Zinc-Iron Flow Batterieswith Common Electrolyte

Steven Selverston231st ECS Meeting, May 30, 2017

Electrochemical Engineering and Energy Laboratory

Department of Chemical and Biomolecular Engineering

Case Western Reserve University, Cleveland, Ohio

Principal Investigators:Robert Savinell, Ph.D.Jesse Wainright, Ph.D.

Motivation: Grid-Scale Energy Storage

30

20

10

12 24

net

load

(GW

)

time of day (h)

solar surplus

energy deficit

batte

ries

0

0.5

1

1.5

2

2.5

3

3.5

4

2014 2016 2018 2020 2022 2024 2026

globa

lmar

kets

ize(b

illion

s) lead acidli-ionflow

sodiumothertotal

• Store solar energy during day, use at night 1

• Replace peaker plants, enable micro-grids• Need improvements in cost and safety

1Adapted from California Association of Independent System Operators (CAISO)

1

“Zinc-Iron” Flow Batteries

• Abundant, low-cost, readily-available materials• Challenges include solids/slurry handling, ion-exchange membranes

ferri/ferrostorage

50 % solids

filter

sump

HEX

H2-Fe(CN)6recomb

cell stackT=60◦ C

ZnOslurry

20 % solid

O2

CEM AEM

negativeelectrolyte

positiveelectrolyte

Zn

Zn(OH)2−4

2Fe2+

2Fe3+

Na+ Cl−

middleelectrolyte

(a) (b)(a) Alkaline zinc-ferricyanide battery, adapted from Magnani et al.1, and(b) a 2-membrane, 3-electrolyte battery, adapted from Gong et al.2

1N. Magnani et al., Sandia Report, SAND86-1266, (1987)2K. Gong et al., Energy Environ. Sci. 8, 2941 (2015)

2

Zinc-Iron Chloride: A New Approach1

• Fe3+ is safer than Br2 (positive redox electrode)• Zn has higher performance than Fe (negative plating electrode)• No ion-exchange membranes, no solids handling, no oil phase, no CN

1.84 V 1.53 V 1.21 V

Zn2+/0

Fe2+/3+

Fe2+/0

Fe2+/3+

zinc all-iron

negative:

positive:

Zn2+/0

Br−/Br2

cell:

name: bromine ironzinc --

FeCl2ZnCl2

FeCl2FeCl3ZnCl2

Zn0

Zn2+

Fe3+

Fe2+

(−) (+)porous

separator

1S. Selverston et al., J. Electrochem. Soc. 164, 1069 (2017)

3

Can We Use a Mixed ZnCl2-FeCl2 Electrolyte?

H+/H2 Fe2+/Fe3+Fe2+/Fe0Zn2+/Zn0

0 0.77-0.44-0.77

• Thermodynamics would predict iron should be electroplated beforezinc

• “Hydrogen evolution [on zinc] is greatly enhanced in the presence ofimpurities such as Fe...” (McBreen, 1984)

• “it is infeasible to use mixed electrolyte containing Zn and Fe as boththe negative electrolyte and positive electrolyte” (Xie et al., 2016)

• Does Zn2+ affect the Fe2+/3+ reaction?

4

Effect of Zinc on Fe2+/3+

• Iron redox kinetics not strongly affected1 by Zn2+

−30

−20

−10

0

10

20

30

0 0.2 0.4 0.6 0.8 1

(a)(b)

204060

3 4 5 6 7 8

(a)(b)

i/mA

cm−

2

E/VAg/AgCl

i p,a/

mA

cm−

2

ν1/2/mV1/2s−1/2

Effect of ZnCl2 on Fe (II)/(III) redox reaction on graphite electrode with1 M NH4Cl at pH=1 at ν = 10 mV s−1. Initial electrolyte (a) contained0.2 M Fe(II) and 0.2 M Fe(III), and the modified electrolyte (b)contained 0.8 M Zn(II).1S. Selverston et al., J. Electrochem. Soc. 164, 1069 (2017)

5

Using “Anomalous Codeposition (ACD)” to our Advantage

• Iron electrodeposition is strongly inhibited in the mixed electrolyte• Deposition and stripping from mixed ZnCl2-FeCl2 almost the same

as ZnCl2-only electrolyte 1

−1.2 −1 −0.8 −0.6 −0.4 −0.2 0

0.5 M ZnCl2

0.5 M ZnCl2 +0.5 M FeCl2

0.5 M FeCl2140 mA cm−2

E / VAg/AgCl

−1

−0.6

−0.2

0.2

0.6

1

−1.2 −1 −0.8 −0.6 −0.4

i ii iiii iv v

i/ip,

a

E/VAg/AgCl

1S. Selverston et al., J. Electrochem. Soc. 164 1069 (2017)

6

Hydroxide Suppression Mechanism (HSM)

• Proposed in 1965 for Ni-Fe 1,2

cathode(Zn)

Zn(OH)2

Zn2+

H+

Fe2+

[OH −]

Zn2+

OH−

Fe

Ni

-600 -700 -800

2

4

6

8

10

2

4

6

8

pH

atelectrodesurface

dep

ositionrate

/mA

cm−2

electrode potential / mV vs NHE

Zn(OH)2 + 2e −→ Zn + 2OH−

1H. Dahms, I. Croll, J. Electrochem. Soc. 112 (1965)2H. Yan, et al., J. Electrochem. Soc. 43, 1577 (1996)

7

Competitive Adsorption

• Emphasis on kinetics rather than thermodynamics1

M(II) + e− k1,M−−→ M(I)ads (1)

M(I) + e− k2,M−−→ M (2)k1,M = ko

1,M exp(b1,MV) (3)

1.0

0.5

0.0

−E/V

θ

θZn

θFe

1M. Matlosz, J. Electrochem. Soc. 140, 2272 (1993)

8

Negative Scan Limit and RPM on Glassy Carbon

• Increasing rotation rate enhances ACD, contrary to HSM1

• 0.5 M ZnCl2, 0.5 M FeCl2, 1 M NH4Cl, ν = 50 mV s−1, bulk pH=1

−40−30−20−10

010203040

−1.2 −1.1 −1 −0.9 −0.8 −0.7 −0.6 −0.5

300 rpm9001500

i/mAc

m−

2

E/VAg/AgCl

−150

−100

−50

0

50

100

150

−1.3 −1.2 −1.1 −1 −0.9 −0.8 −0.7 −0.6 −0.5

300 rpm900 rpm1500 rpm

i/mAc

m−

2

E/VAg/AgCl

-1150 mV -1250 mV

−250−200−150−100−50

050

100150200250300

−1.4 −1.2 −1 −0.8 −0.6 −0.4 −0.2 0

300 rpm900 rpm1500 rpm

i/mAc

m−

2

E/VAg/AgCl

−300

−200

−100

0

100

200

300

400

−1.6 −1.4 −1.2 −1 −0.8 −0.6 −0.4 −0.2 0

300 rpm

900 rpm1500 rpm

i/mAc

m−

2

E/VAg/AgCl

-1350 mV -1450 mV1S. Selverston et al., J. Electrochem. Soc. 164 1069 (2017)

9

Battery Proof-of-Concept

• 42 % relative gain in discharge voltage, 26 % relative gain in voltaicefficiency compared to all-iron at ± 25 mA cm−2

• Consistent with selective zinc plating via anomalous codeposition(ACD)

end

plate

graphite

plate

plastic

ribs

teflon

gasket

thickness = 2 mm

Daramic

separator

carbon

felt

A=6.25 cm2

current

collector

1/8” tubing 0

0.3

0.6

0.9

1.2

1.5

1.8

0 1 2

All-Iron ChlorideDischarge: 0.95 VV.E.: 66 %

Zinc-Iron ChlorideDischarge: 1.35 VV.E.: 83 %

cell

pote

ntial

/V

time / h

10

Accelerated Lifetime Testing

• 30-days, 175 continuous cycles achieved without replacing electrolyte• Room-temperature, no flow fields, no dendrite suppressors used• Two-hour charges at 25 mA cm−2 (loading = 50 mAh cm−2)

00.20.40.60.8

11.21.41.6

0 1 2 3 4

E/V

t/h

0

20

40

60

80

100

0 20 40 60 80 100 120 140 160 180

averages:VE: 82 %CE: 87 %EE: 71 %

efficie

ncy

/%

cycle number

V.E.C.E.E.E.

11

Conclusions

• First demonstration of a zinc-iron hybrid flow battery based usingmixed electrolytes, microporous separators

• Based on anomalous codeposition of zinc; probably not HSM• Demonstrated 30 days of cycling with over 70 % energy efficiency• Extraordinary balance between cost, safety and performance

I would like to thank:

• My advisors: Prof. Robert Savinell and Prof. Jesse Wainright• My colleagues in the EEEL Lab• The project sponsors: US DOE-OE, PNNL

12

Thank you!

12

Supplementary Material

All-Fe Losses

0.100.050.22

0.080.050.20

12

3

12

3

OCV = 1.21 V

discharge 0.87 Vcharge 1.57 V

1 Positive electrode2 Separator IR3 Negative electrode

• Adapted from 1981 study1

1Hruska and Savinell, J. Electrochem. Soc. 128, 18 (1981)

13

Zinc-Iron Chloride Conductivity

• Sufficiently high conductivity, similar to all-iron chloride battery

platinized Pt

leads

50

100

150

200

250

300

0 0.5 1 1.5 2 2.5 3

κ/m

Scm

−1

support concentration / M

NH4ClKCl

NaCl

14

XPS Analysis

0

20000

40000

60000

80000

100000

120000

140000

02004006008001000

1100

1500

1900

2300

304050607080

fe3p (Zn/Fe = 0.5)co

unts

binding energy / eV

coun

ts

binding energy / eV

XPS analysis of three samples deposited at pH=2. The sample withZn/Fe=0.5 showed a pronounced Fe3p peak and an estimated 10 %atiron.

15

Supporting Electrolyte Effects

−100

−50

0

50

100

150

−1.2 −1.1 −1 −0.9 −0.8 −0.7 −0.6 −0.5 −0.4

5 M NH4Cl

3 M NH4Cl1 M NH4Cl

i/mAc

m−

2

E/VAg/AgCl

−100

−50

0

50

100

150

−1.2 −1.1 −1 −0.9 −0.8 −0.7 −0.6 −0.5 −0.4

1 M KCl3 M KCl

Sat’d KCl

i/mA

cm−

2

E/VAg/AgCl

Effect of supporting electrolyte cations on the deposition and dissolution(Zn/Fe=4, pH=1). Total metal ion concentration CMe2+ = 1 M.Titanium sheet substrate. Room-temp, unstirred.

16

EIS during Lifetime Testing

• Color shows number of days• Oscillating resistance- not linearly increasing with time

0

0.5

1

1.5

2

2.5

3

0 0.5 1 1.5 2 2.5 3

-Im(Z

)/Oh

m

Re(Z)/Ohm

0

5

10

15

20

25

30

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

-Im(Z

)/Oh

m

Re(Z)/Ohm

0

5

10

15

20

25

30

Electrochemical impedance spectroscopy (EIS) of zinc-iron chloride cellover 30 days of operation. Scans were carried out from 10 kHz to 10 Hzafter charging.

17