Energetics of a zinc-sulfur fuel cell

Article abstract of DOI:10.1021/jp0135454

Energetics of a novel zinc-sulfur charge storage (generalized as Zn + S a?? ZnS) is explored to access the high (> 1000 Ah/kg) charge capacity of sulfur. At 25 ?°C, the theoretical energy density of the complete Zn/S system is a high 572 Wh/kg, at E^?

Full text of DOI:10.1021/jp0135454

J. Phys. Chem. B 2002, 106, 2989-2995
2989
Energetics of a Zinc-Sulfur Fuel Cell
Tatyana A. Bendikov, Chaim Yarnitzky, and Stuart Licht*
Department of Chemistry, Technion Israel Institute of Technology, Haifa, 32000, Israel
ReceiVed: September 17, 2001; In Final Form: December 20, 2001
Energetics of a novel zinc-sulfur charge storage (generalized as Zn + S f ZnS) is explored to access the
high (>1000 Ah/kg) charge capacity of sulfur. At 25 °C, the theoretical energy density of the complete Zn/S
system is a high 572 Wh/kg, at E° ) 1.04 V. From 273 to 373 K, the thermodynamic cell potential for 29
possible solution-phase reactions of the system were calculated including a variety of known coexisting aqueous
polysulfide species. In this domain zinc sulfide products are thermodynamically preferred over zincate products,
which can be advantageous compared to Al/S storage where aluminate is the favored discharge product.
Initial experimental feasibility of a zinc-sulfur electrochemical system is explored in a zinc anode/aqueous
polysulfide/CoS electrocatalytic cathode cell. Measured sustained cathode capacity approaches the theoretical
2
Faraday/sulfur limit. However, the anode tends to passivate due to the zinc sulfide discharge product,
preventing sustained discharge. Passivation is overcome in potassium polysulfide solutions containing
concentrated (>10 molal) KOH; this results in efficient, sustained 2 Faraday/Zn oxidation and high zinc-
sulfur charge capacities.
Introduction
hydroxide (as KOH within the electrolyte) in the formation of
the aluminum hydroxide discharge product, increasing its
Contemporary technological and societal demands for higher
capacity energy storage provide impetus for the exploration of
alternate electrochemical redox couples capable of higher energy
storage. Conventional cathodic charge capacities such as 292
and 308 Ah/kg, respectively, for NiOOH and MnO2 are
insufficient for future battery and fuel cell applications. The
low weight, low cost, and high two-electron theoretical faradaic
capacity of sulfur (1170 Ah/kg), make it an attractive cathode
candidate for electrochemical energy storage, but sulfur is one
of the most electrically insulating materials. Hence, studies have
generally focused on the use of sulfur in the molten (polysulfide)
state as a cathode material, as limited to high-temperature
batteries and fuel cells.1 However, ambient temperature
aqueous potassium polysulfide solutions are particularly soluble
and effective as cathodes. Indeed, we have shown that aqueous
potassium sulfide solutions can contain, by mass, more sulfur
than water, and also that solid-phase sulfur in contact with a
weight, and decreasing its practical energy density. Although
an alternate (aluminum) sulfide discharge product would prevent
this, the aluminum hydroxide is formed as the thermodynami-
cally favored eq 1 product, and preferred to the sulfide product
reaction which, depending on y ) 1.5/(x - 1), yields only a
-
1 10
∆
G° varying from -228 to -236 kJ mol :
2-
1
-
-
Al(s) + ySx + yH O f / Al S + yOH + yHS (2)
2
2
2 3
where y ) 1.5/(x - 1). Further, the observed potential of the
aqueous Al/S cell is less than 1.4 V, and this diminishes the
maximum energy density of the aluminum-sulfur system to
less than 507 Wh/kg (rather than 659 Wh/kg). This lower Al/S
oxidation potential in aqueous alkaline media, diminished from
the thermodynamic value of 1.8 V, is due to the parasitic
competition from the reaction of aluminum with water:4
-3
,7,10
thin aqueous alkaline polysulfide interface has a cathodic storage
3
Al(s) + 3H O f Al(OH)
+ / H
capacity in excess of 1000 Ah/kg.4
-10
2
3(amorph)
2
2(g)
Fundamental properties
∆
G° ) -426 kJ/mol (3)
of soluble sulfur electrolytes have also been studied to improve
efficiencies in photoelectrochemical solar cells.1
1-13
The intrinsic 820 Ah/kg two-electron theoretical faradaic
capacity of zinc (FW ) 65.39 g mol ), is lower than that of
the three-electron intrinsic 2980 Ah/kg capacity of aluminum
FW ) 26.98 g mol ). However, zinc has been widely used
as a battery anode material for over a century, although without
In 1993 sulfur was proposed as a cathode for aqueous
aluminum-sulfur electrochemical storage.4,7-10 The aqueous
alkaline cell discharge reaction depends on the polysulfide
species, Sx2 , summarized here for the highly soluble (potas-
-
1
-
1
(
-
-
1
sium) tetrasulfide species, with a ∆G° of -527 kJ mol :
1
4-18
sulfur.
In this study, despite the higher isolated anodic
capacity of aluminum compared to zinc, it will be demonstrated
that the zinc-sulfur system can have a considerably higher
intrinsic storage capacity compared to the aluminum-sulfur
system, and an alternate electrochemical storage system with a
generalized discharge reaction is probed:
2
-
-
-
2
Al(s) + S4 +2OH + 4H O f 2Al(OH) + 4HS
2 3
E° ) 1.82 V (1)
In accord with eq 1, the theoretical charge and energy density
of the potassium Al/S electrochemical storage are 362 Ah/kg
and 659 Wh/kg, respectively. This Al/S redox couple consumes
Zn + S f ZnS
(4)
The thermodynamic and initial experimental considerations for
a Zn/S electrochemical storage are explored.
*
Author to whom correspondence should be addressed. E-mail: chrlicht@
techunix.technion.ac.il.
1
0.1021/jp0135454 CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/27/2002

2990 J. Phys. Chem. B, Vol. 106, No. 11, 2002
Bendikov et al.
From this value, S2 activity is negligible over a wide range of
-
pH and sulfide concentrations. The dominant aqueous sulfide
-
-
species are HS and OH , and the aqueous electrolyte ZnS
1
9
reaction is better described by:
-
-
-
Cathode: S(s) + H O + 2e f HS + OH
2
E° ) -0.48 V (vs SHE) (9)
-
-
-
Anode: ZnS(s) + H O + 2e f Zn + HS + OH
2
(s)
E° ) -1.52 V (vs SHE) (10)
Total: Zn(s) + S(s) f ZnS(s) E° ) 1.04 V (11)
Hence, although eqs 9 and 10 involve water, it is not
consumed in the process, and the net eq 11 is identical to eq 7.
From the discharge reactions, eqs 7 and 11, the faradaic capacity
and the theoretical energy density for Zn/S electrochemical
storage are 550 Ah/kg and 572 Wh/kg, respectively. The Zn/S
energy density value is high compared with the theoretical
energy density of conventional aqueous cells, including Pb-
Figure 1. Schematic representation of (left side) generalized Zn/S
electrochemical discharge, and (right side) a Zn/S electrochemical
storage cell utilizing an aqueous polysulfide interface.
Experimental Section
Analytical grade reagents and distilled, doubly deionized
water were used throughout. Similar results were obtained with
anode material propared from 99.9% to 99.999% zinc foil from
Aldrich or Alfa AESAR. KHS solutions were prepared by
saturation of KOH solutions with H2S (gas) as previously
acid (170 Wh/kg), Ni-Cd (217 Wh/kg), alkaline Zn-MnO2
14
(
336 Wh/kg), and zinc-silver (447 Wh/kg).
Compared to Al/S, in the case of Zn/S electrochemical storage
at 298.15 K, the formation of ZnS product (∆G° ) -201.3 kJ
2
0
described. Polysulfide solutions are then formed by addition
of KOH and sulfur to KHS solutions. CoS electrodes were
-1
mol ) is thermodynamically preferred to the formation of
2-
-1
Zn(OH)4 or Zn(OH)2 (∆G° ) -156.3 kJ mol ). Hence unlike
aluminum, the zinc-sulfur reaction is not expected to consume
hydroxide during discharge; leading to more effective utilization
of the electroactive materials. After compensation for the
aqueous voltage loss in aluminum-sulfur, the 507 Wh/kg energy
density of the aluminum-sulfur cell is already less than that of
the alternate zinc-sulfur electrochemical storage. The described
properties make the Zn/S system potentially more attractive than
the Al/S system as an electrochemical power source.
2
formed by electrodeposition of Co at 50 mA/cm from 50 °C 2
m CoSO4, 0.6 m boric acid, 0.2 m NaCl solution onto a brass
substrate, followed by oxidation in polysulfide solution as
previously described.21 Note that molality units, m ) mol/kg
solution, are used in the experiments.
2
2
Both thin (7 cm ) and larger (20 cm ) Zn/S cells were
constructed from sandwiched rectangular Perspex blocks sepa-
rated by flat Teflon spacers. For the separation of anodic and
cathodic compartments Permion HD2291 cation permeable
membrane and conventional alkaline cell separators were used.
Galvanostatic overpotential measurements and linear voltam-
metry experiments were performed with a conventional three-
electrode potentiostatic configuration and Pine AFCBP1 bipo-
tentiostat/galvanostat. In polarization measurements a Ag/AgCl
Even in the absence of zinc, thermodynamic data is available
for the various polysulfide species, or of the related polysulfide
equilibrium or half reactions only in limited temperature ranges
2,5,12,26-33
and conditions.
In this study, the thermodynamic
potentials for the possible electrochemical reactions for the
zinc-sulfur system are calculated over the temperature range
2
reference electrode and a Pt foil (2 cm ) counter electrode were
(273.15-1273.15 K).
used. The constant load discharges of the Zn/S cells were
performed under (1% precision resistors. Experiments were
maintained under thermostatic control to (0.2 °C.
The investigated energetics of the zinc sulfide system focused
on the temperature range from 273.15 to 1273 and in which
reactions are investigated in aqueous media.
Thermodynamic Evaluation of Zinc-Sulfur Redox Stor-
age. The electrochemical processes occurring in a zinc-sulfur
system can be generalized by the cell illustrated on the left side
of Figure 1, and by the traditional half reactions:19
ZnS Thermodynamic Evaluation. The standard reaction
potential E°, in volts, for a temperature, T, in Kelvin, was
calculated from the standard reaction free energy change, in J
-1
4
-1
mol , using the Faraday constant, F ) 9.649 × 10 C mol ,
and the number of the electrons, n:
-
2-
Cathode: S(s) + 2e f S
E° ) -0.58 V (vs SHE) (5)
E°(T) ) -∆G°(T)reaction/nF
(12)
2
-
Anode: ZnS(s) + 2e f Zn(s) + S
(6)
(7)
where the value of ∆G°(T)reaction is the difference of Σ∆G°-
T)products and Σ∆G°(T)reactants, as determined for each species
participating in the reaction from the difference of the standard
(
Total: Zn(s) + S(s) f ZnS(s) E° ) 1.04 V
-
1
enthalpy of formation change, ∆H° (in units of J mol ) and
Zn/S electrochemistry is analyzed here first at ambient
temperatures, in a system consisting of a zinc anode and a highly
concentrated polysulfide cathode in an aqueous electrolyte, and
at higher temperatures in the next section. In aqueous solutions,
the second acid dissociation constant for H2S, is smaller than
-
1
T∆S°(T), using the standard entropy change, in units of J mol
-
1
K . The standard enthalpy change and the standard entropy,
S°, J mol K , were calculated from a constant heat capacity,
Cp, J mol K , respectively, by integrating either as Cp dT or
Cp dT/t over the from 298.15 K.
-
1
-1
-
1
-1
-
17 20-25
generally recognized, and on the order of 10
.
At room-temperature sulfur is an insulator, which despite its
-
1
-1)
high intrinsic two Faraday mol (FW(S) ) 32.07 g mol
storage capacity, presents a challenge to its use as a cathode.
-
-
2-
HS + OH f S + H O pK ) 17.3
(8)
2
2

Energetics of a Zinc-Sulfur Fuel Cell
J. Phys. Chem. B, Vol. 106, No. 11, 2002 2991
TABLE 1: Thermodynamic Data for Different Species at
Category “iii” reactions oxidize a polysulfide reactant to form
a sulfide product:
2
98.15 K26
b
∆
H°
S°
C
p
2
-
2-
-
[J mol-1 K-1]26
[J mol-1 K-1]34
species
[kJ mol 1]26
(x - 1)Zn(s) + Sx
f (x - 1)ZnS(s) + S
(15)
(aq)
(aq)
H+(aq)
Zn(s)
0
0
0
0
0
0
41.6
31.8
205.028
130.684
69.91
0
25.4
22.64
29.35
28.82
75.291
As indicated by the consistently larger cell potentials in Table
2 for reactions of categories i-iii compared to category iv, in
all cases the ZnS product is energetically preferred to the
S
(s)rombic
O
H
2(g)
2(g)
2-
Zn(OH)4 or the Zn(OH)2 products.
H
2
O
(l)
-285.830
-965.567
-643.25
-206.0
-229.994
-17.6
33.05
30.1
a
2-
(aq)
In aqueous polysulfide solutions, the activity distribution of
the coexisting polysulfide species depends strongly on the
Zn(OH)
4
ꢀ-Zn(OH)2(s)
81.6
57.5
-10.75
62.9
-16.3
28.4
65.5
103.4
140.6
72.383
46.02
-148.53
5
,11,31-32
2-
ZnS(s)sphalerite
alkalinity of the solution.
The tetrasulfide species, S4 ,
-
2-
OH (aq)
is the predominant species at pH ) 9-14. S5 dominates at
nearly neutral or slightly alkaline solutions. At high pH values
-
HS (aq)
2
-
S
S
S
S
S
(aq)
(
pH ) 14-16), consistent with over 1 molar hydroxide
2
2
2
2
-
(aq)
2
3
4
5
2
-
-
-
-
concentrations, the main species is S3 . At extremely high pH
(aq)
(aq)
(aq)
25.9
23.0
21.3
(pH>16 occurs in highly concentrated, high activity coefficient,
34
2-
hydroxide solutions ) the predominant species are S3 and
2
-
2-
a
2-
S2 . In accord with eq 8, S is only significantly active in
this highest pH extreme; at smaller pH values, fully reduced
sulfur exists in the hydrolyzed form as HS . On the basis of
For Zn(OH)
4
-
, only the ∆G° value at 298.15 K (∆G° 298.15K )
1 26
-
877.4 kJ mol ) was available. Hence, ∆H° was estimated from
the ∆S° value for the similar Zn(OH) species (∆S°298.15K ) 295. 712
values were not available, ∆G°(T) was
determined using the temperatures and constant values of ∆H°298.15 K
-
2
-
1
-1
b
J mol
K
). When C
p
this information, and the calculated thermodynamic potentials
in Table 2, a prediction of activity favored cell reactions is made
and summarized in Table 3. In addition, albeit to a lesser extent,
the distribution of polysulfide species is also affected by the
ratio of solution phase zero-valent to reduced sulfurs. Hence,
as all sulfur approaches the limit of full discharge, the final
and S°298.15K
.
However, dissolution of zero-valent sulfur in the form of one
of the various soluble aqueous polysulfide species, can provide
an effective interfacial bridge to the use of sulfur at or near
room temperature. This is schematically illustrated on the right
side of Figure 1, in which solid sulfur in contact with aqueous
alkaline solutions dissolves in the form of polysulfide species,
which in the discharge process are reduced while zinc is
oxidized. The controlling cell potential for this process will
depend on the specific polysulfide species, which is reduced,
and on the preferred discharge products. This study predicts
the thermodynamically favored discharge reactions. The kineti-
cally favored discharge mechanism may be more complex,
involving one or more intermediate steps, as exemplified for
-
2-
products must be the fully reduced sulfur species (HS or S ),
these minimum cell potential reactions are also included in Table
3
.
Initial Experimental ZnS Electrochemical Storage. In this
study, the initial experimental demonstration of zinc-sulfur
charge storage is confined to a single temperature range (near
room temperature) in a single medium (aqueous alkaline
electrolytes). Specifically, the viability of facile charge transfer
for a zinc anode/aqueous polysulfide/CoS electrocatalytic
cathode system is probed. We have previously demonstrated
that polysulfide reduction at a CoS electrocatalytic electrode
approaches 100 percent efficiency of the two-electron coulomb
the analogous case of polysulfide redox chemistry probed on a
cadmium chalcogenide surface.11
4
reduction of all available zero-valent sulfur, and that the sulfur
Aqueous alkaline solutions containing sulfur and sulfide salts,
5
cathode may be repeatedly charged. In the present study,
MxSy, are associated with an extensive speciation and a complex
2
y/x
-
2-
2-
2-
2-
2-
cathodic current is collected by a CoS electrode which can
sustain reduce polysulfide oxidization/reduction with a minimal
equilibrium of M , H2S, HS , S , S2 , S3 , S4 , S5 , H2O,
H , and OH .
+
- 5,12,30-33
These species can participate in elec-
2
polarization losses (on the order of 1 mV cm /mA polarization).
trochemical processes with zinc to form different products, and
electrochemical reactions can be predicted that can potentially
occur. The thermodynamic data that was used for the calcula-
tions of the theoretical potentials of the various discharge
reactions at different temperatures is summarized in Table 1.
Table 2 presents the thermodynamic potentials calculated for
Thermodynamically, aqueous polysulfide solutions are unstable,
2-
and sulfur, dissolved as polysulfide species, Sx , x ) 2-4, can
2- 6
decompose to thiosulfate, S2O3 . In that previous investigation,
we concluded that under conditions of high polysulfide con-
centration and at temperatures up to 85 °C, polysulfide solutions
can be extremely stable, and are particularly stable (on the order
29 different aqueous electrochemical reactions at 273.15, 289.15,
6
of years) at temperatures up to 55 °C. Zerovalent sulfur in
and 373.15 K. These reactions may be distributed into four main
categories of zinc discharge. Categories i-iii produce a ZnS
discharge product, while the fourth category, iv, produces a
zincate or zinc hydroxide product. In Table 2, category “i”
reactions oxidize a polysulfide reactant to form a shorter chain
length polysulfide product:
potassium polysufides can be effectively reduced using a CoS
electrocatalyst, either as separated by a cation selective mem-
brane in an isolated cathode half cell, or in a separator free cell.
However, oxidation of the zinc anode can be a challenge in
this medium.
It is unexpected that zinc can be effectively oxidized as an
active anode material in aqueous sulfide solutions. Zinc sulfide
2-
2-
Zn(s) + Sx
f ZnS(s) + S(x-1)
(13)
(aq)
(aq)
is both electrically nonconductive (semiconducting),3
5-38
and
-25 19
is also highly insoluble (the solubility product of ZnS is 10 ).
Hence, as expected, ZnS can passivate the zinc anode. It is
Category “ii” reactions oxidize a polysulfide reactant to form a
hydrosulfide product:
-
6
2
observed that no current (to within 10 A/cm ) occurs in
attempts to oxidize a Zn metal anode in a solution containing
2
-
-
(x - 1)Zn(s) + Sx
+ H O f (x - 1)ZnS + HS
+
(aq)
(aq)
2
(s)
3
m KOH and 3 m K2S4. Figure 2 compares the conventional
OH-
(14)
-2
(aq)
high anodic currents, > > 10 mA cm , zinc will sustain in

2992 J. Phys. Chem. B, Vol. 106, No. 11, 2002
Bendikov et al.
TABLE 2: Possible Reactions Expected for the Zn/S Aqueous Alkaline System, and Their Thermodynamic Potentials at 273.15,
2
89.15, and 373.15 K
reaction
equation
E°reaction
at 273.15 K
E°reaction
at 298.15 K
E°reaction
at 373.15 K
equation
4
i-1
i-2
i-3
ii-1
ii-2
ii-3
ii-4
, 7, 11
Zn(s) + S(s) f ZnS(s)
1.048
1.016
1.021
1.029
0.798
0.907
0.945
0.966
0.826
0.936
0.954
0.973
0.731
0.498
0.716
0.723
0.729
0.607
0.645
0.666
0.761
0.556
0.745
0.751
0.758
0.651
0.684
0.702
0.414
1.043
1.013
1.018
1.026
0.777
0.895
0.936
0.958
0.806
0.924
0.946
0.966
0.810
0.544
0.780
0.786
0.793
0.662
0.703
0.726
0.839
0.602
0.810
0.815
0.822
0.706
0.742
0.762
0.419
1.027
1.005
1.010
1.017
0.714
0.859
0.910
0.937
0.744
0.890
0.919
0.944
1.083
0.682
0.973
0.978
0.985
0.827
0.877
0.904
1.122
0.741
1.002
1.007
1.015
0.872
0.917
0.942
0.427
2
2
2
2
-
-
-
-
2
2-
2-
2-
Zn(s) + S
Zn(s) + S
Zn(s) + S
Zn(s) + S
3
4
5
2
(aq) f ZnS(s) + S
(aq) f ZnS(s) + S
(aq) f ZnS(s) + S
2
3
4
(aq)
(aq)
(aq)
(aq) + H
2
O
(l) f ZnS(s) + HS-(aq) + OH-(aq)
-
-
-
2Zn(s) + S
3Zn(s) + S
4Zn(s) + S
3
(aq) + H
(aq) + H
(aq) + H
2
2
2
O
O
O
(l) f 2ZnS(s) + HS (aq) + OH (aq)
2
2
-
-
- -
4
5
(l) f 3ZnS(s) + HS (aq) + OH (aq)
- -
(l) f 4ZnS(s) + HS (aq) + OH (aq)
2-
2
-
iii-1
iii-2
iii-3
iii-4
iv-1
iv-2
iv-3
iv-4
iv-5
iv-6
iv-7
iv-8
iv-9
iv-10
iv-11
iv-12
iv-13
iv-14
iv-15
iv-16
iv-17
Zn(s) + S
2
(aq) f ZnS(s) + S (aq)
2
-
-
-
2-
2Zn(s) + S
3Zn(s) + S
4Zn(s) + S
3
(aq) f 2ZnS(s) + S (aq)
2
2
2-
4
(aq) f 3ZnS(s) + S (aq)
2-
5
(aq) f 4ZnS(s) + S (aq)
Zn(s) + S(s) + 3OH-(aq) + H
O(l) f Zn(OH)
2-
(aq) + HS-(aq)
2-
2
4
2
2
2
2
-
-
-
-
2
-
-
Zn(s) + S
Zn(s) + S
Zn(s) + S
Zn(s) + S
2
3
4
5
(aq) + 2OH (aq) + 2H
2
O
(l) f Zn(OH)
4
(aq) + 2HS (aq)
-
2-
2-
2-
2-
-
(aq) + 3OH (aq) + H
2
2
2
O
O
O
(l) f Zn(OH)
(l) f Zn(OH)
(l) f Zn(OH)
(l) f 2Zn(OH)
(l) f 3Zn(OH)
4
4
4
(aq) + S
(aq) + S
2
3
4
(aq) + HS (aq)
-
2-
2-
-
(aq) + 3OH (aq) + H
(aq) + HS (aq)
-
-
(aq) + 3OH (aq) + H
(aq) + S
(aq) + HS (aq)
-
-
-
-
2-
-
2Zn(s) + S
3Zn(s) + S
4Zn(s) + S
3
(aq) + 5OH (aq) + 3H
2
O
O
4
4
(aq) + 3HS (aq)
2
2
-
2-
-
4
5
(aq) + 8OH (aq) + 4H
2
(aq) + 4HS (aq)
-
2-
-
(aq) + 11OH (aq) + 5H
2
O
(l) f 4Zn(OH)
4
(aq) + 5HS (aq)
-
2-
2-
Zn(s) + S(s) + 4OH (aq) f Zn(OH)
4
(aq) + S (aq)
2
2
2
2
-
-
-
-
2
-
2-
2-
2-
2-
2-
Zn(s) + S
Zn(s) + S
Zn(s) + S
Zn(s) + S
2
3
4
5
(aq) + 4OH (aq) f Zn(OH)
4
4
4
4
(aq) + 2S (aq)
2-
-
2-
(aq) + 4OH (aq) f Zn(OH)
(aq) + S
(aq) + S
2
3
4
(aq) + S (aq)
2-
-
2-
2-
(aq) + 4OH (aq) f Zn(OH)
(aq) + S (aq)
-
2-
(aq) + 4OH (aq) f Zn(OH)
(aq) + S
(aq) + S (aq)
-
-
-
-
2-
2-
2Zn(s) + S
3Zn(s) + S
4Zn(s) + S
3
(aq) + 8OH (aq) f 2Zn(OH)
4
(aq) + 3S (aq)
2- 2-
2
2
-
4
5
(aq) + 12OH (aq) f 3Zn(OH)
4
4
(aq) + 4S (aq)
-
2-
2-
(aq) + 16OH (aq) f 4Zn(OH)
(aq) + 5S (aq)
2
Zn(s) + 2H O(l) f Zn(OH)2(aq) + H2(g)
TABLE 3: The Dominant Aqueous Zinc-Sulfur Battery
Polysulfide Species, Sulfur Products, and Corresponding
Electrochemical Reactions at Different pH Ranges
dominant
polysulfide
species
final
activity discharged
pH
range
main
product
discharge favored
product
cell
reaction
reaction
2
2
2
-
-
-
2-
-
7
9
1
-9
S
5
S
4
S
3
S
3
S
4
3
2
2
HS
i-3
i-2
i-1
i-1
ii-1
ii-1
ii-1
iii-1
2-
2-
-
-14
4-16
S
S
S
HS
-
HS
2-, S
2-
2-, S2-
S
2-
>
16
2
Figure 3. Zinc anode linear voltammetry at 1 mV/s in 50 °C 8 m
solution also containing various concentrations of KOH.
2 4
K S
electrolyte. Current densities increase on the order of 1000-
fold compared to the less alkaline polysulfide electrolyte, and
-
2
current densities of over 10 mA cm on planar zinc electrodes
are sustained. Evidently, the high hydroxide concentration
permits facile charge transfer, despite competing zinc sulfide
passivation. This effective zinc oxidation, despite the presence
of sulfide, as well as the high capacity due to the zinc sulfide
discharge product, permit demonstration of viable zinc-sulfur
electrochemical storage.
Figure 2. Galvanostatic oxidation polarization of a 99.9% zinc anode
in several aqueous alkaline solutions, both with and without added
polysulfide.
The activation of a zinc anode in aqueous polysulfide solution
is further probed in Figures 3-6 at 50 °C using low sweep rate
(1 mV/s) linear voltammetry. Figure 3 summarizes the results
of a planar zinc anode oxidized in a 8 m K2S4 solution also
containing various concentrations of KOH. Significant oxidation
currents, in excess of 10 mA cm , are sustained in solutions
containing greater than 10 m KOH concentration. No zinc anodic
oxidation current is observed in the 8 m K2S4 solutions
containing less concentrated solutions ([KOH] e 9 m). At lower
sulfur-free alkaline KOH solution (in the absence of polysulfide)
to those in polysulfide solutions. In either a 3 m KOH or an 18
m KOH, at room temperature through 75˚C, high zinc anodic
currents are evident. While as seen in the figure, the same anode
is completely passive in the 3 m KOH and 3 m K2S4 electrolyte
-
2
(to within approximately 3 orders of magnitude).
Unexpectedly, as observed in Figure 2, a high current density
of zinc oxidation occurs in a 18 m KOH and 3 m K2S4

Energetics of a Zinc-Sulfur Fuel Cell
J. Phys. Chem. B, Vol. 106, No. 11, 2002 2993
2
Figure 7. 10mA/cm galvanostatic discharge of a Zn foil (thickness
Figure 4. Zinc anode linear voltammetry at 1 mV/s in 50 °C 8 m
2 4
19.2 µm), working electrode in excess 3 m K S , 14 m KOH electrolyte.
KOH solution also prepared with either 8 m K
or 8 m K
2
S, 8 m K
2
S
2
, 8 m K
2
3
S ,
The horizontal units are calculated from the product of measured time
and constant current density, as normalized to the initial electrode mass.
Counter electrode: CoS. Reference electrode: CoS, and potentials were
then recalculated versus Ag/AgCl using the fixed solution potential
2 4
S .
19
for CoS of -0.68 V vs Ag/AgCl. Inset: recalculated to zero-valent
zinc two-electron capacity.
that no anodic current is observed in the 8 m KOH solution
containing 8 m K2S4.
For the zinc electrode, Figure 5 summarizes the K2S4
passivation concentration in a fixed 3 m KOH solution, also
prepared with 0.12 m, 0.18, 0.20, or 0.75 m K2S4. The zinc is
fully passivated in the latter solution, and only at [K2S4] < 0.2
m are significant anodic currents observed. Finally, Figure 6,
summarizes the electrochemical activity of the planar zinc
electrode in a variety of potassium hydrosulfide or sulfide
solutions, containing no zero-valent sulfur. The small second
acid dissociation constant of K2S, with a pKa = 17, dictates
that in the absence of sulfur, the majority of dissolved sulfide
Figure 5. Zinc anode linear voltammetry at 1 mV/s in 50 °C 3 m
KOH solution also prepared with either 0.12 m, 0.18, 0.20, or 0.75 m
K S .
2 4
2
-
does not form S , but rather is hydrolyzed to form solution-
-
-
phase hydrosulfide, HS , and hydroxide, OH . In the left figure
inset the zinc anode is full passive in the 3 m pure (KOH-free)
KHS solution, whereas a small zinc oxidation current is observed
in the solution nominally prepared as 3 m K2S, but better
described in solution as containing equal concentrations (3 m)
of both KHS and KOH. In the right figure inset each electrolyte
contains 3 m KOH, and as expected, the zinc anode is most
active in the sulfide-free electrolyte and anodic currents are
progressively diminished when prepared with 1 or 3 m K2S.
Analogously, the main portion of Figure 6 illustrates that for a
constant 8 m concentration of K2S, the measured zinc anodic
current is progressively higher for higher KOH concentrations.
Figure 6. Zinc anode linear voltammetry at 1 mV/s in 50 °C in various
aqueous electrolytes. Main figure electrolytes: 8 m K S solution also
prepared with either 2.9, 5.6, or 8.2 m KOH. Left inset electrolytes: 3
m K S or 3 m KHS. Right inset electrolytes: 3 m KOH with either no
S, or either 1 or 3 m K S.
Coulombic efficiency reflects the degree of completion of
an electrode process. The effectiveness of zinc oxidation in
alkaline sulfur electrolytes was experimentally determined in a
galvanostatic measurement. This was accomplished by driving
a planar zinc electrode at constant current to complete anodic
consumption, and comparing the measured coulombs liberated
to the theoretical available two faradays per mole of zinc in the
mass of exposed zinc. As shown in Figure 7, a planar zinc
discharges to a high Coulombic efficiency (in excess of 90%
of the theoretical storage capacity, or more than 750 Ah per
kilogram of zinc). It is also interesting to note in the figure,
that even at this relatively high sustained current density of 10
2
2
K
2
2
concentrations of hydroxide, consisting of 0 to 2.7 m KOH in
curves 1 to 3, no zinc anodic oxidation occur, and the observed
cathodic currents are consistent with polysulfide reduction
occurring on the zinc anode. Figure 4 summarizes the results
of a planar zinc anode oxidized in a fixed 8 m KOH solution
also prepared with either 8 m K2S, 8 m K2S2, 8 m K2S3, or 8 m
K2S4. The first of these solutions contains no reducible sulfur,
while the second, third, and final of these solutions, respectively,
contain 8, 16, and 24 m zero-valent sulfur. In this figure it is
observed that higher concentrations of dissolved zero-valent
sulfur progressively passivate the zinc electrode to the extent
-2
mA cm , the anodic discharge potential is smooth and
decreases by only 50 mV through 90% of the zinc discharge.
This initial result,exemplifies facile anodic charge transfer for
the zinc sulfur system and is expected to be further improved
with optimization and variation of the zinc surface area, cell

2994 J. Phys. Chem. B, Vol. 106, No. 11, 2002
Bendikov et al.
TABLE 4: The Distribution of Species in a Variety of
Electrolytes Calculated Using the Previously Described
Computer Iterative Model and Equilibria Constants12a
nominal
composition
concentration (moles per liter)
K
Sulfur
KOH
2
S
0
0
3
0
0
18
3
0
3
3
9
3
3
9
18
species
in solution
concentration (moles per liter)
[
[
[
[
[
[
[
[
H
2
S]
0
0
0
0
0
0
0
3.0
0
0
0
0
0
0
0
18.0
<0.1
3.0
<0.1
0
0
0
<0.1
<0.1
<0.1
<0.1
0.2
2.6
0.2
<0.1
<0.1
<0.1
<0.1
0.2
2.6
0.2
-
HS ]
2
-
S
S
S
S
S
]
2
2
2
2
-
-
-
-
2
3
4
5
]
]
]
]
Figure 8. Current density/voltage polarization for a zinc-sulfur cell
containing aqueous potassium polysulfide electrolyte. Cells included a
silver chloride reference electrode and a 99.9% zinc anode and CoS
0
6.0
2 4
electrocatalytic cathode in 50(×b1×b1(1) °C 18 m KOH, 3 m K S
-
OH ]
3.0
18.0
electrolyte, and was discharged with a constant load (resistor). Figure
inset: Potential variation with time during discharge of a zinc/sulfur
cell. The top, middle, and lower curves, respectively, represent full
cell, cathode, and anode potentials separately monitored during constant
resistive load discharge.
a
Individual electrolytes are described in each separate column. A
solution with nominal a composition of 3 molar K S and 9 molar sulfur
is equivalent to a solution with a nominal concentration of 3 molar
2
2 4
K S .
configuration, and variations of the zinc and polysulfide
electrolyte composition.
Coupled with the known facile charge transfer for the
polysulfide cathode, the Figures 1-7 anode results provide the
initial basis for viability of zinc/sulfur electrochemical storage
using a zinc anode in contact with aqueous polysulfide. Table
4
presents the calculated concentration variation of dominant
species in aqueous alkaline polysulfide solutions covering the
range of polysulfide solutions in which zinc effectively dis-
charges at high current densities and to high Coulombic
efficiency. The distribution of species is calculated in accordance
with spectrally determined equilibria constants.12 In the high
alkaline domain investigated, it is seen in the table that the
-
predominant reduced sulfide species in solution is HS , and
2
the significant polysulfide species are S32 and mainly S4 . In
-
2-
Figure 9. Constant current load (5 mA/cm ) discharge at 35 °C of a
zin-sulfur cell using the thin cell configuration shown in the figure
particular it should be noted in the table that the dominant
solution-phase polysulfide species in a 3 m K2S4 solution
remains tetrasulfide, S42 , in both the (Zn nonactive) 3 m KOH
electrolyte, as well as in the (Zn active) 18 m KOH electrolyte.
Hence, the range of concentrated polysulfide solutions resulting
in unusually high zinc oxidation faradaic efficiency, cannot be
correlated to substantial variations of the distribution of species
in solution. It is beyond the scope of this initial investigation
to elucidate the specific mechanism by which polysulfide
solutions sustain Zn oxidation, but only in the presence of high
hydroxide activities. However, the thermodynamically favored
ZnS product occurs, and in the above absence of correlation to
the distribution of polysulfide species in solution, it is likely
that the facile zinc oxidation occurs via electrochemical oxida-
tion to zincate followed by chemical reaction to zinc sulfide.
A typical discharge of a Zn/S electrochemical cell is presented
in the inset of Figure 8. The open circuit onset potentials and
discharge potentials of the complete cell, the Zn anode potential,
and the electrocatalytic CoS cathode potential were measured.
Anode, cathode, and reference electrodes were maintained in
the aqueous alkaline polysulfide solution under thermostatic
conditions, and as shown, a discharge is sustained for the bulk
of (∼85%) the discharge without significant variation in cell
voltage. Zn/S cells were discharged under constant load condi-
tions with (1% precision resistors. Such a cell can be discharged
at high current densities while sustaining low polarization losses
as summarized for a variety of current densities in the main
portion of Figure 8.
inset. Current/voltage polarization for a zinc-sulfur cell containing
2
aqueous potassium polysulfide electrolyte. Anode: 7 cm Zn foil.
-
2
Electrocatalytic cathode: 7 cm CoS on thin brass foil. Electrolyte: 5
2 4
m K S , 14 m KOH.
An initial optimization of a prototype zinc-sulfur electro-
chemical fuel cell, was studied in a thin cell configuration. The
thin cell utilized the sub-millimeter inter-electrode dimensions
summarized in the inset of Figure 9. In static (non flow)
discharges, the cell was observed to maintain low polarization
compared to large gap cells, for electrolytes with higher
polysulfide concentrations, and when discharging the cell at a
lower temperature of 35 °C. The cell discharges to high
Coulombic efficiency of ∼80% for both the available zinc and
zero-valent sulfur, and yields a high faradaic capacity of 250
Ah per kilogram of active materials, based on the cell’s total
mass of zinc, polysulfide, and potassium hydroxide. This high
value of faradaic capacity is promising. Future experiments
include an excess zinc anode in which an external reservoir of
polysulfide is pumped into and through the cell.
Conclusions
Thermodynamic considerations of a novel zinc-sulfur elec-
trochemical storage are presented, and the inital experimental
capablities of this system are explored. On the basis of the
thermodynamic calculations, it can be expected for the zinc-
polysulfide aqueous alkaline system, that over a wide range of

Energetics of a Zinc-Sulfur Fuel Cell
J. Phys. Chem. B, Vol. 106, No. 11, 2002 2995
temperatures, from 273 to 373 K, the anodic reaction of zinc
with sulfide (polysulfide) will be preferred to that of zinc with
hydroxide. This minimizes the total system mass and increases
the energy density. A high energy density of storage (572 Wh/
kg at 298.15 K), coupled with the low cost of zinc and sulfur
are competitive with alternate existing electrochemical storage
systems.
Initial experimental feasibility of the zinc-sulfur electro-
chemical system was explored in a zinc anode/aqueous polysul-
fide/CoS electrocatalytic cathode cell. Measured sustained
cathode capacity approaches the theoretical 2 Faraday/sulfur
limit. However, the anode tends to passivate due to the zinc
sulfide discharge product, preventing sustained discharge.
Passivation is overcome in potassium polysulfide solutions
containing concentrated (>10 molal) KOH, resulting in efficient,
sustained 2 Faraday/Zn oxidation and high zinc-sulfur charge
capacities.
(9) Licht, S. J. Electrochem. Soc. 1997, 144, L133.
10) Licht, S.; Jeitler, J. R.; Hwang, J. J. Phys. Chem. B 1997, 101,
959.
11) Licht, S. Nature 1987, 330, 148.
12) Licht, S.; Hodes, G.; Manassen, J. Inorg. Chem. 1986, 25, 2486.
(13) Licht, S. J. Phys. Chem. 1986, 1986, 1096.
14) Linder, D. Handbook of Batteries; McGraw-Hill: New York, 1995.
15) Falk, S. U.; Salking, A. J. Alkaline Storage Batteries; John Wiley
Sons: New York, 1969.
(
4
(
(
(
(
&
1.
(
16) Kordesch, K. V. Batteries; Marcel Dekker: New York, 1974; Vol.
(
(
(
(
17) Allen, P. L.; Hickling, A. Trans. Faraday Soc. 1957, 53, 1626.
18) Licht, S.; Wang, B.; Ghosh, S. Science 1999, 285, 1039.
19) Licht, S. J. Electrochem. Soc. 1988, 135, 2971.
20) Peramunage, D.; Forouzan, F.; Licht, S. Anal. Chem. 1994, 66,
378.
(21) Licht, S.; Longo, K.; Peramunage, D.; Forouzan, F. J. Electroanal.
Chem. 1991, 318, 111.
(
(
(
22) Licht, S.; Forouzan, F.; Longo, K. Anal. Chem. 1990, 62, 1356.
23) Licht, S.; Manassen, J. J. Electrochem. Soc. 1987, 134, 918.
24) Meyer, B.; Ward, K.; Koshlab, K.; Peter, L. Inorg. Chem. 1983,
22, 2345.
(
(
25) Giggenbach, W. Inorg. Chem. 1971, 10, 1333.
26) Bard, A. J.; Parson, R.; Jordar, J. Standard Potentials in Aqueous
Acknowledgment. Acknowledgment. We thank B. Wang
and O. Khaselev who participated in cell discharges and partial
funding for the thermodynamic calculations from the Fund for
the Promotion of Research at the Technion, and the Israel-
Mexico Energy Research Fund, and the BMBF Israel-German
Cooperation.
Solution; Marcel Dekker: New York, 1985.
27) Lessner, P. M.; McLarnon, F. R.; Winnick, J.; Cairns, E. J. J.
(
Electrochem. Soc. 1993, 140, 1847.
(28) Bogomolova, Z. V.; Tikhonov, K. I.; Rotinyan, A. L. Electrokhimiya
979, 15, 1237.
1
(
(
(
(
29) Probst, D. A.; Henderson, G. J. Chem. Educ. 1996, 73, 962.
30) Teder, A. Acta Chem. Scand. 1971, 25, 1722.
31) Giggenbach, W. Inorg. Chem. 1972, 11, 1201.
32) Lessner, P. M.; Winnick, J.; McLarnon, F. R.; Cairns, E. J. J.
References and Notes
(
1) Toboshima, S. I.; Yamamoto, H.; Matsuda, M. Electrochim. Acta
Electrochem. Soc. 1986, 133, 2517.
33) Le Guillanton, G.; Do, Q. T.; Elothmani, D. J. Electrochem. Soc.
996, 143, L223.
1
997, 42, 1019.
(
(
2) Sharivker, V. S.; Ratkye, S. K.; Cleaver, B. Electrochim. Acta 1996,
1
4
1, 2381.
(34) Licht, S. Anal. Chem. 1985, 57, 514.
(
(
(
(
(
(
3) Sangster, J.; Pelton, A. D. J. Phase Equilib. 1997, 18, 89.
4) Peramunage, D.; Licht, S. Science 1993, 261, 1029.
5) Licht, S. J. Electrochem. Soc. 1987, 134, 2137.
6) Licht, S.; Davis, J. J. Phys. Chem. 1997, 101, 2540.
7) Licht, S.; Peramunage, D. J. Electrochem. Soc. 1993, 140, L4.
8) Licht, S.; Hwang, J.; Light, T. S.; Dillon, R. J. Electrochem. Soc.
(35) Cheng, B.; Jiang, W. Q.; Zhu, Y. R.; Chen, Z. Y. J. Mater. Sci.
Lett. 2000, 19, 503.
(36) Zhao, X. K.; Fendler, J. H. J. Phys. Chem. 1991, 95, 3716.
(37) Zhang, Y.; Raman, N.; Bailey, J. K.; Brinker, C. J.; Crooks, R. M.
J. Phys. Chem. 1992, 96, 9098.
1
997, 144, 948.
(38) Wang, E. G.; Ting, C. S. J. Appl. Phys. 1995, 77, 4107.