High throughput electrochemical observation of structural phase changes in LiFe1-x Mnx PO4 during charge and discharge

Article abstract of DOI:10.1149/1.3294564

Electrochemical lithium extraction and insertion in LiFe_1-x Mn_x PO₄ has been investigated by high throughput cyclic voltammetry on an array of samples with different degrees of carbon coating to assist electron transfer. A linear decrease in the capacity was observed with an increase in the value of x and was explained by a gradual loss of electronic conductivity due to the manganese substituent. Slow scan voltammograms show differences in peak shapes corresponding to a heterogeneous (two-phase) reaction in LiFePO₄ and a homogeneous (one-phase) reaction of iron in LiFe_0.2 Mn_0.8 PO₄, illustrating a general interpretation of voltammograms to differentiate these mechanisms.

Full text of DOI:10.1149/1.3294564

Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
A381
0
013-4651/2010/157͑4͒/A381/6/$28.00 © The Electrochemical Society
High Throughput Electrochemical Observation
of Structural Phase Changes in LiFe Mn PO
1
−x
x
4
during Charge and Discharge
a
b,
a
a, ,z
Matthew R. Roberts, Girts Vitins, * Guy Denuault, and John R. Owen *
a
School of Chemistry, University of Southampton, Southampton SO17 1BJ, United Kingdom
b
QinetiQ, Gosport PO12 2AG, United Kingdom
Electrochemical lithium extraction and insertion in LiFe1−xMn PO has been investigated by high throughput cyclic voltammetry
x
4
on an array of samples with different degrees of carbon coating to assist electron transfer. A linear decrease in the capacity was
observed with an increase in the value of x and was explained by a gradual loss of electronic conductivity due to the manganese
substituent. Slow scan voltammograms show differences in peak shapes corresponding to a heterogeneous ͑two-phase͒ reaction in
LiFePO and a homogeneous ͑one-phase͒ reaction of iron in LiFe0.2Mn0.8PO , illustrating a general interpretation of voltammo-
4
4
grams to differentiate these mechanisms.
©
2010 The Electrochemical Society. ͓DOI: 10.1149/1.3294564͔ All rights reserved.
Manuscript submitted September 17, 2009; revised manuscript received November 19, 2009. Published February 9, 2010.
1
6
High throughput techniques have been developed by many
groups as methods of rapidly screening potential lithium battery
altered. This shift was attributed to the inductive effect in the
M–O–P bonding, increasing the ionic character of the transition-
metal atoms.
1
,2
electrodes. This work employs the postsynthesis array transfer
3,4
The aims of the present work are to assess the viability of our
sucrose-based preparation in the synthesis of LiFe1−xMn PO over
the whole composition range, to demonstrate the use of the PoSAT
high throughput method applied to a complex electrochemical sys-
tem with electrochemical properties that change with composition,
to investigate the effect of increasing levels of carbon coating to
compensate for the decrease in electronic conductivity on substitu-
tion of Mn for Fe in LiFePO , and to investigate how the change in
reaction mechanism affects the electrochemical response in high
throughput cyclic voltammetry.
͑
PoSAT͒ high throughput method in the study of LiFe1−xMn PO
x 4
samples with various degrees of carbon coating produced by sucrose
x
4
5
pyrolysis.
LiFePO is of great interest as a safe, environmentally acceptable
4
positive electrode for lithium-ion batteries but suffers from a low
6,7
intrinsic electronic conductivity. Its properties may be consider-
ably enhanced by doping the material or by applying a conducting
carbon coating to the particles. Several authors have successfully
4
8
introduced pyrolytic carbon from sucrose and other precursors after
synthesis as a means of enhancing the surface conductivity and
Experimental
therefore the rate performance of LiFePO . A synthesis method
4
combining the carbon coating step with a pyrolytic sol–gel synthesis
PoSAT array preparation.— Two 500 mL solutions were pre-
pared, the first containing 0.625 M of LiCH COO·2H O ͑Aldrich͒,
of LiFePO ͑with optional doping͒ from a mixed salt solution pre-
4
3
2
4
cursor has been recently investigated in this laboratory. In this case,
Fe͑NO ͒ ·9H O ͑Aldrich͒, and H PO ͑85 wt %, Aldrich͒ and the
3
3
2
3
4
sucrose also acted as a viscosity enhancing additive to suppress the
crystal growth of individual precursor components during the initial
drying process, suppressing elemental segregation before the pyroly-
sis step. A relatively low calcination temperature was found to be
sufficient to produce the required phase without any solid-state mix-
ing or grinding. Results of galvanostatic cycling showed state-of-
the-art values for capacity and rate performance.
second containing 0.4
M
LiCH COO·2H O ͑Aldrich͒,
3
2
Mn͑CH COO͒ ·4H O ͑Aldrich͒, and H PO ͑85 wt %, Aldrich͒.
3
2
2
3
4
Aliquots of the two solutions were mixed down the rows of an array
of quartz tubes so that each row contained a different value of x in
LiFe1−xMn PO ͑x = 0, 0.1, 0.2, 0.3, 0.5, 0.7, 0.9, 1͒. A 100
x
4
mL solution of 2 M sucrose was then prepared and added in differ-
ent amounts to each column so that from left to right each column
contained a sucrose-to-phosphate ratio ͑SPR͒ of 0.10, 0.15, 0.2,
0.22, 0.24, 0.27, 0.30, and 0.35. These solutions were then agitated
using a Fischer “Whirlimix” vortex mixer to ensure homogeneity
before being placed in an oven at 70°C for 12 h to remove the water.
The array of precursor solids was then calcined in an argon atmo-
sphere at 700°C using a large bore ͑80 mm diameter͒ tube furnace
Other systems with the same structure as LiFePO that have also
4
9,10
9
6
been studied include LiCoPO4,
LiNiPO , and LiMnPO . One
4 4
of the most extensively investigated of these is LiMnPO , which
4
gives a higher voltage than LiFePO with approximately the same
4
capacity and therefore has a higher theoretical energy density. How-
ever, this material has extremely low conductivity, and the reported
electrochemical performance has been poor. In the initial work by
͑
Lenton͒. After cooling, the products were crushed to a powder us-
Padhi et al., a capacity of only 6 mAh g−1 was reported, and the
6
ing a glass rod attached to a drill. Composite electrode preparation
began by adding two inks, 4% poly͑vinylidene fluoride͒-co-
hexafluoropropylene ͑PVDF-HFP, Aldrich͒ and 4% acetylene black
capacity was found to linearly decrease with the value of x in
6
11
LiFe1−xMn PO in Padhi et al. and in Yao et al. An improved
x
4
͑
AB, Shawinigan, Chevron Phillips Chemical Co. LP͒ in cyclopen-
−
1
performance of up to 160 mAh g has been reported at slow rates
tanone ͑CP͒, to the active material powders with five 1 mm zirconia
beads placed in each tube. The inks were added to give a final mass
ratio of 10% PVDF-HFP, 25% AB, and 65% active material and
were then mixed using the vortex mixer. Then, 14 ␮L aliquots of
each ink were deposited onto the appropriate position on the array of
aluminum current collectors and spread across the surface to form
an even film of ink. At the same time, a 40 ␮L sample of each ink
was deposited onto a second array of alumina microcrucibles for a
thermogravimetric analysis, as detailed below. The CP was evapo-
rated from both arrays at room temperature before drying at 80°C
followed by evacuation. The samples on both arrays were then ac-
curately weighed using a balance with a digital output for automated
recording.
͑
C/100͒ for some preparations that form extremely small
1
2,13
14
particles.
A detailed study of this material by Yamada et al.
demonstrated 160 mAh g−1 for the composition LiMn0.6Fe0.4PO at
4
slow rates and resolved the reaction mechanism according to a two-
phase charge–discharge reaction in LiFePO moving to a one-phase
behavior on the substitution of Fe for Mn.
Kobayashi et al. has shown a shift in the redox potential of
4
14,15
Further work by
3
+
2+
3+
2+
Fe /Fe and Mn /Mn as the value of x in LiFe1−xMn PO is
x
4
*
Electrochemical Society Active Member.
E-mail: jro@soton.ac.uk
z
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A382
Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
5
4
4
3
3
2
2
1
1
00
306
x in LiFe1-xMn PO4
x
50
00
50
00
50
00
50
00
304
1
.0
.9
3
3
2
2
02
00
98
96
0
0
.7
0
.5
.3
0
294
292
0.2
2
2
2
90
88
86
0
0
.1
.0
5
0
0
0.0
0.2
0.4
x in LiFe1-XMn PO
4
0.6
0.8
1.0
20
25
30
35
40
45
2
theta/ degrees
X
Figure 2. Unit cell volume of LiFe1−xMn PO .
Figure 1. XRD pattern for products formed from selected samples. The SPR
used in this column was 0.35. The value of x in LiFe1−xMn PO is shown.
x
4
x
4
5
Also shown are the standard measurements of ͑b͒ LiFePO
and ͑ϫ͒
4
1
9
LiMnPO .
4
increased from 0 to 1, there is a gradual shifting of the peak position
related to a change in the unit cell parameters. To quantify this
effect, the peaks were indexed and the unit cell volume was calcu-
lated; the results are shown in Fig. 2.
Determination of active masses by thermogravimetric analy-
sis.— The array of samples deposited in alumina microcrucibles was
heated to 800°C in air to burn off carbon and the binder. After
allowing the array to cool to room temperature, each sample was
weighed again to accurately measure the mass of the fully oxidized
active material. The percentage of active materials in each element
of the electrode array was calculated by assuming that the compo-
sitions of samples were identical with corresponding positions in the
two arrays and by considering any weight gain from oxidation of
LiFePO4.
The linear increase in the unit cell volume with x is seen in Fig.
2
, indicating Vegard’s shift and suggesting a solid solution mixing of
14 6
the iron and manganese. Yamada et al. and Padhi et al. reported a
similar result, with slightly different minimum cell volumes of
3
2
91–292 A . The discrepancy is attributed to the problems of scat-
ter due to the GADDS.
High throughput electrochemical measurements.— The cyclic
−
1
voltammograms ͑CVs͒ ͑scan rate 0.05 mV s ͒ for the array are
shown in Fig. 3. Significant currents were not detected for SPR
values less than 0.15. For the high iron-content samples ͑x Ͻ 0.5͒,
this is assumed to be due to the phase impurity, as detected in XRD,
whereas for x Ͼ 0.5, the low currents were attributed to insufficient
High throughput XRD.— For X-ray phase analysis, we used a
combinatorial X-ray diffractometer ͑Bruker AXS C2, Cu K␣ radia-
tion͒ that automatically records X-ray patterns of the 64 composite
electrodes using a general area diffraction detection system
18
electronic conductivity. The highest currents were observed for
samples containing high amounts of carbon ͑prepared using an
SPR Ͼ 0.2͒ and low amounts of Mn ͑x Յ 0.5͒. At high manganese
contents ͑x Ͼ 0.5͒, a larger SPR was required to obtain significant
currents even though the correct phase had formed.
͑
GADDS͒. The scan time was 20 min for each sample, i.e., about 24
h to record diffraction patterns across the whole array.
High throughput electrochemical evaluation of 64 electrode ar-
rays.— The cell construction and instrumentation has been reported
The role of sucrose in this preparation has three functions: It
suppresses the crystallization of the salts during drying, acts to re-
duce Fe͑III͒ to Fe͑II͒ during synthesis, and finally forms a pyrolyzed
carbon residue on the surface of the product. In principle, the thresh-
old amount of sucrose required for a percolating carbon coating can
vary with Fe͑III͒ and nitrate contents. However, the XRD data con-
firmed that all materials prepared at SPR values greater than 0.15
were pure phases, and the electrochemical results showed percola-
tion of carbon even in the sample formed from the precursor with
the highest Fe͑III͒ nitrate. Materials prepared with lower amounts of
Fe͑III͒ nitrate should contain carbon well over the percolation
threshold.
1
7
previously. The electrode array was cycled between 2.5 and 4.5 V
vs Li at the following scan rates: 0.05, 0.1, 0.4, 0.8, and
−
1
1
.6 mV s .
Bulk sample preparation.— Several bulk preparations were also
undertaken for comparison. The same synthesis method was used
for the array except that the volumes were scaled up to give a total
volume of 20 mL. LiFePO , LiFe0.7Mn PO , LiFe0.5Mn PO , and
4
0.3
4
0.5
4
LiMnPO were all prepared with an SPR of 0.25. These solutions
4
were then dried and placed in ceramic crucibles and were heated
under the same conditions. The active materials were then prepared
into composite poly͑tetrafluoroethylene͒-bound films, as described
previously, and were galvanostatically cycled at various C-rates be-
tween 2 and 4.5 V.
The dependence of specific capacity on SPR is shown in Fig. 4,
which can be described as a set of percolation profiles showing the
effect of an increase in the supply of electrons to the olivine par-
ticles via the carbon matrix. The effect of manganese on the con-
ductivity can be seen as a reduction in the maximum capacity as x
increases in LiFe1−xMn PO , as well as an increase in the amount of
Results and Discussion
XRD patterns.— The X-ray diffraction ͑XRD͒ patterns recorded
for the array indicated the minimum SPR required to form the oli-
vine phase by ensuring full reduction of Fe3+ from the iron precur-
sor. Accordingly, a minimum SPR of 0.15 was required for the iron-
rich ͑x Ͻ 0.5͒ samples, whereas an SPR of 0.10 was sufficient for
samples with x Ͼ 0.5.
x
4
sucrose required to give the characteristic increase in capacity. This
result cannot be reconciled with a carbon percolation effect because
the latter should not depend on the conductivity of the olivine. In-
stead, we propose that the effect is due to the increasing surface
coverage of carbon that would be required to make an effective
contact to a material of decreasing conductivity. As a reference, the
Figure 1 shows XRD patterns for materials formed using 0.35
SPR. Selected XRD patterns shown in Fig. 1 indicate that as x
1
8
conductivity of LiFePO is reported by Delacourt et al. as ϳ1.5
4
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
A383
SPR
0.35
0.10
0.15
0.20
0.22
0.24
0.27
0.30
X
200
0
0.00
-200
4
200
0
0.10
-
-
-
-
-
-
-
200
2
00
200
0
0
.20
0
200
-200
4
200
0.30
Figure 3. CVs ͑0.05 mV s−1͒ for the 64
0
200
LiFe1−xMn PO materials.
x
4
200
0.50
0
200
0
.70
.90
200
0
200
0
200
0
200
1.00
200
0
200
3
4
3
4
3
4
3
4
3
4
3
4
3
4
3
4
Potential/ V vs. Li
−
9
−1
−1
ϫ 10 S cm
at 25°C, whereas LiMnPO4 is ϳ3
Expanded versions of the CVs at various scan rates are shown in
Fig. 6 for selected electrodes on the array. The CVs in Fig. 6 show
−
9
ϫ 10 S cm at 300°C. Using the quoted activation energy of 1.4
eV, we estimate a conductivity of 1.2 ϫ 10− S cm at 25°C.
Figure 5 compares data from samples with an SPR value of 0.27,
which is more than sufficient to form a percolating carbon network
in each case. It shows a remarkably linear trend of decreasing ca-
20
−1
2+
3+
charge and discharge peaks around 3.5 and 4.0 V for Fe /Fe and
2
+
3+
Mn /Mn , respectively, with varying peak heights dependent on
the degree of manganese substitution. The ratio between the Mn and
Fe peak heights corresponds to the composition. This is confirmed in
Fig. 7, where a linear trend is seen, which has a gradient of approxi-
mately 1, and intercepts the axis at the origin.
−
1
pacity with Mn content, from 160 to 10 mAh g , with x changing
from 0 to 1. For comparison and validation of the high throughput
method, a series of points recorded using a standard small cell test
method from materials prepared in the bulk is shown. These points
confirm the results seen using the high throughput method.
2
+
3+
A change in the Fe /Fe peak shape with increased substitution
of Mn is also noted in Fig. 6 and explained in Fig. 8 with schematic
180
160
140
120
100
80
180
160
140
120
100
60
40
20
0
80
60
40
20
0
0.0
0.2
0.4
0.6
x
0.8
1.0
x in LiFe Mn PO
4
1
-x
0.10
0.15
0.20
0.25
0.30
0.35
Figure 5. Specific capacity vs x in LiFe1−xMn PO measured using standard
x
4
Sucrose:LiFe1-XMn PO Ratio
x
4
battery techniques ͑SPR 0.27͒ ͑᭡͒ and combinatorial techniques ͑SPR 0.3͒
᭿͒. Also shown is a line of best fit for the combinatorial measurements ͑—͒.
͑
Figure 4. Specific capacity vs sucrose:LiFe1−xMn PO ratio for x values 0
The standard materials were cycled between 2.5 and 4.5 V under galvano-
static control at a rate of C/10.6. The CVs used to make the combinatorial
measurements were cycled between 2.5 and 4.5 V at a scan rate of
x
4
͑
-᭿-͒, 0.3 ͑-᭡-͒, 0.7 ͑-૽-͒, and 1 ͑-b-͒ to investigate the percolation thresh-
old of these materials. These capacities were calculated from the CVs shown
in Fig. 3.
−
1
0.05 mV s ; this equates to an approximate rate of C/11.1.
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A384
Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
0.006
0.004
0.002
0.000
0.006
0.004
0.002
0.000
LiFePO4
LiFe0.8Mn0.2PO4
-
0.002
0.004
-0.002
-
2.5
3.0
3.5
4.0
4.5
-0.004
2.5
3.0
3.5
4.0
4.5
Potential/ V
Potential/ V
0
.003
0.0020
0.0015
0.0010
0.0005
0.0000
LiFe0.5Mn0.5PO4
0.002
LiFe0.7Mn0.3PO4
0
.001
.000
0
-
-
-
0.0005
0.0010
0.0015
-
-
0.001
0.002
2.5
3.0
3.5
4.0
4.5
2
.5
3.0
3.5
4.0
4.5
Figure 6. ͑Color online͒ LiFe1−xMnxPO4
CVs normalized by scan rate and shown
for four different scan rates: 0.05 ͑black͒,
Potential/ V
Potential/V
0
.2 ͑red͒, 0.4 ͑green͒, and 0.8 ͑blue͒
0.0006
0.0004
0.0002
0.0000
−
1
mV s . The SPR used for these prepara-
tions was 0.35.
0
.0008
.0006
0.0004
LiFe0.3Mn0.7PO4
0
LiFe0.1Mn0.9PO4
0
.0002
.0000
0
-
-
-
-
0.0002
0.0004
0.0006
0.0008
-
-
-
-
0.0002
0.0004
0.0006
0.0008
-0.0010
2.5
3.0
3.5
Potential/ V
4.0
4.5
2.5
3.0
3.5
4.0
4.5
Potential/ V
0
.0008
.0006
.0004
.0002
.0000
0
0
0
0
LiMnPO4
-0.0002
-0.0004
-0.0006
-0.0008
2.5
3.0
3.5
4.0
4.5
Potential/ V
CVs for one- and two-phase insertion reactions. The one-phase re-
action shows each redox reaction as a bell-shaped peak according to
the lattice gas model of insertion/extraction in a solid solution ͑Fig.
reaction near 4 V. The cathodic peak at 3.3 V, however, shows a
shoulder at 3.6 V, characteristic of a one-phase reaction before the
main peak for the two-phase reaction, which can be extrapolated
back to an onset of 3.5 V. ͑The peak position depends on a scan rate
and should not be used as a measure of the phase-change potential.
Therefore, the onset potential that does not depend on the scan rate
should be used.͒ These results are consistent with Yamada’s phase
diagram, which we have reproduced in Fig. 9 to show the single-
phase region in the charged materials containing manganese. The
anodic reaction, heterogeneous lithium extraction with phase change
8a͒, with some displacements due to kinetic and diffusional overpo-
tentials as in Fig. 8b. The two-phase reaction shows a rapid current
increase just beyond the equilibrium potential in each direction to a
peak followed by a decay determined by mass transport processes as
1
4
in Fig. 8c. The first CV, for LiFePO , indicates a phase change as
4
expected. In LiFe0.8Mn0.2PO , the anodic peak at 3.6 V is similar to
4
the pure material except for the small peak due to the manganese
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
A385
Li MnPO
1.0
0.8
0.6
0.4
0.2
y
4
2.5
2.0
1.5
1.0
0.5
0.0
two phase
2+
3+
Mn /Mn
single phase
2+
3+
Fe /Fe
two phase
2
+
3+
Fe /Fe
Li FePO
0.0
0.5
1.0
[
1.5
2.0
2.5
y
4 0.0
0
.0
0.2
0.4
0.6
0.8
1.0
Fe]/[Mn]
Charged
y in Li (Fe Mn )PO
4
Discharged
y
1-x
x
Figure 7. The linear relationship of the ratio of peak heights for Fe ͓I ͑Fe͔͒
p
−
1
and Mn ͓I ͑Mn͔͒ from CVs recorded at 0.05 mV s and the ratio of com-
Figure 9. ͑Color online͒
A
schematic phase diagram of the
14
p
position. This relationship is shown for both the charge ͑-᭿-͒ and discharge
Li ͑Mn Fe ͒PO ͑0 Յ x,y Յ 1͒ system according to Yamada et al.
y
1−x
x
4
͑
-b-͒ peaks.
Using the information from Fig. 8 and 9, the CVs from the array
is shown by the dotted arrow and followed by homogeneous inser-
tion into a single phase, which is inconspicuous due to the similar
decay transients in both cases. The cathodic reaction, following the
arrows in reverse, shows an initial homogeneous reaction followed
by a sharp increase in current due to a phase nucleation at y = 0.6;
the bold arrow indicates the heterogeneous two-phase insertion at
can be interpreted to observe these changes in the reaction mecha-
nism, as shown in Fig. 10. This shows selected voltammograms for
LiFe1−xMn PO samples superimposed on simulations based on lin-
x
4
ear combinations of the voltammograms of the end members
LiFePO and LiMnPO , both showing a two-phase behavior. The
4
4
comparison between the data and simulations emphasizes the
changes toward the one-phase behavior.
3
.5 V.
According to the phase diagram, the manganese redox reactions
Figure 10 shows that iron oxidation always begins by the two-
phase mechanism with an approximately constant onset potential of
3.5 V. However, the reduction of iron is increasingly pre-empted by
the one-phase reaction on substitution with manganese. The manga-
nese reduction peaks also concur with the phase diagram, showing
relatively sharp transients before the peaks. The surprising result is
should all be expected to follow a two-phase mechanism. Con-
versely, the proportion of the respective phase behaviors for the iron
reactions should change with x until 60% of Fe is replaced by Mn,
whereupon the charge from Fe2+/Fe is extracted by a complete
one-phase reaction.
3+
2.0
1.5
1.0
0.5
0.0
0.8
0.6
0.4
0.2
0.0
-
-
-
-
0.5
1.0
1.5
2.0
-
0.2
-0.4
-0.6
-0.8
2.5
3.0
3.5
4.0
4.5
2.5
3.0
3.5
4.0
4.5
Potential/ a.u.
Potential/ a.u.
a
b
Figure 8. Schematic drawings of ͑a͒ ide-
alized one-phase, ͑b͒ typical one-phase,
and ͑c͒ two-phase behaviors.
1
1
0
0
.5
.0
.5
.0
-
0.5
1.0
1.5
-
-
2.0
2.5
3.0
3.5
4.0
4.5
Potential/ a.u.
c
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A386
Journal of The Electrochemical Society, 157 ͑4͒ A381-A386 ͑2010͒
0.8
0.6
0.4
0.2
0.0
0
.6
LiFe0.5Mn0.5PO
4
LiFe0.7Mn0.3PO4
0.4
0
.2
.0
0
-
0.2
0.4
-
-
-
0.2
0.4
0.6
-
Figure 10. LiFe1−xMn PO CVs ͑exact
x
4
compositions indicated on the graph͒. To
-0.6
2+
3+
indicate the change in shape of Fe /Fe
2.5
3.0
3.5
Potential/ V vs. Li
4.0
4.5
2.5
3.0
3.5
Potential/ V vs. Li
4.0
4.5
2+
3+
and Mn /Mn , discharge peaks in the
solid solution CVs were simulated by
combining the CVs for the pure LiFePO4
and LiMnPO4 and by normalizing the
peak heights of each according to the
amount of Fe or Mn. Simulated CVs are
shown by the dotted ͑----͒, and real CVs
are shown in the solid ͑——͒.
0
0
0
0
0
.8
.6
.4
.2
.0
LiFe0.3Mn0.7PO4
-
-
-
-
-
0.2
0.4
0.6
0.8
1.0
2
.5
3.0
3.5
4.0
4.5
Potential/ V vs. Li
that a substantial degree of iron substitution of manganese causes a
pre-emption of the oxidation of manganese with an initially shallow
transient characteristic of a one-phase insertion reaction instead of a
two-phase reaction with a composition dependent equilibrium po-
tential, as reported by Yamada. A comparison with Yamada’s study
also shows a much lower separation between the anodic and ca-
thodic peaks for manganese, especially close to x = 1, where our
carbon coating and possibly a smaller particle size would have given
the greatest benefit.
References
1
. M. D. Fleischauer, T. D. Hatchard, G. P. Rockwell, J. M. Topple, S. Trussler, S. K.
Jericho, M. H. Jericho, and J. R. Dahn, J. Electrochem. Soc., 150, A1465 ͑2003͒.
2. T. D. Hatchard, J. M. Topple, M. D. Fleischauer, and J. R. Dahn, Electrochem.
Solid-State Lett., 6, A129 ͑2003͒.
. M. R. Roberts, A. D. Spong, G. Vitins, and J. R. Owen, J. Electrochem. Soc., 154,
A921 ͑2007͒.
. M. R. Roberts, G. Vitins, and J. R. Owen, J. Power Sources, 179, 754 ͑2008͒.
. A. D. Spong, G. Vitins, and J. R. Owen, J. Electrochem. Soc., 152, A2376 ͑2005͒.
. A. K. Padhi, K. S. Nanjundaswamy, and J. B. Goodenough, J. Electrochem. Soc.,
3
4
5
6
1
44, 1188 ͑1997͒.
Conclusions
7
. A. K. Padhi, K. S. Nanjundaswamy, C. Masquelier, S. Okada, and J. B. Good-
enough, J. Electrochem. Soc., 144, 1609 ͑1997͒.
. M. Armand, M. Gauthier, J.-F. Magnan, and N. Ravet, U.S. Pat. 7,601,318 ͑2004͒.
. F. Zhou, M. Cococcioni, K. Kang, and G. Ceder, Electrochem. Commun., 6, 1144
The synthesis of LiFe1−xMn PO with varying degrees of carbon
x
4
8
9
coating has been successfully demonstrated using the sucrose sol–
gel method. A phase pure material was obtained from each case
when the appropriate amount of sucrose was used, and a larger
carbon coating was found to improve the capacity according to an
increased contact area with the particles. This work has shown that
LiFe1−xMn PO reacts with a heterogeneous two-phase mechanism
at the extremes of charge and discharge but tends toward a homo-
geneous one-phase behavior in the intermediate region. The analysis
also illustrates a general method of interpretation of results of cyclic
voltammetry to complement phase information gained by diffraction
methods.
͑2004͒.
10. K. Amine, H. Yasuda, and M. Yamachi, Electrochem. Solid-State Lett., 3, 178
͑2000͒.
1. J. Yao, S. Bewlay, K. Konstantionv, V. A. Drozd, R. S. Liu, X. L. Wang, H. K. Liu,
and G. X. Wang, J. Alloys Compd., 425, 362 ͑2006͒.
2. N. H. Kwon, T. Drezen, I. Exnar, I. Teerlinck, M. Isono, and M. Graetzel, Elec-
trochem. Solid-State Lett., 9, A277 ͑2006͒.
1
x
4
1
1
3. H. S. Fang, L. P. Li, and G. S. Li, Chem. Lett., 36, 436 ͑2007͒.
14. A. Yamada, Y. Kudo, and K. Y. Liu, J. Electrochem. Soc., 148, A1153 ͑2001͒.
15. M. Kope¢, A. Yamada, G. Kobayashi, S. Nishimura, R. Kanno, A. Mauger, F.
Gendron, and C. M. Julien, J. Power Sources, 189, 1154 ͑2009͒.
6. G. Kobayashi, A. Yamada, S. Nishimura, R. Kanno, Y. Kobayashi, S. Seki, Y.
Ohno, and H. Miyashiro, J. Power Sources, 189, 397 ͑2009͒.
7. A. D. Spong, G. Vitins, S. Guerin, B. E. Hayden, A. E. Russell, and J. R. Owen, J.
Power Sources, 119–121, 778 ͑2003͒.
8. C. Delacourt, L. Laffont, R. Bouchet, C. Wurm, J. B. Leriche, M. Morcrette, J. M.
Tarascon, and C. Masquelier, J. Electrochem. Soc., 152, A913 ͑2005͒.
1
1
1
Acknowledgment
This work was supported by the EPSRC “Accelerated Discovery
of Lithium Battery Materials ͑Adlib͒ grant GR/S27238/01.
University of Southampton assisted in meeting the publication costs of
this article.
19. O. García-Moreno, M. Alvarez-Vega, F. Garcia-Alvarado, J. García-Jaca, J. M.
Gallardo Amores, M. L. Sanjuán, and U. Amador, Chem. Mater., 13, 1570 ͑2001͒.
Downloaded on 2014-05-06 to IP 156.56.100.223 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).