Performance analysis of Lithium-ion-batteries: status and

www.iam.kit.edu/wetKIT – The Research University in the Helmholtz Association

Performance analysis of Lithium-ion-batteries: status and prospectsDPG conference Erlangen

March 2018

Ellen Ivers-Tiffée, Philipp Braun, Michael Weiss

Karlsruhe Institute of Technology (KIT), Germany

Institute for Applied MaterialsMaterials for Electrical and Electronic Engineering

DPG_Erlangen_E_Ivers-Tiffée.ppt, Slide: 2, 19.03.2018

www.iam.kit.edu/wet

MotivationRagone Diagram

100 101 102 103 104100

101

102

103

lead acid

supercaps

Ni-MH

Wgr

av /

Wh

kg-1

Pgrav / W kg-1

Li-ion

measurement

high power

Ragone diagram on cell level

high energy

high energy&

high powerD

urat

ion

Performance



www.iam.kit.edu/wet

The Lithium-Ion CellResistance Contributions

copp

er

alum

iniu

m

– +

anodeand electrolyte

cathodeand electrolyte

separatorand electrolyte

e-

Li+

e-

20-100 µm 20-100 µm20-40 µm

load



www.iam.kit.edu/wet

The Lithium-Ion Celloverpotential

Uce

ll / V

Q / As

Umin

Loss processes

• Transport (Li+, e-)

• Charge transfer (Li+)

• Diffusion (Li0)

overpotential

η

copp

er

alum

iniu

m

– +

chargetransfer

e-

Li+

e-

20-100 µm 20-100 µm20-40 µm

chargetransfer

Li+ transport

e- transport

diffusiondiffusion

e- transport

Umax

working voltage(discharge)

Open Circuit PotentialOCV = ϕcathode - ϕanode

discharge ii

I R



www.iam.kit.edu/wet

The Lithium-Ion Cellenergy- and power density

Uce

ll / V

Q / As

Umin

copp

er

alum

iniu

m

– +

chargetransfer

e-

Li+

e-

20-100 µm 20-100 µm20-40 µm

chargetransfer

Li+ transport

e- transport

diffusiondiffusion

e- transport

Umax Open Circuit PotentialOCV = ϕcathode - ϕanode

arg

arg

disch e

gravcell cell

disch e

gravdischarge cell discharge cell

ii

ii

OCV I dQenergyW

m m

OCV I dQenergy

m m

RP

t t

R

Energy- and Power-density

energy U dQ



www.iam.kit.edu/wet

Characterization of Lithium-Ion CellsElectrochemical Impedance Spectroscopy (EIS)

chargetransportelectrochemical

reactions

degradation diffusion

formation

1 µs1 ms1 s1 min1 h1 d1 M1 a10 a

f

Electrochemical Impedance Spectroscopy (EIS)

Characteristic time constants of different loss mechanisms in Lithium-ion cells:

frequency

time

1 MHz1 kHz1 Hz1 mHz 1 µHz1 nHz



www.iam.kit.edu/wet


Nyquist Plot:AC Small Signal Perturbation: Frequency Response Analyzer:

System Requirements:

linearitycausality time invariance

Kramers-Kronig Validity Test:

pert

urba

tion

t tEISt

tem

pera

ture

T

startT

endTstart endT T!

ImRe 2 2

0

' ( ')2( ) ''

ZZ d

ImIm 2 2

0

( ')2( ) ''

ZZ d

M. Schönleber, D. Klotz, and E. Ivers-Tiffée, “A Method for Improving theRobustness of linear Kramers-Kronig Validity Tests,” Electrochim. Acta, vol.131, pp. 20–27, 2014.

t

t

( )u t

( )i t

1 2

U

I

0

Z΄

-Z΄΄

12

( )ˆ ( )( ) ˆ( )

jUZ eI

'( ) ''( )Z jZ

synthetic battery-like spectrum



www.iam.kit.edu/wet


Impedance Analysis:

0 5 10 15 20 25 30 35 400

-5

-10

-15

-20Z'

' /

cm

2

Z' / cm2

Rohmic Rpolarization

What is the physical origin of these contributions?



www.iam.kit.edu/wet


Temperature variation at SoC 80 %: State-of-charge (SoC) variation at 25 °C:

How to predict the performance of lithium-ion cells?

Strong dependency of Rpolarization on SoC

0 50 100 150 200 250

0

-20

-40

-60

-80

-10030 °C 0 °C20 °C 10 °C

Z'' /

c

m²

Z' / cm²

40 °C

0 20 40 60 80 100

0

-20

-40

Z'' /

c

m²

Z' / cm²

0% 25% 50% 75% 100%

Rpolarization

Rohmic

Rpolarization

Strong dependency of Rohmic and Rpolarization on temperature



www.iam.kit.edu/wet

Modelling of Lithium-Ion Cells

Behavior Model

“Simple” Equivalent Circuit Model:no physical meaning

simple parameterization

very short computation time

R0R1

C1OCV Ucell (t)

Icell(t)

Cdiff

“Advanced” Equivalent Circuit Model

Transmission Line Models:

physically motivated

feasible parameterization

short computation time

Multiphysics Model

0

1cLi / cLi,max

Finite Element Method:

physically correct

challenging parameterization

long computation time+

+

-

+

+

+

-

-

+



www.iam.kit.edu/wet

Modeling of Lithium-Ion CellsTransmission Line Model (TLM)

[1] J. Bisquert, G. Garcia-Belmontea, F. Fabregat-Santiagoa, and A. Compteb,“Anomalous transport effects in the impedance of porous film electrodes,”Electrochemistry Commun., vol. 1, no. 9, pp. 429–435, 1999.

electrolyte

Current collector

1 1 1 1

2 2 2 2 2

1electrode thickness

A

TCT

M

CZ lV a

charge transfer

A

CT

M

Va specific charge transfer resistance

active surface area per unit volume

electrode volume

solid-state diffusion0

0 1

1Diff

diff

Diff IZC D I

l

Finite-Space Warburg Impedance

2Diff

DiffDlj

differential capacity

diffusion length

0

Di

Di

f

ff

fD

l

C

diffusion coefficient

ionic path

21 1

ion A

A

tortuosity (pore)

volume fraction of pores

electrode area

Volume fractions

Tortuosity

Active Surface Area

Particle Size

Diffusion Length

Microstructure parameters:



www.iam.kit.edu/wet

Characterization of Lithium-Ion CellsMicrostructure: Anode

Anode: Graphite X-ray nano-tomography:

X-raysource

X-raycamera

sample

M. Ender, J. Joos, A. Weber, and E. Ivers-Tiffée, “Anode microstructures from high-energyand high-power lithium-ion cylindrical cells obtained by X-ray nano-tomography,” J. PowerSources, vol. 269, pp. 912–919, 2014.

electrode thickness lelectrode = 90 µm

volume fraction εgraphite = 0.75

εpore = 0.25

tortuosity pore = 5.12

active surface area agraphite = 0.314 µm-1

particle size dgraphite,vol-av = 12.07 µm



www.iam.kit.edu/wet

Characterization of Lithium-Ion CellsMicrostructure: Cathode

Focused ion beam (FIB) tomography: Cathode: NCA/LCO Blend

electron beamGa+ ion beam

sample

e- Ga+

M. Ender, J. Joos, T. Carraro, and E. Ivers-Tiffée, “Three-dimensional reconstruction of acomposite cathode for lithium-ion cells,” Electrochem. commun., vol. 13, no. 2, pp. 166–168, 2011.M. Ender, J. Joos, T. Carraro, and E. Ivers-Tiffée, “Quantitative Characterization of LiFePO4Cathodes Reconstructed by FIB/SEM Tomography,” J. Electrochem. Soc., vol. 159, no. 7,pp. A972–A980, Jan. 2012.

electrode thickness lelectrode = 75 µm

volume fraction εAM = 0.57

εcarbon = 0.17

εpore = 0.26

tortuosity pore = 4.29

active surface area aAM = 0.73 µm-1

particle size dAM,vol-av = 4.06 µm

LiNixCoyAl1-x-yO2 = NCALiCoO2 = LCO



www.iam.kit.edu/wet

Characterization of Lithium-Ion CellsParametrization of transition line models

transmission line model (TLM)

microstructure e- conductivity impedance measurement

M. Ender, A. Weber, and E. Ivers-Tiffée, “A novelmethod for measuring the effective conductivityand the contact resistance of porous electrodesfor lithium-ion batteries,” Electrochem. commun.,vol. 34, pp. 130–133, 2013.

M. Ender, J. Joos, T. Carraro, and E. Ivers-Tiffée,“Quantitative Characterization of LiFePO4Cathodes Reconstructed by FIB/SEMTomography,” J. Electrochem. Soc., vol. 159, no.7, pp. A972–A980, Jan. 2012.

ionVa

pse electronicconductivity

volume fractions

tortuosity

electrode thickness

active surface area

particle sizel

conclusion

Electrochemical and microstructure parameters for anode and cathode are

determined

0 5 10 15 20 25 30 35 400

-5

-10

-15

-20

Z'' /

c

m2

Z' / cm2



www.iam.kit.edu/wet

Li-Ion Battery ModelHomogenized 1D Model

charge-transferohmiclosses

anode

diffusion

-Z''

/ c

m2

Z' / cm2

calculated impedance

cathode

Cur

rent

col

lect

or

22222χ1χ1χ1χ1

2 2 2 2 2χ1 χ1 χ1 χ1 χ1 SepC

urre

nt c

olle

ctor

χ1anodeZ cathodeZ

“Advanced” Equivalent Circuit Model:

Impedance: Discharge curve: Ragone Plot:

cathode thickness anode thickness separatorthickness

Uce

ll / V

Q / As 100 101 102 103 104100

101

102

103

lead acid

supercaps

Ni-MH

Wgr

av /

Wh

kg-1

Pgrav / W kg-1

Li-ion

measurement model calculation

Open Circuit PotentialOCV = ϕcathode - ϕanode

energy U dQ

total anode Sep cathodeZ Z R Z



www.iam.kit.edu/wet

Model-Based Optimization

energy density

power density

cycle durability calendar life

costs

fast-charging capability

safety

LFP NMC NCA LMO

0 100 200 300 400 1000 2000 3000 40000

1

2

3

4

5

LTO

graphite

Si

LMO

LFP

LCO

NCA

NMC

pote

ntia

l / V

spec. capacity / mAh g-1

HV Spinell

Energy Optimization Relevant Design FactorsExample: cathode active materials

cathode anodeenergy dQ

y-axis x-axis

LFP: LiFePO4NMC: LiNi1-x-yCoxMnyO2HV Spinell: 0.3Li2MnO3 0.7LiNi0,5Mn1,5O4

NCA: LiNi1-x-yCoxAlyO2LMO: LiMn2O4



www.iam.kit.edu/wet

Model-Based Optimization change parameters of “state-of-the-art” active materials

Increase capacityReduce mass

Anode Cathode Passive Components

High EnergyOptimization

High PowerOptimization

Increase surface area

Reduce electrode thickness

Specific Capacity:

graphite

graphite + 15% Si

+ 30%

Specific Capacity:LCO / NCA

NMC811

+ 40%

Volume Fractions:57% 70%

+ 20%

(350 mAh g-1)

(450 mAh g-1)

(145 mAh g-1)

(200 mAh g-1)

NCA/LCO NMC811

current collectors:

lcc = 20 µm

lcc = 10 µm

- 50%

Thickness:

90 µm 40 µm

Thickness:

75 µm 30 µm- 55% - 60%

Particle Size:

4.1 µm 0.5 µm- 88%

Particle Size:

12.1 µm 0.5 µm- 95% +

all high energy optimization steps



www.iam.kit.edu/wet


100 101 102 103 104100

101

102

103

Li-ion

lead acid

supercaps

Ni-MH

Wgr

av /

Wh

kg-1

Pgrav / W kg-1

measurement high energy optimization high power optimization




www.iam.kit.edu/wet


101 102 103

102

2x102

3x102W

grav

/ W

h kg

-1

Pgrav / W kg-1

measurement high energy optimization high power optimization

BMW i3

Tesla Model S

Renault Zoe

consumer cell

250 Wh kg-1

175 Wh kg-1

240 Wh kg-1




www.iam.kit.edu/wet

State-of-the-arte electric vehiclescomparison from cell to car

250 Wh kg-1240 Wh kg-1

175 Wh kg-1

150 Wh kg-1

139 Wh kg-1 146 Wh kg-1133 Wh km-1

181 Wh km-1

131 Wh km-1

Renault Zoe

Tesla Model SBMW i3

0

50

100

150

200

250

300

Wgr

av /

Wh

kg-1

Renault Zoe

Tesla Model SBMW i3

0

50

100

150

200

250

300

Wgr

av /

Wh

kg-1

Renault Zoe

Tesla Model SBMW i3

0

50

100

150

200

250

cons

umpt

ion

/ Wh

km-1

Energy-Density: Cell-Level Energy-Density: Battery-Level ConsumptionN

MC

622

NC

A

NM

C11

1

pouchcells

cylindricalcells

prismaticcells

192 7104 96



www.iam.kit.edu/wet

Electric Range Prediction

State-of-the-art Battery

200 km

800 km(calculated by volume*)Future Prospects



www.iam.kit.edu/wet

Acknowledgements:Research Partners and Team IAM-WET

Documents

Performance analysis of Lithium-ion-batteries: status and