Silver mediation of Fe(VI) charge transfer: Activation of the K2FeO4 super-iron cathode

Article abstract of DOI:10.1021/jp014720j

An unexpectedly large Ag(II) mediation of Fe(VI) redox chemistry improves alkaline Fe(VI) cathodic charge transfer. Combined with a Zn anode, this results in a cell with 3- to 5-fold higher energy capacity than the conventional high-power Zn/MnO₂ alkaline battery, and twice that previously observed for Zn/BaFeO₄. Both experimental results and a model of this phenomenon are presented. The Ag(II) salt may be introduced as a simple composite of AgO with the Fe(VI) salt. The Fe(VI) super-iron salt K₂FeO₄ has a high 3e^- intrinsic charge capacity (406 mA/g), and is more environmentally benign than the Fe(VI) salt BaFeO₄, but had exhibited comparatively poor charge transfer. Successful AgO cathodic activation of both K₂FeO₄ and BaFeO₄ redox chemistry are presented. Various other K₂FeO₄ activators are also studied. An observed interaction of Fe(VI) with Mn(VII/VI) can improve charge efficiency of a K₂FeO₄ composite with KMnO₄ or BaMnO₄, albeit not to the extent observed in an K₂FeO₄/AgO composite cathode. The extent of an activation effect of oxides, hydroxides, and titanates salts, as well as KMnO₄, BaMnO₄, AgMnO₄, and fluorinated graphites, on the cathodic discharge of K₂FeO₄ are probed.

Full text of DOI:10.1021/jp014720j

J. Phys. Chem. B 2002, 106, 5947-5955
5947
Silver Mediation of Fe(VI) Charge Transfer: Activation of the K2FeO4 Super-iron Cathode
Stuart Licht,* Vera Naschitz, and Susanta Ghosh
Department of Chemistry and Institute of Catalysis Science, Technion - Israel Institute of Technology,
Haifa, 32000, Israel
ReceiVed: December 31, 2001; In Final Form: March 29, 2002
An unexpectedly large Ag(II) mediation of Fe(VI) redox chemistry improves alkaline Fe(VI) cathodic charge
transfer. Combined with a Zn anode, this results in a cell with 3- to 5-fold higher energy capacity than the
conventional high-power Zn/MnO2 alkaline battery, and twice that previously observed for Zn/BaFeO4. Both
experimental results and a model of this phenomenon are presented. The Ag(II) salt may be introduced as a
simple composite of AgO with the Fe(VI) salt. The Fe(VI) super-iron salt K2FeO4 has a high 3e- intrinsic
charge capacity (406 mA/g), and is more environmentally benign than the Fe(VI) salt BaFeO4, but had exhibited
comparatively poor charge transfer. Successful AgO cathodic activation of both K2FeO4 and BaFeO4 redox
chemistry are presented. Various other K2FeO4 activators are also studied. An observed interaction of Fe(VI)
with Mn(VII/VI) can improve charge efficiency of a K2FeO4 composite with KMnO4 or BaMnO4, albeit not
to the extent observed in an K2FeO4/AgO composite cathode. The extent of an activation effect of oxides,
hydroxides, and titanates salts, as well as KMnO4, BaMnO4, AgMnO4, and fluorinated graphites, on the cathodic
discharge of K2FeO4 are probed.
1. Introduction
are its stability and environmental impact. The stability of
BaFeO4 is less than that for K2FeO4 (which degrades by less
than 0.1% per year).10,11 Furthermore, K2FeO4 salts will not
generate the regulatory issues that can arise for BaFeO4
(recently, as a general class barium salts have been restricted
by the U.S. EPA).15 Yet, compared to BaFeO4, cathodic charge
transfer to pure K2FeO4 is considerably less efficient, and only
a small fraction of the intrinsic 3e- capacity is generated at
high discharge rates.1,5-7,10,12,14 SrFeO4 was observed to sustain
cathodic current densities intermediate to those of K2FeO4 and
BaFeO4.8
An electrochemical storage based on a zinc anode, an aqueous
electrolyte, and a manganese dioxide cathode has been a
dominant primary battery chemistry for over a century. The Zn/
MnO2 cell remains one of the most widely distributed products
in the world and the most widely used consumer battery. This
is despite its low intrinsic energy capacity and a particularly
low practical energy capacity in high power applications, which
is insufficient for developing electronic, lighting, and medical
devices. Energy, power, mass, safety, and cost of this electro-
chemical system are constrained by its MnO2 cathode.
In this study a co-cathode redox and electronic mediation
which can enhance Fe(VI) charge transfer is modeled and
demonstrated with AgO activation of the K2FeO4 cathode. A
substantial increase in electrochemical energy storage is re-
ported, resulting in an additional 2-fold increase in high-power
discharge storage capacity.
Recently we introduced a series of battery types, based on
an unusual Fe(VI) cathodic charge storage.1 Iron typically occurs
as a metal or in the valence states Fe(II) or Fe(III). Fe(VI)
species have been known for over a century, although its
chemistry remains relatively unexplored.2 By the mid 20th
century, even comprehensive texts such as Cotton & Wilkinson’s
Inorganic Chemistry no longer mention +6 valence iron,
although more recently highly oxidized irons have been explored
as an alternate oxidant for synthesis,2 and for chlorination
purification of water,3 and as intermediate in biological charge-
transfer processes.4 The term “ferrate” has been variously
applied to both Fe(II) and Fe(III) compounds. Instead, due to
their highly oxidized iron basis, multiple electron transfer, and
high intrinsic energy, we refer to cells containing iron com-
pounds in a greater than three valence state as “super-iron”
batteries. The charge insertion or reduction of Fe(VI) represents
an energetic and high-capacity source of cathodic charge.
We have synthesized, characterized, and compared several
Fe(VI) compounds to introduce and study their functionality
for primary, secondary, and nonaqueous electrochemical energy
storage.1,5-14,22,23 An alkaline battery cathode based on BaFeO4
contains intrinsic iron in the valence state Fe(VI) and exhibits
a facile, high-potential 3e- reduction. The cathodic discharge
potential of alkaline Fe(VI) reduction is 0.25 V favorable to
that of MnO2, and the intrinsic 3 F/mole charge capacity of
Fe(VI) salts such as K2FeO4 of 406 mA g-1 and BaFeO4 of
313 mA g-1 compares favorably with the 1 F/mole charge
capacity of the conventional MnO2 cathode.
2. Experimental Section
In the high-power discharge domain the BaFeO4 cathode
provides over a 2-fold increase in storage capacity compared
to the conventional alkaline MnO2 cathode.1 Two of the
main challenges to the implementation of the BaFeO4 cathode
Preparation and analysis of K2FeO4 and BaFeO4 have been
recently detailed elsewhere.10,11 K2FeO4 of 97-98.5% is
prepared according to
Fe(NO3)3‚9H2O + 3/2KClO + 5KOH f
* Corresponding author. E-mail: chrlicht@techunix.technion.ac.il.
URL: http://www.technion.ac.il/technion/chemistry/staff/licht/
K2FeO4 + 3/2KCl + 3KNO3 + 23/2H2O (1)
10.1021/jp014720j CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/21/2002

5948 J. Phys. Chem. B, Vol. 106, No. 23, 2002
Licht et al.
The dried K2FeO4 product has been found to be stable in time
on the order of years and may be used for BaFeO4 synthesis
directly or after storage. In this synthesis, dissolved Ba(OH)2‚
8H2O is mixed with a solution of dissolved K2FeO4 precipitating
BaFeO4 as previously described, yielding 96-98% purity
BaFeO4, according to10,11
K2FeO4 + Ba(OH)2 f BaFeO4 + 2KOH
(2)
AgO is prepared by standard methods (the reaction at 85 °C of
an alkaline AgNO3 solution with K2S2O8) in accord with
4AgNO3 + 2K2S2O8 + 8NaOH f 4AgO + K2SO4 +
3Na2SO4 + 2NaNO3 + 2KNO3 + 4H2O (3)
This study probes cathodic charge transfer through galvano-
static three-electrode measurements (utilizing a Pine Instruments
AFCBP1 bipotentiostat in the constant current mode, with a
Ag/AgCl reference electrode), as well as the preparation and
use of two-electrode sealed cylindrical alkaline cells. Although
not isolating the cathode redox couple as effectively as in a
galvanostatic or potentiostatic three-electrode configuration, this
two-electrode configuration has several advantages; the same
type of anode was used throughout the study (as removed from
the commercial Zn alkaline cell). As this anode exhibits facile
(low polarization, efficient) oxidation, the charge-transfer at-
tributes of the cathode are effectively demonstrated. It is the
cathode that limits the observed discharge characteristics of these
cells. Advantages of this two-electrode cell are its reproduc-
ibility, sealing (preventing competing oxygen reduction effects),
and that the cell requires a minimum of electrolyte and provides
a direct comparison to a widely distributed commercial cell.
Components are removed from standard commercial AAA
alkaline cells (a cylindrical cell configuration with diameter 10.1
mm and a 42 mm cathode current collector case height), and
the outer MnO2 mix is replaced with any cathode mix used in
these experiments, followed by reinsertion of the separator, Zn
anode mix, gasket, and anode collector and resealing of the cell.
The cathode composites contain various cathode salts, 1 µm
graphite (Leico Industries), and 13.5 M KOH electrolyte added
to support the cathode reduction. Cell potential variation over
time was measured via LabView Data Acquisition on a PC,
and cumulative discharge, as ampere hours, was determined by
subsequent integration.
Figure 1. Galvanostatic reduction on Pt for 2 mM K2FeO4 in 13.5 M
KOH or the indicated pressed Fe(VI) powders on Pt in 13.5 M KOH.
Galvanostatic reduction is measured as the current density sustained
at various potentials versus an Ag/AgCl reference electrode and shifted
by 0.2 V to the standard hydrogen electrode potential, SHE.
degree of Fe(VI) discharge; coexisting product stoichiometries
may include
Fe2O3 + xH2O a 2FeOOH‚(x - 1)/2H2O
(4)
Fe2O3 + xBa(OH)2 a BaxFe2O(3+x)‚yH2O + (x - y)H2O (5)
Fe2O3 + 2xKOH a K2xFe2O(3+x)‚yH2O + (x - y)H2O (6)
It is evident from the observed higher relative Coulombic
efficiencies of BaFeO4 cathodes,1,5-7,10,12,14 and from the high
current densities sustained in Figure 1, that the barium product
of a BaFeO4 reduction does not inhibit charge transfer to the
degree of inhibition of the potassium product of K2FeO4
reduction.
K2FeO4 and BaFeO4 permanganate (as in KMnO4) and
manganate (as in K2MnO4) exhibit energetically similar alkaline
rest potentials, which in functioning cells may exhibit a
Nernstian shift of over 0.1 V, depending on the relative activities
of the cell constituents. The conventional MnO2 cathode exhibits
an alkaline rest potential approximately 250 mV lower.
3. Results and Discussion
FeO42- + 5/2H2O + 3e- f 1/2Fe2O3 + 5OH-
3.1. Inhibition of Fe(VI) Charge Transfer. The alkaline
galvanostatic reduction of solution phase Fe(VI) on Pt generates
an observable solid Fe(III) overlayer and sustains cathodic
current densities of only less than 100 µA cm-2. As seen in
comparison in Figure 1, solid Fe(VI) cathodes can sustain 2
orders of magnitude higher current density, highest when mixed
with several percent of graphite. As represented in the scheme
included within the figure, Fe(VI) charge transfer inhibition,
when occurring, appears directly related to buildup of a low
conductivity Fe(III) reduction product at the Fe(VI)/cathode
current collector interface. We have used FTIR, ICP, and X-ray
powder diffraction to study the Fe(III) products.11 From the
FTIR, these Fe(VI) discharge products contain hydroxide and
Fe(III) salts, but the amorphous nature of the observed spectra
does not lead to identification of the specific MaFe(III)Ox (OH)y-
(H2O)z product (M ) K2 or Ba). It is reasonable to assume that
the product will vary with pH, the extent of hydration, and the
E ) 0.6 V vs SHE
(7)
(8)
2-
MnO4- + e- f MnO4
E ) 0.6 V vs SHE
MnO42- + 2H2O + 2e- f MnO2 + 4OH-
E ) 0.6 V vs SHE
(9)
MnO2 + 1/2H2O + e- f 1/2Mn2O3 + OH-
E ) ∼0.35 V vs SHE (10)
3.2. Chemical Mediation of Fe(VI) Charge Transfer. In a
recent letter to this journal, we observed a permanganate,
Mn(VII), improvement of (both potassium and barium) Fe(VI)
charge transfer.14 In Figure 2 we propose a mechanism
consistent with this facile charge transfer. With a similar alkaline
potential, an analogous description for a manganate, Mn(VI),

Silver Mediation of Fe(VI) Charge Transfer
J. Phys. Chem. B, Vol. 106, No. 23, 2002 5949
Figure 2. Co-cathode redox mediated Fe(VI) charge transfer, exemplified by Mn(VII) addition. Composites with near lying redox processes
provide multiple pathways to facilitate Fe(VI) charge transfer; illustrated here by energy, mechanism, and chemical schematic representation.
facilitated Fe(VI) process will be evident from eq 12, but is not
included in the Figure 2 description for clarity. The process
utilizes the overlapping energetics of Mn(VII) and Fe(VI) redox
chemistry to provide alternate pathways to minimize the effect
summarized in Figure 1 of Fe(III) charged transfer inhibition
and promoting Fe(VI) regeneration. The driving force for the
Fe(VI) regeneration is the chemical and potential gradient that
will spontaneously arise as the cathode discharges and, which
as described in Figure 2, creates an anodic shift in more highly
reduced portions of the cathode. Fe(III) at these sites is
spontaneously regenerated to nonpassiviating Fe(VI) by Mn-
(VII), as expressed by
potentials of the Fe(VI/III) and Mn(VII/IV) half reactions. As
detailed in Figure 2, the process may be summarized by the
co-cathode chemical mediation of the Fe(VI/III) redox reaction
to prevent Fe(VI) depletion near the cathode current collector
and provide facile charge-transfer according to
Fe(VI)collector + 3e- f Fe(III)collector
(13)
Fe(III)collector + Mn(VII) f Fe(VI)collector + Mn(IV) (14)
Fe(VI)bulk + Mn(IV) f Fe(III)bulk + Mn(VII) (15)
3.3. Chemical and Electronic Mediation of Fe(VI) Charge
Transfer. The multiple pathway charge transfer model proposed
in Figure 2 for chemically mediated facile Fe(VI) charge transfer
is consistent with the improvement in charge collection and
current density observed in the presence of permanganate co-
cathodes and can be further probed by future impedance and
chronopotentiostatic and galvanostatic measurements. For this
study, the immediate specific significance is that refinement of
the process also indicates means for further enhancements of
Fe(VI) cathodic charge which are presented here. Again the
proposed process observes that the low cathodic efficiency and
current densities sustainable for an unmodified K2FeO4/graphite
cathode are related to Fe(VI) depletion and the concurrent
2MnO4- + Fe2O3 + 2OH- a 2MnO2 + 2FeO42- + H2O
(11)
(or via manganate as expressed by)
2-
3MnO42- + H2O + Fe2O3 a 3MnO2 + 2OH- + 2FeO4
(12)
This same potential gradient will will drive Mn(IV) regeneration
to Mn(VII) via interior (bulk) Fe(VI) and the reverse reaction
described in eq 11. This provides a charge shuttle to access
bulk Fe(VI), which is only possible due to the near lying redox

5950 J. Phys. Chem. B, Vol. 106, No. 23, 2002
Licht et al.
Figure 3. Co-cathode electronic and redox mediated Fe(VI) charge transfer, exemplified by Ag(II) addition. Composites with near lying redox
processes provide multiple pathways to facilitate Fe(VI) charge transfer; illustrated here by energy, mechanism, and chemical schematic representation.
localized buildup of less conductive Fe(III). Also again the
proposed process stipulates that an energetically similar co-
cathode can alleviate this by spontaneous regeneration of this
Fe(III). In addition, we now stipulate that the co-cathode also has
sufficient electronic conduction to provide a parallel conductive
matrix which further enhances Fe(VI) charge transfer.
The refined Figure 2 model is detailed in Figure 3 in a process
for co-cathode electronic and redox mediated Fe(VI) charge
transfer in which conductive composites with energetically near
lying redox processes provide multiple pathways to facilitate
Fe(VI) charge transfer. In Figure 3, the process is exemplifed
by addition of an Ag(II) salt, in which the cathode/collector
interfacial conductivity (a) increases with Ag(I) and Ag(0)
formation, (b) diminishes with Fe(VI) reduction, and (c) recovers
with Fe(VI) regeneration. An Ag(II) salt, such as AgO, as an
Fe(VI) co-cathode is analogous to Mn(VII) and Mn(VI) in that
it exhibits intrinsic two separate alkaline cathodic redox couples
in the same potential domain as the single 3e- Fe(VI) redox
couple described in eq 7:
In comparison to Figure 2, in Figure 3 the additional pathway
for Fe(VI) charge transfer is evident in the mid-left chemical
schematic labeled electronic mediated charge transfer. Concur-
rent with increasing cathode consumption is an increasing
buildup of conductive Ag, providing electronic access to bulk
Fe(VI). Silver is a superlative metallic conductor; as the AgO
discharges, the concentration of reduced silver grows and
provides a growing conductive matrix to increasingly facilitate
the Fe(VI) reduction.
3.4. Mn Mediation of Fe(VI) Charge Transfer. Supporting
experimental evidence for the manganese (chemical) mediated
Fe(VI) charge transfer has been detailed elsewhere.14 This
synergistic improvement of a K2FeO4 cathode with either
KMnO4 or BaMnO4 is presented in Table 1. In the table it is
seen that a pure K2FeO4 exhibits poor charge-transfer kinetics
compared to a BaFeO4 cathode at both low and high rate. These
constant load discharges are respectively equivalent to constant
current discharges of approximately 1.5 and 35 mA cm-2. At
low rate, the K2FeO4 discharges to a Coulombic efficiency,
AgO + 1/2H2O + e- f 1/2Ag2O + OH
η
efficiency, of only 35% of the intrinsic 3e- charge storage
capacity, compared to ηefficiency(BaFeO4) ) 85% under the same
discharge conditions. In this determination, the intrinsic capaci-
ties again calculated from the active cathode mass, assuming a
3 F mole-1 Fe(VI f III) reduction, and subsequently compared
E ) 0.6 V vs SHE
1/2Ag2O + 1/2H2O + e- f Ag + OH-
E ) ∼0.35 V vs SHE (17)
(16)

Silver Mediation of Fe(VI) Charge Transfer
J. Phys. Chem. B, Vol. 106, No. 23, 2002 5951
TABLE 1: Comparison of the Discharge Behavior in an Alkaline AAA Cell of Either a Pure K2FeO4 with a BaFeO4 or a
a
BaMnO4 Cathode, or 50% Composite Cathodes with KMnO4 or BaMnO4
measured storage capacity, discharged to 0.8 V cutoff
dry cathode composition
high rate (2.8 Ω load)
low rate (75 Ω load)
wt % Fe salt
wt % Mn salt
charge (Ah)
ηefficiency
energy (Wh)
charge (Ah)
ηefficiency
energy (Wh)
100% BaFeO4
0.67
0.29
0.24
0.41
0.56
51%
23%
17%
22%
41%
0.88
0.34
0.28
0.47
0.67
1.11
0.80
0.50
0.98
0.85
85%
62%
35%
52%
62%
1.78
0.96
0.68
1.41
1.20
100% BaMnO4
100% K2FeO4
50% K2FeO4
50% K2FeO4
50% KMnO4
50% BaMnO4
a The Coulombic efficiency, ηefficiency, is determined by normalizing the measured charge by the theoretical capacity (based on intrinsic charge
equivalents in the cathode mass). The 100% cathodes contain either 4.2 g BaFeO4, 4.1 g BaMnO4, or 3.5 g K2FeO4. The solid cathode mix containing
50% BaMnO4 has a mass of 3.8 g, while that containing 50% KMnO4 has a mass of 3.5 g. In addition to the solid cathode mix indicated, the final
cathode mix includes 9 wt % graphite and 9 wt % of 18 M KOH.
Figure 5. Cell potential and energy capacity of alkaline cells with
K2FeO4 cathode composites containing various weight fractions of AgO
during discharge at a high constant load rate of 2.8 Ω. Cells use an
alkaline AAA configuration including in the cathode 9 wt % graphite
and 18 M KOH electrolyte.
Figure 4. Cell potential and energy capacity of alkaline cells with
BaFeO4 cathode composites containing various weight fractions of AgO
during discharge at a high constant load rate of 2.8 Ω. Cells use an
alkaline AAA configuration including in the cathode 9 wt % graphite
and 18 M KOH electrolyte. Inset: Energy at various powers during
discharge of these and comparative cells.
AgO, a high intrinsic storage capacity compared to other active
alkaline cathode materials. Consistent with their formula weight,
the salts have respective 1 or 2 F mole-1 capacities of 231 Ag2O
mAh g-1 and 433 mAh g-1 AgO. Yet as with MnO2, AgO
exhibits ineffective alkaline charge transfer at high discharge
rate. Even in the presence of a significant conductive matrix,
such as provided by the 9 wt % 1 µm graphite added to the
cathode mix, the cathodes discharge respectively to only 35%
or 37% of their intrinsic charge under 2.8 Ω load in an optimized
AAA cell configuration and determined from the active cathode
mass, assuming a 1 F mole-1 Mn(IV f III) or a 2 F mol-1
Ag(II f 0) reduction. Due to the low cell discharge potential,
the MnO2 discharge capacity of 0.6 Wh observed in the top
portion of Figure 6 diminishes by a further factor of 2 to 0.3
Wh, when the cell is discharged at a constant, higher power of
0.7 W. The constant 2.8 Ω load discharge capacity also falls to
0.4 Wh when 6 wt %, rather than 9 wt % of 1 µm graphite is
utilized in the MnO2 cathode.
In the top portion of Figure 6, the large intrinsic AgO capacity
still results in a substantial measured energy capacity of 1.2
Wh despite the low AgO charge efficiency, although the high
mass (over 5 g) of silver in the cell is generally a prohibitive
cost factor. The Coulombic efficiency of Ag2O is better
(ηefficiency(Ag2O) ) 46%) but yields a lower discharge energy
of 0.9 Wh due to the lower 1 F mol-1 Ag(I) intrinsic capacity
of Ag2O, compared to AgO. Consistent with eqs 16 and 17,
Ag2O and AgO, exhibit distinct cathodic potentials in alkaline
solution, and the expected single and double discharge voltage
plateaus are observed in Figure 6 for the respective pure
cathodes.
to the measured (Ah converted to F) discharged. High rates of
charge transfer increase inhibition and under high rate condi-
tions, the K2FeO4 discharges to a Coulombic efficiency of only
17%, and only one-third that observed for ηefficiency(BaFeO4)
under the same conditions. At low rate, a 50 wt % composite
cathode improves ηefficiency(K2FeO4) from 35% to 52%, or 62%,
respectively with either KMnO4 or BaMnO4, while at high rate
ηefficiency(K2FeO4) is improved from 17% to 22% or 41%. A
better measure of the storage capability is the energy generated
during discharge. As seen in the table, at high rate the BaFeO4
cell exhibits a discharge energy of 0.88 Wh, compared to only
0.28 Wh for the K2FeO4 cell, which improves to 0.47 and 0.67
Wh, with respective addition of 50 wt % KMnO4 or BaMnO4.
3.5. Ag Mediation of Charge Transfer. The activation of
BaFeO4 and K2FeO4 by AgO in Figures 4 and 5 is substantial
compared to that by KMnO4 or BaMnO4 observed in Table 1.
As seen in Figure 5, as little as 7 wt % AgO composite with
K2FeO4 yields a discharge energy comparable or larger than
the 50 wt % the KMnO4/K2FeO4 composite cathode. At larger
AgO fractions, high rate discharge energies as great as 1.5 Wh
are observed. In comparison with the top portion of Figure 6,
it is seen that these discharge energies are substantially higher
than conventional alkaline MnO2 and are also higher than AgO
(or BaFeO4) alone.
Silver oxide cathodes, such as Ag(I)2O, and silver peroxide,
Ag(II)O, although costly, are characterized by attractive,
relatively flat, discharge potential profiles, and in the case of

5952 J. Phys. Chem. B, Vol. 106, No. 23, 2002
Licht et al.
As with K2FeO4, an unusually high energy discharge also
occurs for an AgO cathode composite with the BaFeO4 Fe(VI)
salt. In Figure 4, a BaFeO4/AgO cathode mix maintains the
unusual high-power characteristic known for the BaFeO4
cathode without AgO. Hence, the BaFeO4 cathode both with
and without AgO generates a power of at least 0.7 W over a
constant 2.8 Ω load. However, in addition, the Fe(VI) composite
cathode unexpectedly discharges for ∼170 min and generates
1.5 Wh, whereas under the same conditions, the BaFeO4 cathode
without AgO discharges for ∼80 to 90 min and generates 0.9
Wh. The BaFeO4/AgO composite cathode exhibits a maximum
discharge energy higher than either component alone. The
discharge capacity is ∼5-fold higher than the equivalent constant
power discharge of the conventional alkaline MnO2 cell, or ∼3-
fold higher than a constant resistive load discharge.
Compared to the BaFeO4/AgO composite electrode effect,
the activation of a K2FeO4 cathode by AgO is even more
dramatic. The discharge time increase from ∼30 min to 150
min, and as seen in Figure 5, yields discharges in excess of
1.4 Wh. Incorporating a relatively low mass of added AgO,
a K2FeO4 composite cathode, without other activators, will
generate a significant alkaline energy discharge, and this low
Ag amount is expected to further decrease upon introduction
of additional alternate K2FeO4 activators. This study includes
exploration of such alternate activators for K2FeO4, with and
without added KMnO4 and BaMnO4.
Figure 6. Cell potential and energy capacity of alkaline cells with
either a pure 4.7 g MnO2, 3.5 g KMnO4, 4.1 g BaMnO4, 6.0 g AgO, or
6.0 g Ag2O cathode, or a composite cathode containing 50 wt % AgO
with 50 wt % of either MnO2, KMnO4, BaMnO4 during discharge at a
high constant load rate of 2.8 Ω. Cells use an alkaline AAA
configuration including in the cathode 9 wt % graphite and 18 M KOH
electrolyte.
3.6. Non-Ag Activation of K2FeO4. The K2FeO4 cathode
can be improved by inclusion of various activators including
manganese salts, although the observed magnitude of such
effects is small compared to that in the AgO composite. Also a
number of other composite K2FeO4 cathodes were explored
without indications of charge-transfer activation. Hence, the
AgO activation of K2FeO4 charge transfer is substantial and
unexpected. When an alternate K2FeO4 activation with alternate
composites did occur, it could be attributed to a nonmediative
physical chemical phenomenon. Figure 7 summarizes the
observed energy of the best of these alternative composites,
which are further detailed in Tables 2 and 3.
The lower portion of Figure 6 presents the ineffective
discharge of a pure KMnO4, Mn(VII), cathode. The generated
0.4 Wh reflects only 15% of the intrinsic 4 F mol-1 capacity of
the KMnO4 in the cathode. As also seen in the lower section of
the figure, the pure BaMnO4 (VI) cathode with a lower intrinsic
3 F mol-1 charge capacity discharged to a lower capacity than
KMnO4. Composite cathodes were also prepared containing both
an Ag(II) salt and either MnO2, BaMnO4, or KMnO4. Addition
of 50 wt % AgO to each of the non-silver pure cathodes in
Figure 6 leads to an improvement in the high rate cathodic
charge generation, although each of these composite cathodes
yields an energy capacity of less than 1.0 Wh, which is
substantially less than that of the pure AgO cathode, and to a
first approximation is the average of the separate cathode
components.
Composite cathodes were also prepared containing both an
Ag(II) salt and either K2FeO4 or BaFeO4. Unlike MnO2,
KMnO4, or BaMnO4, the AgO has a synergistic activation on
BaFeO4 or K2FeO4 in which the combined discharge capacity
of the composite Ag(II)/Fe(VI) cathode is larger than that of
either cathode alone. A pure K2FeO4 cathode does not exhibit
the high power density and Coulombic efficiencies evident in
a BaFeO4 cathode, although several factors (toxicity, availability,
and intrinsic capacity) favor the K2FeO4 electrode. As seen
comparing Figures 4 and 5, under comparable high load
discharge conditions, in the absence of AgO, the BaFeO4 cell
generates 3-fold higher energy, and higher average power,
compared to the K2FeO4 cell. Consistent with the observed flat
discharge potential, a similar energy of 0.9 Wh is generated
when the BaFeO4 cell is discharged at either a constant 2.8 Ω
(as shown) or a constant 0.5 A current, or discharged at a
constant 0.7 W power (not shown).
As detailed in Table 2, added salts, such as LiOH, NaOH, or
KOH do not increase a K2FeO4 cathode discharge, and the high
rate discharge of K2FeO4 with and without 10% KOH is
compared in the top section of Figure 7. Also included in Table
2 are relatively small, but significant, improvements of the
K2FeO4 discharge energy with 10% addition of CsOH and
Ba(OH)2. We have investigated a related CsOH effect on a
BaFeO4 cathodic chemistry,14 while the Ba(OH)2 improvement
is due to a partial conversion of K2FeO4 to the (facile charge
transfer) barium salt in accord with eq 2. As also detailed
in Table 2, a variety of added oxide, hydroxide, and salts do
not increase a BaMnO4 cathode’s observed charge capacity.
These include added KOH, LiOH, NaOH, Ba(OH)2, Sr(OH)2,
Ca(OH)2, Mg(OH)2, Al2O3, BaTiO3, or Co2O3. The addition of
these possible activators is in lieu of the intrinsic storage capacity
of the cathode. Hence as also seen in the table, if the activation
effect is not substantial, then high additive levels (33 wt %)
significantly diminish the observed energy capacity.
A composite K2FeO4/BaMnO4 cathode yields a significantly
higher energy capacity than either pure cathode. A pure BaMnO4
cathode discharges to a significantly higher fraction of the
intrinsic charge than does a KMnO4 cathode,14 although not
approaching the high values observed for alkaline BaFeO4 or
MnO2 cathodes. At low rate (75 Ω load/AAA configuration),
the pure BaMnO4 cathode cell yields (0.96 Wh) and an observed
Vav ) 1.19 V. As compared in Table 2, this manganate potential

Silver Mediation of Fe(VI) Charge Transfer
J. Phys. Chem. B, Vol. 106, No. 23, 2002 5953
TABLE 2: Comparison of the Discharge Behavior in an
Alkaline AAA Cell of Either a K2FeO4, KMnO4, or a
BaMnO4 Cathode with or without Added Hydroxidea
dry cathode composition,
by mass discharge to 0.8 V (constant load)
2.8 Ω
Vav
(Wh) (V) (Wh) (V)
75 Ω
E Vav
Fe
salt
wt
%
Mn
salt
wt
%
wt
%
E
salt
K2FeO4 100
K2FeO4 90
K2FeO4 90
K2FeO4 90
K2FeO4 90
K2FeO4 90
0.28 1.17 0.68 1.36
10 0.27 1.08 0.55 1.36
Ba(OH)2 10 0.35 1.15 0.79 1.44
KOH
LiOH
NaOH
CsOH
10 0.30 1.09 0.63 1.31
10 0.24 1.05 0.60 1.35
10 0.34 1.17 0.83 1.44
0.34 1.16 0.96 1.19
BaMnO4 100
BaMnO4 90 KOH
BaMnO4 90 Al2O3
10 0.32 1.10 0.98 1.18
10 0.27 1.08 0.78 1.19
BaMnO4 90 Ba(OH)2 10 0.29 1.11 0.76 1.20
BaMnO4 90 Sr(OH)2 10 0.28 1.09 0.70 1.17
BaMnO4 90 Ca(OH)2 10 0.29 1.07 0.73 1.18
BaMnO4 90 Mg(OH)2 10 0.32 1.14 0.67 1.19
BaMnO4 90 LiOH
BaMnO4 90 NaOH
10 0.27 1.08 0.73 1.18
10 0.23 1.11 0.75 1.20
BaMnO4 90 BaTiO3 10 0.28 1.08 0.93 1.16
BaMnO4 90 Co2O3
BaMnO4 67 KOH
BaMnO4 67 Al2O3
10 0.30 1.07 0.77 1.17
33 0.22 1.10 0.50 1.19
33 0.25 1.06 0.69 1.18
BaMnO4 67 Ba(OH)2 33 0.15 1.07 0.49 1.14
a Relative amounts of the indicated solid cathode mix are presented.
In addition, the final cathode mix includes 9 wt % graphite and 9 wt
% of 18 M KOH. The LiOH salt cell gave similar results with a
saturated LiOH electrolyte. The CsOH salt cell utilizes an 18 M CsOH
electrolyte.
Figure 7. Cell potential and energy capacity of alkaline cells with
K2FeO4 composite cathodes containing various relative amounts of
BaMnO4, KMnO4, AgMnO4, Ba(OH)2, or KOH, compared to K2FeO4
in the cathode mix, during discharge at a high constant load rate of 2.8
Ω. In the composite cells, the combined mass of the K2FeO4 and other
salts is intermediate to the mass of the pure salts (a pure cathode
contains 3.5 g K2FeO4, 4.2 g BaFeO4, 4.1 g BaMnO4, 3.5 g KMnO4,
or 4.6 g AgMnO4). Cells use an alkaline AAA configuration including
9 wt % graphite and 18 M KOH electrolyte in the cathode mix.
TABLE 3: Comparison of the Discharge Behavior in an
Alkaline AAA Cell of a Cathode Composite Containing
a
K2FeO4 and BaMnO4 or K2FeO4 and KMnO4
dry cathode composition,
by mass discharge to 0.8 V (constant load)
is considerably less than observed for the K2FeO4 cathode in
Table 2, consistent with the manganate’s greater fraction of the
cathodic process constrained to the lower Mn(IV f III) potential
of eq 10. However, the respective high and low rate discharge
capacity of the pure K2FeO4 cathode, at 0.28 and 0.68 Wh, are
somewhat lower than the capacities measured for the pure
BaMnO4 cathodes. As indicated in the top section of Figure 7
as the open or solid small circles, added barium manganate can
significantly enhance the discharge energy of the K2FeO4
cathode. This is a synergistic effect, increasing the energy of
either pure cathode alone. For the K2FeO4/BaMnO4 composites,
a maximum 2.8 Ω discharge energy of 0.78 Wh is measured
for the cell containing 45 wt % K2FeO4 and 55 wt % BaMnO4,
which is more than double that seen for either the pure K2FeO4
or the pure BaMnO4 cathode. As detailed in Table 3 at the low
rate (constant 75 Ω load), the K2FeO4/BaMnO4 composite
cathode exhibits a nearly constant maximum energy capacity
of 1.2 Wh over a wide composition range varying from
33%:67% to 67%:33%. KOH or Al2O3 added to BaMnO4,
impairs the discharge effectiveness both of the pure BaMnO4
cathode (Table 2) and also of the BaMnO4/K2FeO4 composite
cathode (Table 3). Ba(OH)2 added to BaMnO4 also impairs the
discharge effectiveness of the pure BaMnO4 cathode, but
modestly increases the discharge energy of the composite
BaMnO4/K2FeO4 cathode (Table 3) due to the improvements
of Ba(OH)2 on the pure K2FeO4 electrode (Table 2).
2.8 Ω
Vav
(Wh) (V) (Wh) (V)
75 Ω
E Vav
Fe
salt
wt
%
Mn
salt
wt
%
wt
%
E
salt
K2FeO4
K2FeO4
0 BaMnO4 100
5 BaMnO4 95
0.37 1.16 0.96 1.19
0.44 1.16 1.03 1.26
0.54 1.17 1.14 1.31
0.59 1.17 1.16 1.39
0.65 1.19 1.20 1.40
0.78 1.20 1.20 1.41
0.67 1.20 1.20 1.41
0.66 1.20 1.19 1.44
0.57 1.22 1.12 1.45
0.38 1.21 0.98 1.45
0.38 1.19 0.71 1.43
0.28 1.17 0.68 1.36
K2FeO4 10 BaMnO4 90
K2FeO4 25 BaMnO4 75
K2FeO4 33 BaMnO4 67
K2FeO4 45 BaMnO4 55
K2FeO4 50 BaMnO4 50
K2FeO4 67 BaMnO4 33
K2FeO4 75 BaMnO4 25
K2FeO4 90 BaMnO4 10
K2FeO4 95 BaMnO4
K2FeO4 100 BaMnO4
5
0
K2FeO4 33 BaMnO4 57 Al2O3
K2FeO4 57 BaMnO4 33 Al2O3
K2FeO4 33 BaMnO4 57 KOH
K2FeO4 50 BaMnO4 25 KOH
K2FeO4 57 BaMnO4 33 KOH
10 0.62 1.19 1.10 1.37
10 0.56 1.17 1.13 1.43
10 0.61 1.20 1.17 1.39
25 0.52 1.24 0.96 1.43
10 0.65 1.23 1.00 1.43
K2FeO4 10 BaMnO4 57 Ba(OH)2 33 0.44 1.26 0.68 1.28
K2FeO4 33 BaMnO4 34 Ba(OH)2 33 0.60 1.28 0.92 1.51
K2FeO4 33 BaMnO4 57 Ba(OH)2 10 0.81 1.27 1.17 1.41
K2FeO4 50 BaMnO4 25 Ba(OH)2 25 0.78 1.28 1.15 1.51
K2FeO4 57 BaMnO4 33 Ba(OH)2 10 0.81 1.29 1.21 1.46
K2FeO4 57 BaMnO4 10 Ba(OH)2 33 0.76 1.29 1.27 1.61
K2FeO4 57 KMnO4
10 Ba(OH)2 33 0.63 1.26 1.04 1.16
a The cathode mix also includes graphite and electrolyte.
The discharge of K2FeO4/KMnO4 composite cathodes was
optimized with various additives. Consistently the Ba(OH)2 was
the most effective additive, and the high rate discharge is
summarized in the midsection of Figure 7. In the presence of
both Mn(VII) salts and Fe(VI), competing alkali earth hydroxide
effects are complex. For example, in the reaction with Ba(OH)2,
the KMnO4 reaction product is highly soluble (the product

5954 J. Phys. Chem. B, Vol. 106, No. 23, 2002
Licht et al.
yields a 75 Ω load discharge of 1.0 Wh, compared to 0.7 Wh
for the comparable cell containing one-third the graphite. This
effect is limited in that it also removes cathode active material
from the cell. Recently, we have probed a series of fluorinated
graphite materials that can serve not only as a conductive matrix
but also which have an intrinsic cathodic capacity in alkaline
media. Fluorinated graphite polymers of the form (CFx)n, which
are also described by the weight percent of fluorine contained
in the carbon, are prepared by the reaction of graphite or
carbon materials under a variety of temperature and reaction
conditions and have been widely studied as lithium intercalation
electrodes.17-21 As we have recently noted, and as presented in
Figure 8 in this paper, replacement of the regular graphite
fraction of the cathode mix with a fluorinated graphite can
increase the aqueous alkaline cell capacity.12 As additionally
shown in this figure, further optimization of this cathode mix
may be accomplished by variation of the degree of fluorination,
and addition of hydroxide or permanganate. However, as evident
in the figure, compared to nonfluorinated graphite, fluorinated
graphite’s relatively high resistance leads to a low voltage (and
hence low power) discharge at a high rate, constant 2.8 Ω
discharge.
Figure 8. Cell potential and energy capacity of K2FeO4 cathode
alkaline cells with various relative amounts, by weight, of regular or
fluorinated graphite in the cathode mix, during discharge at the indicated
load of either 2.8 Ω or 75 Ω. Cells use an alkaline AAA configuration
including in the cathode 9 wt % graphite and 18 M KOH electrolyte.
3.7. Constant Power Comparison of MnO2, BaFeO4, and
AgO/K2FeO4 Cathodes. The unusually high specific energy/
specific power of various Fe(VI) alkaline batteries is sum-
marized in the inset of Figure 4. Of relevance to both practical
electronics and as a fundamental energy comparison, a constant
power density, rather than constant load or constant current
density, is a more stringent comparison of cathode capabilities.
In this discharge the lower average cathode potential of the
MnO2 cathode (eq 10) compared to Fe(VI) (eq 7) must be
compensated by a higher average current density, and this
will further impair the MnO2 charge transfer. As previously
observed,1 under conditions of constant, rapid 0.7 W discharge
in an AAA cell configuration, the MnO2 discharges to a
maximum of 0.52 h (0.36 Wh), whereas a 5% KMnO4/95%
BaFeO4 cathode (containing (4.0 g BaFeO4) discharges for 1.26
h to 0.88 Wh.11 Under the same conditions, for the composite
AgO/K2FeO4 cathodes, a 8 wt % (0.3 g) AgO/92 wt % K2FeO4
cell discharges for 1.28 h to 0.90 Wh, a 20 wt % (0.7 g)
AgO/80 wt % K2FeO4 cell discharges for 1.58 h to 1.11 Wh,
and a 39 wt % (1.5 g) AgO/61 wt % K2FeO4 cell discharges
for 2.13 h to 1.49 Wh.
Ba(MnO4)2 has an aqueous solubility of 18 molal). However,
the alternative reaction K2FeO4 with Ba(OH)2 forms BaFeO4,
which is insoluble in water. As seen in the midsection portion
of Figure 3, the addition of Ba(OH)2 to the K2FeO4/KMnO4
cathode results in a significant increase in discharge energy,
and at an average discharge potential greater than that observed
for the K2FeO4/KMnO4 composite without Ba(OH)2. At both
high and low rate, a maximum discharge energy is observed
with the 33:57:10 wt % K2FeO4:KMnO4:Ba(OH)2 composition
which provides 0.73 and 1.62 Wh respectively over either 2.8
Ω or 75 Ω load discharges.
AgMnO4 provides an unusual salt in that the Ag valence acts
in a manner intermediate to Ag(I) and Ag(II), that is as for
Ag(I + x)Mn(VII - x)O4, where 0 < x < 1.13,16 Of the
permanganate and manganate salts explored to date, AgMnO4
promotes one of the larger increases in the K2FeO4 alkaline
cathode discharge, a phenomenon consistent with the observed
Ag activation of Fe(VI), but the AgMnO4 activation phenom-
enon is only substantial in the presence of KOH (added as a
solid salt to the mix).13 This is observed in the lowest section
of Figure 7. In the absence of KOH, the 2.8 Ω discharge of the
K2FeO4 cathode increases from to 0.3 Wh to ∼0.4 Wh with
addition of 18% AgMnO4, but is enhanced to 0.5 Wh using
only 12 wt % AgMnO4 with KOH (6 wt %). This increases
to ∼0.8 Wh with inclusion of 38 wt % AgMnO4 and 12 wt %
KOH. Finally, as also seen in the figure, a K2FeO4, partially
converted to the barium salt with a Ba(OH)2 wash and mixed
with AgMnO4 and KOH, provides a cathode with a high rate
discharge similar to the desired capacity of the BaFeO4 cathode,
exhibiting a higher energy capacity, but lower average discharge
potential. Aspects of the interesting KOH activation of the pure
AgMnO4 alkaline cathode (without K2FeO4) are explored in a
recent study,13 and the AgMnO4 activation of K2FeO4 is still
lower than that observed in Figure 5 for the AgO mediation of
K2FeO4 charge transfer.
4. Conclusion
An activated cathodic reduction is demonstrated for super-
iron cathodes, to increases the modest but expanding foundation
of understanding of charge transfer of these unusual Fe(VI)
charge storage salts. A model presented for co-cathode chemical
and electronic mediation of Fe(VI) suggests charge transfer
enhancement by Ag(II). This enhancement is experimentally
observed in the form of AgO/K2FeO4 composite cathodes and
provides a step toward an environmentally benign alkaline
cathode with power and storage characteristics superior to the
widely used conventional MnO2 cathode. In a Zn anode alkaline
cell, an AgO/K2FeO4 composite cathode provides a high rate
discharge with 3- to 5-fold higher high power energy capacity
(for example 1.5 Wh, compared to 0.36 Wh, at 0.7 W in a AAA
cell configuration) than the conventional Zn/MnO2 alkaline.
The chemical mediation of Fe(VI)/Mn(VII) or Fe(VI)/Mn(VI)
can also improve charge of efficiency of both the composite
K2FeO4/KMnO4 or K2FeO4/BaMnO4 cathodic discharge through
inclusion of both salts in the cathode mix. The effect of oxide,
hydroxides, titanates, fluorinated graphites, and AgMnO4 on the
cathodic process is also explored. In the absence of Ag(II)
The cathode reduction is supported by a conductive matrix
provided through inclusion of graphite in the cathode mix. The
K2FeO4 cathode can be enhanced by replacing a fraction of this
active cathode material with a larger fraction of graphite to
increase conductivity and thereby improve utilization of the
remaining intrinsic Fe(VI f III) charge utilization. Hence in
Figure 8, a 75 wt % K2FeO4 and 25 wt % graphite cathode mix

Silver Mediation of Fe(VI) Charge Transfer
J. Phys. Chem. B, Vol. 106, No. 23, 2002 5955
(6) Licht, S.; Wang, B.; Ghosh, S.; Li, J.; Naschitz, V. Electrochem.
Commun. 1999, 1, 527.
(7) Licht, S.; Wang, B.; Ghosh, S.; Tel-Vered, R.; Naschitz, V.
Electrochem. Commun. 2000, 2, 535.
(8) Licht, S.; Naschitz, V.; Ghosh, S.; Lin, L.; Lui, B. Electrochem.
Commun. 2001, 3, 340.
(9) Licht, S.; Wang, B. Electrochem. Solid State Lett. 2000, 3, 209.
(10) Licht, S.; Naschitz, V.; Liu, B.; Ghosh, S.; Halperin, N.; Halperin,
L.; Rozen, D. J. Power Sources 2001, 99, 7.
(11) Licht, S.; Naschitz, V.; Lin. L.; Chen, J.; Ghosh, S.; Liu, B. J.
Power Sources 2001, 101/2, 167.
(12) Licht, S.; Ghosh, S.; Dong, Q. J. Electrochem. Soc. 2001, 148,
A1072.
(13) Licht, S.; Naschitz, V.; Liu, B.; Ghosh, S. Electrochem Solid. State
Lett. 2001, 4, A209.
(14) Licht, S.; Ghosh, S.; Naschitz, V.; Halperin, N.; Halperin, L. J.
Phys. Chem. B 2001, 105, 11933.
(15) U.S. Federal Register (January 3, 1997) Vol. B, No. 2, p 367.
(16) Mehne, L. F.; Wayland, B. B. J. Inorg. Nucl. Chem. 1975, 37, 1371.
(17) Nakajima, T.; Koh, M.; Gupta, V.; Zemva, B.; Lutar, K. Electro-
chim. Acta 2000, 45, 1655.
(18) Yazami, R.; Hany, P.; Masset, P.; Hamwi, A. Mol. Cryst. Liq. Cryst.
Sci. Technol. 1998, 310, 397.
(19) Hamwi, A.; Al Saleh, I. J. Power Sources 1994, 48, 311.
(20) Bubyreva, N. S.; Mararenko, B.; Yu, K.; Pavlo, P.; Perederi, I.
M.; Rozenblyum, N. D.; Tomin, O. A Elektrokhimiya 1982, 18, 1105.
(21) Watanabe, N. Solid State Ionics 1980, 1, 87.
(22) Licht, S.; Naschitz, V.; Wang, B. J. Power Sources 2002, in press.
(23) Licht, S.; Ghosh, S. J. Power Sources 2002, in press.
activation, alkaline K2FeO4 AAA cells containing specific
activators can improve the discharge energy and discharge
behavior, approaching that previously observed for the BaFeO4
cathode (0.88 Wh at high rate AAA discharge). Addition of
33% KMnO4 and 10% Ba(OH)2 to the pure K2FeO4 cathode
mix increases the 2.8 Ω discharge 2-fold to 0.54 Wh. Alterna-
tively, addition of 33% BaMnO4 increases this to 0.58 Wh and
to 0.81 Wh with 33% BaMnO4 and 10% Ba(OH)2. Compared
to the pure K2FeO4 cathode, addition of 34% AgMnO4 and 11%
KOH increases the discharge energy to 0.78 Wh, which is
further increased to 0.97 Wh when the K2FeO4 is pretreated
with a Ba(OH)2 wash.
Acknowledgment. We thank B. Wang, D. Rozen, and N.
Halperin and M. Zuidman who participated in the analyses.
References and Notes
(1) Licht, S.; Wang, B.; Ghosh, S. Science 1999, 285, 1039.
(2) Johnson, M. D.; Read, J. F. Inorg. Chem. 1996, 35, 6795.
(3) Sharma, V. K.; Smith, J. O.; Millero, F. J. EnViron. Sci. Technol.
1997, 31, 2486.
(4) Rryter, M.; Tyrell, R. M. Free Radical Biol. Med. 2000, 28, 289.
(5) Licht, S.; Wang, B.; Ghosh, S.; Li, J.; Naschitz, V. Electrochem.
Commun. 1999, 1, 522.