Electrodeposition of mesoporous tin films

Article abstract of DOI:10.1039/a808775j

Mesoporous metallic tin has been electrodeposited, from the homogeneous hexagonal mesophase of a series of amphiphilic non-ionic surfactants, with a controllable repeat structure in the range of 5-10 nm.

Full text of DOI:10.1039/a808775j

Electrodeposition of mesoporous tin films
Adam H. Whitehead, Joanne M. Elliott, John R. Owen* and George S. Attard
Dept. of Chemistry, University of Southampton, Southampton, UK SO17 1BJ. E-mail: jro@soton.ac.uk
Received (in Bath, UK) 10th November 1998, Accepted 12th January 1999
Mesoporous metallic tin has been electrodeposited, from the
homogeneous hexagonal mesophase of a series of amphi-
philic non-ionic surfactants, with a controllable repeat
structure in the range of 5–10 nm.
Table 2 Repeat distances of electrolytes and corresponding tin films based
on XRD measurements. Also shown are the second cycle lithium extraction
capacities of the films from electrochemical measurements
Repeat distance
of electrolyte/Å
Repeat distance Lithium extraction
2
1
Electrolyte
tin film/Å
capacity/C g
It has been reported that mesoporous silica may be prepared by
a sol–gel route from silicon alkoxides in the presence of a low
E1
E2
E3
E4
E5
N1
58 ± 3
61 ± 3
82 ± 4
63 ± 3
84 ± 4
—
60 ± 6
66 ± 6
81 ± 8
63 ± 5
85 ± 9
—
2510
2080
2060
2570
2120
1260
1
concentration of a cationic surfactant. Similarly the formation
2
of several mesoporous metal oxides has been reported. It has
also been shown that nanostructured silica may be formed in the
presence of surfactant concentrations high enough to form
homogeneous liquid crystalline mesophases.3 Under these
conditions the nanostructure of the silica was a cast of the
architecture of the liquid crystalline phase in which it was
formed. In other words the liquid crystal phase acted as a direct
template for the silica.
,4
technique. The repeat distance was noted to depend on the type
of surfactant used and to be increased by the presence of n-
heptane. As may be expected, if a direct templating mechanism
were operating, the repeat distances of the tin films were similar
to those obtained for the respective electrolytes from which they
were deposited, Table 2. In addition, the repeat distance for tin
There have been very few reports of the formation of
nanostructured metals although Attard et al. have used a direct
templating method to prepare mesoporous platinum as both a
5
6
powder and coherent film. The platinum film was electro-
deposited from a homogeneous hexagonal (H ) mesophase of a
I
deposited from E1 was of the same order of magnitude as that
range of octaethylene glycol monoalkyl ethers with aqueous
hexachloroplatinic acid. The following work reports on the
electrochemical synthesis of mesoporous tin films (H -eSn)
I
using a route similar to that employed for platinum. Mesoporous
tin could be expected to offer advantageous properties over
existing negative electrodes in lithium ion batteries.
Tin was electrodeposited from five electrolytes containing
oligoethylene glycol monoalkyl ether surfactants (E1–E5, Table
6
I 8
observed for H -ePt using C16EO .
Transmission electron microscopy (TEM) revealed meso-
porosity in the samples, although a regular structure could not
be assigned unambiguously, Fig. 1.
22
Further analysis of tin samples with a loading of 1.2 mg cm
was undertaken after drying under vacuum at 100 °C for 24 h.
The dried samples were used as the working electrodes of non-
aqueous electrochemical cells with lithium foil counter and
1
) and one electrolyte without surfactant (N1).† After deposi-
tion the cells were disassembled and the electrodes bearing the
electrodeposited tin (H -eSn) were repeatedly washed with ca.
00 ml of absolute ethanol to remove the electrolyte. Gravi-
reference electrodes and a 1 M LiCF
diethyl carbonate electrolyte.
3 3
SO –ethylene carbonate–
I
Using these cells ac impedance measurements were carried
out over the frequency range 5 kHz to 5 Hz with a dc bias
2
metric measurements of the deposits revealed that the current
efficiency was > 94%, based on the two electron reduction of
+
potential of 2 V vs. Li–Li and oscillator level of 15 mV p–p.
II
Sn to Sn.
Small angle X-ray diffractograms of the tin deposits were
recorded over the range 2q 0.7–3.0° (Cu-Ka radiation), which
revealed a single peak for each film prepared from the
templating electrolytes E1–E5, but not for the reference
electrolyte N1, Table 2. Hence the X-ray peak may be taken to
indicate that the templated electrodes possessed a spatially
periodic structure with a characteristic repeat distance in the
range 5–10 nm which was not an artefact of the X-ray
Table 1 Electrolyte compositions
Electrolyte composition (w/o)
4
0.2 M SnSO ,
Electrolyte
Surfactant
0.3 M H
2
SO
4
n-Heptane
Surfactant
E1
E2
E3
E4
E5
a
C
C
C
C
C
16EO
18EO10
18EO20
8
50
50
40
47
50
0
0
0
5
50
50
60
48
a
a
16EO
8
a
18EO10
3.5
46.5
Nominal surfactant formula.
Fig. 1 Typical TEM image of a tin sample deposited from a templating
electrolyte (E4).
Chem. Commun., 1999, 331–332
331

These measurements showed the tin to act as a blocking
electrode at low frequencies, as would be expected in the
absence of faradaic reactions. From the values of the electrode
capacitance it appeared that the samples prepared from the
Notes and references
†
Electrodeposition procedure: of the surfactants, octaethylene glycol
monohexadecyl ether (C16EO , > 98% Fluka), had a narrow distribution of
both alkyl and oxyethylene chain lengths, whereas the others contained a
broader mixture of chain lengths: Brij 76 (average composition of C18EO10
8
C
8
16EO -based electrolytes (E1 and E4), had a surface area 35 ±
,
2
2
10 times greater than that of rolled tin foil (9.0 mF cm , Advent
Aldrich) and Brij 78 (average composition of C18EO20, Aldrich). Templat-
ing electrolytes were prepared by mixing the surfactants with an aqueous
metals) of the same geometric surface area. However, an E1
sample dried at 180 °C under vacuum had a surface area
enhancement of only 2 ± 1. Such a decrease in surface area may
be understood in terms of a reduction in the number of
accessible mesopores, which might occur by a surface melting
phenomenon at below the bulk melting point of tin (232 °C).9
For comparison tin deposited from the Brij-based electrolytes
solution of ≈ 0.2 M SnSO
n-heptane which was expected to act as a swelling agent, increasing the radii
of the cylindrical aggregates of the H phase and hence the pore size in the
templated electrodeposit. The concentrations of surfactant, aqueous solu-
tion and n-heptane (Table 1) were chosen such that a H phase of each
4 2 4
and ≈ 0.3 M H SO . E4 and E5 also contained
I
,10
1
lyotropic liquid crystal was formed, as verified from the characteristic
optical textures when viewed through a polarising microscope.7
Although the C16EO -based electrolytes (E1 and E4) were used at room
8
temperature (22 ± 2 °C) the C18EO10-based electrolytes (E2 and E5)
required heating to 45 °C, and the C18EO20-based electrolyte (E3) to 50 °C,
in order to improve ionic conductivity and ensure that a homogeneous H
I
phase was formed. All solutions and electrolytes were freshly prepared
immediately prior to commencing electrodeposition of tin.
Reference samples were prepared from a non-templating plating
electrolyte, N1, with the composition, 0.15 M SnSO
(
E2, E3 and E5) had surface areas only slightly larger than the
reference sample (N1), with surface enhancements of 8 ± 3 and
6
± 2 respectively.
It has been noted previously that the electrochemical alloying
of lithium may cause tin to undergo a volume expansion of up
to 259%.11 This expansion and any subsequent contraction, on
removal of lithium, tends to create stresses within the tin alloy,
leading to disintegration of the structure.12 As loose tin–lithium
particles would not be expected to be in good electrical contact
with the bulk of the sample the amount of lithium that may be
extracted should give an indication of the extent of pulverisa-
tion. It would be expected that extensive mesoporosity would
significantly reduce internal stresses during expansion and thus
decrease the mechanical degradation of the electrodes.
4 2 4
, 0.6 M H SO , 0.28 M
8
4
-hydroxybenzene sulfonic acid and 0.055 M p-cresol. Tin was deposited
2
2
galvanostatically from N1 with a current density of 5 mA cm . Tin was
deposited potentiostatically from E1–E5 by applying 2100 mV across a
two-electrode cell, with a tin counter electrode and copper or gold working
2
electrode of area ca. 0.5 cm .
1
J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge,
K. D. Schmitt, C. T-W. Chu, D. H. Olson, E. W. Sheppard, S. B.
McCullen, J. B. Higgins and J. L. Schenkler, J. Am. Chem. Soc., 1992,
114, 10 834.
In order to examine the electrodeposited tin, lithium was
inserted and extracted using a pulsed coulometric titration
13
regime, between the potential limits of 0.05 and 0.95 V vs. Li–
2
3
N. Ulagappan and C. N. R. Rao, Chem. Commun., 1996, 14, 1685.
G. S. Attard, M. Edgar, J. W. Emsley and C. G. Göltner, MRS Symp.
Proc., 1996, 425, 149.
+
Li . The magnitude of the insertion and extraction current
2
2
21
densities was 350 mA cm (equivalent to 290 mA g of tin).
The second cycle extraction capacities are shown in Table 2,
4
G. S. Attard, J. C. Glyde and C. G. Göltner, Nature, 1995, 378, 366.
from which it may be noted that the H
I
-eSn samples all showed
5 G. S. Attard, C. G. Göltner, J. M. Corker, S. Henke and R. H. Templer,
Angew. Chem., Int. Ed. Engl., 1997, 36, 1315.
6 G. S. Attard, P. N. Bartlett, N. R. B. Coleman, J. M. Elliott and J. R.
Owen and J.-H. Wang, Science, 1997, 278, 838
7 D. J. Mitchell, G. J. T. Tiddy, L. Waring, T. Bostock and M. P.
McDonald, J. Chem. Soc., Faraday Trans. 1, 1983, 79, 975.
higher extraction capacities than the non-templated tin (from
N1). The samples prepared from the mixed chain length
surfactants (E2, E3 and E5) had similar extraction capacities to
2
1
one another, ca. 2100 C g . However, the single chain length
surfactants (E1 and E4) provided tin samples with the highest
extraction capacities, although still below the theoretical
8
D. Pletcher, Industrial Electrochemistry, Chapman and Hall, 1982 p.
07.
4
2
1 14
maximum of 3570 C g
the H -eSn extraction capacities were much higher than those of
commercial lithium battery negative electrode materials (coke
.
Nevertheless, it may be noted that
9
E. Søndergård, R. Kofman, P. Cheyssac, F. Celestini, T. Ben David and
I
Y. Lereah, Surf. Sci., 1997, 388, L1115.
10 Y. Oshima and K. Takayanagi, Surf. Rev. Lett., 1996, 3, 1199.
11 I. A. Courtney and J. R. Dahn, J. Electrochem. Soc., 1997, 144, 2045.
12 J. Yang, M. Winter and J. O. Besenhard, Solid State Ionics, 1996, 90,
2
1 15
and graphite) which are typically < 1350 C g
In conclusion mesoporous H -eSn films were prepared from
a series of templating electrolytes. In addition tin deposited
from the C16EO -based electrolytes (E1 and E4) holds promise
.
I
2
81.
3 A. H. Whitehead, M. Perkins and J. R. Owen, J. Electrochem. Soc.,
997, 144, L92.
4 J. Wang, I. D. Raistrick and R. A. Huggins, J. Electrochem. Soc.,1986,
33, 457.
5 M. Winter, J. O. Besenhard, M. E. Spahr and P. Novak, Adv. Mater.,
998, 10, 725.
1
1
1
8
1
for use as a negative electrode material for rechargeable lithium
ion batteries.
We thank Mr N. R. B. Coleman and Prof. P. N. Bartlett for
their support and advice, and Dr C.G. Göltner for the TEM
studies. This work was supported by Southampton Innovations
Ltd and the EPSRC.
1
1
Communication 8/08775J
332
Chem. Commun., 1999, 331–332