carlo segre - csrri.iit.edusegre/presentations/2014_iit_mmae.pdf · potential, energy density, and...

$Page 1: Carlo Segre - csrri.iit.edusegre/presentations/2014_iit_mmae.pdf · Potential, energy density, and power determined by the chemistry Illinois Institute of Technology IIT { MMAE/Chemistry$

Making a nanoelectrofuel flow battery

Carlo Segre

Physics Department&

Center for Synchrotron Radiation Research and InstrumentationIllinois Institute of Technology

November 19, 2014

Illinois Institute of Technology IIT – MMAE/Chemistry Colloquium November 19, 2014 1 / 33

Outline

• Batteries 101

• Nanoelectrofuel concept

• Prototype design

• Electrochemical characterization

• EXAFS studies

• Conclusions

Solid state batteries

Anode - negatively chargedelectrode

Cathode - positively chargedelectrode

Separator - allows ions to passwithout short circuit

Electrolyte - medium throughwhich ions move

Consider a Li-ion battery

+

e-

Charge - Li+ ions move from cathode to anode and electrons also flow to theanode externally, anode is reduced

Discharge - Li+ ions move back to cathode and electrons flow through theexternal load, anode is oxidized

Potential, energy density, and power determined by the chemistry

+

e-

+

e-

+

e-

+

e-

Common solid state battery chemistries

Lead-acid battery: Eoc = 2.05 VCathode: PbO2 + SO4

2− + 4H+ + 2 e− ←→ Pb2SO4 + 2 H2OAnode: PbSO4 + 2 e− ←→ Pb + SO4

2−

NiMH battery: Eoc = 1.28 VCathode: NiOOH + H2O + e− ←→ Ni(OH)2 + OH−

Anode: M + H2O + e− ←→ MH + OH−

Li-ion battery: Eoc = 4.00 VCathode: CoO2 + Li+ + e− ←→ LiCoO2

Anode: Li+ + C6 + e− ←→ LiC6

Characteristics

• Medium to high energy density

• Limited cycle life (<1000)

• Large packaging overhead

Flow batteries

Characteristics

• Low packaging overhead

• Unlimited cycle life

• Low energy density

Vanadium: Eoc = 1.26 VCathode: V3+ + e− ←→ V2+

Anode: VO2+ + 2 H+ + e− ←→ VO2+ + H2O

Zinc-Bromine: Eoc = 1.67 VCathode: Br2(aq) + 2 e− ←→ 2 Br−(aq)Anode: Zn2+

(aq) + 2 e− ←→ Zn(s)

Nanoelectrofuel battery

Suspended electroactive nanoparticles

Advantages of flow batteries

Energy density of solid state

Chemistry agnosticaqueous or non-aqueous

3 year funded program

Prototype: 1 kWh total energy stored40 V, C/3 discharge rate

Develop commercialization plan

Nanoelectrofuel challenges

• What is the intrinsic performanceof active materials innanoparticle form?

• Can suspended nanoparticles beeffectively charged anddischarged during flow?

• How much loading can bestabilized in suspension?

• Will these nanoelectrofuels bepumpable and not destroy theenclosure materials?

• Can physics graduate studentson the project get a Ph.D. doingthis very applied project?

40 V aqueous chemistry stack

25 kWh using 4.5 L of nanoelectrofuel

26 kg stack, 10 kg 50% loaded fluid

70 Wh/kg (compare to 40 Wh/kg forPb-acid)

Charging & discharging nanoelectrofuel

Charging and discharging in a flow can be achieved byproper design of the electrode but all these ideas haveto be validated through computation and experiment.

• Porous electrodefor high contactprobability

• Turbulent flowto maximizeelectrodecontact

• Moderatepressure dropacross the cell

• Must haveelectron transferwith transientcontact

First charging results

December 2012 data comparing x-ray absorption spectroscopy results on Cu6Sn5

anode material in a coin cell and flowing through a metal frit.

0 2 4

R (Ao

)

0

0.2

0.4

0.6

0.8

χ(R

) [A

o-3

]

lithiated @ 0.5V

unlithiated

lithiated @ 0.0V

lithiation

coin cell

0 2 4

R (Ao

)

0

0.1

0.2

0.3

0.4

0.5

0.6

χ(R

) [A

o-3

]

1 hr @ 0.01V vs. Li/Li+

unlithiated

1 hr @ 0.23V vs. Li/Li+

lithiation

flow cell

Similar trends indicate that nanoparticles in the flow cell are charging, albeitslowly and inefficiently.

Initial prototype cell

Made from metal with machined postsfor increased contact area

Future designs manufactured with 3Dprinting & metal electrode inserts

CFD modeling

Inject 5000 particles and evolve for 15 s

Extend to repeated injections of 5000 particles and add electrochemical modelingIllinois Institute of Technology IIT – MMAE/Chemistry Colloquium November 19, 2014 11 / 33

Initial CFD results

Pressure field −→

Velocity field ↑Illinois Institute of Technology IIT – MMAE/Chemistry Colloquium November 19, 2014 12 / 33

Nanofluid settling

TiO2 pristine (left & center) and sulfonated (right) in 0.4M KOH/LiOH (left) and0.04M KOH/LiOH (center & right)

0 hours

45 hours

18 hours

1 month

Nanoparticle electrochemical performance

LiNixMnyCozO2 non-aqueous cathode material

As received micron-sized particles (MTI Inc.)

LiNixMnyCozO2 non-aqueous cathode material

Ball milled ∼400 nm particles (MTI Inc.)

0 10 20 30 40 50 60Cycle #

0

50

100

150

200

250

Specific

Capacity [m

Ah/g

]

C/10

C/3

C/4

MTI as-received (2.5V - 4.3V)

MTI ball milled (2.5V-4.6V)

MTI ball milled (2.5V-4.3V)

Nanoparticle-sizedLiNixMnyCozO2 haslower capacity and morefading

Cycling to 4.6 V yieldsslightly higher initial ca-pacity but faster fading

Solid electrolyte inter-face (SEI) layer is asignificant problem athigh potentials and fornanoparticles

Initial nanofluid charging tests

Fe2O3 coin cell

Hydrogen evolution atpotentials below -1.0V

Fe2O3 cyclic voltam-metry shows Li interca-lation

Fe2O3 coin cell

Theoretical capacity forFe+3 →Fe+2 is ∼335mAh/g

Fe2O3 nanofluid

5% wt suspension ofFe2O3 nanoparticles inKOH/LiOH solution

Performance ofnanofluid equivalent orto solid nanoparticleelectrode

Capacity increase withcycles indicates that itis limited by suboptimalcurrent collector

Need to move to flow-through current collec-tor design

Fe2O3 nanofluid – overcharged

5% wt suspension ofFe2O3 nanoparticles inKOH/LiOH solution

The EXAFS equation

The EXAFS oscillations can be modelled and interpreted using a conceptuallysimple equation (the details are more subtle!)

χ(k) =∑j

NjS20 fj(k)

kR2j

e−2k2σ2j e−2Rj/λ(k) sin [2Rj + δj(k)]

The sum could be over shells of atoms (Pt-Pt, Pt-Ni) or over scattering paths forthe photo-electron.

fj(k): scattering factor for the pathλ(k): photoelectron mean free pathδj(k): phase shift for the jth path

Nj : number of paths of type jRj : half path lengthσj : path “disorder”

Ni

Pt

Ni Pt

EXAFS analysis

11500 12000 12500

E(eV)

-2

-1.5

-1

-0.5

0

0.5

ln(I

o/I

)

remove background

andapply k-weighting

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

extract structural pa-rameters for first shell ormore distant atoms asappropriate

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

take Fourier Transform

and fit with a structuralmodel

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

remove background

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

remove background

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

remove background

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

remove background andapply k-weighting

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

extract structural pa-rameters for first shell

ormore distant atoms asappropriate

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

take Fourier Transformand fit with a structuralmodel

EXAFS analysis

11500 12000 12500

E(eV)

0

0.5

1

ln(I

o/I

)

0 5 10 15

k(Å-1

)

-0.2

-0.1

0

0.1

0.2

χ

0 2 4 6

R(Å)

0

5

10

15

20

25

30

χ(R

)

take Fourier Transformand fit with a structuralmodel

Synthesis of Sn-graphite nanocomposites

One-pot synthesis pro-duces evenly distributedSn3O2(OH)2 nanoparticleson graphite nanoplatelets

XRD shows a small amountof Sn metal in addition toSn3O2(OH)2

10 20 30 40 50 60 70

0

100

200

300

Inte

nsity

Scattering Angle (2 )

GnP

Sn3O2(OH)2

Sn3O2(OH)2/GnP

In situ XAS studies of lithiation

1 2 3 4 50.0

0.3

0.6

0.9

1.2Sn-O

|x

(R)|

(Å-3

)

R (Å)

OCV 1st Charge 1st Discharge

Sn-Sn

1 2 3 4 50.0

0.1

0.2

In situ battery box

Pouch cell clamped against front window in helium environment

In situ battery box

Suitable for both transmission and fluorescence measurements

0 5 10 15 20 25 300

100

200

300

400

500

600

700

2100

2800

In situ

C

apac

ity (m

Ah/

g)

Cycle Number

Charge Discharge

Ex situ

-0.5

0.0

0.5

1.0

0 2 4 6 8 10 12

-0.5

0.0

0.5

1.0

0 1 2 3 4 50.0

0.5

1.0

k (Å-1)

k 2

(k) (Å-2

)

Re[

(R)]

(Å-3

)

|x(R

)| (Å

-3)

R (Å)

Fresh electrode can be fitwith Sn3O2(OH)2 structurewhich is dominated by thenear neighbor Sn-O dis-tances

Only a small amount ofmetallic Sn-Sn distances canbe seen

-0.5

0.0

0.5

1.0

0 2 4 6 8 10 12

-0.5

0.0

0.5

1.0

0 1 2 3 4 50.0

0.5

1.0

k (Å-1)

k 2

(k) (Å-2

)

Re[

(R)]

(Å-3

)

|x(R

)| (Å

-3)

R (Å)

-0.5

0.0

0.5

1.0

0 2 4 6 8 10 12

-0.5

0.0

0.5

1.0

0 1 2 3 4 50.0

0.5

1.0

k (Å-1)

k 2

(k) (Å-2

)

Re[

(R)]

(Å-3

)

|x(R

)| (Å

-3)

R (Å)

-0.2

0.0

0.2

0 2 4 6 8 10 12

-0.1

0.0

0.1

0 1 2 3 4 50.0

0.1

0.2

k (Å-1)

k 2

(k) (Å-2

)

Re[

(R)]

(Å-3

)

|(R

)| (Å

-3)

R (Å)

Reduction of number of Sn-O near neighbors and 3 Sn-Li paths characteristic of theLi22Sn5 structure

-0.2

0.0

0.2

0 2 4 6 8 10 12

-0.2

-0.1

0.0

0.1

0.2

0 1 2 3 4 50.0

0.1

0.2

k (Å-1)

k 2

(k) (Å-2

)

Re[

(R)]

(Å-3

)

|(R

)| (Å

-3)

R (Å)

Metallic Sn-Sn distances ap-pear but Sn-Li paths are stillpresent, further reduction inSn-O near neighbors.

OCV1s

t Cha

rge1s

t Disc

harge

2nd C

harge

2nd D

ischa

rge3rd

Cha

rge3rd

Disc

harge

4th C

harge

0

2

4

6

8

10

12

14

Med. Li

Long Li

Short Li

C/7.5C/2.5

Num

ber o

f Nei

ghbo

rs

Total Li

Number of Li nearneighbors oscil-lates with thecharge/dischargecycles but neverreturns to zero

In situ cell promotesaccelerated aging be-cause of Sn swellingand the reducedpressure of the thinPEEK pouch cellassembly

OCV1s

t Cha

rge1s

t Disc

harge

2nd C

harge

2nd D

ischa

rge3rd

Cha

rge3rd

Disc

harge

4th C

harge

0

2

4

6

8

10

12

14

Med. Li

Long Li

Short Li

C/7.5C/2.5

Num

ber o

f Nei

ghbo

rs

Total Li

OCV1s

t Cha

rge1s

t Disc

harge

2nd C

harge

2nd D

ischa

rge3rd

Cha

rge3rd

Disc

harge

4th C

harge

0

2

4

6

8

10

12

14

Med. Li

Long Li

Short Li

C/7.5C/2.5

Num

ber o

f Nei

ghbo

rs

Total Li

Acknowledgements

Illinois Institute of Technology• John Katsoudas – MRCAT staff

• Vijay Ramani - Chemical Engineering

Argonne National Laboratory• Elena Timofeeva – Energy Systems Division

• Dileep Singh – Energy Systems Division

• John Zhang – Chemical Sciences & Engineering Division

Graduate Students• Chris Pelliccione – Ph.D. Physics

• Yujia Ding – Ph.D. Physics

• Yue Li – Ph.D. Chemical Engineering

• Nathaniel Beaver – Ph.D. Physics

• Shankar Aryal – Ph.D. Physics

Supported by the DOE ARPA-e programIllinois Institute of Technology IIT – MMAE/Chemistry Colloquium November 19, 2014 32 / 33

Abstract

We are currently in the first year of an ARPA-e project to pro-duce a prototype nanoelectrofuel flow battery. This new bat-tery concept marries the traditional solid state battery with aflow battery to obtain higher energy densities and reduction inpackaging weight. Successful development of this new batteryformat requires the ability to charge and discharge nanoparti-cle suspensions by transient contact with the current collectors.While the basic effect has been demonstrated, many challengeslie ahead. Notably the ability to make efficient and high capac-ity battery materials in nanoparticle form. In order to under-stand the differences between battery materials in macroscopic(micron-sized) and nanoparticle form, we are using x-ray ab-sorption spectroscopy to probe the structure of materials asthey are electrochemically cycled. I will present some initial re-sults on our in-situ studies of anode lithiation and discuss ourfuture plans.

carlo segre - csrri.iit.edusegre/presentations/2014_iit_mmae.pdf · potential, energy density, and...

Documents