The super-iron boride battery

Article abstract of DOI:10.1149/1.2839554

A high-capacity alkaline redox storage chemistry is explored based on an environmentally benign zirconia-stabilized Fe6+ B2- chemistry. This super-iron boride battery sustains an electrochemical potential matched to the pervasive, conventional MnO2 -Zn battery chemistry, but with a much higher electrochemical storage capacity. Whereas a conventional alkaline battery pairs the 2 e- zinc anode with a 1 e- MnO2 cathode, the new alkaline cell couples an 11 e- boride anode, such as VB2, with a 3 e- storage hexavalent iron cathode. The cell has an open circuit and discharge potential comparable to the conventional, commercial alkaline battery. Based on VB2 (72.6 g mol-1) and the Fe(VI) salt K2 FeO4 (198.0 g mol-1), the super-iron boride cell has an 11 Faraday theoretical capacity of 369 mAh g-1. Added AgO mediates and further facilitates the 3 e- K2 FeO4 reductive charge transfer, and we demonstrate for super-iron boride that over 300 mAh g-1 is approached experimentally, which is substantially higher than the conventional Zn MnO2 alkaline battery with an experimental capacity (to 0.8 V) of 160 mAh g-1 and a theoretical capacity of 224 mAh g-1.

Full text of DOI:10.1149/1.2839554

Journal of The Electrochemical Society, 155 ͑4͒ A297-A303 ͑2008͒
A297
0013-4651/2008/155͑4͒/A297/7/$23.00 © The Electrochemical Society
The Super-Iron Boride Battery
,z
*
*
Stuart Licht, Xingwen Yu, Yufei Wang, and Huiming Wu
Department of Chemistry, University of Massachusetts, Boston, Massachusetts 02125, USA
A high-capacity alkaline redox storage chemistry is explored based on an environmentally benign zirconia-stabilized Fe6+/B2−
chemistry. This super-iron boride battery sustains an electrochemical potential matched to the pervasive, conventional MnO2–Zn
battery chemistry, but with a much higher electrochemical storage capacity. Whereas a conventional alkaline battery pairs the 2e−
zinc anode with a 1e− MnO2 cathode, the new alkaline cell couples an 11e− boride anode, such as VB2, with a 3e− storage
hexavalent iron cathode. The cell has an open circuit and discharge potential comparable to the conventional, commercial alkaline
battery. Based on VB2 ͑72.6 g mol−1͒ and the Fe͑VI͒ salt K2FeO4 ͑198.0 g mol−1͒, the super-iron boride cell has an 11 Faraday
theoretical capacity of 369 mAh g−1. Added AgO mediates and further facilitates the 3e− K2FeO4 reductive charge transfer, and we
demonstrate for super-iron boride that over 300 mAh g−1 is approached experimentally, which is substantially higher than the
conventional Zn/MnO2 alkaline battery with an experimental capacity ͑to 0.8 V͒ of 160 mAh g−1 and a theoretical capacity of
224 mAh g−1
.
© 2008 The Electrochemical Society. ͓DOI: 10.1149/1.2839554͔ All rights reserved.
Manuscript submitted November 9, 2007; revised manuscript received December 26, 2007.
Available electronically February 20, 2008.
VB2 + 11/3FeO24− + 5/6H2O → 11/6Fe2O3 + VO43− + 1/3BO33−
Aqueous Zn–MnO2 redox charge storage chemistry has been es-
tablished for over a century and continues as the dominant mecha-
nism of primary electrochemical storage. Capacity, power, cost, and
safety factors have led to annual global use on the order of 1010
primary Zn–MnO2 batteries.1 This chemistry is increasingly chal-
lenged in meeting the growing energy and power demands of con-
temporary optical, electromechanical, electronic, and medical con-
sumer devices. Hence, the search for new energy storage chemistry
systems, with higher capacity and energy density, has been increas-
ingly emphasized.1-7
+ 5/3HBO23−
Here we provide an expanded study of these energetic, super-iron
cathode/boride anode alkaline redox couples.
Experimental
K2FeO4 of 97–98.5% was prepared according to
Fe͑NO3͒3 · 9H2O + 3/2KClO + 5KOH
In 1999, we introduced a cathodic chemistry,1 which is based on
a class of Fe͑VI͒ or “super-iron” oxides with a 3e− intrinsic capacity
͑e.g., K2FeO4: 406 mAh g−1͒ higher than MnO2 ͑308 mAh g−1͒. In
2004 it was reported that metal borides could be used as anodic
alkaline charge storage materials.8,9 However, two obstacles to
implementation of this alkaline boride anodic chemistry were evi-
dent, in which boride undergoes a 5e− oxidation from the −2 to +3
valence state. There is a significant domain in which the boride
anode materials corroded spontaneously, generating hydrogen gas,
and the electrochemical potential of the boride anodes was less
negative than that of zinc. Therefore, a boride MnO2 cell was sub-
ject to decomposition, and the voltage of this cell was low compared
to the electrochemical potential of the pervasive Zn–MnO2 redox
chemistry. In a recent communication, we introduced a battery
chemistry based on Fe͑VI͒–boride redox charge transfer, overcom-
ing both obstacles to utilization of the boride alkaline anode.7 The
super-iron cathode provides the requisite additional electrochemical
potential, and our recently demonstrated zirconia hydroxide-shuttle
overlayer, which stabilizes alkaline charge-transfer chemistry,6 is
also shown to prevent corrosion of the boride anode. Sustaining
→ K2FeO4 + 3/2KCl + 3KNO3 + 23/2H2O
͓1͔
Preparation and analysis of K2FeO4 has been detailed elsewhere.10,11
The dried K2FeO4 product is highly stable over a period of
years.10,11 In this study, the K2FeO4 was used as synthesized or,
when indicated, after size separation using polypropylene grids
͑MacMaster–Carr͒ of various mesh sizes, as salts with various con-
strained particle sizes, in the size range Ͻ28, Ͻ35, 35–73, or
73–149 ␮m.
AgO was prepared by the reaction of an alkaline AgNO3 solution
with K2S2O8 at 85°C, in accordance with standard methods12
4AgNO3 + 2K2S2O8 + 8NaOH → 4AgO + K2SO4 + 3Na2SO4
+ 2NaNO3 + 2KNO3 + 4H2O
͓2͔
The boride anode was generally either TiB2 ͑Aldrich 10 ␮m
powder͒ or VB2 ͑Aldrich 10 ␮m/325 mesh powder͒ ͑coated or un-
coated͒ as indicated and mixed with the graphite. In addition, TaB
͑Aldrich͒, TaB2 ͑Aldrich͒, MgB2 ͑Aldrich͒, CrB2 ͑Aldrich͒, CoB2
͑Aldrich͒, Ni2B ͑Acros͒, and LaB6 ͑Acros͒ were also considered.
We have established that a zirconia overlayer stabilizes the alka-
line anode and cathode materials.6,7 A 1% ZrO2 coating on a powder
of either K2FeO4, TiB2, or VB2 was prepared as follows: 8 mg
ZrCl4 ͓analytical reagent grade, ACROS͔ dissolved in 8 mL ether
͑Fisher͒ and stirred with 0.8 g of the solid powder in air for 30 min,
followed by vertex suction, then vacuum removal of the remaining
organic solvent and drying overnight. Analysis of the coating was
performed with attenuated total reflectance Fourier transform infra-
red ͑ATR/FTIR͒ spectroscopy using a Nicolet 4700 FTIR with the
accessory Smart Orbit diamond crystal and using the Smart Orbit
tip. The powder samples ͑coated material͒ were compressed into
thin pellets, and these pellets were cycled in contact with the dia-
mond crystal by the tip and removed three times to ensure the
sample pellet was firmly contacted with the diamond crystal.
Performance, and electrochemical enhancement, of the super-
iron boride battery was studied through preparation of button cells
containing electrodes with various cathode and anode compositions.
much higher electrochemical capacity than
a conventional
MnO2–Zn battery, this Fe͑VI͒–boride alkaline battery chemistry has
individual electrode and overall discharge reactions, such as
anode:
VB2 + 20OH− → VO43− + 2BO33− + 10H2O + 11e−
cathode:
FeO24− + 5/2H2O + 3e− → 1/2Fe2O3 + 5OH−
cell:
*
Electrochemical Society Active Member.
z E-mail: stuart.licht@umb.edu
Downloaded on 2014-11-21 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A298
Journal of The Electrochemical Society, 155 ͑4͒ A297-A303 ͑2008͒
TaB + 6OH− → Ta + BO33− + 3H2O + 3e−
͓4͔
In accordance with Eq. 3 and 4, the intrinsic n = 6e− capacity of
TaB2 and the n = 3e− capacity of TaB anodes are, respectively,
793.8 and 419.3 mAh g−1, as determined from a net intrinsic capac-
ity of nF/W ͑F ϭ the Faraday constant͒. The experimental anodic
charge storage constraints of the TaB and TaB2 anodes are shown in
Fig. 1. The anodic capacity is measured by discharge in electro-
chemical coin cells configured to contain excess ͑K2FeO4͒ cathode
capacity. In a balanced manner, later cases in this study probing
cathode constraints are measured in cells instead containing excess
anode capacity. Unless stated otherwise, the coin cells are generally
discharged at low current ͑constant, high load resistance͒ to maxi-
mize coulombic efficiency and to probe the discharge potential at
low overpotential, although significant higher rate discharges are
also included later in this study. As seen in Fig. 1, the TaB2 exhibits
an anodic storage capacity comparable to the widely used conven-
tional alkaline zinc anode ͑820 mAh g−1͒. Compared to the theoret-
ical anodic capacity determined in accordance with Eq. 3, approxi-
mately 85% coulombic efficiency was obtained for this TaB2 anode,
reflecting experimental access to a high degree of the 6e− intrinsic
capacities. Compared to the theoretical capacity determined from
Eq. 4, 93% coulombic efficiency was obtained for the TaB anode.
Although the 6e− TaB2 provides higher anodic capacity than the
3e− TaB, the 6e− titanium boride anode provides even higher acces-
sible capacity.7 The analogous anodic discharge of the boride TiB2,
rather than the tantalum diboride-containing alkaline cell, provides a
significantly higher gravimetric charge storage capacity to the metal
product due to the lighter weight of this metal ͑47.87 g mol−1 tita-
nium compared to 180.95 g mol−1 tantalum͒, and cell products are
generalized as Ti and BO33−
Figure 1. Tantalum boride anodes. Super-iron/boride cells which contain
either a TaB or a TaB2 anode are discharged. Anode: 75 wt % TaB2 ͑or
TaB͒, 20% graphite, 4.5% KOH, and 0.5% binder ͑T-30, 30% teflon͒, com-
bined anode mass: 90 mg. Anode limited conditions are studied by packing
each cell, respectively, with excess intrinsic cathode capacity. Cathode: 75%
K2FeO4 and 25% graphite. Electrolyte: saturated KOH. The 1 cm button
cells are anode capacity limited ͑cells are configured with a cathode capacity
to probe the anode’s charge storage limit͒. The cells are discharged at a
constant load of 3 k⍀ at 22°C.
Cathodes were composed of K2FeO4 ͑uncoated or zirconia coated͒,
either with or without indicated additives, and mixed with graphite
as a conductor ͑1 ␮m graphite, Leico Industries Inc.͒. For compari-
son, cells were also alternatively prepared with cathodes containing
MnO2 ͑electrolytic manganese dioxide, EraChem K60͒, NiOOH
͓from a commercial Powerstream Ni–metal hydride ͑MH͒ button
cell͔, or KIO4 ͑ACROS͒. Anodes were composed of the indicated
borides ͑uncoated or zirconia coated͒, either with or without indi-
cated additives, and mixed with the graphite as a conductor ͑1 ␮m
graphite, Leico Industries Inc.͒. For comparison, cells were also al-
ternatively prepared with anodes containing metal hydride ͑from a
commercial Powerstream Ni–MH button cell͒ or Zn gel ͑from a
commercial alkaline Energizer cell͒. Saturated KOH was used as the
electrolyte. Cells were discharged over constant, described load re-
sistances. The potential variation over time was recorded via Lab
View Acquisition on a PC. The cumulative discharge was deter-
mined by subsequent integration.
TiB2 + 12OH− → Ti + 2BO33− + 6H2O + 6e−
͓5͔
The small third acid dissociation constant of boric acid ͑H3BO3
with pK1,2,3 = 9.1, 12.7, and 13.8͒ drives spontaneous free hydrox-
ide formation from borate and water in the pH 14 domain. Hence,
Eq. 5 is in simultaneous equilibrium with cell products which also
include HBO23−. In this case, the equilibrium between BO33− and
HBO23− acts as a buffer against hydroxide depletion during anodic
discharge
TiB2 + 10OH− → Ti + 4H2O + 2HBO23− + 6e−; E° = 1.74 V
͓6͔
Results and Discussion
BO33− + H2O ⇔ OH− + HBO32−
;
K = Kw/K3 = 10−0.2 ͓7͔
FeB and CoB were previously considered as alkaline boride an-
odes; however, each was observed to exhibit an anodic capacity less
than that of TiB2 and VB2.9 In this study, we have also considered
the borides TaB, TaB2, MgB2, CrB2, CoB2, Ni2B, and LaB6 as
alternative active anode materials for alkaline electrochemical stor-
age cells. While the two tantalum boride salts exhibit a degree of
accessible anodic charge storage capacity, each of the other borides
did not exhibit significant primary discharge behavior due to their
high solubilities in alkaline solution or their reactivity with alkaline
electrolyte.
The electrochemical discharge characteristics of TaB and TaB2,
as candidate alkaline boride anodes, are summarized in Fig. 1. The
alkaline reduction of the TiB2 anode had been previously shown to
be consistent with formation of zero valent titanium,8,12 and simi-
larly, we expect the reduction of TaB2 can yield tantalum. According
to the half-cell oxidation reaction of TaB2 ͑formula weight W
= 202.57 g mol−1͒ or TaB ͑W = 191.76 g mol−1͒ to a Ta product,
the theoretical ͑intrinsic͒ anodic reactions of the salts in an alkaline
medium are
The products vary with hydroxide concentration, degree of hydra-
tion, and depth of discharge, including the anhydride product B2O3
and cation containing species, such as in aqueous KOH:
KxHzBO3−x−z ͑where x ranges from 0 to 3 and z from 0 to 3 − x͒, as
well as p3olymeric species, such as those related to the boric conden-
sation reaction forming borax species KyB4O2−y
.
In accordance with Eq. 5 or 6 and a7 formula weight, W
= 69.5 g mol−1, TiB2 has a net intrinsic 6e− anodic capacity of
2314 mAh g−1. A significant advantage of the titanium boride anode
is that it contains higher capacity compared with the conventional
alkaline zinc anode ͑820 mAh g−1͒. As seen in the top of Fig. 2, the
observed TiB2 anode capacity is in excess of 2000 mAh g−1 at low
discharge rates ͑over a 100 k⍀ load discharge in the 1 cm diameter
cell͒ and still in excess of 1300 mAh/g at moderate discharge rates
͑over a 3 k⍀ load͒.
VB2 has the highest anode capacity studied. Unlike TiB2, the
alkaline VB2 undergoes oxidation of the tetravalent transition metal
ion with V4+ ⇒ V5+, as well as oxidation of each of the two borides
with B2− ⇒ B3+, for a net 11e− anodic process
TaB2 + 12OH− → Ta + 2BO33− + 6H2O + 6e−
͓3͔
Downloaded on 2014-11-21 to IP 134.129.182.74 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, 155 ͑4͒ A297-A303 ͑2008͒
A299
Figure 3. High-capacity boride or conventional alkaline anodes. Electrolyte
and cell conditions are as described in Fig. 1, discharged under the indicated
constant ohmic load conditions. Comparison of TiB2 and Zn or MH alkaline
anodes. Cells contain excess cathode capacity and are limited to 10 mAh
anode capacity. Cells discharged include MnO2–Zn and NiOOH–MH con-
ventional alkaline cells, as well as the Zn or MH anodes, instead of with a
Fe͑VI͒ cathode ͑containing 75% K2FeO4, 20% graphite, and 5% KOH͒.
Other cells contain the same cathodes but a TiB2 anode instead of Zn or MH
anodes. TiB2 anode: 75% TiB2, 20% graphite, 4.5% KOH, and 0.5% binder
͑T-30, 30% Teflon͒.
Figure 2. Discharge of high-capacity TiB2 and VB2 boride alkaline anode
cells with K2FeO4 cathodes. Electrolyte and cell conditions are as described
in Fig. 1, discharged under the indicated constant ohmic load conditions.
Cells contain excess cathode capacity and are limited to either 5 mAh
͑100 k⍀ load͒ or 10 mAh ͑Ͻ100 k⍀͒ intrinsic anode capacity to insure
discharge in a short time frame ͑3 weeks or less͒. ͑Top͒ TiB2 anode: 75%
TiB2, 20% graphite, 4.5% KOH, and 0.5% binder ͑T-30, 30% Teflon͒. Cath-
ode: 75% K2FeO4, 20% graphite, and 5% KOH. ͑Bottom͒ Same cathode;
anode: 75 wt % VB2, 20% graphite, 4.5% KOH, and 0.5% binder ͑T-30,
30% Teflon͒.
MnO2/Zn cells are both seen to generate an initial discharge voltage
of 1.2–1.5 V. The metal hydride cell has an anode capacity of
ϳ300 mAh g−1, while the MnO2/Zn alkaline battery approaches
800 mAh g−1. It is evident in the figure that a new, alternative
MnO2/boride cell can sustain a substantially larger anode capacity.
The potential of this alkaline MnO2/boride cell is approximately
0.25 V lower than that of the conventional MnO2/Zn or metal hy-
dride cells. However, the alkaline thermodynamic potential of the
super-iron cathode, which undergoes a 3e− reduction Fe͑VI ⇒ III͒,
can compensate for this loss, as it occurs at a value of 0.25 V higher
than the 1e− reduction of MnO2 ͑SHE = the standard H2 electrode
potential͒:
VB2 + 20OH− → VO34− + 2BO33− + 10H2O + 11e−; E° = 1.67 V
͓8͔
Therefore, in accordance with Eq. 8, VB2 has an unusually high,
intrinsic 11e− anodic capacity of 11 F/͑W = 72.6 g mol−1
͒
= 4060 mAh/g, rivaling the high anodic capacity of lithium
͑3860 mAh/g͒. The vanadium boride anode VB2 has five times the
alkaline capacity of zinc. The cell products are generalized in Eq. 8
as VO34− and BO33− but, as with the titanium boride discharge, vary
with the hydroxide concentration, hydration, and depth of discharge.
The speciation of borate and the anhydride salt of boric acid has
been previously described ͑H3BO3, H2BO−3, HBO32−, BO33−, B2O3−,
KxHzBO3−x−z,KyB4O2−y, and B2O3͒. In addition, a distribution of
vanadate3 species ari7ses. Vanadate is the deprotonated form of a
triprotonic acid, vanadic acid ͑H3VO4 with pK1,2,3 = 3.8, 7.8, and
13.0͒. Equilibrium-driven speciation of VO34−, HVO24−, and H2VO4−,
as well as the anhydride product, V2O5, and potassium cation spe-
cies occur.
FeO24− + 5/2H2O + 3e− → 1/2Fe2O3 + 5OH−;
E = 0.60 V vs SHE
͓9͔
MnO2 + 1/2H2O + e− → 1/2Mn2O3 + OH−; E = 0.35 V vs SHE
͓10͔
The additional potential of the Fe͑VI͒ cathode in Eq. 9 compared to
Eq. 10 is sufficient to compensate for the low voltage of the
MnO2/boride cell observed in Fig. 3. As seen, the super-iron/boride
couple, respectively, with either a TiB2 ͑Fig. 2 top and Fig. 3͒ or
VB2 ͑Fig. 2 bottom͒ anode, generates a discharge potential similar
to the conventional alkaline MnO2/Zn cell.
A standard oxidation potential for the TiB2 and VB2 reactions
has been included in Eq. 6 and 8, which have not been previously
available and which we have calculated based on literature thermo-
dynamic values.13,14 The 0.07 V differential between these values is
consistent with the slightly lower potentials we measure in VB2
compared to TiB2 cells, although the respective calculated standard
reduction potentials for these cells of −1.67 and −1.74 V vs SHE is
considerably lower than the value of approximately −1.0 V vs SHE
that we would estimate based on our measured value of the open-
circuit potential of these half reactions. This difference is equivalent
to a large overpotential, and we observe that the oxidation reaction
is highly irreversible. The smooth, steady discharge voltage through-
out discharge of VB2 is fascinating; this may be the only example of
a molecule that appears to undergo an observed, full 11e− electro-
A vanadium, compared with a titanium, diboride salt yields the
highest alkaline anodic capacity. As seen in the lower portion of Fig.
2, the VB2 anode approaches this high capacity even at moderate
discharge rates ͑over 3 or 10 k⍀ load͒. In addition to the higher
anodic capacity, it is seen that in comparison with the upper portion
of the figure, the VB2 anode can sustain a higher coulombic effi-
ciency ͑a higher fraction of the theoretical charge capacity͒ at mod-
erate current density than the TiB2 anode. Normalized by the intrin-
sic capacities in accordance with Eq. 6, the TiB2 anode sustains 56%
of the intrinsic capacity over the 3 k⍀ discharge load and ap-
proaches 70% at a 10 k⍀ discharge, whereas the VB2 sustains 91%
of the intrinsic capacity at 3 k⍀, which increases to 94% of the
intrinsic 11e− capacity during a 10 k⍀ discharge.
Figure 3 compares the discharge of alkaline electrolyte cells con-
taining various cathode and anode couples, including MnO2/Zn and
metal hydride cells. One societal driving force for the continued use
of MnO2/Zn chemistry is the several generations of devices which
have been designed for the normal operative domain of the
MnO2/Zn battery. The conventional metal hydride and alkaline
Downloaded on 2014-11-21 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A300
Journal of The Electrochemical Society, 155 ͑4͒ A297-A303 ͑2008͒
chemical oxidation at a singular potential. Further studies are needed
to understand the large difference between the thermodynamic and
experimental potentials, but we suggest it is related to an averaging
effect of the highly multielectron character of the oxidation reaction.
A variety of alkali and alkaline earth Fe͑VI͒ ͑super-iron͒ salts,
which sustain a 3e− reduction in alkaline media at potentials consis-
tent with Eq. 9, have been considered as alkaline cathodes.15-17 Of
these salts, pure BaFeO4 sustains the highest rates of cathodic re-
duction but is susceptible to decomposition in the solid state; fur-
thermore, the barium salts are problematic from a toxicological
perspective.17,18 The potassium Fe͑VI͒ salt, K2FeO4, provides sev-
eral advantages. It is stable in the solid state, and its discharge prod-
uct is environmentally benign. Furthermore, it has a higher cathodic
capacity ͑406 compared to 313 mAh g−1͒ than BaFeO4, which is
significantly higher than the conventional alkaline cathode MnO2
͑308 mAh g−1͒. Pure K2FeO4 does not sustain the high rates of
cathodic reduction observed with BaFeO4 cathodes. However,
K2FeO4 cathodes with hydroxide or manganese salts as possible
additives, and in particular with silver oxide additives, can substan-
tially enhance the rate of Fe͑VI͒ reduction and facilitate and sustain
high cathodic current densities.6,18,19 An AgO additive strongly fa-
cilitates Fe͑VI͒ cathodic charge transfer. The pure ͑AgO-free͒
K2FeO4 cathode in Fig. 1 required 25 wt % graphite to ensure con-
ductivity and effective cathode charge transport. The K2FeO4 cath-
ode with 5 wt % added KOH to facilitate charge transfer in Fig. 2
required only 20 wt % graphite to ensure effective cathode charge
transport. The addition of AgO substantially facilitates Fe͑VI͒ ca-
thodic charge transfer and necessitates considerably less graphite in
the cathode mix. With only 8.5 wt % of AgO added to the K2FeO4
cathode, only 10 wt % graphite is required as the conductive media
and still yields a greater discharge capacity ͑over 80% of the intrin-
sic cathode capacity͒ than a pure K2FeO4 cathode containing 25%
graphite ͑which yields only 70% coulombic efficiency͒. A pure
K2FeO4 cathode containing only 10% graphite yields only 50% cou-
lombic efficiency under the same discharge conditions. Also, at high
discharge rates, these low levels of AgO retain this higher relative
Figure 4. High current density discharge of composite K2FeO4 with AgO
cathode–TiB2 anode cells. The cathodes are discharged under a constant
100 ⍀ load in 1 cm coin cells with excess intrinsic anode capacity. As indi-
cated in the figure, the fraction of the AgO cathode additive ranges from 0 to
59.5 wt %. The cathode is prepared with a total of 29 mg of the K2FeO4
+ AgO and also contains 5% KOH and 10% graphite. Electrolyte: saturated
KOH. TiB2 anode: 75% TiB2, 20% graphite, 4.5% KOH, and 0.5% binder
͑T-30, 30% Teflon͒. The measured cathode capacity to 0.6 V discharge for
the 8.5, 17, 34, 42, 51, or 59.5% AgO cells is 115, 212, 253, 298, 321, 343,
or 355 mAh/g.
1/2Ag2O + 1/2H2O + e− → Ag + OH−; E = 0.35 V vs SHE
͓13͔
As seen on the right sides of Fig. 6 and 7, the Fe͑VI͒ cathode
discharges to higher capacity compared to MnO2, NiOOH, and
KIO4 alkaline cathode under equivalent cell conditions. Both figures
present results utilizing the composite super-iron cathode ͑76.5%
K2FeO4, 8.5% AgO, 5% KOH, and 10% g͒. As seen on the left side
of the figures, this Fe͑VI͒ cathode also provides a higher potential
match for the boride anode than other cathodes and results in a
discharge voltage domain which is similar to that of conventional
alkaline and metal hydride batteries. The anodic capacity of the
fraction of the intrinsic capacity compared to AgO-free cells.18
A
composite cathode containing a 9:1 ratio of K2FeO4 to AgO has an
intrinsic cathodic capacity of 408 mAh g−1, as determined from
K2FeO4’s 406 mAh g−1, and the 432 mAh g−1 capacity of pure
AgO, determined from its 2e− alkaline cathodic reduction of
AgO + H2O + 2e− → Ag + 2OH−
͓11͔
In Fig. 4, it is seen that increasing an AgO additive in a compos-
ite K2FeO4 cathode also substantially improves its charge transfer at
high discharge rates. This is measured in cathode-limited cells
through discharge at a constant load of 100 ⍀, rather than at 1000 to
100,000 ⍀ loads as previously examined in the cells in Fig. 2. At
this high rate of discharge, the composite cathode yields over four
times the cathode capacity with a large fraction of added AgO as the
capacity realized without the additive.
In Fig. 5 it is seen that a relatively low level of added AgO
͑8.5 wt %͒ sustains a high coulombic efficiency over a wide range
of current densities. The cathode contains a fixed composition of
76.5% K2FeO4, 8.5% AgO, and 5% KOH. The beneficial character-
istics of this cathode can be effectively coupled with the high-
capacity TiB2 anode. The accessible depth of discharge of this com-
posite cathode is over 80% and drops to 45% of the intrinsic
capacity in the most rapidly discharged cell ͑at the 100 ⍀ load͒. The
elevated voltage plateau evident in the initial Ͻ10% of the dis-
charge, particularly at lower rates ͑higher discharge load conditions͒,
occurs only in the AgO composite Fe͑VI͒ cathodes. This plateau
reflects the higher alkaline redox potential of the Ag͑II͒ to Ag͑I͒,
compared to the Ag͑I͒ to Ag͑0͒ reduction
Figure 5. Discharge-voltage profiles, at a range of discharge loads, of fixed-
composition composite K2FeO4 ͑zirconia-coated͒/AgO–TiB2 button cells.
Excess capacity TiB2 anode is prepared as 75% TiB2, 20% graphite, 4.5%
KOH, and 0.5% bender ͑T-30, 30% Teflon͒. Composite cathode is prepared
with 29.5 mg 76.5% K2FeO4, 8.5% AgO, 5% KOH, and 10% graphite. The
cells are discharged at various constant loads at 22°C. The measured cou-
lombic efficiency to 0.6 V discharge for the cell discharge at these loads of
100, 500, 1000, 3000, or 10,000 ⍀ is 52, 67, 80, 83, or 87%.
AgO + 1/2H2O + e− → 1/2Ag2O + OH−; E = 0.65 V vs SHE
͓12͔
Downloaded on 2014-11-21 to IP 134.129.182.74 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, 155 ͑4͒ A297-A303 ͑2008͒
A301
Figure 6. Titanium boride anodes dis-
charged coupled with a variety of alkaline
cathodes under ͑left͒ anode-limited and
͑right͒ cathode-limited conditions. Anode:
75 wt % TiB2, 20% g, 4.5% KOH, and
0.5% binder ͑T-30, 30% Teflon͒,
g
= 1 ␮m graphite. The cathode is either
K2FeO4, MnO2, NiOOH, or KIO4 or, more
specifically, 76.5% K2FeO4, 8.5% AgO,
5% KOH and 10% g; 90% MnO2 and 10%
g; NiOOH; or 75% KIO4 and 25% g.
Anode- and cathode-limited conditions are
studied by packing each cell, respectively,
with excess intrinsic cathode or anode ca-
pacity. Electrolyte and cell conuration are
as described in Fig. 1; cells are each dis-
charged at a constant 3 k⍀ load.
titanium boride anode in Fig. 6 is substantially higher than that of
zinc. The discharge reaction for the super-iron/titanium boride full
cell is derived through Eq. 6 and 9
As seen in Fig. 8, similar to the coated cathode materials,6 ATR/
FTIR analysis reveals the zirconia 1396 and 1548 cm−1 peaks on a
variety of coated materials. This includes the coated active anode
materials TiB2 and VB2, as well as coated cathode materials, such as
K2FeO4 and AgO. Pure ZrO2 is prepared ͑as a colloid͒ for compari-
son, and the two absorption peaks coincide with the absorption spec-
tra of zirconia.6,22 The conversion of ZrCl4 to zirconia forms a pure
ZrO2/Zr͑OH͒4, depending on the extent of hydration
TiB2 + 2FeO24− + H2O → Ti + Fe2O3 + 2HBO23−
͓14͔
The observed VB2 anode capacity of ϳ4000 mAh g−1 represents
an increase even higher than that of TiB2. The electrochemical en-
ergy storage of super-iron boride including the 11e− reduction per
molecule of VB2, is diagrammatically represented in Scheme 1. As
seen on the left side of Fig. 7 for an VB2, rather than a TiB2 anode,
the Fe͑VI͒ again provides a good potential match with conventional
alkaline systems, and the anodic capacity of the vanadium boride
anode is now severalfold higher than that of zinc ͑800 mA g−1͒. The
overall discharge reaction for the super-iron/vandium boride full cell
is derived from Eq. 8 and 9
ZrCl4 + O2 → ZrO2 + 2Cl2; ZrO2 + 2H2O ⇔ Zr͑OH͒
4
͓16͔
The discharge of the complete super-iron titanium boride redox
chemistry is investigated in Fig. 9 using cells with balanced TiB2
anode and Fe͑VI͒ cathode capacity ͑based on the intrinsic capacity
of the anode and cathode components͒. The capacity is determined
from the integrated charge, normalized by the sum of the total active
anode ͑TiB2͒ and cathode ͑K2FeO4 and AgO͒ components. Alkaline
redox couples are prepared in an electrochemical coin cell configu-
ration and are studied with or without a zirconia coating in this
figure. The reaction products depend on the depth of discharge, pH,
and the degree of dehydration of the boric and ferric products, and
the cell may be generalized in the representative deep discharge
reaction. As determined from Eq. 14, the theoretical capacities for
the complete super-iron/TiB2 ͑6F per TiB2 + 2K2FeO4͒ cell is
VB2 + 11/3FeO24− + 5/6H2O → 11/6Fe2O3 + VO43− + 1/3BO33−
+ 5/3HBO23−
͓15͔
Boride salts can be unstable in alkaline solution. TiB2 visibly
reacts on contact with KOH electrolyte ͑evolving H2 gas͒. This is
not only a chemical loss of the electrochemical capacity but is flam-
mable. In addition, due to the evolved gas, a sealed battery swells or
even cracks during storage. Unlike uncoated titanium boride, we
recently demonstrated that a 1% zirconia-coated TiB2 does not
evolve hydrogen.7 A 1% zirconia-coated TiB2 is equivalent to a
0.1 ␮m ZrO2 overlayer on a 30 ␮m diameter TiB2 particle. Zirconia
is insoluble and extremely stable in aqueous alkaline media.20,21
Based on our previous experience ͑Mn coating10͒, we recently de-
veloped a zirconia overlayer which is derived from an organic
soluble zirconium salt ͑ZrCl4͒ via an organic medium.6 In this study,
high-capacity boride anodes are modified with this zirconia coating
methodology.
345 mAh g−1
.
The electrochemical stability of K2FeO4 is dramatically im-
proved by a 1% ZrO2 coating, and an 8.5 wt % AgO additive further
facilitates the cathode charge transfer. The zirconia overlayer pro-
vides an ionic-conductive, alkaline stable coating which is capable
of mediating hydroxide transport from the electrolyte to the elec-
trode, and a zirconia-modified K2FeO4/AgO composite cathode ex-
hibits longevity and high charge-storage capacity.6
Figure 7. Vanadium boride anode batter-
ies discharged with a variety of alkaline
cathodes under ͑left͒ anode-limited and
͑right͒ cathode-limited conditions. Anode:
75 wt % VB2, 20% g, 4.5% KOH, and
0.5% binder ͑T-30, 30% Teflon͒,
g
= 1 ␮m graphite. The cathode is either
K2FeO4, MnO2, NiOOH, or KIO4. Cath-
ode details and cell conuration conditions
are described in Fig. 5; cells are each dis-
charged at a 3 k⍀ load.
Downloaded on 2014-11-21 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A302
Journal of The Electrochemical Society, 155 ͑4͒ A297-A303 ͑2008͒
Figure 9. Capacity ͑anode + cathode͒ of the super-iron TiB2 boride alkaline
battery compared to that of the conventional ͑manganese dioxide/zinc͒ alka-
line battery. The super-iron boride cell is prepared with a 76.5% K2FeO4,
8.5% AgO, 5% KOH, and 10% 1 ␮m graphite cathode and a TiB2 anode, as
indicated in the figure. The cells are discharged at a constant load of 3 k⍀ at
22°C. Charge retentions ͑stability͒ of the cells are compared when freshly
discharged and after 1 week of storage at 22°C, with or without a 1% zir-
conia coating applied to the Fe͑VI͒ or boride salts.
Scheme 1. An 11e− anodic alkaline redox storage chemistry is explored
based on an environmentally benign zirconia-stabilized Fe6+/B2− chemistry,
which sustains an electrochemical potential matched to the pervasive, con-
ventional MnO2–Zn battery chemistry, however with a higher electrochemi-
cal capacity.
than that of contemporary alkaline rechargeable cells. In addition to
the stability of the super-iron VB2 cell presented in Fig. 10 at room
temperature, we observe that the vanadium boride anode exhibits
higher stability than the titanium boride anode. Accelerated long-
term stability is measured at higher temperatures. Without the zirco-
nia overlayer, after 1 week of storage the vanadium boride anode
retains 90% of the original charge capacity at 45°C ͑100% after 1
week with the zirconia coating͒ and 65% of the charge capacity at
70°C ͑85% with zirconia͒. We have previously demonstrated the
extended long-term solid-state stability of several alkalai super-iron
salts at these temperatures16 and have recently studied the variation
of the coating level and particle size and demonstrated longer-term
Fe͑VI͒ cathode stability conditions.23
As recently demonstrated for the cathode,6 this zirconia over-
layer also substantially stabilizes the alkaline anode, such as the
TiB2 anode.7 As seen in Fig. 9, after storage, the uncoated super-iron
titanium boride cell generates only 10–15% of the 3 k⍀ discharge
capacity of the fresh cell. In marked contrast to this, the full charge
capacity is retained when zirconia-coated super-iron and zirconia-
coated boride are utilized. Instead of the uncoated electrodes, if
either anode or cathode ͑but not both͒ is coated, then a fraction, but
not all, of the charge capacity is lost. Also evident in the figure, the
zirconia coated super-iron vanadium boride cell retains its substan-
tial charge capacity after storage. The fundamental chemistry of con-
ventional alkaline primary4 and metal hydride rechargeable2 batter-
ies are understood and continue to be of widespread interest. The
charge retention for the super-iron boride cells is already compa-
rable to that observed in early alkaline primary cells and is better
The capacity range of the conventional alkaline Zn–MnO2 cell is
included as dashed vertical lines in Fig. 10, which show the range
from a maximum experimental ͑Ͻ160 mA/g͒ to a theoretical ͑2F
per Zn + 2MnO2 = 222 mAh g−1͒ charge storage capacity. The
super-iron boride chemistry exhibits significantly higher charge stor-
Figure 10. Capacity ͑anode + cathode͒ of the super-iron TiB2 and a VB2
boride alkaline battery. The super-iron boride cell is prepared with a 76.5%
K2FeO4, 8.5% AgO, 5% KOH, and 10% 1 ␮m graphite cathode and either a
TiB2 or a VB2 anode, as indicated in the figure. Charge retentions ͑stability͒
of the cells are compared when freshly discharged and after 1 week of
storage, with a 1% zirconia coating applied to the Fe͑VI͒ and boride salts.
Figure 8. ATR/FTIR spectra of zirconia-coated and uncoated TiB2, VB2,
K2FeO4, and AgO compared with the pure ZrO2 spectra. Spectra of 5%
coated salts are included for emphasis; a 1% zirconia coating exhibits evi-
dent but proportionally smaller 1396 and 1548 cm−1 peaks.
Downloaded on 2014-11-21 to IP 134.129.182.74 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, 155 ͑4͒ A297-A303 ͑2008͒
A303
higher charge storage than conventional alkaline primary storage
chemistry while sustaining an electrochemical potential ͑under both
open-circuit and discharge conditions͒ matched to this pervasive,
conventional MnO2–Zn battery chemistry. The VB2 anode capacity
͑4060 mAh g−1͒ and K2FeO4 cathode capacity ͑408 mAh g−1͒ are
considerably higher than that of Zn and MnO2, or NiOOH, in con-
ventional alkaline batteries. Based on VB2 ͑72.6 g mol−1͒ and the
Fe͑VI͒ salt K2FeO4 ͑198.0 g mol−1͒, the complete super-iron boride
cell has an 11 Faraday theoretical capacity of 369 mAh g−1 ͑or
371 mAh g−1 when a cathode also contains 1 part AgO to 9 parts
K2FeO4; the added AgO mediates and facilitates the 3e− K2FeO4
charge transfer͒. We demonstrate that this intrinsic capacity is ap-
proached experimentally and is substantially higher than the conven-
tional Zn/MnO2 alkaline battery with a theoretical capacity of
224 mAh g−1 and an experimental capacity ͑to 0.8 V͒ of
160 mAh g−1
.
A further optimization of both the boride and super-iron salt
particle size, coupled with study and variation of the zirconia coat-
ing, may further enhance cell performance. Alternate metal borides,
as well as alternate super-irons, also affect characteristics of the
super-iron boride cell capacity, and ongoing studies of the charge
transfer of these unusual cathode and anode salts will impact, and
may be expected to lead to, further enhancements of charge transfer,
retention, and capacity of the super-iron boride chemistry.
Figure 11. Discharge of the 90 mAh vandadium boride super-iron battery at
various discharge loads. Anode ͑90 mAh͒: VB2 ͑1% Zr coated; 90 wt %͒ +
graphite ͑1 ␮m, 10 wt %͒; cathode ͑100 mAh͒: K2FeO4 ͑1% Zr-coated;
35–73 ␮m; 76.5 wt %͒ ϩ AgO ͑8.5 wt %͒ + KOH ͑5.0 wt %͒ + graphite
͑1 ␮m, 10 wt %͒; electrolyte: saturated KOH solution.
Acknowledgment
This program is supported in part by the U.S. Department of
Energy and the University of Massachusetts through the CVIP Tech-
nology Development Fund.
age than conventional alkaline primary storage chemistry. As deter-
mined from Eq. 15, the theoretical capacity for the complete super-
iron/VB2͑11F per VB2 + 11/3K2FeO4͒ cell is 369 mAh g−1. As seen
in the figure, the super-iron titanium boride cell combined anode and
cathode capacity is experimentally in excess of 250 mAh g−1, and
that of the super-iron VB2 cell is over 310 mAh g−1, which is two-
fold higher than that of the conventional alkaline battery chemistry
͑MnO2/Zn͒.
Figure 11 explores the super-iron boride ͑VB2͒ battery as pre-
pared with a higher, more functional storage capacity. The anode
and cathode active components of the 1 cm button ͑coin͒ cell con-
figuration are increased by an order of magnitude to 90 and
100 mAh of the respective intrinsic capacities of the anode and cath-
ode capacities. The higher capacity ͑100 mAh͒ of the super-iron
cathode is in line with the marginally lower coulombic efficiency ͑in
excess of 80% of the intrinsic capacity, compared to in excess of
90% for the anode͒ observed at the 3 k⍀ discharge. The anode is
composed of 90 wt % VB2 and 10% 1 ␮m graphite. The cathode is
composed of 76.5 wt % 2K2FeO4, 8.5% AgO, 5% KOH, and 10%
graphite. As seen in the figure, the capacity of the cell exceeds
320 mA g−1 at the 3 k⍀ discharge, based on these active materials.
Further enhancement, particularly in terms of optimization of the
particle size within the anode and cathode, and also the electrolyte
composition, is expected to further improve the capacity and sus-
tainable current density, while a determination of capacity inclusive
of nonactive components ͑graphite and zirconia mass, case, etc.͒
will decrease the practical capacity.
References
1. S. Licht, B. Wang, and S. Ghosh, Science, 285, 1039 ͑1999͒.
2. S. R. Ovshinsky, M. Fetcenko, and J. Ross, Science, 260, 176 ͑1993͒.
3. J. R. Dahn, T. Zheng, Y. Liu, and J. S. Xue, Science, 270, 590 ͑1995͒.
4. Z.-R. Tian, W. Tong, J.-Y. Wang, N.-G. Duan, V. Krishnan, and S. L. Suib, Science,
276, 926 ͑1997͒.
5. S. Licht and R. Tel-Vered, Chem. Commun. (Cambridge), 2004, 628.
6. S. Licht, X. Yu, and D. Zheng, Chem. Commun. (Cambridge), 2006, 4341.
7. S. Licht, X. Yu, and Y. Wang, Chem. Commun. (Cambridge), 2007, 2753.
8. H. X. Yang, Y. D. Wang, X. P. Ai, and C. S. Cha, Electrochem. Solid-State Lett., 7,
A212 ͑2004͒.
9. Y. D. Wang, X. P. Ai, Y. L. Cao, and H. X. Yang, Electrochem. Commun., 6, 780
͑2004͒.
10. S. Licht, V. Naschitz, B. Liu, S. Ghosh, N. Halperin, L. Halperin, and D. Rozen, J.
Power Sources, 99, 7 ͑2001͒.
11. S. Licht, V. S. Licht, V. Naschitz, L. Halperin, N. Halperin, L. Lin, J. Chen, S.
Ghosh, and B. Liu, J. Power Sources, 101, 167 ͑2001͒.
12. S. Licht, L. Yang, and B. Wang, Electrochem. Commun., 7, 931 ͑2005͒.
13. A. J. Bard, R. Parsons, and J. Jordan, Standard Potentials in Aqueous Solutions,
Marcel Dekker, New York ͑1985͒.
14. Property
and
Thermochemistry
Database,
http://mole.ipe.ac.cn/
all_thermochemistry/ ͑last accessed Feb. 6, 2008͒.
15. S. Licht, V. Naschitz, D. Rozen, and N. Halperin, Electrochem. Commun., 3, 350
͑2001͒.
16. S. Licht, V. Naschitz, D. Rozen, and N. Halperin, J. Electrochem. Soc., 151, A1147
͑2004͒.
17. S. Licht, D. Rozen, R. Tel-Vered, and L. Halperin, J. Electrochem. Soc., 151, A31
͑2004͒.
18. S. Licht, S. Ghosh, V. Naschitz, L. Halperin, and N. Halperin, J. Phys. Chem. B,
106, 5947 ͑2002͒.
19. S. Licht, S. Ghosh, V. Naschitz, L. Halperin, and N. Halperin, J. Phys. Chem. B,
105, 11933 ͑2001͒.
Conclusion
20. W. F. Linke, Solubilities of Inorganic and Metal-Organic Compounds, 4th ed., Van
Nostrand, Princeton, NJ ͑1958͒
21. S. Hettiarachchi, P. Kedzierzawski, and D. D. Macdonald, J. Electrochem. Soc.,
Whereas a conventional alkaline battery pairs the 2e− zinc anode
and 1e− MnO2 cathode, the alkaline super-iron/boride cell pairs up
to an 11e− boride anode, such as VB2, and a 3e− hexavalent iron
cathode. The super-iron boride chemistry exhibits significantly
132, 1866 ͑1985͒.
22. X. Fang, C. Yang, and J. Chen, J. Chin. Ceram. Soc., 6, 732 ͑1996͒.
23. S. Licht, X. Yu, and Y. Wang, J. Electrochem. Soc., 155, A1 ͑2008͒.
Downloaded on 2014-11-21 to IP 134.129.182.74 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).