Colloidal-crystal-templated synthesis of ordered macroporous electrode materials for lithium secondary batteries

Article abstract of DOI:10.1149/1.1590324

This paper presents a general method of preparing three-dimensionally ordered macroporous (3DOM) electrode materials, including both cathode materials (V₂O₅ and LiNiO₂) and an anode material (SnO₂). The method is based on templated precipitation of inorganic precursors within a colloidal crystal of poly(methyl methacrylate) spheres and subsequent chemical conversion. 3DOM electrodes possess several features of interest in the design of novel battery materials, such as high accessible surface areas, continuous networks, and structural features on the nanometer scale. Optimal synthesis conditions and structural features of 3DOM electrode materials are described on the basis of X-ray diffraction, scanning electron microscopy, nitrogen adsorption, and chemical analysis.

Full text of DOI:10.1149/1.1590324

A1102
Journal of The Electrochemical Society, 150 ͑8͒ A1102-A1107 ͑2003͒
0
013-4651/2003/150͑8͒/A1102/6/$7.00 © The Electrochemical Society, Inc.
Colloidal-Crystal-Templated Synthesis of Ordered Macroporous
Electrode Materials for Lithium Secondary Batteries
Hongwei Yan, * Sergey Sokolov, Justin C. Lytle, Andreas Stein, Fan Zhang,b,*
a,
a
a
a,z
b,
and William H. Smyrl **
a
b
Department of Chemistry and Department of Chemical Engineering and Materials Science,
University of Minnesota, Minneapolis, Minnesota 55455, USA
This paper presents a general method of preparing three-dimensionally ordered macroporous ͑3DOM͒ electrode materials, includ-
ing both cathode materials (V O and LiNiO ) and an anode material (SnO ). The method is based on templated precipitation of
2
5
2
2
inorganic precursors within a colloidal crystal of poly͑methyl methacrylate͒ spheres and subsequent chemical conversion. 3DOM
electrodes possess several features of interest in the design of novel battery materials, such as high accessible surface areas,
continuous networks, and structural features on the nanometer scale. Optimal synthesis conditions and structural features of
3
DOM electrode materials are described on the basis of X-ray diffraction, scanning electron microscopy, nitrogen adsorption, and
chemical analysis.
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1590324͔ All rights reserved.
©
Manuscript submitted July 8, 2002; revised manuscript received February 19, 2003. Available electronically June 24, 2003.
Conventional lithium secondary batteries experience large capac-
ity losses when they are charged/discharged at high rates. This be-
havior is attributed to the rate-limiting step during the electrochemi-
cal processes, i.e., slow diffusion of lithium ions through the
ides, metals, alloys, metal chalcogenides, carbon allotropes, silicon,
2
4
germanium, and polymers. This kind of ordered macroporous
structure with spherical interconnected voids provides unique ad-
vantages for battery material applications. First, the macropores,
with pore sizes in the range of several hundred nanometers, enable
the organic electrolyte solution to infiltrate the electrode freely so
1
ϩ
electrode materials. At high charge/discharge rates, large Li inser-
ϩ
tion or extraction fluxes at the surface, and slow Li transport in the
2
ϩ
that a relatively large contact area ͑typically 10-50 m /g͒ decreases
bulk result in concentration polarization of Li within the electrode
the current density. Second, since the wall thickness of the 3DOM
material. This causes a rise/drop in battery voltage, which leads to
termination of the charge/discharge before the maximum capacity of
ϩ
structure is only several tens of nanometers, the Li diffusion dis-
the electrode material is utilized.2,3 Therefore, optimization of the
tances can be reduced dramatically. Third, previous transmission
electron microscopy ͑TEM͒ studies of the 3DOM metal oxides in-
dicated that the walls are composed of fused grains.25 Such a con-
tinuous network of electrode materials should have better electrical
conductivity than nanocrystalline materials in which the grains are
loosely aggregated together. Most importantly, this kind of assembly
has less chance of penetrating the separator because the dimensions
of 3DOM structures are much larger than the channel sizes in the
porous membrane, so that safety problems can be effectively con-
trolled. Although some practical limitations remain, such as low
energy density and poor mechanical strength, it is of scientific and
theoretical significance to synthesize and electrochemically evaluate
electrode material parameters that influence ion kinetics is an impor-
tant subject in battery research.
Recently, numerous investigations have focused on processing of
4
-20
electrode materials with submicrometer grain sizes.
By several
4
-13
low-temperature preparation techniques ͑sol-gel methods,
wet
chemical Pechini techniques1
4-16
17-20
͒ or fast microwave synthesis,
various ultrafine electrode materials have been prepared. Since the
grain sizes ͑Ͻ1 ␮m͒ of these electrode materials are much smaller
ϩ
than those produced from solid-state techniques ͑Ͼ10 ␮m͒, the Li
diffusion distances are correspondingly shorter; hence, less time is
needed to achieve full charge or discharge at the same current den-
sity. In addition, the larger surface areas of these electrode materials
lower the current density, resulting in a decrease of concentration
polarization. Although the utilization of submicrometer-grained ma-
terials has the significant advantage of increasing rate capacity, it
also leads to several drawbacks for practical applications. First, re-
ducing average grain size, while keeping the mass constant, in-
creases the grain boundary resistances of the electrode ͑internal re-
sistance͒. To maintain the internal resistance at the same level,
conductive additives ͑such as carbon black͒ are used, which de-
crease the energy density of the electrode. In addition, it is possible
for very small grains to penetrate the porous membrane separating
the electrodes, thereby short-circuiting the system and resulting in
severe safety problems. Therefore, it is necessary to assemble the
ultrafine electrode materials into continuous structures to enhance
contacts and to suppress free particle movement.
3DOM electrode materials. In this paper, we focus on the synthesis
of 3DOM materials for secondary lithium batteries, including cath-
ode (V O and LiNiO ) and anode precursor (SnO ) compositions.
2
5
2
2
The effects of synthesis conditions on the structure and texture are
discussed based on the results of X-ray diffraction ͑XRD͒, scanning
electron microscopy ͑SEM͒, N adsorption, and chemical analysis.
2
Experimental
Sample preparation.—Monodisperse poly͑methyl methacrylate͒
͑PMMA͒ spheres with a narrow size distribution and a diameter of
495 nm were synthesized via an emulsifier-free emulsion polymer-
2
6
ization technique and packed into colloidal crystals by gravita-
tional settling. For some experiments, large PMMA spheres with a
diameter of 3-5 ␮m were used, which were synthesized by disper-
2
7
sion polymerization in polar media. To prepare macroporous elec-
trode materials, centimeter-scale, dry colloidal crystals were soaked
in the corresponding precursor solution described in Table I. When
all the colloid crystals sunk to the bottom of the solution ͑which
implies that air in the interstitial spaces of the colloid crystals was
replaced with precursor solution͒, excess solution was removed by
vacuum filtration. The samples were dried in air for 0.5-2 h. For
some systems (LiNiO2 and SnO2), another soak-filtration cycle in
precipitant solution was performed to convert the soluble precursor
into a pyrolytic precipitate ͑see Table I, footnote͒. Then the samples
were calcined at the desired temperature ͑400-800°C͒ in a controlled
atmosphere for 1-15 h. All calcinations were conducted in a 22 mm
i.d. quartz tube oven. To remove the PMMA templates gently and to
Novel colloidal crystal templating methods make the assembly of
nanoscale materials possible.2
1-23
By using close-packed arrays of
uniform spheres ͑polymer or silica͒ as templates, infiltrating them
with appropriate precursors, and processing these composites to re-
move the template and form solid wall skeletons, various materials
with three-dimensionally ordered macroporous ͑3DOM͒ structures
have been prepared, including silicates, organosilicates, metal ox-
*
Electrochemical Society Active Member.
*
* Electrochemical Society Fellow.
z
E-mail: stein@chem.umn.edu
Downloaded on 2014-11-19 to IP 128.82.252.58 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, 150 ͑8͒ A1102-A1107 ͑2003͒
A1103
Table I. Synthesis conditions for 3DOM electrode materials.
Calcination
Electrode materials
Precursor
Solvent
Temp. ͑°C͒
Time ͑h͒
Atmosphere
a
V O
NaVO3
H O/MeOH
10 mL EtOH
20 mL MeOH
400
400-800
400-700
1
3
1-15
Air/N2
Air
Air or O2
2
5
2
b
SnO2
LiNiO2
2.25 g SnCl •2H O
2
2
c
1.638 g Li(CH COO)•2H O/
3
2
4
g Ni(CH COO) •4H O
3
2
2
a
A 0.5 M sodium metavanadate solution was prepared with ultrapure deionized water. The solution was passed through an ion-exchange column
containing Dowex 50WX2-100 ion-exchange resin. The precursor solution was prepared by mixing 10 mL of methanol and 20 mL of fresh solution
Ϫ1
Ϫ1
obtained from the ion-exchange column. The calcination was conducted in a mixed atmosphere of air and N ͑0.2 L min air/1.0 L min N ).
2
2
b
c
An ammonium hydroxide solution ͑20 mL, NH content 28-30 wt %͒ was used as precipitant.
3
An oxalic acid solution ͑ca. 2.4 g of oxalic acid in 20 mL of ethanol͒ was used as precipitant. Calcination between room temperature and 400°C was
Ϫ1
Ϫ1
carried out in a mixed atmosphere ͑0.2 L min air/1.0 L min N ) at a heating rate of 2°C/min; at temperatures greater than 400°C, the atmosphere
2
was switched to air or O and the heating rate changed to 10°C/min.
2
maintain an ordered macroporous inorganic structure, all the heating
rates between room temperature and 400°C were kept at 2°C/min.
pore entrances are visible throughout the surface of the product.
Figure 1B shows the detailed morphology of the same surface at a
higher magnification. It can be seen that uniform pores and inter-
connected walls formed a periodic pore structure. In most cases, two
or more layers are visible, confirming that ordered porosity exists in
three dimensions. Average pore diameters ͑center-to-center͒ of 385
Ϯ 10 nm and wall thicknesses of 50-80 nm were measured using
Product analysis.—Powder XRD measurements were performed
on a Siemens D5005 wide-angle XRD spectrometer with Cu K␣
radiation. Average grain size determinations were based on the
Scherrer equation by fitting the XRD patterns using JADE software.
A Hitachi S-800 scanning electron microscope was used to examine
the morphology of the samples. Samples for SEM were prepared by
supporting 3DOM particles on an SEM stub with double-sided car-
bon conductive tape and crushing the particles with a spatula to
expose internal features. All the samples for SEM were coated with
495 Ϯ 10 nm PMMA spheres as templates. Figure 2 shows the
XRD pattern of 3DOM V O . After calcination at 400°C for 1 h,
2
5
the walls are composed of a pure crystalline phase ͑Shcherbinaite,
PDF no. 41-1426͒. The average grain size, calculated from diffrac-
tion line widths using the Debye-Scherrer equation, was 33.4
Ϯ 0.6 nm. Elemental analyses of carbon ͑0.15 wt %͒ and hydrogen
͑0.11 wt %͒ indicated that all the PMMA was removed after calci-
nation at 400°C.
5
nm Pt to reduce charging. Nitrogen adsorption measurements were
performed on an RXM-100 catalyst characterization instrument
Advanced Scientific Design, Inc.͒ and specific surface areas were
͑
calculated by the Brunauer-Emmett-Teller ͑BET͒ method. Chemical
analyses for carbon and hydrogen were carried out at Atlantic Mi-
crolabs, Inc., Norcross, GA.
Because most of the PMMA colloidal crystals prepared in this
study have a face-centered cubic ͑fcc͒ structure,25 the interstitial
space between PMMA spheres available for infiltration with elec-
trode materials is about 26% of the total template volume. For
Results and Discussion
3
DOM V O the precursor solution was prepared by protonation of
3
DOM V O .—Recently macroporous V O with amorphous
2 5
2
5
2
5
0.5 M sodium metavanadate solution; hence, the theoretical filling
walls has been prepared by colloidal crystal templating, and revers-
ible insertion/deinsertion of lithium was demonstrated. Previous
2
8
volume of V O was only 2.7% of the total interstitial space. In
2
5
order to increase the volume fraction, the effect of multiple filling
cycles was investigated. Contrary to our expectations, neither the
pore sizes nor the wall thicknesses were affected significantly by the
number of filling cycles. Furthermore, the BET surface areas of
attempts to form 3DOM V O with crystalline walls using a vana-
2
5
dium propoxide precursor were unsuccessful because V O crystals
2
5
grew to dimensions larger than the void spaces at a calcination tem-
perature of 450°C, thereby destroying any templating effects; at
lower template removal temperatures the products were
25
macroporous but not well ordered. In this study, use of a proton-
exchanged metavanadate solution as precursor permitted formation
of well-ordered 3DOM products.
Figure 1 shows SEMs of a 3DOM V O sample calcined at
2
5
400°C for 1 h. Figure 1A is a lower magnification image in which
Figure 1. SEM images of 3DOM V O calcined at 400°C for 1 h: ͑A͒
2
5
low-magnification image of a V O particle, ͑B͒ high-magnification image
Figure 2. Powder XRD pattern for 3DOM V2O5 calcined at 400°C for 1 h
2
5
of the same particle.
and line pattern for PDF no. 41-1426, V O in the Shcherbinaite form.
2
5
Downloaded on 2014-11-19 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1104
Journal of The Electrochemical Society, 150 ͑8͒ A1102-A1107 ͑2003͒
Table II. Structural parameters and surface areas for 3DOM
V O prepared by multiple fillings.
2
5
Macropore diametera Wall thicknessa BET surface area
2
Filling cycles
͑nm͒
͑nm͒
͑m /g͒
1
2
3
385 Ϯ 10
384 Ϯ 7
390 Ϯ 12
50-80
60-75
50-75
30.6
36.9
33.0
a
Macropore diameters and wall thicknesses were estimated from SEM
images.
2
these samples all fell within 30-37 m /g with no obvious trends of
decreasing or increasing ͑see Table II͒. These results demonstrate
that multiple filling has little influence on the 3DOM structure. The
reason for this phenomenon may be that formation of a relatively
dense layer of V O gel on the surface of the templates during the
2
5
Figure 3. Powder XRD pattern for 3DOM SnO2 samples calcined at 500-
first filling cycle limits subsequent penetration by the precursor, so
that the filling fraction cannot be further increased. SEM studies
provided direct evidence to support this proposition. A cross section
of a 3DOM V O particle prepared by three filling cycles revealed a
800°C for 3 h: ͑A͒ 500, ͑B͒ 600, ͑C͒ 700, and ͑D͒ 800°C.
2
5
The calcination temperature also has a significant effect on the
clear difference between surface morphology and inner structure
image not shown͒. While the macropores inside the particle were
morphology of 3DOM SnO . Figure 6 shows high-magnification
͑
2
SEM images of 3DOM SnO2 calcined at different temperatures.
When calcined at 500°C, the structure can be regarded as a three-
dimensionally ordered array of air spheres surrounded by SnO2
well ordered and opened throughout the whole particle, the surface
appeared less ordered and macropores were partially closed, prob-
ably due to deposition of excess precursor. Similar results have also
been reported by Ni et al.29 for the infiltration of the Ti(OC H )
shells that are continuous except at window openings. With increas-
2
5 4
ing calcination temperature, the SnO shells become rougher due to
into a silica opal. Therefore, the use of multiple filling cycles is not
an efficient way to increase the volume fraction of the electrode
materials studied here.
2
grain growth. At 700°C many large grains can be distinguished on
the surface of the shell. The sample calcined at 800°C underwent
significant structural shrinkage due to sintering. A ‘‘skeleton-like’’
wall structure evolved and the windows widened significantly.
Based on these results, the 3DOM structure could be fine-tuned by
adjusting the calcination temperature.
3
DOM SnO .—Anodes derived from SnO have been studied for
2
2
possible use in lithium-ion batteries because of their high
capacities.3
0-32
It was reported that these Sn-based materials experi-
ence large volume changes during charging and discharging, which
results in internal damage to the electrode and loss of capacity and
3
DOM LiNiO .—3DOM lithium metal oxides, such as LiCoO ,
2
2
3
3,34
cyclability.
Prior work has demonstrated that electrodes with
LiNiO2 , and LiMn2O4 , can be prepared using a similar salt precipi-
tate procedure. Because these oxides are binary compounds and are
synthesized at high temperature, impurity control, phase control and
grain size control are more critical. The focus here is on 3DOM
LiNiO2 to reveal the synthesis methodology.
smaller particles and grains have better electrochemical
3
3-35
performance.
Therefore, it is of interest to study 3DOM SnO2
because the electrochemical performance of Sn-based anodes might
be further improved with such a unique structure.
Instead of the sol-gel method employed to synthesize 3DOM
In order to synthesize pure LiNiO while keeping a 3DOM struc-
2
V O , a salt precipitation reaction was applied in the synthesis of
ture, appropriate calcination conditions needed to be determined.
2
5
3
DOM SnO . This method is simple: first, stannous chloride is
2
loaded into the interstitial space of the PMMA templates via solvent;
then stannous hydroxide precipitates are formed in situ by soaking
the impregnated, dried templates in ammonium hydroxide solution.
The precipitates are further converted to SnO upon heating. Figure
2
3
5
shows the XRD patterns of 3DOM SnO2 samples calcined at
00-800°C for 3 h. All the samples can be identified as the crystal-
line SnO phase ͑Cassiterite, PDF no. 41-1445͒. Increasing calcina-
2
tion temperature resulted in a gradual increase in the diffraction
peak intensities accompanied by sharpening of the peaks, indicating
enhanced crystallinity. Figure 4 shows the dependence of average
grain size on the calcination temperature. The average grain size of
3
DOM SnO increased nearly linearly with temperature, from 8 nm
2
at 500°C to 17 nm at 800°C. Correspondingly, the specific surface
area of the 3DOM SnO samples decreased steeply with tempera-
2
2
2
ture, as shown in Fig. 5, from 55.2 m /g at 500°C to 12.3 m /g at
00°C. In a plot of 1/ͱ͑specific surface area͒ vs. temperature, a linear
8
relationship could be obtained, which agrees well with the depen-
dence of average grain size on temperature. These data indicate that
the specific surface area of the sample depends only on the grain
surfaces exposed to the macropores, and that no additional textural
mesoporosity is present. Therefore, the complete surface is expected
to be easily accessible to a liquid electrolyte.
Figure 4. Average grain size of 3DOM SnO2 as a function of calcination
temperature.
Downloaded on 2014-11-19 to IP 128.82.252.58 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, 150 ͑8͒ A1102-A1107 ͑2003͒
A1105
Figure 5. Specific surface area of 3DOM SnO as a function of calcination
2
temperature. The inset in Fig. 5 shows the temperature dependence of
1/ͱ(specific surface area).
XRD analysis was applied to investigate the evolution of phases
under various calcination conditions. Figure 7 shows XRD patterns
of 3DOM LiNiO samples (Li:Ni ϭ 1:1 mol ratio͒ calcined in air at
2
4
6
6
00-700°C for 1-15 h. At calcination temperatures lower than
00°C, only Li CO and NiO phases were found in the products. At
Figure 7. Powder XRD patterns for 3DOM LiNiO2 (Li:Ni ϭ 1:1) calcined
in air under the following conditions: ͑A͒ 400°C 1 h, ͑B͒ 500°C 1 h, ͑C͒
2
3
00°C, crystalline LiNiO began to appear, accompanied by Li CO
3
2
2
600°C 1 h, ͑D͒ 600°C 3 h, ͑E͒ 600°C 5 h, ͑F͒ 600°C 10 h, and ͑G͒
and NiO impurities. The intensity ratio of the LiNiO2 diffraction
peaks compared to those of impurities increased with calcination
time. Although at higher temperature, such as 700°C, pure LiNiO2
could be obtained, it was found that the 3DOM structure was lost
due to excessive grain growth. Therefore, the interplay between
grain growth and structural order provided an upper limit for the
700°C 5 h.
packed ͑ccp͒ oxide ions. Experimentally, it is difficult to prepare
stoichiometric compositions because LiNiO decomposes readily to
2
calcination temperature.
Li1ϪxNi1ϩxO at high temperature due to the higher vapor pressure
ϩ
2
Stoichiometric LiNiO has a layered structure in which Li and
2ϩ
3ϩ 36-38
2
of lithium salts and formation of Ni
from Ni .
Ni3 occupy octahedral sites in alternate layers of cubic close
ϩ
Li1ϪxNi1ϩxO2 exhibits poor electrochemical performance because
ϩ
3ϩ
Li ion sites are partially substituted by Ni ions and lithium-ion
3
9
diffusion is hindered. Cation mixing in the layers also has a sig-
nificant effect on X-ray intensity ratios, such as the I͑003͒/I͑104͒
5
,40
and I͑006,102͒/I͑101͒ ratios.
LiNiO with larger rechargeable ca-
2
pacity shows a higher I͑003͒/I͑104͒ ratio in the XRD patterns. Table
III summarizes the I͑003͒/I͑104͒ ratios of the 3DOM LiNiO2
samples calcined under the conditions shown in Fig. 7. All the ratios
are less than 1.0, predicting relatively poor electrochemical perfor-
mance. In order to improve the structure of LiNiO , the preparation
2
was optimized by two modifications. First, the lithium loss was
5
compensated by adding excess lithium salt. Second, an oxygen gas
Table III. Observed intensity ratios of the d003 and d104 peaks
†
I„003…ÕI„104…‡ for 3DOM LiNiO2 synthesized under different
conditions.
Synthesis conditionsa
I͑003͒/I͑104͒
Li:Ni ϭ 1:1, air, 600°C, 1 h
Li:Ni ϭ 1:1, air, 600°C, 3 h
Li:Ni ϭ 1:1, air, 600°C, 5 h
Li:Ni ϭ 1:1, air, 600°C, 10 h
Li:Ni ϭ 1:1, air, 700°C, 5 h
0.66
0.58
0.70
0.77
0.70
1.38
1.11
1.08
Li:Ni ϭ 1.05:1, O , 600°C, 5 h
2
Li:Ni ϭ 1.05:1, O , 600°C, 10 h
2
Li:Ni ϭ 1.05:1, O , 600°C, 15 h
2
Figure 6. SEM image of 3DOM SnO2 calcined at 500-800°C for 3 h: ͑A͒
a
500, ͑B͒ 600, ͑C͒ 700, and ͑D͒ 800°C.
Li:Ni mol ratios, calcination atmosphere, temperature, and time.
Downloaded on 2014-11-19 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1106
Journal of The Electrochemical Society, 150 ͑8͒ A1102-A1107 ͑2003͒
phology as the calcination time was varied. When calcined for 5 h,
DOM LiNiO had a relatively uniform, skeleton-like wall struc-
3
2
ture. With increasing calcination time, the intersecting points in the
skeleton grew in size. Grain size analysis based on XRD linewidths
revealed an average grain size increase from 25 nm for a sample
calcined for 5 h, to 31 nm for a sample calcined for 15 h. Corre-
spondingly, the specific surface area was reduced from 12 to 7.5
2
m /g. An attempt to prepare 3DOM LiNiO at higher temperature
2
was not successful. Figure 9C shows the morphology of LiNiO2
calcined at 650°C for 5 h. Due to severe sintering, a worm-like
porous structure was obtained instead of ordered macroporous
structure.
From these results, it can be concluded that grain size control is
a critical factor in the synthesis of 3DOM structures for battery
applications and other applications. Once the grain size exceeds a
critical value, the ordered arrangement of pores disappears. How-
ever, because the tolerable value depends on the sphere size of the
templates, more severe synthesis conditions may be allowed when
larger spheres are used. Figure 9D shows an SEM image of LiNiO2
calcined at 650°C for 5 h using 3-5 ␮m PMMA spheres as tem-
plates. Although the structure was not quite ordered due to the poly-
dispersity of these spheres, the template features were reproduced in
the replica structure, resulting in interconnected macropores with
thicker walls. These data indicate that in the design of 3DOM elec-
trode materials multiple factors must be taken into consideration,
including template size and composition, precursors, and calcination
conditions.
Figure 8. Powder XRD patterns for 3DOM LiNiO2 (Li:Ni ϭ 1.05:1) cal-
cined in O at 600°C for different times: ͑A͒ 5, ͑B͒ 10, and ͑C͒ 15 h.
2
flow was applied in the calcination because a higher oxygen partial
3
ϩ
pressure in the atmosphere can suppress the reduction from the Ni
to the Ni2ϩ state.
38
Figure 8 shows the XRD patterns of 3DOM LiNiO2 (Li:Ni
ϭ 1.05:1) calcined in an oxygen gas flow at 600°C for 5-15 h. No
impurity phases were observed, indicating that under these condi-
Conclusions
Colloidal crystal templating is a general method that can be ap-
plied to prepare 3DOM electrode materials. This study focused on
V O , SnO , and LiNiO systems, but other electroactive materials,
tions, the synthesis of LiNiO was much faster. Compared to the
2
previous samples, the I͑003͒/I͑104͒ ratio of LiNiO2 calcined at
2
5
2
2
600°C for 5 h increased to 1.38, predicting a better electrochemical
as well as doped compositions, should also be feasible. The synthe-
sis conditions, such as calcination temperature, time, and atmo-
sphere, have significant effects on the fabrication of 3DOM struc-
tures. By adjusting the grain size of electrode materials, the
morphology of macroporous structures can be tuned precisely.
performance. Increased calcination times led to a gradual decrease
in the I͑003͒/I͑104͒ ratios. Therefore, a long calcination time is not
desirable under optimized conditions. Figure 9 shows the SEM im-
ages of these samples. Although all the samples have highly ordered
macroporous structures, one can observe some differences in mor-
3
DOM electrode materials possess several features that are expected
to benefit their electrochemical performance. These features include
i͒ the size of the overall 3DOM particle ͑up to millimeter size͒,
͑
which should prevent penetration of wall-nanoparticles into porous
separation membranes; ͑ii͒ the large open pores which permit facile
transport of liquid electrolyte; ͑iii͒ the thin walls which significantly
shorten ion diffusion distances; ͑iv͒ the continuous network which is
expected to improve electrical conductivity; and ͑v͒ the relatively
large accessible surface. There are also some potential drawbacks of
3
3
DOM electrodes. First, because of low solid volume fractions,
DOM electrode materials have an inherently low volumetric en-
ergy density. Second, their high porosity might lead to increased
ohmic losses. To address this problem, one might incorporate con-
ductive additives within the structure. Even with these limitations,
advantages of 3DOM electrode structures have already been
2
8
demonstrated. Further research to investigate the overall perfor-
mance of 3DOM electrode materials is underway.
Acknowledgments
This research was funded by the Office of Naval Research ͑grant
no. N00014-01-1-0810, subcontracted from NWU͒, the David and
Lucile Packard Foundation, the NSF ͑DMR-9701507͒, the MRSEC
program of the NSF ͑DMR-9809364͒, the U.S. Army Research
Laboratory, and the U.S. Army Research Office ͑DAAD 19-01-1-
0512͒.
The University of Minnesota assisted in meeting the publication costs of
this article.
References
Figure 9. SEM images of 3DOM LiNiO (Li:Ni ϭ 1.05:1) calcined in O2
2
under the following conditions: ͑A͒ 600°C 5 h, ͑B͒ 600°C 15 h, and ͑C͒
1
. B. Scrosati, J. Electrochem. Soc., 139, 2276 ͑1992͒.
650°C 5 h, samples in ͑A͒-͑C͒ were prepared using 495 nm PMMA spheres
2. J. R. Owen, Chem. Soc. Rev., 26, 259 ͑1997͒.
as a template; ͑D͒ 650°C 5 h, prepared using 3-5 ␮m PMMA spheres.
3. C. J. Patrissi and C. R. Martin, J. Electrochem. Soc., 146, 3176 ͑1999͒.
Downloaded on 2014-11-19 to IP 128.82.252.58 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, 150 ͑8͒ A1102-A1107 ͑2003͒
A1107
4
5
. S. Bach, J. P. Pereiraramos, and N. Baffier, J. Electrochem. Soc., 143, 3429 ͑1996͒.
. Y. M. Choi, S. I. Pyun, S. I. Moon, and Y. E. Hyung, J. Power Sources, 72, 83
1998͒.
. C. Julien, L. El-Farh, S. Rangan, and S. Massot, J. Sol-Gel Sci. Technol., 15, 63
1999͒.
22. B. T. Holland, C. F. Blanford, and A. Stein, Science, 281, 538 ͑1998͒.
23. J. E. G. J. Wijnhoven and W. L. Vos, Science, 281, 802 ͑1998͒.
24. A. Stein and R. C. Schroden, Curr. Opin. Solid State Mater. Sci., 5, 553 ͑2001͒.
͑
6
2
2
5. B. T. Holland, C. F. Blanford, T. Do, and A. Stein, Chem. Mater., 11, 795 ͑1999͒.
6. D. Zou, S. Ma, R. Guan, M. Park, L. Sun, J. J. Aklonis, and R. Salovey, J. Polym.
Sci., Part A: Polym. Chem., 30, 137 ͑1992͒.
͑
7
8
. Y. S. Lee, Y. K. Sun, and K. S. Nahm, Solid State Ionics, 109, 285 ͑1998͒.
. W. F. Liu, X. J. Huang, Z. X. Wang, H. Li, and L. Q. Chen, J. Electrochem. Soc.,
2
2
2
7. K. Cao, J. Yu, B. Li, B. Li, and Z. Pan, Chem. Eng. J., 78, 211 ͑2000͒.
8. J. S. Sakamoto and B. Dunn, J. Mater. Chem., 12, 2859 ͑2002͒.
9. P. Ni, P. Dong, B. Cheng, X. Li, and D. Zhang, Adv. Mater. (Weinheim, Ger.), 13,
1
45, 59 ͑1998͒.
9
. C. H. Lu and S. K. Saha, Mater. Sci. Eng., B, 79, 247 ͑2001͒.
1
0. L. Rabinovich, J. Gun, O. Lev, and D. Aurbach, Adv. Mater. (Weinheim, Ger.), 10,
77 ͑1998͒.
1. Y. K. Sun and S. H. Jin, J. Mater. Chem., 8, 2399 ͑1998͒.
2. J. J. Xu, A. J. Kinser, B. B. Owens, and W. H. Smyrl, Electrochem. Solid-State
Lett., 1, 1 ͑1998͒.
4
37 ͑2001͒.
5
3
0. Y. Idota, M. Mishima, M. Miyaki, T. Kubota, and T. Miyasaka, Eur. Pat. Appl.
1
1
1
1
65140 A1 94116643.1 ͑1994͒.
31. K. Tahara, H. Ishikawa, F. Iwasaki, S. Yahagi, A. Sakkata, and T. Sakai, Eur. Pat.
Appl. 93111938.2 ͑1993͒.
32. Y. Idota, T. Kubota, A. Matsufuji, Y. Mackawa, and T. Miyasaka, Science, 276,
1395 ͑1997͒.
3. W. S. Yoon, K. K. Lee, and K. B. Kim, Electrochem. Solid-State Lett., 4, A35
͑
2001͒.
4. W. Liu, G. C. Farrington, F. Chaput, and B. Dunn, J. Electrochem. Soc., 143, 879
1996͒.
5. W. Liu, K. Kowal, and G. C. Farrington, J. Electrochem. Soc., 143, 3590 ͑1996͒.
6. S. P. Simner, P. W. Wu, and B. Dunn, J. Mater. Res., 13, 866 ͑1998͒.
7. H. W. Yan, X. J. Huang, Z. H. Lu, H. Huang, R. J. Xue, and L. Q. Chen, J. Power
Sources, 68, 530 ͑1997͒.
8. H. W. Yan, X. J. Huang, and L. Q. Chen, J. Power Sources, 82, 647 ͑1999͒.
9. M. H. Bhat, B. P. Chakravarthy, P. A. Ramakrishnan, A. Levasseur, and K. J. Rao,
Bull. Mater. Sci., 23, 461 ͑2000͒.
͑
3
3
3
3
3
3
3. I. A. Courtney and J. R. Dahn, J. Electrochem. Soc., 144, 2943 ͑1997͒.
4. J. D. Besenhard, J. Yang, and M. Winter, J. Power Sources, 68, 87 ͑1997͒.
5. N. Li and C. R. Martin, J. Electrochem. Soc., 148, A164 ͑2001͒.
6. J. Morales, C. Perez-Vicente, and J. L. Tirado, Mater. Res. Bull., 25, 623 ͑1990͒.
7. W. Li, J. N. Reimers, and J. R. Dahn, Phys. Rev. B, 46, 3236 ͑1992͒.
8. A. Hirano, R. Kanno, Y. Kawamoto, Y. Takeda, K. Yamaura, M. Takano, K.
Ohyama, M. Ohashi, and Y. Yamaguchi, Solid State Ionics, 78, 123 ͑1995͒.
9. T. Ohzuku, H. Komori, M. Nagayama, K. Sawai, and T. Hirai, Chem. Express, 6,
161 ͑1991͒.
1
1
1
1
1
3
2
2
0. V. Subramanian, C. L. Chen, H. S. Chou, and G. T. K. Fey, J. Mater. Chem., 11,
3
348 ͑2001͒.
1. O. D. Velev, T. A. Jede, R. F. Lobo, and A. M. Lenhoff, Nature (London), 389, 447
1997͒.
40. T. Ohzuku, A. Ueda, M. Nagayama, Y. Iwakoshi, and H. Komori, Electrochim.
Acta, 38, 1159 ͑1993͒.
͑
Downloaded on 2014-11-19 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).