Synthesis of NaxMnO2+δ by a reduction of aqueous sodium permanganate with sodium iodide

Article abstract of DOI:10.1006/jssc.2000.9003

Reduction of sodium permanganate with sodium iodide in aqueous solutions has been investigated systematically. The products formed have been characterized by X-ray diffraction, wet-chemical analysis, and surface area and magnetic susceptibility measurements after firing at various temperatures. The results reveal that the sodium content x in the reduction products Na_xMnO_2+δ depends strongly on the reaction pH and mildly on the relative concentrations of the reactants. Na_0.7MnO_2+δ obtained at pH>11 followed by firing at T>500°C adopts the P2 layer structure (hexagonal) with cation vacancies arising from a δ≈0.3. Na_0.7 MnO_2+δ crystallizing in a distorted P2 structure (orthorhombic) without cation vacancies (δ≈0) could be obtained by annealing the hexagonal Na_0.7MnO_2+δ (δ≈0.3) in N₂ atmosphere around 600°C. While the orthorhombic Na_0.7MnO_2+δ (δ<0.05) is stable during ion-exchange reactions with lithium salts at 25≤T≤180°C, the hexagonal Na_0.7MnO_2+δ (δ≈0.3) transforms to spinel-like phases due to the presence of cation vacancies. Na_0.5MnO_2+δ obtained at a controlled pH of 9.3 adopts a metastable layer structure on firing at 500°C and a tunnel structure isostructural with Na₄Mn₄Ti₅O₁₈ on firing at T≥600°C. The tunnel structure is stable to ion-exchange reactions without transforming to spinel-like phases. In addition, washing the reduction products with various organic solvents before firing at higher temperatures is found to influence the reaction kinetics, composition, and crystal chemistry.

Full text of DOI:10.1006/jssc.2000.9003

Journal of Solid State Chemistry 156, 331}338 (2001)
doi:10.1006/jssc.2000.9003, available online at http://www.idealibrary.com on
Synthesis of NaxMnO2؉d by a Reduction of Aqueous Sodium
Permanganate with Sodium Iodide
Y. U. Jeong and A. Manthiramꢀ
Texas Materials Institute, ETC 9.104 The University of Texas at Austin Austin, Texas 78712
Received March 20, 2000; in revised form October 5, 2000; accepted October 13, 2000; published online December 21, 2000
such as the lithium-ion batteries. In a drive to develop
Reduction of sodium permanganate with sodium iodide in relatively inexpensive and environmentally benign electrode
aqueous solutions has been investigated systematically. The
products formed have been characterized by X-ray di4raction,
wet-chemical analysis, and surface area and magnetic suscepti-
bility measurements after 5ring at various temperatures. The
results reveal that the sodium content x in the reduction products
NaxMnO2؉d depends strongly on the reaction pH and mildly on
the relative concentrations of the reactants. Na0.7MnO2؉d ob-
tained at pH > 11 followed by 5ring at T > 5003C adopts the P2
layer structure (hexagonal) with cation vacancies arising from
a d+0.3. Na0.7MnO2؉d crystallizing in a distorted P2 structure
(orthorhombic) without cation vacancies (d+0) could be ob-
tained by annealing the hexagonal Na0.7MnO2؉d (d+0.3) in N2
atmosphere around 6003C. While the orthorhombic Na0.7MnO2؉d
(d < 0.05) is stable during ion-exchange reactions with lithium
salts at 254T41803C, the hexagonal Na0.7MnO2؉d (d+0.3)
transforms to spinel-like phases due to the presence of cation
vacancies. Na0.5MnO2؉d obtained at a controlled pH of 9.3
adopts a metastable layer structure on 5ring at 5003C and a
tunnel structure isostructural with Na4Mn4Ti5O18 on 5ring at
T56003C. The tunnel structure is stable to ion-exchange reac-
tions without transforming to spinel-like phases. In addition,
washing the reduction products with various organic solvents
before 5ring at higher temperatures is found to in6uence the
reaction kinetics, composition, and crystal chemistry. ( 2001
Academic Press
materials for lithium-ion batteries, manganese oxides have
drawn much attention. The interest in manganese oxides as
electrode hosts has created intensive worldwide activity on
the synthesis and characterization of manganese oxides par-
ticularly by low-temperature, soft chemistry procedures
(1, 7).
Sodium manganese oxides Na MnO
several phases with allotropic modi"cations depending on
the values of x and d (8, 9). Some examples are Na MnO ,
(04x41) form
V
ꢁ>B
ꢂꢃꢁ
ꢁ
Na MnO , Na
MnO , Na MnO
(04d40.25),
ꢂꢃꢄ
ꢁ
ꢂꢃꢄꢄ
ꢁ
ꢂꢃꢅ
ꢁ>B
and NaMnO . Sodium manganese oxides have been
ꢁ
synthesized both by conventional solid-state reactions
(8}11) between sodium salts and manganese oxides or salts
as well as by soft chemistry synthesis procedures (12}15).
For example, they have been synthesized by a sol}gel reac-
tion between NaMnO and glucose (12), reaction between
ꢄ
MnCl and NaOH solutions followed by heating the sus-
ꢁ
pension at 853C (13), reaction between NaMnO and Mnꢁ>
ꢄ
under alkaline conditions (14), and hydrothermal treatment
of c-MnO with NaOH solution in an autoclave (15).
ꢁ
We present in this paper the synthesis of Na MnO
by
V
ꢁ>B
reducing sodium permanganate by sodium iodide in aque-
ous solutions. The in#uence of the reduction reaction pH,
concentration of reactants, washing agents, and post-heat-
treatment temperature on the crystal chemistry and com-
position of the products is presented. The procedure gives
1. INTRODUCTION
layered Na MnO
(04d40.3) and layered and tunnel
ꢂꢃꢅ
ꢁ>B
ꢁ>B
Manganese oxides exhibit a rich variety of crystal chem-
istry (1, 2) and physical properties including the recently
discovered colossal magnetoresistance (3). Binary and
ternary manganese oxides "nd several applications such as
electrode materials in primary and secondary batteries (4),
mixed ionic}electronic conductors in solid oxide fuel cells
(5), and catalysts (6). The exponential growth in portable
electronic devices such as cellular phones and laptop com-
puters has created enormous interest in lightweight batteries
Na MnO
ꢂꢃꢆ
(04d40.15). The structural stability of the
phases during ion exchange with lithium salts to obtain
cathode materials for lithium cells is presented. In addition,
the magnetic properties of the Na MnO
presented.
phases are
V
ꢁ>B
2. EXPERIMENTAL
The reduction of sodium permanganate with sodium iod-
ide was performed by two procedures. In the "rst procedure,
100 mL of 0.5 M aqueous solution of NaMnO )H O was
ꢀ To whom correspondence should be addressed.
ꢄ
ꢁ
331
0022-4596/01 $35.00
Copyright ( 2001 by Academic Press
All rights of reproduction in any form reserved.

332
JEONG AND MANTHIRAM
mixed with 100 mL of various concentrations (0.05}3M) of thiosulfate (17). Surface area was evaluated by the BET
aqueous NaI and stirred for 24 h on a magnetic stirrer; this method with a Qunatachrome Autosorb-1 instrument.
procedure is designated as &&reduction reactions without pH Thermogravimetric analysis (TGA) data were collected with
control.'' In the second procedure, 50 mL of 0.1 or 0.5 M a Perkin-Elmer Series 7 instrument at a heating rate of
aqueous solution of NaI was added slowly from a burette 23C/min in a #owing mixture of 80% N and 20% O .
ꢁ
ꢁ
into 50 mL of aqueous solution of 0.1 or 0.5 M NaMnO
Magnetic properties were evaluated with a SQUID mag-
ꢄ
that was kept under constant stirring on a magnetic stirrer netometer and an applied "eld of 1000 Oe. There were no
while maintaining the reaction pH at a predetermined value signi"cant di!erences between the zero-"eld-cooled and
by adding dilute HCl; this procedure is designated as &&re- "eld-cooled data and the data presented later in this paper
duction reactions with pH control.'' The products formed refer to "eld-cooled data.
were "ltered, washed with deionized water unless otherwise
speci"ed, and allowed to dry in air. The products were then
"red in air for 24 h at 300}9003C unless otherwise speci"ed.
Ion-exchange reactions of the "red Na MnO samples
were carried out both at ambient temperature and at higher
temperatures. The ambient temperature ion exchange was ing 100 mL of 0.5 M NaMnO )H O with 100 mL of vari-
3. RESULTS AND DISCUSSION
3.1. Reduction Reactions without pH Control
Wet chemical analysis of the products obtained by reduc-
V
ꢁ
ꢄ
ꢁ
carried out by stirring the samples for 24 h with a 4-fold ous concentrations (0.05}3 M) of NaI without controlling
excess LiCF SO solution in acetonitrile. The high-temper- the pH indicated a Na/Mn ratio of around 0.7 irrespective
ꢇ
ꢇ
ature ion exchange was carried out by re#uxing the sample of the concentration of NaI. The "nal pH in these reactions
for 12 h with a 10-fold excess LiBr solution in hexanol at after mixing the two reactants was found to be above 11.
1803C. Oxidative extraction of sodium from the "red TGA data reveal that the as-prepared samples contain
Na MnO samples was carried out by stirring for 24 h with signi"cant amount of water ((25 wt%), which is lost on
V
ꢁ
a "vefold excess iodine solution in acetonitrile. The prod- heating at ¹(3003C. Table 1 gives the compositions of the
ucts formed after the ion-exchange and sodium extraction products obtained after "ring at various temperatures the
reactions were "ltered, washed with acetonitrile, and dried. precipitate formed by reducing 100 mL of 0.5 M
The samples were characterized by X-ray powder di!rac- NaMnO ) H O with 100 mL of 0.5M NaI. The analytical
ꢄ
ꢁ
tion. Lithium, sodium, and manganese contents were deter- data show that the products "red at ¹45003C contain
mined by atomic absorption spectroscopy. Oxygen contents a small amount of iodine, which decreases with increasing
were determined by a potentiometric redox titration "ring temperature, and those "red at ¹56003C do not
employing vanadyl sulfate (16). Iodine content was deter- contain any iodine; the decrease in iodine content with
mined by oxidizing I\ with hydrogen peroxide followed by increasing "ring temperature was also con"rmed by a con-
extracting the iodine with CCl and titrating with sodium tinuous weight loss in TGA up to 6003C. The observed
ꢄ
TABLE 1
Wet Chemical Analysis and the Surface Area Data of Products Obtained by Reducing 100 mL of 0.5 M NaMnO4 H O with
'
2
100 mL of 0.5 M NaI
Magnetic susceptibility
Composition after
Sodium extraction
Firing temperature
Surface area
(3C)
Composition?
(mꢁ/g)
l
(calc) (meas)
l
Ion exchange@
%&&
%&&
300
400
500
Na
Na
Na
MnO
MnO
MnO
I
I
ꢂꢃꢅꢀ
ꢂꢃꢅꢀ
ꢂꢃꢅꢂ
ꢁꢃꢇꢈ ꢂꢃꢂꢈꢉ
3.793
ꢁꢃꢇꢉ ꢂꢃꢂꢁꢉ
I
59
39
3.956
3.986
3.997
4.059
4.110
3.940
Li
ꢂꢃꢆꢈ ꢂꢃꢀꢊ
(Li Mn O #Na MnO
ꢂꢃꢅ
MnO
ꢂꢃꢊꢂ ꢂꢃꢀꢄ ꢁꢃꢇꢇ
(Li Mn O #Na MnO
ꢂꢃꢅ
Na
MnO
I
Na
Na
MnO
MnO
I
ꢁꢃꢇꢀ ꢂꢃꢂꢁꢊ
ꢁꢃꢇꢇ ꢂꢃꢂꢁꢁ
ꢂꢃꢊꢇ
ꢁꢃꢁꢅ ꢂꢃꢂꢇꢊ
)
)
ꢁ
ꢁ
Na
ꢄ
ꢁ>B
ꢁ>B
600
700
800
900
Na
MnO
Mn
3.965
Li
ꢂꢃꢊꢈ
ꢁꢃꢁꢈ
ꢂꢃꢊꢄ
ꢁꢃꢁꢅ
[Na
Na
O ]
ꢂꢃꢊꢂ
ꢂꢃꢉꢅ
ꢁ
ꢁ
ꢁ ꢄ
MnO
Mn
ꢂꢃꢊꢉ
ꢁꢃꢁꢉ
O ]
[Na
Na
ꢂꢃꢊꢂ
ꢂꢃꢉꢉ ꢁ
MnO
Mn
9
3.959
4.016
(Li Mn O #Na MnO
ꢂꢃꢅ
)
Na
MnO
ꢂꢃꢊꢉ
ꢁꢃꢁꢆ
O ]
ꢁ
ꢁ
ꢄ
ꢁ>B
ꢂꢃꢊꢄ
ꢁꢃꢁꢇ
[Na
Na
ꢂꢃꢊꢂ
ꢂꢃꢉꢈ ꢁ
MnO
Mn
ꢂꢃꢊꢅ
ꢁꢃꢁꢁ
O ]
[Na
ꢂꢃꢊꢂ
ꢂꢃꢈꢂ ꢁ
?Crystallographic compositions based on an oxygen content of 2.00 are given in square brackets.
@Phase identi"cations are given in parantheses; data correspond to ion-exchange reactions carried out at room temperature.

SYNTHESIS OF Na MnO
333
d
V
ꢁ>
iodine content could be due to an ine$cient removal by Table 1). The as prepared sample does not show any dis-
water washing of the reaction product I formed during the cernible re#ections (Fig. 1a). The sample "red at 6003C show
ꢁ
reduction reaction. The removal of small amount (&0.03) of re#ections (Fig. 1c) corresponding to hexagonal
iodine for ¹'4003C is accompanied by a small decrease in Na MnO
, (0.05(d(0.3) that has a layer structure
ꢂꢃꢅ
ꢁ>B
Na content, which could be due to the elimination of iodine designated as P2 structure (8, 9). Further "ring of this hex-
as NaI since NaI has a melting point of 6603C and volatil- agonal Na MnO
in N atmosphere at 6003C (Fig. 1f)
ꢂꢃꢅ
ꢁ>B
ꢁ
izes at ¹'4003C. The analytical data also show that the gives orthorhombic Na MnO
(d(0.05) having a dis-
ꢂꢃꢅ
ꢁ>B
samples have an oxygen content of '2 and the oxygen torted P2 structure (8, 9). On the other hand, the sample
content decreases with increasing "ring temperature due to "red at 5003C (Fig. 1b), although showing resemblance to
a decrease in the oxidation state of manganese. However, the hexagonal Na MnO
in Fig. 1c, exhibits poorly
ꢂꢃꢅ
ꢁ>B
density measurements of Na MnO
have shown that the de"ned peaks excepting the (00l) re#ections. The poorly
V
ꢁ>B
excess d oxygen atoms are accommodated by cation de"ned, broad (hkl) re#ections could be due to the cation
vacancies in the lattice rather than as interstitial oxygen disorder arising from a high concentration of cation va-
atoms (18). Accordingly, the crystallographic compositions cancies, which in turn results from a high oxygen content
based on an oxygen content of 2.00 are given in square (d'0.3).
brackets in Table 1 for the samples that do not contain
It should be noted that the hexagonal Na MnO
ꢂꢃꢅ
ꢁ>B
iodine and are "red at ¹'6003C.
(0.05(d(0.3) could be obtained only by using an oxygen
Figure 1 shows the X-ray di!raction patterns of the atmosphere in the case of conventional solid-state reaction
sample obtained by reducing 100 mL of 0.5 M between sodium carbonate and manganese oxide, which is
NaMnO ) H O with 100 mL of 0.5 M NaI (sample in consistent with the literature data (8, 9). The solid-state
ꢄ
ꢁ
reaction in air gives impurity phases at 6003C due to incom-
plete reaction and a mixture of Na MnO
and tunnel
ꢂꢃꢅ
ꢁ>B
Na MnO
ꢂꢃꢆ
(see below) at 8003C. Interestingly, with the
solution-based procedure described in this study, hexagonal
ꢁ>B
Na MnO
(0.05(d(0.3) without impurity phases,
ꢂꢃꢅ
ꢁ>B
could be obtained in ambient atmosphere (air) at
600}8003C.
Table 1 also gives the surface area of the products after
"ring at various temperatures. The surface area decreases
from 59 to 9 mꢁ/g as the "ring temperature increases from
500 to 8003C due to a coarsening of the grains. Figure 2
gives the magnetic data of the samples given in Table 1. All
the samples are paramagnetic above about 150 K. The
sample "red at 5003C shows a broad antiferromagnetic
FIG. 1. X-ray powder di!raction patterns of the samples obtained by
reducing 100 mL of 0.5 M NaMnO ) H O with 100 mL of 0.5 M NaI
ꢄ
ꢁ
(sample in Table 1): (a) as-prepared sample, (b) after heating at 5003C, (c)
after heating at 6003C, (d) after ion-exchanging sample c with LiCF SO in
ꢇ
ꢇ
acetonitrile, (e) after ion-exchanging sample c with LiBr in hexanol at
FIG. 2. Variation of inverse molar magnetic susceptibility of
1803C, (f) after heating sample c in N atmosphere at 6003C, and (g) after
ꢁ
Na MnO
(sample in Table 1) with temperature.
ion-exchanging sample f with LiBr in hexanol at 1803C.
ꢂꢃꢅ
ꢁ>B

334
JEONG AND MANTHIRAM
of around 0.7, we turned to pursue the reactions as a func-
tion of pH with an aim to obtain products with various
Na/Mn ratios. Accordingly, 50 mL of 0.1 and 0.5 M
NaMnO ) H O was reduced with 50 mL of 0.1 and 0.5 M
ꢄ
ꢁ
NaI at a speci"ed pH value. Figure 3 gives the variation of
Na/Mn ratio (x value in Na MnO
) in the reduction
V
ꢁ>B
product with reaction pH. The amount of Na is negligible or
very small in the reduction products synthesized under
acidic conditions with pH(5. For the range 5(pH(12,
the Na content increases rapidly with increasing pH and
reaches a maximum value of around 0.7. The curves in
Fig. 3 show a dependence on the concentration of the react-
ants and the curves are shifted to higher pH values as the
ratio of NaI/NaMnO ) H O decreases. The data in Fig. 3
ꢄ
ꢁ
explain why products with a Na/Mn ratio of around 0.7
(x+0.7) were obtained invariably with all concentrations of
NaI for the reactions described in the previous section (3.1)
without pH control, in which the reaction pH was found to
be above 11. It should be noted that water also participates
in the reduction of permanganate ion depending on the pH
FIG. 3. Variation of sodium content x in the reaction product
Na MnO
with reaction pH for various concentrations of the reactants
V
ꢁ>B
(50 mL each). The analysis of the phases formed after "ring the products at
6003C is also indicated for various values of x.
ordering around 120 K followed by a possible weak in addition to the iodide ion. As a result, the products
ferromagnetic ordering around 20 K. The antiferromagnetic obtained are not simply due to a stoichiometric reaction
ordering is suppressed as the "ring temperature increases between permanganate and iodide ions. Additionally, re-
and the samples "red at ¹58003C show only the weak duction of permanganate was found to be incomplete with
ferromagnetic component at ¹(50 K. Table 1 compares lower concentrations of NaI (0.1 M) in Fig. 3 as indicated by
the measured magnetic moments in the temperature range the color of the "ltrate.
200}300 K with that calculated for high-spin Mnꢇ>ꢋꢄ> ions.
Figure 3 also gives the phase analysis data for various
The measured values are in close agreement with the cal- values of Na content x. The phase analysis was carried out
culated values. The magnetic moment increases with by X-ray di!raction after "ring the reduction products at
increasing "ring temperature due to an increase in the 6003C in air. Binary manganese oxides are formed for low
number of unpaired electrons arising from a decrease in the values of x(0.07. Sodium manganese oxide with a tunnel
oxidation state of manganese.
structure is formed for x"0.5$0.02, and with a hexagonal
layer structure ( P2 structure discussed in the previous
section 3.1) is formed for x"0.7$0.03. Two phase regions
are found for other x values as indicated in Fig. 3.
3.2. Reduction Reactions with pH Control
Since the reduction reactions described in the previous
Table 2 gives the compositions of the products obtained
section without pH control gave invariably a Na/Mn ratio after "ring at various temperatures the precipitate formed
TABLE 2
Wet Chemical Analysis and Surface Area Data of Products Obtained by Reducing 50 mL of 0.5 M NaMnO4 H O with 50 mL
'
2
of 0.5 M NaI at a Controlled pH Value of 9.3
Magnetic susceptibility
Composition after
Sodium extraction
Firing temperature
Surface area
(3C)
Composition?
(mꢁ/g)
l
(calc) (meas)
l
Ion exchange@
%&&
%&&
300
500
600
Na
MnO
I
4.027
4.079
4.264
ꢂꢃꢆꢀ
ꢁꢃꢀꢊ ꢂꢃꢂꢇꢀ
I
Na MnO
48
21
3.836
Na
Na
Li
MnO
MnO
I
Na
Na
MnO
MnO
I
ꢂꢃꢆ
ꢁꢃꢀꢆ ꢂꢃꢂꢂꢈ
ꢂꢃꢂꢊ ꢂꢃꢄꢊ
Li
ꢁꢃꢀꢊ ꢂꢃꢂꢂꢊ
ꢂꢃꢇꢅ
ꢂꢃꢇꢆ
ꢁꢃꢀꢇ ꢂꢃꢂꢁꢀ
Na MnO
ꢂꢃꢆ
ꢁꢃꢂꢊ
ꢂꢃꢀꢊ ꢂꢃꢇꢉ
ꢁꢃꢂꢉ
ꢁꢃꢂꢆ
[Na
Mn
MnO
MnO ]
O ]
ꢂꢃꢄꢉꢆ
ꢂꢃꢈꢅ
ꢁꢃꢂꢂ
ꢁ
800
Na
6
4.387
4.099
Na
Li
MnO
Na
MnO
ꢂꢃꢄꢈ
[Na
ꢂꢃꢄꢈ
ꢂꢃꢀꢊ ꢂꢃꢇꢅ
ꢁꢃꢂꢀ
ꢂꢃꢇꢆ
ꢁꢃꢂꢂ
ꢁ
?Crystallographic compositions based on an oxygen content of 2.00 are given in square brackets.
@Data correspond to ion-exchange reactions carried out at room temperature.

SYNTHESIS OF Na MnO
335
d
V
ꢁ>
by reducing 50 mL of 0.5 M NaMnO ) H O with 50 mL of
ꢄ
ꢁ
0.5M NaI at a controlled pH of 9.3. As found in Table 1, the
products "red at ¹45003C contain a small amount of
iodine, which decreases with increasing "ring temperature,
and those "red at ¹56003C do not contain any iodine.
The iodine content in the samples was also found to depend
on the reaction pH. The iodine content was found to de-
crease as the reaction pH decreases from 11 to 7 and then to
increase slightly as the pH decreases from 7 to 1. Thus the
samples synthesized close to neutral pH had the least
amount of iodine content. The oxygen content in the prod-
ucts decreases with increasing "ring temperature due to
a decrease in the oxidation state of manganese. As in
Table 1, the crystallographic compositions based on an oxy-
gen content of 2.00 are given in square brackets in Table 2
for the samples that do not contain iodine. However, the
oxygen nonstoichiometry in Table 2 is to a lesser degree
than that in Table 1 and the sample "red at 8003C has an
oxygen content of 2.00 without cation vacancies.
Figure 4 shows the X-ray di!raction patterns of the
sample in Table 2 with a Na/Mn ratio of around 0.5. The
as-prepared sample does not show any discernible re#ec-
tions (Fig. 4a). The sample "red at 5003C (Fig. 4b) shows
three broad re#ections, which could be indexed on a hexag-
onal layer structure with a"2.839 A and an unde"ned
c parameter (19). The three re#ections centered around
approximately 2h"163, 373, and 663 refer, respectively, to
(00l), (100), and (110) re#ections. Since there is no observable
(hkl) re#ection, the c parameter could not be de"ned and the
material could thus be turbostratic. For example, the c para-
meter could be 5.56 A (l"1), 11.12 A (l"2), 16.68 A (l"3)
or multiples of 5.56 A depending on the value of l for the
"rst re#ection centered around 2h"163. Although the
X-ray pattern in Fig. 4b shows some resemblance to that
reported in the literature (20) for a close composition
FIG. 4. X-ray powder di!raction patterns of the samples obtained by
reducing 50 mL of 0.5 M NaMnO ) H O with 50 mL of 0.5M NaI at
ꢄ
ꢁ
a controlled pH of 9.3 (sample in Table 2): (a) as-prepared sample, (b) after
heating at 5003C (layer structure), (c) after ion exchanging sample b with
LiCF SO in acetonitrile, (d) after extracting sodium from sample b with
ꢇ
ꢇ
iodine in acetonitrile, (e) after heating at 6003C (tunnel structure), (f) after
ion exchanging sample e with LiCF SO in acetonitrile, (g) after extracting
ꢇ
ꢇ
sodium from sample e with iodine in acetonitrile, and (h) after washing the
reaction product with acetonitrile and heating at 6003C (P2 layer structure
Na
MnO
that was obtained by washing with water
similar to that found for hexagonal Na MnO
).
ꢂꢃꢄꢆ
ꢁ>B
ꢂꢃꢅ
ꢁ>B
a sol}gel synthesized Na MnO
be established unambiguously that the two samples are
identical due to poor crystallinity and a fewer number of
)0.3NaOH, it could not
ꢂꢃꢅ
ꢁꢃꢀꢄ
re#ections. The sample "red at 6003C (Fig. 4e), on the other form to spinel-like phases. However, the Na
ꢂꢃꢄꢄ
hand, show several sharp re#ections, which correspond to composition reported in the literature is invariably con-
MnO
ꢁ
the orthorhombic Na
MnO (8, 9) that has a tunnel taminated with Mn O impurity phase. Interestingly, the
ꢂꢃꢄꢄ ꢁ ꢁ ꢇ
ꢄ
structure isostructural with Na Mn Ti O (21). The tun- 6003C sample in Fig. 4 with a Na/Mn ratio of 0.5 consists of
nel structure is maintained on raising the "ring temperature single phase tunnel material without Mn O or other impu-
further to 8003C. The 600 and 8003C samples have a lattice rities as indicated by Rietveld analysis. This observation
ꢄ ꢆ ꢀꢉ
ꢁ ꢇ
parameters of, respectively, a"9.091 (1), b"26.422 (6), and immediately raises the question that the tunnel structure
c"2.825(1) and a"9.076(1), b"26.445(6), and may be formed for x"0.5 rather than for x"0.44 in
c"2.8232(5), which are in close agreement with that re- Na MnO . In order to clarify this issue, we have carried out
V
ꢁ
ported in the literature for Na
MnO (8).
solid-state reactions at 8003C in air with various ratios of
ꢂꢃꢄꢄ
ꢁ
Recently, the tunnel Na
MnO has drawn consider- Na CO and Mn O . Since there is a possibility of volatil-
ꢂꢃꢄꢄ
ꢁ
ꢁ
ꢇ
ꢁ ꢇ
able attention as a cathode material in lithium cells ization of sodium at higher temperatures, the sodium and
(10, 11, 22, 23) as it does not transform to the spinel-like manganese contents of the products obtained by solid-state
phases during the electrochemical charge}discharge cycling reaction were also analyzed by atomic absorption spectros-
while most other nonspinel manganese oxides tend to trans- copy. The analytical data reveal that there is no loss of

336
JEONG AND MANTHIRAM
FIG. 6. Crystal structure of tunnel Na MnO (previously thought to
ꢂꢃꢆ
MnO in the literature) isostructural with Na Mn Ti O
ꢂꢃꢄꢄ ꢀꢉ
ꢁ
be Na
.
ꢁ
ꢄ
ꢄ ꢆ
Table 2 also gives the surface area of the products after
"ring at various temperatures. The surface area decreases
from 48 to 6 mꢁ/g as the "ring temperature increases from
500 to 8003C. The magnetic data of the Na MnO
ꢂꢃꢆ
ꢁ>B
samples are given in Fig. 7 and Table 2. While the tunnel
sample obtained after "ring at 8003C shows an antiferro-
magnetic order around 30 K, the layered sample obtained
after "ring at 5003C shows a slight increase in susceptibility
at lower temperatures ¹(40 K. The measured magnetic
moments in the temperature range 200}300 K are in close
agreement with the calculated values. The magnetic mo-
FIG. 5. X-ray di!raction patterns of Na MnO that were obtained by
V
ꢁ
"ring various ratios of Na CO and Mn O at 8003C in air. The unmarked
ꢁ
ꢇ
ꢁ ꢇ
re#ections correspond to the tunnel structure. The re#ections correspond-
ing to Mn O and Na MnO
impurities are indicated by arrows.
ꢁ
ꢇ
ꢂꢃꢅ
ꢁ>B
sodium. The X-ray patterns of the samples obtained by ment increases with increasing "ring temperature due to
solid-state reaction are given in Fig. 5 (24). The data clearly decreasing oxidation state of manganese.
reveal that the tunnel structure is formed for x"0.5 in
Na MnO and not for x"0.44. Compositions with x(0.5
V
ꢁ
3.3. Ewect of Washing on the Reduction Products
show Mn O impurity while those with x'0.5 show
ꢁ ꢇ
ꢁ>B
Na MnO
impurity. Since the samples "red at 8003C do
Since I is formed as one of the reaction products and its
ꢂꢃꢅ
ꢁ
not contain excess oxygen (d"0) or cation vacancies (Table solubility in water is very low, the e!ect of washing the
2), the results unambiguously demonstrate that the tunnel
structure is formed for x"0.5 in Na MnO .
V
ꢁ
The formation of the tunnel structure for x"0.5 in
Na MnO rather than for the expected value of x"0.44 in
V
ꢁ
analogy with Na Mn Ti O could be understood by con-
sidering the crystal structure (Fig. 6). The structure of
Na Mn Ti O consists of MnO square pyramids and
MnO octahedra, which are connected together by sharing
edges and corners to give two types of tunnels: one large
ꢄ
ꢄ ꢆ ꢀꢉ
ꢄ
ꢄ
ꢆ ꢀꢉ
ꢆ
ꢊ
S-shaped tunnel and two pentagonal tunnels per unit cell.
While each S-shaped tunnel contains four Na sites, each
pentagonal tunnel contains one Na site. In Na Mn Ti O
ꢄ
ꢄ ꢆ ꢀꢉ
(x"0.44), all the Na sites in the pentagonal tunnels (corre-
sponding to x"0.22) and 50% of the Na sites in the
S-shaped tunnels (corresponding to x"0.22) are occupied.
In Na MnO with x"0.5, all the Na sites in the pentag-
ꢂꢃꢆ
ꢁ
onal tunnels (corresponding to x"0.22) and 62.5% of the
Na sites in the S-shaped tunnels (corresponding to x"0.28)
will be occupied.
FIG. 7. Variation of inverse molar magnetic susceptibility of
Na MnO
(sample in Table 1) with temperature.
ꢂꢃꢆ
ꢁ>B

SYNTHESIS OF Na MnO
337
d
V
ꢁ>
reduction products with other solvents such as acetonitrile, comparison of the intensity of the re#ections in Figs. 1c}1g.
acetone and ethanol was investigated. While iodine is highly The stability of the orthorhombic form is consistent with
soluble in acetonitrile and acetone, it is only slightly soluble that recently found with the layered Na
[Li Mn
]O
ꢂꢃꢊꢅ ꢂꢃꢀꢅ ꢂꢃꢉꢇ
ꢁ
in ethanol. Washing of the reduction products obtained by that has the P2 structure without cation vacancies (25). It is
reducing 50 mL of 0.5 M NaMnO ) H O with 50 mL of remarkable that the P2 structure with no or little cation
ꢄ
ꢁ
0.5M NaI at a controlled pH of 9.3 (sample in Table 2) with vacancies is quite stable while most manganese oxides with
water or ethanol followed by "ring at 6003C gave close-packed structures tend to transform to spinel-like
Na MnO
(d40.06) having the tunnel structure as seen phases. The distorted P2 structure thus has the potential to
ꢂꢃꢆ
ꢁ>B
in Fig. 4e. On the other hand, washing of the products with emerge as a candidate for lithium-ion cells (25, 26). We
acetonitrile or acetone followed by "ring at the same tem- believe that the instability of the hexagonal form during
perature of 6003C gave Na MnO
a hexagonal layer structure (Fig. 4h) similar to that of vacancies arising from a high oxygen content. The cation
(d'0.1) having ion-exchange reactions is due to the presence of cation
ꢂꢃꢆ
ꢁ>B
Na MnO
(P2 layer structure). The higher oxygen con- vacancies seem to promote the migration of the Li> ions
ꢂꢃꢅ
ꢁ>B
tent in the acetonitrile-washed product seems to favor the into the manganese layer and thereby favor the formation of
layer structure instead of the expected tunnel structure. This spinel phases. However, only a small amount of sodium
metastable layer structure was found to be maintained re- could be extracted by oxidation with iodine from both the
markably up to a "ring temperature of 8003C. For "ring forms (Table 1).
temperatures ¹59003C, the tunnel phase begins to appear.
With the Na MnO
composition, the layer structure
ꢂꢃꢆ
ꢁ>B
The washing with di!erent solvents thus appears to a!ect formed at 5003C undergoes a larger degree of ion-exchange
the reaction kinetics and thereby in#uence the oxygen con- than the tunnel structure (Table 2). The smaller degree of
tent and the crystal chemistry of the product. Although ion exchange in the latter is consistent with the di$culty of
washing with acetonitrile removes completely the reaction removal of Na> ions from the smaller pentagonal tunnels as
product iodine and thereby avoids a possible decrease in Na found by Doe! et al. (11). Ion exchange under more drastic
content on "ring at ¹'4003C due to the removal of iodine conditions (2703C) with eutectic mixtures is required to
as NaI, the sodium contents were found to be similar exchange all the Na> by Li> in the tunnel structure (23). On
(x"0.5$0.02) in the water-washed and acetonitrile- the other hand, both the structures undergo small but nearly
washed products after "ring at 6003C. Thus the di!erences the same degree of sodium extraction (0.14$0.01 Na per
observed in the crystal chemistry are not due to di!erences Na MnO
in Na contents.
formula) with iodine. While no impurity
ꢂꢃꢆ
ꢁ>B
phases are formed during the ion exchange reactions
(Figs. 4c and 4f), unknown impurity phases are formed
during the sodium extraction reactions (Figs. 4d and 4g), as
indicated by extra re#ections and Rietveld analysis. Al-
3.4. Ion Exchange and Sodium Extraction Reactions
Due to a widespread interest in manganese oxides as though some of the extra re#ections seem to correspond to
cathode hosts in lithium cells, both ion-exchange reactions Na MnO , it could not be established unambiguously
ꢂꢃꢁꢆ
ꢁ
with lithium salts and sodium-extraction reactions with due to the poor crystallinity in Fig. 4d and possible overlap
iodine of the products synthesized were investigated. The of peaks in Fig. 1g. More importantly, the tunnel phase does
compositions of the products obtained after the ion- not transform to the spinel-like phases during ion exchange
exchange and sodium-extraction reactions are given in or sodium extraction as has been found before by Doe! et
Tables 1 and 2. The X-ray di!raction patterns of the prod- al. (10, 11, 22).
ucts are given in Figs. 1 and 4.
With the Na MnO
(d+0.3) with undistorted P2 structure and cation vacancies
(Figs. 1c, 1d, and 1e) undergoes ion exchange more readily
than the orthorhombic form (d(0.05) with distorted P2 sodium permanganate with sodium iodide in aqueous solu-
structure (Figs. 1f and 1g). With the hexagonal form, the ion tions followed by "ring the reduction products at higher
exchange occurs to a greater degree at higher temperatures temperatures. The sodium content, oxygen content, and
than at ambient temperature (Figs. 1d and 1e). More impor- crystal chemistry of the products are in#uenced by the
tantly, the ion-exchange reactions of the hexagonal form reaction pH, solvents used for washing the products, and the
composition, the hexagonal form
4. CONCLUSIONS
ꢂꢃꢅ
ꢁ>B
Na MnO
oxides have been synthesized by reducing
V
ꢁ>B
leads to the formation of the spinel LiMn O (Fig. 1e) or "ring temperature. Na MnO
crystallizing in a hexag-
ꢁ ꢄ ꢂꢃꢅ
ꢁ>B
lithiated spinel Li Mn O (Fig. 1d), indicating the instabil- onal P2 structure (d+0.3) and in a distorted orthorhombic
ꢁ
ꢁ ꢄ
ity of the hexagonal form. On the other hand, the distorted P2 structure (d+0) and Na MnO
crystallizing in two
ꢂꢃꢆ
ꢁ>B
orthorhombic form (Fig. 1f) is quite stable during the ion metastable layer structures and in a thermodynamically
exchange reaction without transforming to the spinel-type stable tunnel structure have been obtained. Both the
phases (Fig. 1g), as indicated by a careful analysis and orthorhombic Na MnO
(d+0) and the tunnel
ꢂꢃꢅ
ꢁ>B

338
JEONG AND MANTHIRAM
are stable to ion-exchange reactions with 7. A. Manthiram and J. Kim, Chem. Mater. 10, 2895 (1998).
Na MnO
ꢂꢃꢆ
ꢁ>B
8. J. Parant, R. Olazcuaga, M. Devalette, C. Fouassier, and P. Hagenmul-
ler, J. Solid State Chem. 3, 1 (1971).
lithium salts, while the hexagonal Na MnO
(d+0.3) is
ꢂꢃꢅ
ꢁ>B
unstable to ion-exchange reactions forming spinel-like
phases due to the migration of manganese ions promoted by
the presence of cation vacancies. This observation may help
in designing and developing structurally stable manganese
oxide cathode hosts for lithium-ion batteries. Additionally,
while it is di$cult to synthesize Na MnO
tional solid-state reactions in air, the solution-based
approach described in this paper accesses readily Na
MnO
9. A. Mendiboure, C. Delmas, and P. Hagenmuller, J. Solid State Chem.
57, 323 (1985).
10. M. M. Doe!, M. Y. Peng, Y. Ma, and L. C. De Jonghe, J. Electrochem.
Soc. 141, L145 (1994).
11. M. M. Doe!, T. J. Richardson, and L. De Kepley, J. Electrochem. Soc.
143, 2507 (1996).
12. S. Ching, D. J. Petrovay, M. L. Jorgensen, and S. L. Suib, Inorg. Chem.
36, 883 (1997).
by conven-
ꢂꢃꢅ
ꢁ>B
13. P. Le Go!, N. Ba$er, S. Bach, and J. P. Pereira-Ramos, Mater. Res.
Bull. 31, 63 (1996).
ꢂꢃꢅ
in ambient air over a range of temperatures
ꢁ>B
14. Y. F. Shen, R. P. Zerger, R. N. DeGuzman, S. L. Suib, L. McCurdy,
D. I. Potter, and C. L. O'Young, Science 260, 511 (1993).
15. Q. Feng, H. Kanoh, Y. Miyai, and K. Ooi, Chem. Mater. 7, 1722
(1995).
5004¹49003C.
ACKNOWLEDGMENT
16. C. Tsang and A. Manthiram, Solid State Ionics 89, 305 (1996).
17. J. Kim and A. Manthiram, Nature 390, 265 (1997).
18. C. Delmas and C. Fouassier, Z. Anorg. Allg. Chem. 420, 184
(1976).
This work was supported by the Welch Foundation Grant F-1254 and
Texas Advanced Technology Program Grant 003658-0488-1999.
19. Y. U. Jeong and A. Manthiram, Electrochem. Solid State ¸ett. 2, 421
(1999).
REFERENCES
20. P. Le Go!, N. Ba$er, S. Bach, J. P. Pereira-Ramos, and R. Messina,
Solid State Ionics 61, 1 (1971).
21. W. G. Mumme, Acta Crystallogr. B 24, 1114 (1968).
1. P. Strobel and J. Charenton, Rev. Chim. Miner. 23, 125 (1986).
2. M. M. Thackeray, Prog. Solid State Chem. 25, 1 (1997).
3. R. Von Helmolt, J. Wecker, B. Holzapfelh, L. Schultz, and K. Samwer, 22. M. M. Doe!, L. Ding, and L. C. De Jonghe, Mater. Res. Soc. Symp.
Phys. Rev. ¸ett. 71, 2331 (1993). Proc. 393, 107 (1995).
4. R. Huber, K. V. Kordesch, A. Kozawa, and D. B. Wood, in &&Batteries'' 23. A. R. Armstrong, H. Huang, R. A. Jennings, and P. G. Bruce, J. Mater.
(K. V. Kordesch, Ed.), Vol. 1. Dekker, New York, 1974. Chem. 8, 255 (1998).
5. H. J. M. Bouwmeester and A. J. Burggraaf, in &&The CRC Handbook of 24. A. Manthiram and Y. U. Jeong, &&Electrochemical Society Proceedings''
Solid State Electrochemistry'' (P. J. Gellings and H. J. M. Bouwmees-
ter, Eds.), p. 481. CRC Press, New York, 1997.
6. C. G. Vayenas, S. I. Bebelis, I. V. Yentekakis, and S. N. Neophytides, in
&&The CRC Handbook of Solid State Electrochemistry'' (P. J. Gellings
and H. J. M. Bouwmeester, Eds.), p. 445. CRC Press, New York, 1997.
(G. A. Nazri, T. Ohzuku, and M. M. Thackeray, Eds.), Vol. 99}24,
p. 29. The Electrochemical Society, Pennington, NJ, 2000.
25. J. M. Paulsen, C. L. Thomas, and J. R. Dahn, J. Electrochem. Soc. 146,
3560 (1999).
26. J. M. Paulsen and J. R. Dahn, Solid State Ionics 126, 3 (1999).