coordination sites for the lithium ions since 1 possessed eight
oxygen atoms on its glyme chains. Therefore, a part of the
borate anion in 1 did not interact with the lithium ion.
Conduction of the free anions took place, decreasing the
tLi+ of 1.
Interestingly, 1-0.1 showed an increased tLi+ value of 0.7 at
40 1C (Fig. S8, ESIw). Since the DSC curve of 1-0.1 (Fig. S10,
ESIw) was almost identical to that of 1 (Fig. 3), 1-0.1 behaved
as a crystal below 74 1C, indicating that the ionic conduction
through 1-0.1 occurred both under the crystalline state and the
plastic crystalline phase. The conduction mechanism through
the crystal and the PC phases of 1-0.1 should have been
different. In the crystal lattice of 1-0.1, defects were probably
introduced by the addition of small amounts of LiTFSI.
Hence, lithium ion conduction in the crystal phase of 1-0.1
likely occurred via a hopping mechanism through the vacant
space in the crystal lattice. Previous reports proposed a similar
conduction mechanism in regards to the crystalline polymer
electrolytes.18–22 Since the volume of the lithium ion was small
compared with that of the anions in 1-0.1, the lithium ion
conduction proceeded preferentially to give a large tLi+ value.
In general, the electrolytes consisted of an ionic matrix and
lithium salt possessed a low tLi+, which was approximately
0.01–0.35.23–26 Compared with the reports, 1-0.1 showed very
large tLi+ values since this electrolyte was a solid composed of
only lithium salts.
Fig. 3 Conductivities of 1 (black), 1-0.1 (red), 1-0.5 (green), and 1-1.0
(blue) as a function of temperature.
to plastic crystal (PC) and the second peak to melting. The
small entropy value, 1.6 J Kꢀ1 molꢀ1, for the second peak for 1
supported the existence of a PC phase in 1 since these
values were less than the criterion set by Timmermans:15
DSm o 20 J Kꢀ1 molꢀ1. To the best of our knowledge, 1 is
the first example of a plastic crystalline lithium salt.
To analyze the ionic conductivity of the plastic crystalline
lithium salts, the powder of the lithium salt was dried under
reduced pressure for at least 48 h, followed by pressing into a
disk. The disk was placed between a pair of stainless-steel
electrodes in a cell for ac impedance measurements. The ac
impedance data showed a well-defined semicircle and a
low-frequency spike (Fig. S5, ESIw). These data indicated that
the grain boundary resistances of the electrolyte were quite small.
The conductivities of lithium borate 1 were measured over a
temperature range from 20 to 120 1C (Fig. 3). The ionic
conductivities were calculated from the touchdown point on
the Z0-axis in a Nyquist plot of the products, which indicated
the resistance of the sample. The conductive behavior clearly
correlated to the observations from thermal analysis. With
increasing temperature, the ionic conductivity of 1 increased
from 1.51 ꢁ 10ꢀ9 S cmꢀ1 to 3.14 ꢁ 10ꢀ8 S cmꢀ1 in the plastic
crystalline phase.
The tLi+ of 1-0.5 and 1-1.0 at 40 1C were 0.6 and 0.1,
respectively. Because the addition of large amount of LiTFSI
to 1 increased the TFSI anion content in the electrolyte, the
increased conduction of the TFSI anion reduced the
tLi+ value. 1 existed as a crystal at room temperature, and
changed to a plastic product above ca. 80 1C. The addition of
more than 0.5 equivalents of LiTFSI to 1 led to a decrease in
the melting point. The melting points of 1-0.5 and 1-1.0 were
51.6 and 54.1 1C, respectively, as shown in Fig. S11 and S12
(ESIw). On the other hand, 1-2.0 was a room-temperature
ionic liquid with high viscosity, although it was composed of
only two lithium salts without matrix materials.
We demonstrated a new concept in the construction of
solid lithium electrolytes based on PCs with large tLi+. We
developed the novel lithium salt 1 using the simple ligand
To improve the conductive property, small amounts of
lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) were
added to 1 to give 1-n (n = amount of LiTFSI). Fig. 3 shows
ionic conductivities of the obtained electrolytes. This figure
clearly indicates that the addition of LiTFSI to 1 led to a
dramatic increase in the ionic conductivity. Conductivities
of 1 and 1-0.1 measured at 90 1C were 1.02 ꢁ 10ꢀ8 and
7.41 ꢁ 10ꢀ5 S cmꢀ1, respectively. Surprisingly, the addition
of only 0.1 equivalents of LiTFSI increased the ionic
conductivity approximately 7000 times. The electrochemical
stability of 1-0.1 is shown in Fig. S6 (ESIw). The measurement
revealed a stability window between 1.9 and 3.6 V at 80 1C.
The lithium transport numbers of electrolytes were determined
via a DC impedance polarization method using lithium foils as
non-blocking electrodes.16 Borates 1 and 1-0.1 showed similar
tLi+ values, which were ca. 0.3 at 80 1C (Fig. S7, ESIw).
The literature reports a common coordination number of
lithium complexes from 4 to 6.17 Borate 1 provided multiple
exchange reaction of borate. Borate
1 showed plastic
crystallinity and solid-state ionic conductivity. With the
addition of small amounts of LiTFSI to 1, we succeeded not
only in drastically increasing ionic conductivity but also in
inducing a large tLi+
.
This work was supported by JST PREST program and
Tatematsu Foundation.
Notes and references
1 D. R. MacFarlane, J. Huang and M. Forsyth, Nature, 1999, 402,
792.
2 D. R. MacFarlane and M. Forsyth, Adv. Mater., 2001, 13,
957.
3 Y. Abu-Lebdeh, P. J. Alarco and M. Armand, Angew. Chem., Int.
Ed., 2003, 42, 4499.
4 P. J. Alarco, Y. Abu-Lebdeh, A. Abouimrane and M. Armand,
Nat. Mater., 2004, 3, 476.
5 Y. Abu-Lebdeh, A. Abouimrane, P.-J. Alarco and M. Armand,
J. Power Sources, 2006, 154, 255.
6 J. M. Tarascon and M. Armand, Nature, 2001, 414, 359.
c
6312 Chem. Commun., 2011, 47, 6311–6313
This journal is The Royal Society of Chemistry 2011