochemistry: Li + insertion
~ : 7Nm
In the Li+-contaming electrolyte solutions and at potentials below the flatband, the insertion of Li + into anatase takes place [ 1,2,4.5]. TiO2 + x( Li + + e I ~ Li,TiOz
(2 )
Fig. 3 displays the galvanostatic charge/discharge characteristics of the TiO, and LiCoO2 electrodes versus lithium metal in I M LiN(CF3SO2) 2 + E C / D M E electrolyte solution. Screening of many combinations of lithium salts and aprotic solvents such as propylene carbonate (PC), DME, EC, dimethyl carbonate ( DMC ), diethyl carbonate I DEC ),
,
M e a n intercalation time: (t) = R 2 = 1.56x103s ~D where
R=7nm, D~10 -10 cm2/s
N a n o c r y s t a l l i n e e n h a n c e m e n t factor =
Ftg. 4. D f f f u m o n m o d e l for c o n v e n t m n a l battery
5000 _ = 3.2x106 1.56x10 3
and nano crystalline Ll+-[on
722
1. Exnar et al. / Journal of Power Sources 68 (1997) 720-722
Volts 30
tical TiO/LiConsNiosO2 button cells we found that the optimal anatase particle size is about 10-15 nm. The galvanostatic charge/discharge behaviour of the TiOz/LiNin 5Coo 502 battery is shown in Fig. 5. The completed crimp-sealed coin-type cell (diameter: 9.5 mm, height: 2.0 ram) delivered a reversible capacity of about 50 m A h / g (based on the mass of active electrode materials at _+0.33 m A / c m 2 in the 1-2.2 V region ) tbr over 300 cycles as shown in Fig. 6.
+
2,0 1.5 1.0 0.S
i corlsf
0
-- 2'7
0
i 5&
l 81
=
0,66 m A l c m 2
i 108
135
Time (hl
Fig. 5. Oalvanostatic charge/dmcharge of the TiOz/L1Ni. ~Co. ~O: button cell. 50
~. ......
i .......
i .......
~o
, . . . . . . .
, . . . . . . .
i
. . . . . . .
i . . . . . . .
i . . . . . . .
--
3o m (11
-~
@ 0,33male@ (~) o,66mAlcd
~ I0
~.
@ ill
....
i ....
/+0
iiiiii1~,1=1
80
.......
i .......
I .......
120 160 200 Cycle number
0,99
mA/cm
i .......
240
2 r .......
280
5. Conclusions The investigated nano-crystalline TiO2 (anatase) and LiNin 5Coo 502 system is an example of promising rechargeable 2 V rocking-chair batteries for low drain devices such as watches, computer memory backup, etc. A standard size R921 coin cell supplied typically 3-5 mAh at _+0.1-0.3 mA. The obtained charge density is about 50 mAh/g. The advantage of the nano-crystalline porous morphology of insertion materials is demonstrated in enhancement of the reversibility and insertion ratios.
320
Fig 6 Cychng perlormance of the TiO2/LiNk~ ~O2 cell. The cell was periodically discharged ( 1.0-2.2 V ) at constant current densmes of 0.33, 0.66 and 0.99 mA/cm-'.
where D is diffusion coefficient of Li + in anatase 1,81 × 10-13 c m 2 / s and R is the particle radius = 7 rim. Nevertheless, the real insertion time of Li + is in order of magnitude higher than predicted by Eq. (3). We have recently suggested that the solid-state diffusion of Li + is not the rate-determining process and that the Li + insertion is controlled by the slow formation of an accumulation layer due to ohmic drops and electron trapping at the surface. Screening of many TiO2-types with different morphology by galvanostatic chronopotentiometry showed that at relatively high charging rate there is some enhancement of the insertion ratio, x, and the reversibility ( Eq. 1) with decreasing particle size of the insertion material. Kavan and co-workers [3,7] observed this eftect in detail by electrochemical mesurements on thin TiO2 films supported on SnOz-coated glass. In prac-
Acknowledgements This work was partially supported by the Swiss Commission for the encouragement of scientific research.
References [ 1 ] S.Y Huang, L Kavan, A. Kay, I. Exnar and M. Gratzel, Active and Pmswe Elec. Comp., 18 ( 19951 23-30. 12] S.Y. Huang, L. Kavan, I Exnar and M. Gratzel. J. Ele~trochem. Soc.. 142 (1995) L142-144. [ 3 ] L. Kavan, M Gr~ttzel, J. Rathousky and A. Zukal, J. Electrochem. Soc, 143 (1996) 394 [4] T. Ohzuku and T Hwm, Electrochim. Acta. 27 (1982) 1263. [5] T. Ohzuku and T Kodama, J. Power Sources. 14 11985 ) 153. [ 6 ] K. Kanamura, K. Yuasa and Z. Takehara, J Power Sources, 20 ( 1987 ) 127. [7 ] L. Kavan, K. Kratochvilovfi and M. Gratze[. J Electroanal Chem., 394 { 1995) 93