Cell Analysis – High-Energy Density Cathodes and Anodes

Thomas J. RichardsonLawrence Berkeley National Laboratory

June 8, 2010

This presentation does not contain any proprietary, confidential, or otherwise restricted information

LBNL

Project ID: ES053

OVERVIEWTimeline• PI Joined BATT 2001

• Cathodes Task Started 2001

• Anodes Task Started 2006

Barriers Addressed• Available Energy

• Cycle life

• Abuse tolerance

Budget• FY09 $475K

• FY10 $530K

Partners• Collaborations: Grey (Stony Brook)

Kostecki, Srinivasan, Doeff, Cabana (BATT-LBNL), Radmilovic (NCEM), Kunz, Rotenberg (ALS), Misra (SSRL, APS)

• Interactions: Zaghib (HQ)

• Project lead: John Newman

OBJECTIVES

• Synthesize and evaluate new electrode materials with improved energy density.

• Investigate the relationships of structure, morphology and performance of cathode and anode materials.

• Explore kinetic barriers and utilize the knowledge gained to design and develop electrodes with improved energy density, rate performance and stability.

MILESTONESJune 2009 Report rate and cycling performance of Li alloy and/or

intermetallic electrodes with capacity exceeding 500 mAh/g.

July 2009 Report on mechanisms governing LiMnPO4performance and measures to improve utilization and rate.

September 2009 Report experimental results on new cathode materials.

March 2010 Report capacities and charge-discharge potentials for new cathode materials.

July 2010 Report on reduction of anode irreversible capacities.

APPROACH• Identify candidate cathode compositions by systematic

analysis of phase diagrams and literature reports.

• Synthesize novel materials and/or unique structures and employ XRD, electron microscopy, vibrational spectroscopies, and electroanalytical techniques to determine their applicability to BATT goals.

• Develop and test strategies to reduce capacity losses in conventional and alternative anodes.

• Explore kinetic barriers, and utilize the knowledge gained to design and develop electrodes with improved energy density, rate performance and stability.

• Beam size can be as small as 1 μm x 1 μm

(b)Visualization of Charge Distribution in an Electrodeby Synchrotron Microdiffraction

with Martin Kunz, Kai Chen and Nobumichi Tamura, Advanced Light Source, LBNL

TECHNICAL ACCOMPLISHMENTS

16 18 20 22 24 26 28 30 32 34 36 38

301

301

020

020

200

LFP+FP

FP

Inte

nsity

(arb

itrar

y un

its)

Cu Kα 2θ (°)

LFP

200

LFP 200

FP 200

LFP 020

FP 020

LFP 301

FP 301

020 peaks were used to determine the phase distribution, which is “frozen” due to the two-phase nature of the LiFePO4 system

XRD patterns of LiFePO4, FePO4 and a 1:1 mixture

Stereographic XRD patterns collected by CCD at 6 keV

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

(b)12.5 mm diameter Swagelok cell electrodes

0 100 200 300 4002.4

2.6

2.8

3.0

3.2

3.4

3.6

3.8

4.0

4.2

I = 20 mA/gI = - 20 mA/g

Volta

ge (V

)

Capacity (mAh/g)

I = 20 mA/g

Slow Charging

0 100 200 300 4002.42.62.83.03.23.43.63.84.04.24.4 Fast Charging

Volta

ge (V

)

Capacity (mAh/g)

I = 3 A/g

I = - 20 mA/gI = 20 mA/g

Both cells charged to 50% of first discharge capacity

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

(b)(a) (b)

Electrode cross-section is ~ 50 μm Fe fluorescence locates the active layer

XRD data collected from cross section at 2 μm intervals

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

Phase distribution in electrode cross-section after slow charging is quite uniform

0 10 20 30 40 500

10

20

30

40

50

60

FePO

4 con

cent

ratio

n (%

)

Distance (µm)

(a) Vertical scan

0 10 20 30 40 500

10

20

30

40

50

60(b) Horizontal scan

FePO

4 con

cent

ratio

n (%

)Distance (µm)

20 mA/g charge

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

(b)

0 10 20 30 40 50

30

40

50

60

70(a)

FePO

4 con

cent

ratio

n (%

)

Distance (µm)

0 10 20 30 40 5055

60

65

70

75(b)

FePO

4 con

cent

ratio

n (%

)Distance (µm)

Phase distribution in electrode cross-section after fast charging is uniform in-plane

but varies dramatically from top to bottom

Top surface

Near current collector

3000 mA/g charge

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

Liquid Phase Transport Model (V. Srinivasan, LBNL)Porous electrode model including electrolyte transport properties,

thickness, porosity, and tortuosity used to predict the charge distribution

• Model predicts data well• Slight overprediction of the impact of electrolyte transport at higher rates

• Data provides a means of tuning the model, especially for lower porosities- Future work

C/10 charge 20C charge

Data from Liu and RichardsonModel prediction

Data from Liu and RichardsonModel prediction

TECHNICAL ACCOMPLISHMENTS

Large polarization when charged at high current density

(b)4 cm x 4.5 cm pouch cell electrode

0 50 100 150 200 250 300 3502.42.62.83.03.23.43.63.84.04.24.4

Volta

ge (V

)

Capacity (mAh/g)

I = 2 A/g

I = - 20 mA/gI = 20 mA/g

45 mm

Si

40 mm

TECHNICAL ACCOMPLISHMENTSVisualization of Charge Distribution

The charge distribution in lithium battery electrodes has been visualized for the first time using synchrotron microdiffraction

For swagelok-type cells, the non-uniform distribution of the delithiated phase (FePO4) in the cross-section of an electrode charged with high current density is due to electrolyte polarization in the pores of the electrode

Calculations by V. Srinivasan agree well with the observed distribution

For a rectangular electrode in a pouch cell configuration, the non-uniform charge distribution is due to electronic resistance in the Al substrate and porous composite

Charge Distribution by Microdiffraction

SUMMARY – CELL ANALYSIS

APPROACH - ANODES

Merits Good cyclabilityRelatively low cost

DemeritsLimited capacity: Graphite (LiC6: 372 mAh/g)Slow rate capabilityPoor initial coulombic efficiency SEI layer formation

SpeciesGraphite (MCMB, Natural graphite), Soft carbon, and etc.

Carbon anodesMerits

High capacity: Li4.4Si (4100 mAh/g) & Li4.4Sn (990 mAh/g) Good rate capabilityLess safety problems

DemeritsPoor cyclability Volume expansionPoor initial coulombic efficiency SEI layer formation

SpeciesSn, Si, and etc.

Alloying AnodesAlloying anodes

Si SnGraphitea

b

a

b

a

c

b

Lithium nitride is a convenient and inexpensive source of lithium for reduction of irreversible capacity losses in:

Solid solution alloys:2Li3N + 3Mg → 6Li + Mg3N2

2Li3N + (3+x)Mg → Li6Mgx + Mg3N2

Intermetallics:

Li3N + Al → 3Li + AlN4Li3N + 4Al → 3LiAl + AlN

Li3N + xSi(Sn) → 3LiSi(Sn)0.33x + ½N2

Carbon anodes:Li3N + Al + 18C→ 3LiC6 + AlN

APPROACH - ANODES

Prelithiated Sn/C composites

20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

: Li2Sn5

**

[ (Li3N+3Sn)_70wt% + C_30wt%]

: Be (window)*

: Li13Sn5

: Li7Sn3

Inte

nsity

(a.u

.)

2θ (degree)

: Sn

TECHNICAL ACCOMPLISHMENTS

Prelithiated Si synthesized from Li3N & SiTECHNICAL ACCOMPLISHMENTS

20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 502θ (degree)

Inte

nsity

(a.u

.)

: Si3N4

: Li13Si4: Li14Si6

: Li12Si7

BeBe

BeBe

Be

Li3N·xSi (x=3)

Li3N·xSi (x=2)

Li3N·xSi (x=1) : SiBe

Li3N + xSi (x=1,2,3)

Si + LiySiz (Li12Si7, Li14Si6 , Li13Si4) + Si3N4

0 500 1000 1500 2000 25000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

- Voltage range: 0-1.5 V- Current rate: 100 mA/g- OCV: 1.4 V- 1st Coulombic efficiency: 73 %

Pure Si (no binder & no conducting agent)5 2 1

Po

tent

ial (

V vs

. Li+ /L

i)

Capacity (mAh/g)

0 200 400 600 800 1000 1200 1400 1600 1800

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6 Li3N + 3Si (1:3 mol)

- Voltage range: 0 - 1.5 V- Current rate: 100 mA/g- OCV: 0.51 V- 1st Coulombic efficiency: 86 %

5 2 1

Po

tent

ial (

V vs

. Li+ /L

i)

Capacity (mAh/g)

20 25 30 35 40 45 50

Si3N4

Li12Si7Li13Si4

Inte

nsity

(a.u

.)

2θ (degree)

Be

Be

[(Li3N+3Si)_70wt% + C_30wt%]

Si

TECHNICAL ACCOMPLISHMENTSPrelithiated Si/C composites

SUMMARY - ANODES

Reactions of lithium nitride with Sn and Si are simple, safe, effective

Carbon can be added before reaction, products are free-flowing powders

Prelithiated Sn & Si electrodes have low open circuit voltages

Stable SEI formation with little or no irreversible capacity losses

Small volume changes enhance cycle life

Prelithiated Anodes

APPROACH - CATHODES• Explore alternative (non-olivine) phosphates of Co, Cu

• Prepare Na+, K+, NH4+ phosphates

• Convert to Li+ phosphates by ion exchange

• Utilize Cu(II)/Cu(III) couple?

Olivine LiCoPO4 α-NaCoPO4 NaCuPO4

• Ion exchange at 160 °C in hexanol, small lattice shrinkage• ~ 0.3 Li was extracted at 5 V during the first charge• Only 0.08 Li was inserted during discharge• No significant high voltage capacity on subsequent cycles

(b)LixNa1-xCoPO4 from α-NaCoPO4

2.0 2.5 3.0 3.5 4.0 4.5 5.0

0.0

0.2

0.4

0.6Li0.3Na0.7CoPO4 synthesized by RIE route

I (m

A)

E (V vs. Li/Li+)

2nd cycle

1st cycle

4.93 V

TECHNICAL ACCOMPLISHMENTS

“LiCuPO4”

10 20 30 40 50 60 70

LiCuPO4 synthesized at 700 °C

Inte

nsity

(arb

itrar

y un

it)

Cu Kα 2θ (degree)

10 20 30 40 50 60 70

Inte

nsity

(arb

itrar

y un

it)Cu Kα 2θ (degree)

LiCuPO4 synthesized at 850 °C

TECHNICAL ACCOMPLISHMENTS

(a) (b)

Electrochemical behavior of “LiCuPO4”(b)

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5LiCuPO4 synthesized at 700 °C

I (m

A)

E (V vs. Li/Li+)

2nd cycle

1st cycle

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5 LiCuPO4 synthesized at 850 °C

2nd cycle

I (m

A)E (V vs. Li/Li+)

1st cycle

TECHNICAL ACCOMPLISHMENTS

• Ex-situ XRD indicates that LiCuPO4 decomposes during the first discharge

• The redox couple in second cycle appears to be Li1.5CuO2 and Li1.5-xCuO2

SUMMARY - CATHODESNon-olivine Phosphates

• NaCuPO4, KCuPO4, NH4CuPO4 decomposed during ion exchange

• Direct electrochemical extraction of Na+ did not occur below 5.5 V

• Attempts to prepare LiCuPO4 by solid state synthesis resulted in adifferent phase from that reported by Amine (US Patent 6,319,632)

• The new phase had no charge capacity below 5 V

• The new phase decomposed to Li1+xCuO2 and Li3PO4 on lithiation

COLLABORATIONS• Grey, Cabana (Stony Brook): NMR studies of solid

solutions in iron and manganese phosphates

• Kostecki (LBNL): Raman and FTIR characterization of materials and electrodes

• Srinivasan (LBNL): visualization and modeling of current and charge distribution in composite electrodes

• Chen, Doeff, Cabana (LBNL) and Misra, Toney (SSRL): in situ X-ray diffraction and absorption

• Radmilovic, Lee et al. (NCEM): High resolution TEM of anode and cathode materials

• Kunz, Tamura, Rotenberg (ALS): XAS and XPS of anode materials

Each of these collaborations has already or soon will lead to a major publication.

31

FUTURE WORKCELL ANALYSIS

• Extend charge distribution study to electrodes with a range of different compositions, thicknesses, and porosity.

• Extend study to electrodes salvaged from commercial cells.

ANODES

• Identify binders and solvents that are compatible with partially prelithiated anode materials.

• Establish optimum prelithiation levels for tin, silicon, graphite.

CATHODES

• Continue to search for alternative cathode materials with high energy density, low cost, and safety.

• Increase electrode loading of conventional materials through improved microstructures and intrinsic conductivity.

PROJECT SUMMARYCELL ANALYSIS

• A new technique was developed to visualize charge distribution in composite electrodes.

• The observed distributions in plane and in cross section are consistent with expectations from modeling.

ANODES

• Partial prelithiation of anodes leads to reduced capacity losses, more efficient cycling.

• Excellent results for tin and silicon anodes.

CATHODES

• Continued to search for non-olivine phosphate cathode materials with high energy density, low cost, and safety.

• In common with all other groups, the challenge remains.