18 3. EZectrochem. Soc.

: E L E C T R O C H E M I C A L S C I E N C E A N D T E C H N O L O G Y January 1981

B. M. L. Rao, in " l l t h I n t e r s o c i e t y E n e r g y Con- 11. A. Sequeira, I. Bernal, I. D. Brown, a n d R. F a g g i -

version and E n g i n e e r i n g Conference (197~), E x - ani, Ac$a Crystallogr. B, 31, 1735 (1975).
t e n d e d Abstracts," pp. 418-423 (1976). 12. D. T. Hurd, J. Org. Chem., 13, 711 (1948).
4. G, H. Newman, R. ~W. Francis, L. H. Gaines, and 13. R. Damico, ibid., 29, 1971 (1964).
B. M. L. Rao, This Journal, 127, 2025 (1980). 14. J. C. Cessna, Corrosion, 27, 244 (1971).
5. M. F. Lappert~ Chem. Rev., 56, 1035 (1956). 15. W. K. Busfield, R. M. Lee, and D. Merigold, Die
6. E. L. Muetterties, Editor, "The C h e m i s t r y of Boron Macromolek. Chem., 156, 183 (1972).
and its Compounds," pp. 443-582, J o h n W i l e y and 16. E. C. Ashby, L. C. Chao, and J. Laemmle, J. Org.
Sons, Inc., New Y o r k (1967). Chem., 39, 3258 (1974).
7. E. Negishi, J. Organomet. Chem., 1@8, 281 (1976). 17. D. Nicholls and M. Szwarc, J. Phys. Chem., 71, 2727
8. D. N. B h a t t a c h a r y y a , C. L. Lee, J. Smid, and M. (1967).
Szwarc, J. Phys. Chem., 69, 608 (1965). 18. E. C. Ashby, F. R. Dobbs, and H. P. Hopkins, Jr.,
9. P. J a g o d z i n s k i and S. Petrucci, ibid., 78, 917 (1974). J. Am. Chem. Soc., 95, 2823 (1973).
10. W. E. Rhine, G. Stucky, and S. W. Peterson, J. Am. 19. A. H. Thompson, Phys. Rev. Lett., 40, 1511 (1978).
Chem. Soc., 97, 6405 (1975). 20. M. B. Dines, Mater. Res. Bull., 10, 287 (1975).

Investigation of Factors Affecting Performance

of the Iron-Redox Battery
L. W. Hruska*
Diamond Shamrock Corporation, Painesville, Ohio 44077

and R. F. Savinell*
Department of Chemical Engineering, University of Akron, Akron, Ohio 44325

ABSTRACT
The iron*redox b a t t e r y is a low p o w e r d e n s i t y e n e r g y storage device t h a t
m a y be a t t r a c t i v e for applications such as load leveling and solar e n e r g y
storage. D u r i n g the charge cycle of this b a t t e r y , the ferrous ion f r o m an
aqueous chloride e l e c t r o l y t e plates onto the n e g a t i v e electrode and is oxi-
dized to f e r r i c ion at the positive electrode. A solid g r a p h i t e or t i t a n i u m p l a t e
can be used for the iron electrode while a high specific a r e a m a t e r i a l is neces-
s a r y for the r e d o x electrode. A 10O cm 2 single cell w i t h a microporous plastic
s e p a r a t o r has been cycled at a c u r r e n t d e n s i t y of 60 m A / c m 2 at 60~ T h e
cell was c a p a b l e of storing a n d discharging e n e r g y at a r o u n d trip c u r r e n t
efficiency of 90% and an e n e r g y efficiency of 50%, w i t h a m a x i m u m discharge
p o w e r density of 50 mVV/cm 2. The largest voltage losses o c c u r r e d at the iron
electrode a n d to a lesser e x t e n t at the r e d o x electrode. F a c t o r s affecting
voltaic and coulombic losses including electrolyte composition and t e m p e r a -
ture, cell separators, electrode materials, and e l e c t r o l y t e additives h a v e been
studied a n d the results reported.

S e v e r a l electrochemical e n e r g y storage b a t t e r i e s are Discharge

u n d e r d e v e l o p m e n t for b u l k e n e r g y storage (1, 2).
These b a t t e r i e s a r e often characterized b y their low F e~ --> Fe +2 -5 2e Negative electrode
p o w e r density, b u t are a t t r a c t i v e as s t a t i o n a r y storage 2Fe +s -5 2 e - --> 2Fe +2 Positive electrode
devices for l o a d ]eveling a n d solar applications b e -
cause of t h e i r p o t e n t i a l l y low cost. One such b a t t e r y Electrochemically, this b a t t e r y can be v i e w e d as a
which has u n d e r g o n e r a p i d d e v e l o p m e n t over the h y b r i d of m o r e common b a t t e r y types. The n e g a t i v e
p a s t s e v e r a l y e a r s is the r e d o x b a t t e r y (3-6). In the or iron electrode utilizes plating and dissolution of
r e d o x b a t t e r y , both the products and r e a c t a n t s of iron as a solid plate, s i m i l a r to Zn/C12 or F e / N i b a t -
e l e c t r o l y s i s a r e soluble salts. In o r d e r to p r e v e n t teries. The positive electrode, however, uses a carbon
chemical shorting a n d i r r e v e r s i b l e loss of reactants, s t r u c t u r e to oxidize and reduce iron cations which
a h i g h l y efficient anion p e r m e a b l e m e m b r a n e s e p - r e m a i n it~ solution. Ideally, upon complete discharge,
a r a t o r m u s t be employed. A n a l t e r n a t i v e method of the b a t t e r y would be b r o u g h t to its initial state with
p r e v e n t i n g the i r r e v e r s i b l e loss of reactants is t h r o u g h ferrous ion as the o n l y active cation in the system.
utilization of a r e a c t a n t couple which involves three Other a d v a n t a g e s of the i r o n - r e d o x b a t t e r y include:
different valence states of a single metallic element. low t e m p e r a t u r e aqueous electrolytes w i t h i n e x p e n -
One such device is the i r o n - r e d o x b a t t e r y (FeO/Fe+2/ sive r e a c t a n t s that are r e a d i l y a v a i l a b l e in large q u a n -
Fe+ ~), which is the subject of this paper. tities; the use of i n e x p e n s i v e m a t e r i a l s of construc-
The h a l f - c e l l reactions for the i r o n - r e d o x b a t t e r y tion; the use of electrodes which are not subject to
can be expresse d as follows: shape change or dendrites; the distinction of being
the least hazardous of a n y a d v a n c e d b a t t e r y system.
Charge A d i s a d v a n t a g e w i t h the i r o n - r e d o x b a t t e r y is its
Fe+2 -5 2 e - --> F e 0 Negative electrode i n h e r e n t l y low cell voltage (1.21V as c o m p a r e d to
2.12V for Zn/C12 and 1.98V for P b / a c i d ) which tends
2Fe +2 -> 2Fe +s -5 2e Positive electrode to limit the p o w e r density and efficiency of the
* Electrochemical Society Active Member. battery. The actual o p e r a t i o n a l characteristics of the
Key words: membrane, separators, energy storage. b a t t e r y m u s t be m e a s u r e d e x p e r i m e n t a l l y . D e v e l o p -

Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 VoL 128, No. 1 IRON=REDOX BATTERY 19

m e n t a l w o r k on a n i r o n - r e d o x b a t t e r y has b e e n re. l I l I I
p o r t e d p r e v i o u s l y (7, 8), b u t d a t a on factors c o n t r o l -
OPEN SYMBOLS - EXPERIMENTAL DATA
ling e n e r g y efficiency and on cell o p e r a t i o n a r e limited. CLOSED SYMBOLS - DATA FROM REF, (9)
The results o b t a i n e d in this e x p e r i m e n t a l s t u d y 9 - BOTH EXPERIMENTAL AND
indicate the e n e r g y levels t h a t can be o b t a i n e d f r o m PUBLISHED DATA
t h e i r o n - r e d o x b a t t e r y a n d i d e n t i f y factors w h i c h
[3
l i m i t its p e r f o r m a n c e . 5i 1~ --

Experimental
The studies r e p o r t e d h e r e i n w e r e p r i m a r i l y e x p e r i -
mental. S e p a r a t o r s a n d electrode m a t e r i a l s w e r e c o m - ~FECL3 ~O /
mercial, r e a d i l y a v a i l a b l e items. The salts w e r e all 9 9 0 m
of r e a g e n t g r a d e a n d deionized w a t e r was used for
electrolyte makeup.
The e l e c t r o l y t e resistivity m e a s u r e m e n t s w e r e t a k e n
w i t h a s t a n d a r d H - t y p e c o n d u c t i v i t y cell w i t h p l a t i n -
=
c~ lo -- \ ./
9 FECL2 9
ized p l a t i n u m electrodes (cell constant of 0.1). The
cell was i m m e r s e d in a t e m p e r a t u r e b a t h a n d the
resistance was m e a s u r e d w i t h a B e c k m a n I n s t r u m e n t s
a - c b r i d g e o p e r a t i n g at a f r e q u e n c y of 1000 Hz.
m
D a i l y cycle tests w e r e p e r f o r m e d in a 100 c m ~ 5 p

r e c t a n g u l a r acrylic l a b o r a t o r y cell. Each h a l f - c e l l was I I I ] ]

e q u i p p e d w i t h e l e c t r o l y t e i n l e t a n d outlet manifolds 200 400
a n d v i t o n O - r i n g s w e r e used for sealing. C h l o r i d e d ELECTROLYTE CONCENTRATION, GPL
s i l v e r wires w e r e e m p l o y e d as r e f e r e n c e electrodes. Fig. 2. Measured and published resistivity data of iron chloride
T h e e l e c t r o l y t e circuit is depicted s c h e m a t i c a l l y in aqueous solutions at 18~
Fig. 1. A l l components of the test system w e r e m a d e
of p o l y p r o p y l e n e , PVC, o r glass.
F o r example, considering an initial concentration of
Results and Discussion both catholyte and a n o l y t e to be 500 gpl ferrous
Electrolyte properties.~The p r i m a r y ionic r e a c t a n t s chloride a n d a f t e r full charge t h e r e is 50 gpl ferrous
of the i r o n - r e d o x b a t t e r y a r e the ferric a n d ferrous chloride r e m a i n i n g in the anolyte, the charge d e n s i t y
ions. Since these r e a c t a n t s a r e soluble in aqueous of the b a t t e r y is 63.5 A - h r / l i t e r .
solution t h e i r salts serve as the b a t t e r y electrolyte. As will be shown later, p H control is essential in
I n o r d e r to o p e r a t e a cell at h i g h voltaic efficiency o r d e r to m a i n t a i n a h i g h f a r a d a i c efficiency a n d good
the e l e c t r o l y t e resistivity m u s t be m a i n t a i n e d as low plating morphology. Since p a r a s i t i c reactions in this
as possible. F o r this reason we chose the chloride salts b a t t e r y do result in a n a l t e r a t i o n of the electrolyte
of i r o n as o u r solute (the chloride salts w i l l also pH, it is quite i m p o r t a n t to use an e l e c t r o l y t e h a v i n g
have g r e a t e r charge c a r r y i n g c a p a c i t y p e r g r a m of a p H response t h a t is m o d e r a t e a n d not a b r u p t w h e n
salt as c o m p a r e d to other salts such as sulfates or the concentration is p e r t u r b e d . W e h a v e f o u n d the
n i t r a t e s ) . A n e x a m i n a t i o n of solution conductivity ferrous chloride electrolyte to be w e l l b e h a v e d as
d a t a is useful for identifying a p r o m i s i n g c o n c e n t r a - d e m o n s t r a t e d b y the d a t a r e p o r t e d in Fig. 3.
tion r a n g e for cell operation. M e a s u r e d r e s i s t i v i t y In o r d e r to r e d u c e voltaic losses, an indifferent b u t
d a t a for b o t h ferric a n d ferrous chloride solutions conductive solute can be a d d e d to the electrolyte. The
at 18~ a r e given as a function of concentration in resistivities of a 250 gpl f e r r o u s chloride e l e c t r o l y t e
Fig. 2. The d a t a a g r e e w i t h p u b l i s h e d d a t a (9). The w i t h additions of s e v e r a l indifferent solutes a r e shown
useful concentration r a n g e can be established from in Fig. 4. The top curve is for the a d d i t i o n of calcium
these d a t a as ca. 100-500 gpl of i r o n chloride. A t l o w e r chloride. A l t h o u g h this a d d i t i v e is c o m m o n in i r o n
concentrations the charge c a r r y i n g d e n s i t y becomes p l a t i n g baths, it has no a d v a n t a g e as a c o n d u c t i v i t y
too l o w a n d concentration polarizations become too
high. A t high concentrations the forces r e l a t e d to
interionic a t t r a c t i o n p r e d o m i n a t e and ionic m o b i l i t y
decreases. By k n o w i n g the w o r k a b l e concentration 3 --
range, the m i n i m u m v o l u m e of e l e c t r o l y t e for a given ~ 20~
coulombic c a p a c i t y of the b a t t e r y can be estimated.
o

FE+2 ~ FE0
S SEPARATOR
I~FE+2~-~.:-FE+3
\o
TITANIUM I \
" CURRENT I %e O~
~o
~ o FECL2
-2 \o
\. + i,o
~PoRous
REOOX
ELECTRODE
@- I 200
I I 400
l
FECL 2 CONCENTRATION, GPL
Fig. I. Schematic of iron-redox laboratory cell and electrolyte
circuit. Fig. 3. Measured pH of ferrous chloride solutions at 20~
Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 20 J. Electrochem. Soc.: E L E C T R O C H E M I C A L SCIENCE AND TECHNOLOGY January 1981

anode to produce the ferric ion. Simultaneously, a t

the cathode the ferrous ion is reduced to metallic
10-- iron. The ferric ion produced at the anode is then
5 capable of reacting directly with the iron plate ac-
cording to the chemical reaction
Fe 0 + 2Fe+S-~ 3Fe+~
This reaction is limited by mass transport of the
ferric ion to the electrode surface and can be a
KCL p r i m a r y source of coulombic efficiency loss (other
losses are discussed in later sections). The current
i I "~" NHaCL -- efficiency (based on iron plate) when discharging an
| 250 GPL FECL 2 unseparated cell was found to be about 60%. This
2[." , 20oc corresponds to the level of efficiency loss that would
!
be expected from mass transfer control of the above
reaction (see Table I).
0 200 400 In order to reduce the coulombic efficiency loss
caused by the oxidation of the iron plate, microporous
SUPPORTCONCENTRATION,
GPL materials were evaluated as cell separators. Ferric
Fig. 4. Measured effect of supporting electrolyte on the resistivity ion diffusion data and membrane resistivity data are
of a 250 gpl ferrous chloride aqueous solution at 20~ presented in Fig. 6 and 7 for two commercial materials.
Using these data, calculations were made on the ex-
pected performance of these separators in cell opera-
support. The addition of sodium or potassium chloride, tion. The Celgard| separator (Celanese Corporation)
however, lowers the resistivity, significantly. One dif-
ficulty found with using a sodium or potassium chloride
additive was the poor quality and nonadherence of
the iron plate deposited from electrolytes having these
salt additives. 30_ 0733 -
The most effective conductivity support tested was
ammonium chloride. The addition of 3.5 moles of
ammonium chloride to a liter containing 2.0 moles ~" ~ A R D 5511
ferrous chloride (nearly saturated solution) will r e - 20 SOLUTION VOLUME - 0.28L
duce the electrolyte's resistivity from 7.8 to 4.0 ~2-cm
r~EMBRANE AREA - 45.6 CM 2
at 20~ F u r t h e r testing has shown that good quality
iron plates can be obtained in the presence of am-
monium chloride additives. In fact, the iron deposit
had very little tendency to dendrite, powder, crater, ~ !0-- --
peel-off, or exhibit the other problems that are gen-
erally associated with many of the metallic plates
used for b a t t e r y applications. A further reduction in / ~ ! ' , 0 7 5 3
resistivity of the ferrous-ammonium chloride solution
can be achieved at elevated temperatures. This effect g--'- I I I
is shown in Fig. 5. The resistivity of 4 s at 20~ 2 ~ 6
can be reduced to 2 12-cm at 77~ For a cell with a DIFFUSION TIIiE, IIOL'.RS
2 mm gap operating at a current density of 60 mA/cm~,
this represents an overall voltaic loss of 4% (as com- Fig. 6. Measurements of ferric ion transport rate across two
pared to 16% at 20~ with no addition of indifferent commercial microporous separators using a well-stirred membrane
electrolyte). The ferrous chloride/ammonium chloride cell.
electrolyte was chosen for further battery studies.
Cell separators.--At the fully discharged state iron
is not in metallic form and the ferrous concentration
is at its maximum (perhaps 500 gpl FeC12 and 50 gpl 2M FECL2
FeCl~). During charge, ferrous ion is oxidized at the 0
5C -- 2,8M NH4LL I
25 ~C / DARA~IC
/ W0733
4 ,%1 I I
5 \ 2MFECLz
/
0
0
\ 3.5M NH4CL
ok
o N-- 3O
3-- O% 0
o
m
%0%

N
2--
0~0~0~ i0 - - / ~ / / 5511

_ ( I I
25 50 75 i00 50 i00 150
TEMPERATURE, ~ CURRENT DENSITY, MA/CM2

Fig. 5. Measured temperature dependence of the resistivity of a Fig. 7. Measured IR drop across two commercial microporous
2M FeCI~/3.SM NH4C/aqueous solution, separators.

Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Vol. I28, ~o. i IRON-REDOX BATTERY 21

Table I. Porous separator performance at 60 mA/cm 2 utilizing diffusion and resistivity data
of Fig. 6 and 7

Separator
ico,,* CEc = CEv = ID/ CEt = thickness P o r e size Resistivity IRloss
( m A / c m 2) Ic - icorr I v + icorr CEc X CED (mm) (/~m) (~-cm~) (mV)

No s e p a r a t o r 29.0 0.52 0.67 0.35 -- -- 0 0

C e l g a r d * * 5511 14.6 0.76 0.80 0.61 0.20 0.04 0.155 9.3
D a r a m i c ~ W0733 1.94 0.97 0,97 0.94 0.58 0.05 0,41 24.6

* T h i s v a l u e w a s calculate.d u s i n g F i e k ' s l a w of s i m p l e d i f f u s i o n ( N = - - D C ~ / 6 ) and assuming an electrolyte linear velocity of 2

c m / s e c , e l e c t r o d e l e n g t h o f 7 c m , a d i f f u s i o n coefficient of 10-~ c m e / s e c , a m e m b r a n e g a p of 0.1 c m , a n d a c h a n g e of f e r r i c ion concen-
t r a t i o n d u r i n g c h a r g e f r o m 0 to 127 gpl.
** P o r o u s p o l y p r o p y l e n e m a t e r i a l m a n u f a c t u r e d b y C e l a n e s e C o r p o r a t i o n .
7 P o r o u s p o l y e t h y l e n e b a s e d m a t e r i a l m a n u f a c t u r e d b y W. It. G r a c e C o r p o r a t i o n .
C E r = c h a r g e c o u l o m b i c efficiency; C E D = d i s c h a r g e c o u l o m b i c efficiency.
I~ = c h a r g e c u r r e n t ; /D = d i s c h a r g e current.
ieorr : c a l c u l a t e d d i f f u s i o n c o n t r o l l e d c o r r o s i o n c u r r e n t .

offers some i m p r o v e m e n t in coulombic efficiency. On i n Fig. 8. Although the data reported are for a t i t a n i u m
the other hand, the Daramic| separator (W. R. Grace substrate, similar data were obtained for the F e + 2 / F e 0
Corporation) should limit the ferric ion diffusional reaction on a graphite substrate. With a n o m i n a l cur-
loss to about 3% for both charge and discharge modes rent density of 60 m A / c m 2, at 90~ and pH : 1, the
of operation. Daily operation of laboratory size (109 cathodic and anodic polarizations are 95 and 135 mV,
cm 2) redox cells with the Daramic separators gave respectively. This represents a voltaic efficiency for
coulombic efficiencies of 93%-96% d u r i n g charge and the iron electrode of 70%. The iron electrode polariza-
90% d u r i n g discharge. The chemical resistances of tion can be decreased by adjusting the electrolyte pH
both Celgard and Daramic separators i n iron elec- to higher values. For example, at the same conditions
trolytes were excellent at pH ----O. as above but with a pH _-- 3, the voltaic efficiency
Studies on electrode polarizations (which are re- would be 83%. The pH dependence of the electrode
ported later i n this paper) indicate the o p t i m u m polarization is in agreement with the following reac-
catholyte (on charge) to have a pH of 3-4 while the tion mechanism as put forth in the l i t e r a t u r e (12)
anolyte m u s t be kept at or below a pH of a p p r o x i m a t e l y
Fe +2 + H20 ~ FeOH + -F H+
1.0 to m a i n t a i n stability of the ferric ion. I n order
to m a i n t a i n a pH difference of this m a g n i t u d e across FeOH + + e - --> FeOH (rds)
the cell, a highly efficient anion permeable m e m b r a n e
m u s t be used as the separator. Two commercial m e m - FeOH+H + +e-~Fe ~
branes were evaluated i n this study: Neosepta ACH- Finally, as would be expected, the electrode polariza-
45T ( T o k u y a m a Soda Company, Limited) and Sele- tion was found to be t e m p e r a t u r e sensitive (see Fig.
mion| DMV (Asahi Glass Company Limited). Both 8). For example, reducing the electrolyte t e m p e r a t u r e
of these m e m b r a n e s had an initial resistivity of ap- from 70 ~ to 60~ will decrease the voltaic efficiency
p r o x i m a t e l y 7 ~ - c m 2 i n a ferric-ferrous chloride
to 75% (at 60 m A / c m 2, pH ----3.7).
solution (420 mV loss at 60 mA/cm2). Upon e v a l u a -
tion i n laboratory cells, the m e m b r a n e s discolored Plating current e~ciency.--The coulombic efficiency
giving evidence of deterioration and the resistivity of the i r o n - r e d o x b a t t e r y is controlled by reactions
of the m e m b r a n e increased significantly. The increased at the iron plate (the ferric/ferrous redox reaction
resistivity is a t t r i b u t e d to FeC14- fouling and is not is n e a r l y 100% efficient). One source of efficiency loss
an u n c o m m o n p h e n o m e n a among anion m e m b r a n e s is the reaction of ferric ion with the iron plate
(2, 3, 10). I n spite of the poor voltaic performance

-i'01 t
of the m e m b r a n e , the overall coulombic efficiency
was found to be exceptionally good (CE ~ 96%).
9 90~ PEI 1,O
Characteristics o] the iron electrode.--Electrode 60~C PH 3,7
polarization.--During cell charging the ferrous ion is [] 70~ PH 3,5
reduced at the iron electrode o 90~ PH 3,0
Fe + 2 + e - --> Fe ~
-0,9
The plating reaction is carried out on a nonreactive 2[~ FECL 2 ~o ~~
substrate. I n this way complete discharges eliminate 2,811 N H 4 C L o ~~ .
shape change of iron plate, limited cycle life, etc. -0,8 o~ A A~/X
Both t i t a n i u m and graphite are suitable substrates
and have been utilized in l a b o r a t o r y cell studies. % ^~...9
Although both of these materials are resistant to ~" -0,7
corrosion and good iron plate morphology has been
obtained on both of them, they do suffer from degrada-
tion if polarized anodically to a great extent (as d u r -
ing overdischarge). Graphite will anodically corrode
excessively at higher pH's (pH 3-4) (11) while bare
taJ

-0,6 '{" ~" 9 "'- 5~" " U " " 0 ~ 0~ 0 "'0

"~" 9 - ~ " ~ A - ~ , ^ " ~ ~.- D .,,..
~'"o~ z-x'--A~ []
""0
A I

t i t a n i u m will dissolve anodically at sufficiently posi- -0,5 __ ~ ~A~ .__~

tive potentials. A t i t a n i u m substrate used in our lab- O..~,, O ,,,," '/X.
oratory cell had experienced massive pitting d u r i n g 9 ,,,,,, O
an overdischarge. One of the positive aspects of u t i l - -0,4
izing graphite as a substrate is the low cost. On the
other hand, t i t a n i u m is more conductive and is weld- 50 lOO 200
able (for bus bar connections), thus capable of higher
loads for a given level of s u b s t r a t e / b u s b a r voltage CURRENT DENSITY, MA/CM 2
loss.
The polarization curves (steady state) of the iron Fig. 8. Measured polarization of an iron electrode on a titanium
electrode i n 250 gpl FEC12/150 gpl NH4C1 are given substrate.

Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 22 J. EZectrochem. Soc.: E L E C T R O C H E M I C A L SCIENCE AND TECHNOLOGY January 198I

FeO + 2Fe+S--> 8Fe+S 1.0

This reaction has been discussed earlier and c a n b e
minimized by the use of a suitable microporous sep- 2M FECL2 /
arator.
A second source of coulombic efficiency loss can I'gM NHqCL t /--
be represented by the following reaction
/ /
Fe o + 2HC1--> FeCIs + H2
Chemical corrosion of this type results in h y d r o g e n
gassing which has been confirmed by analysis of gases
above electrolyte storage bottles. A lowered coulombic
efficiency has been measured even when the cell was
at open circuit. It is therefore desirable to minimize
the rate of this reaction. Studies in our laboratory
have shown the above reaction to have a corrosion o
rate strongly dependent on electrolyte pH and on c_)
plate composition (i.e., additives in the plate).
The pH effect on chemical corrosion is demonstrated
by the results presented in Fig. 9. According to these
results the desired electrolyte pH should be greater 0,2 - -

than pH 1-1.5. (Note: the higher pH electrolyte also

contributes to more favorable polarizations of the iron
electrode.) On the other hand, in order to prevent
the formation of insoluble ferric oxides and hydroxides,
the electrolyte must be kept below pH = 1.5. Since
protons are highly mobile, the electrolyte pH must 0.01 0.i0 1.0 i0.0
be similar on both sides of the cell. Consequently,
chemical corrosion can be minimized by altering the ADDITIVE CONCENTRATION,GPL
composition and morphology of the iron plate (open- Fig. 10. Measured corrosion rate for iron samples plated from
circuit corrosion can be avoided by draining the ceil solutions containing metal ion additives.
of electrolyte, a feature of the iron-redox battery
system).
It is well known that small quantities of metal ions cathodic than iron. When added to the plating bath,
in the electrolyte will affect the properties of the all were generally effective in reducing the corrosion
iron plate. Tests have shown that some additives result rate as measured by weight loss. Four other metal
in iron with a lower corrosion rate in acidified ferrous ion additives tested, cobalt, nickel, cadmium, and lead,
chloride solution. The measured corrosion rates of have plating potentials less cathodic than iron and
samples plated from solutions containing 0.01-10.0 gpl were generally ineffective in reducing the chemical
of several metal ion additives along with a standard corrosion mechanism. One exception was a 1 gpl
base rate for pure electrolytic iron are presented in nickel ion addition which resulted in an iron dissolu-
Fig. 10. Three of the metal ion additives reported tion rate 91% lower than the standard. However, this
(Zn, Mn, Cr) have electroplating potentials more beneficial effect was not present at the lower concen-
trations, and chemical analysis of the plating solutions
showed the nickel ion to be present at a much higher
I f f concentration in the plate than in the initial solution.
This concentrating effect would remove the nickel ion
from the electrolyte at the start of plating, leaving
2r'~ FECL2 an ineffectively low concentration of nickel ion in the
electrolyte. Consequently, nickel cannot be considered
i, 9r~o[~II4CL
a good additive.
Characteristics of the redox electrode.--In a single
cell, counter to the iron electrode is the ferrous/ferric
redox electrode. The reaction at this electrode can
be written as follows (during charge)
Fe+9-~ Fe+8 + e-
c~ This reaction has been selected previously for applica-
N tion in Fe/Ti and Fe/Cr redox batteries and its polar-
ization has been shown to be small (13-15). One of
c~
these studies (15) used a rotating disk electrode of
edge-on pyrolytic graphite and has shown the reaction
1-- mechanism in FeCls/3M HCI to be by simple electron
_J transfer. Of course, in order to minimize unfavorable
diffusional effects, a high surface area porous elec-
trode must be used.
In Fig. ii, a comparison between smooth graphite
and porous carbon electrodes is shown for the oxida-
tion of ferrous ion in the FeCI2/NH4CI electrolyte.
Although the single electrode potentials are not re-
ported, the beneficial effect that the porous electrode

0.5 LO
I I
1.5
has on the ferrous ion oxidation is demonstrated. This
is especially true at lower temperatures and higher
current densities.
SOLUTIONPH A polarization curve of the ferrous-ferric reaction
Fig. 9. Measured dependence of electrolytic iron corrosion rate on porous carbon, during charge, is given in Fig. 12.
on solution pH. Although not reported, the Fe +8 reduction curve is

Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Vot I28, No. 1 IRON-REDOX B A T T E R Y 23

e 25oc-I I I I The coulombic efficiency of the ferrous/ferric couple

i was always measured to be near 100%; that is, no
9 50~ ~ UNION CAP~IDE
9 90~ ATJ side reactions (Os or C12 evolution during charge)
O o ~ e were observed. The only coulombic loss was due to
25 C~UNION C A B I D E ~ *
2,5 -- o 50oc| PC_25e / ferric ion migration through the separator. The high
/S 90~ C.J / 9 eoulombic efficiency of the ferrous/ferric reaction
coupled with the competing proton reaction at t h e
iron electrode lead to an accumulation of ferric ion.
This charge/discharge imbalance causes operational
2,0- 2M~,FE2L2pL 9 -- problems of a b a t t e r y and is discussed in the n e x t
section.

/ /o..o Daily cycle performance.--In order to measure t h e

actual current and energy efficiencies of an iron-redox
cell, daily cycling operation is necessary. Cycle tests
were performed in our laboratory with a 100 cm~
cell and using a catholyte and anolyte volume of 2
liters each. Generally the cell was charged for 3 hr
then discharged until a lower limit of zero volts
was reached. Both the charging and the discharging
50 iO0 150 were at constant current density of 60 m A / c m 2. Dur-
CURRENT DENSITY, MA/CM 2 ing cell cycling, solution pH (pH -- 1) and tempera-
ture (60~ were held constant.
Fig. 11. Comparison of measured charge voltage between labora- A graph showing the current and energy efficiencies
tory cells having a smooth graphite redox electrode (Union Carbide for 57 cycles is shown in Fig. 13. I t is obvious that
ATJ) and one having a porous carbon redox electrode (Union Car- cell performance is highly variable from one cycle
bide PC-25). to the next, not attributable to any intentional change
in operating parameters. The major changes that were
symmetrical when a clean electrode is used. Some made during this period are given in Table II. The
other observations made during our studies m a y be effects from many of these changes have already been
worth mentioning at this time. For instance, the discussed. Current efficiencies of 80%-90% could be
voltaic efficiency of the ferrous/ferric electrode was routinely obtained but the best energy efficiency was
greatest on a fresh carbon surface. Oxides accumulat- around 50%. The low energy efficiency then, of course,
ing on the electrode surface quickly deteriorated the was due to the poor voltaic efficiency (55% at best).
voltage performance of this electrode. These oxides Measurements with silver/silver chloride refer-
precipitated when solution pH became greater than ence electrodes identified the source of electrode polar-
1.5 and they have been identified as mixed oxides and izations that occurred during cell cycling. A typical
hydroxides, noncrystalline, and p r i m a r i l y of ferric voltage breakdown at the start of cell cycling is given
ion. The solution pH would increase because of proton in Fig. 14. As can be seen, during charge 50% of the
discharge at the iron cathode during charge and be- voltage loss is due to the iron electrode, 27% due to
cause of air oxidation. The second factor was elimi- the redox electrode, and 14% due to the separator
nated by sealing the electrolyte tanks and nitrogen resistance. During the first cycle the voltage loss across
blanketing. The activity of the electrode could be the Daramic separator was about 50~ mV. This
restored to its original state by applying an acid wash. polarization increased rapidly to about 120 mV shortly
Unlike the graphite electrode for the zinc/chloride thereafter. The increase in diaphragm resistivity has
b a t t e r y (16), the surface activity of the electrode in been attributed to Fe(OH)2 blockage of the separator
this system was not important and no activation pro- pores. The conductivity of the separator could be
cedures were required. Poor electrode performance returned to its original state by applying an acid
was noted for the ferrous ion oxidation when Fe +2 wash. Eventually, a thicker (2.4 mm) Daramic sep-
concentration dropped below 50 gpl. A reticulated
vitreous carbon showed acceptable activity although
it had poor electronic conductivity. A ruthenized po-
rous titanium electrode also showed good activity and Table II. Notes on iron-redox cell cycle tests
has a very conductive substrate.
Electrolyte: FeCI~/NH4C1 (2M/2.8M). Redox electrode: PC-25.
l l I Te mpe ra t ure : 60~ Iron electrode s ubs t ra t e : t i t a n i u m sheet.
Current density: 60 mA/cmS. El e c t rol yt e flow ra t e : I0 cm/sec.
FE+2 ~ FE+3 + E Acid additions: re qui re d almost daily to maintain pH of 1.
150 - POROUS CARBON ELECTRODE Coulombic Ene rgy
UNION CARBIDE PC-25 Cycle efficiency efficiency
number (%) (%) Action t a k e n prior to cycle
2M FECL2
2,8M NHaCL o/~ 1 58 36 Separator, Celgard 5511
2s~ o / 5 70 35 New separator, Daramte W0733
-- laO - o / - 6 72 35 FeCla addition
11 84 29 FeCI~ addition
12 83 44 Cell dra i ne d of electrolyte and
..-1 0 acid washed
17 77 37 FeC12 addition
N 50 21 75 29 Cell drained of electrolyte a n d
acid washed, flow rate i n c r e a s e d
N to 40 c m / s e c
22 85 38 Flow ra t e decreased to original
value
27 53 24 W a t e r addition
0 29 63 21 Cell dra i ne d of electrolyte, solu-
I t I tions replaced
50 100 150 35 59 11 Electrolytes filtered to r e m o v e ox-
ides
CURRENT DENSITY, MA/CM 2 39 76 15 N e w separator, electrodes, and so-
lutions
Fig. Measured polarization of ferrous ion oxidation at
12. a 42 59 23 New redox electrode, .reticulated
porous carbon electrode. vitreous carbon

Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 24 J. E~ectrochem. Soc.: E L E C T R O C H E M I C A L S C I E N C E A N D T E C H N O L O G Y January 1981

i00 100
0

80 80

60 60 -<
Fig. 13. Measured efficiency ,,~
during cycling-of a 100 cm 2
iron-redox cell. Details of op- • A I
u.J
eratlng conditions are given in ~- 40 40 -.<
Table il.

20

t W 20

' '
io ' " 5'0 ' 65
CYCLE NUMBER

a r a t o r w i t h a r i b b e d s t r u c t u r e was e v a l u a t e d . a n d g r a m (at a b o u t 100 cycles) two m e t a l ion additives

found to have a longer life ( g r e a t e r t h a n 20 cycles). were e v a l u a t e d in t h e cell a f t e r obtaining the results
The largest v o l t a g e loss is caused b y the polarization from the i r o n corrosion tests. The zinc ion caused a
a t t h e iron electrode. This polarization m a y be r e - r e d u c t i o n in o b s e r v e d c u r r e n t efficiency to 70% as
duced b y a d j u s t i n g the e l e c t r o l y t e p H to h i g h e r values. c o m p a r e d to the usual 90%. A l t h o u g h not confirmed
We a t t e m p t e d to m a i n t a i n a p H d i f f e r e n t i a l in l a b - at this time, we speculate the zinc addition caused
o r a t o r y cells b y using an anion selective m e m b r a n e a shift to m o r e negative potentials of the iron p l a t i n g
separator. But, as discussed previously, the s e p a r a t o r s reaction and, consequently, increased the significance
q u i c k l y fouled. The fouling p r o b l e m m a y be avoided of the p r o t o n reduction reaction. The addition of
in the f u t u r e b y i n c o r p o r a t i n g r e c e n t advances in m a n g a n e s e ion i m p r o v e d b a t t e r y operation s u b s t a n -
m e m b r a n e technology f r o m the N A S A r e d o x p r o - tially. A l t h o u g h no increase i n c u r r e n t efficiency was
g r a m (2). obvious, the electrolyte pI-I r e m a i n e d stable thus e l i m -
T h e second most significant voltage loss was at inating the need for acid addition.
the r e d o x electrode. A l t h o u g h not large initially, the Conclusions
polarization increased d u r i n g cell operation. The loss A n i r o n - r e d o x single cell, at 60~ a n d using a
of electrode p e r f o r m a n c e was associated w i t h a p a s - microporous plastic separator, has been shown to be
sivating film a c c u m u l a t i n g on the electrode surface. capable of storing and discharging e n e r g y at a r o u n d
As s t a t e d earlier, electrode p e r f o r m a n c e was r e s t o r e d t r i p c u r r e n t efficiency of 90% and e n e r g y efficiency
b y a p p l y i n g an acid wash. of 50%. A discharge p o w e r d e n s i t y of 50 mW/cm2
The o p e r a t i n g cycle of 100%-83% depth of dis- has been obtained.
charge in these tests do not r e p r e s e n t a p r a c t i c a l o p e r -
S e v e r a l factors affecting the cell p e r f o r m a n c e have
a t i n g cycle such as 9.0%-10% depth of discharge. been investigated. A f e r r o u s - a m m o n i u m chloride elec-
However, it was sufficient to note the occurrence of t r o l y t e was found to be a h i g h l y conductive m e d i u m
electrolyte i m b a l a n c e and f u r t h e r charging o n l y w o u l d capable of y i e l d i n g good q u a l i t y electrolytic iron
compound the p r o b l e m at this time. The electrolyte plate. Both t i t a n i u m a n d g r a p h i t e w e r e found to be
i m b a l a n c e is b r o u g h t a b o u t b y proton r e d u c t i o n ( d u r - suitable substrates for iron plating. The largest voltage
ing charge) which in t u r n affects electrolyte p H and loss o b s e r v e d occurred at the iron electrode for both
causes an a c c u m u l a t i o n of ferric ion. As a result, charge and discharge. A t t e m p t s at l o w e r i n g this p o l a r -
hydrochloric acid and ferrous chloride h a d to be ization b y m a i n t a i n i n g a p H differential across the
r o u t i n e l y a d d e d to the electrolyte. L a t e r in the p r o - cell w i t h commercial anion m e m b r a n e s were unsuc-
O, i0 cessful because of m e m b r a n e fouling. To a lesser
I extent, but" not inconsequential, was the polarization
0.05
at the high specific a r e a porous r e d o x electrode. The
Q 0,22 addition of s m a l l amounts of m a n g a n e s e ion was
Q 0 oO:8 beneficial to cell p e r f o r m a n c e in t h a t ferrous chloride
@ 0,20 and acid additions d u r i n g cycle tests become unnec-
essary.
Acknowledgment
OCV i, 20V T h e financial s u p p o r t for this p r o j e c t was received
from D i a m o n d S h a m r o c k Corporation. Technical dis-
cussions w i t h J a c k B e n n e t t and Nick F a t i c a and e x -
p e r i m e n t a l assistance p r o v i d e d b y P. J. P a l m e r w e r e
i n v a l u a b l e and g r e a t l y appreciated.
CHARGE 1,57 DISCHARGE 0.87 V
M a n u s c r i p t s u b m i t t e d J u l y 8, 1980; revised m a n u -
(~) FE+2/FE +3 ELECTRODE script received Aug. 25, 1980.
<Z) SEPARATOR iR A n y discussion of this p a p e r w i l l a p p e a r in a Discus-
(~) IRON ELECTRODE sion Section to be p u b l i s h e d in the D e c e m b e r 1981
JOURNAL. A l l discussions for the D e c e m b e r 1981 Discus-
VOLTAIC EFFICIENCY = 0,87 = 55 % sion Section should be s u b m i t t e d b y Aug. 1, 1981.
!,57
Publication costs of this article were assisted by the
Fig. 14. Measured voltage breakdown for the initial cycle of a University o] Akron and the Diamond Shamrock Cor-
100 cm2 laboratory cell at 60mA/cm% 60~ and pH = 1. poration.
Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Vol. 128, No. 1 I R O N - R E D O X BATTERY 25

REFERENCES 9. E. W. Washburn, "International Critical Tables of

Numerical Data, Physics, Chemistry and Tech-
1. "Energy Storage," J. Berkowitz and H. Silver- no.ogy," Vol. 6, pp. 229-233. Published for Na-
man, Editors, The Electrochemical Society, Soft- tional Research Council by McGraw-Hill, New
bound Proceedings Series~ Princeton, N.J. (1976). York (1929).
2. "Proceedings of the 14th Intersociety Energy Con- 10. R. F. Savinell, Ph.D. Dissertation, University of
version Engineering Conferences," Boston, Mas- Pittsburgh, Pittsburgh, Pa. (1977).
sachusetts, Aug. 5-10, 1979, American Chemical 11. J. S. Sconce, "Chlorine: Its Manufacture, Proper-
Society, Washington, D.C. (1979). ties and Uses," ACS Monograph, pp. 81-166, Rein-
3. L. H. Thaller, NASA Technical Memorandum TM- holt Publishing Corp., New York (1962).
79186, 1979; also in Ref. (2), Vol. 1, p. 715. 12. J. O'M. Bockris and A. K. N. Reddy, "Modern
4. L. H. Thaller, U.S. Pat. 3,996,064 (1976). Electrochemistry," Vol. 2, p. 1091, Plenum Pub-
5. L.H. Thaller, Abstract 85, p. 229, The Electrochemi- lishing Corp., New York (1970).
cal Society Extended Abstracts, Atlanta, Georgia, 13. J. Giner, L. Swette, and K. Cahil, NASA CR-134705,
Oct. 9-14, 1977. 1976. (Available through NTIS, Springfield, Va.).
6. R. F. Savinell, C. C. Liu, R. T. Galasco, S. H. Chiang, 14. M. A. Reid and R. F. Gahn, NASA Technical Memo-
and J. F. Coetzee, This Journal, 126, 357 (1979). randum TM X-73669 (1977).
7. R. Zito, U.S. Pat. 3,719,526 (1973). 15. R. O. Miher, NASA Technical Memorandum TM-
8. R. Zito and C. M. Harman, Presented at Depart- 73716 (1977).
ment of Energy Second Annual Battery and 16. P . C . Symons, Report EM-711, pp. 2.2-2.17 (1978).
Electrochemical Technology Conference, Arling- (Available from Electric Power Research Insti-
ton, Va., Jl~ne 5-7, 1978. tute, Palo Alto, Cal.).

The Effect of Zirconium on the Cyclic

Oxidation of NiCrAI Alloys
C. A. Barrett, A. S. Khan,1 and C. E. Lowell*
N A S A - L e w L s R e s e a r c h Center, Cleveland, Ohio 44186

ABSTRACT
Cyclic oxidation tests of Ni-(9-2O)Cr-(15-30)Al-xZr alloys were carried
out at 1100~ and 1200~ in static air. In these alloys the concentration of zir-
conium varied from 0 to 0.63 atomic percent (a/o). The phases found in the
cast alloys were % -~', ~, and ~. The oxidized surfaces were characterized by
metallography, XRD, and electron microscopy. Although the alloys were basi-
cally alumina/nickel aluminate formers, the zirconium-containing alloys
also formed small amounts of ZrO2. The metallographic and electron micro-
scopic examination of oxidized surfaces revealed significant aluminum
oxide (s) penetration. The depth of attack and the degree of oxide penetration,
however, strongly depended on the zirconium content of the alloy. The zir-
conium-free alloy had massive penetration of aluminum oxide(s). Small
amounts of zirconium led to minimal penetration. However, as the zirconium
content increased, the oxide penetration again became more pronounced. All
of the weight change ( A W / A ) - t i m e data were fitted to the paralinear oxida-
tion model: A W / A -- kl~/2t ~/~ -- k2t where kl~/~ represents the rate of scale
formation and k2 the rate of oxide spalling. The Zr-containing alloys had
relatively good cyclic oxidation resistance as a result of intraoxide spalling
at low rates. The Zr-free alloys had relatively poor cyclic oxidation resistance
due to massive alumina spalling to bare metal, which resulted'in a shift to
NiO-controlled oxidation and even greater rates of attack. The derived
parameters Ka (which equals kill2 § 10 k2) and Wm (metal consumption)
which indicate the degree of total oxidation attack were also estimated as a
function of zirconium content. All four parameters showed minimums in the
range of 0.03 to about 0.20 a/o zirconium. Further addition of zirconium be-
yond 0.20 resulted in increased values of k l ~/2, Ka, k2, and Win. Alloys with
zirconium contents between 0.03 and 0.20 a/o seemed to have the best cyclic
oxidation resistance over the range of compositions tested, while oxidatiQn
resistance was essentially independent of Cr and A1 content over the com-
positional range evaluated for these alloys.

The improvement of the oxidation resistance of al- posed that the internal rare earth oxide which is
loys by the addition of reactive elements such as Y, formed during oxidation mechaniCally keys the ex-
Si, etc. has been known for several years (1-8). ternal oxide to the alloy and also prevents void forma-
However, the mechanism by which these elements tion at the alloy-oxide interface. Allam et al. make a
improve oxidation resistance has not yet been satis- similar case for oxide adhesion by pegging in the
factorily explained. In this respect several mechanisms CoCrA1 system(s). Their work indicated not only an
have been proposed assuming that the added elements improved adhesion of the scale by the addition of Y
produce beneficial effects due to some interaction or Hf, but also made a case for an optimum concen-
either at the alloy-oxide interface, in the oxide layer, tration of these elements below 1 weight percent
or in the matrix. Pettit and co-workers (7, 8, 11) pro- (w/o). On the other hand, in a recent paper Golightly
et al. (9) prorosed that the addition of rare earth
* E l e c t r o c h e m i c a l Society A c t i v e M e m b e r .
elements prevents oxide formation within the existing
Present address: P r a t t & Whitney Aircraft, West Palm Beach, oxide layer, probably by altering the outward diffu-
F l o r i d a 33402. sion of aluminum. Subsequent oxide growth occurs
Key w o r d s : s c a l e a d h e s i o n , o x i d e spalling, A12Os f o r m a t i o n , oxi-
d a t i o n attack parameters, at the alloy-oxide interface forming an adherent scale.
Downloaded on 2014-11-05 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).