Ji-Guang Zhang

Pacific Northwest National Laboratory Richland, Washington, U.S.A.

The 4th Symposium on Energy Storage: Beyond Lithium Ion

June 8, 2011

1

Stable Operation of Li-Air Batteries

Outline

2

1. Development of High Capacity Primary Li-air Batteries

2. Stability of Li-air Batteries Using Aqueous Electrolyte & NASICON Glass

3. Stability of Li-air Batteries Using Non-Aqueous Electrolyte 3.1 Reaction Products During Discharge Process

3.2 Reaction Products During Charge Process

4. Summary

3

0

0.4

0.8

1.2

1.6

2

2.4

2.8

3.2

3.6

0 0.2 0.4 0.6 0.8 1 1.2 1.4

Volta

ge (V

)

Cell capacity (Ah)

Operated in ambient air (~20% RH) for 33 daysTotal weight of the complete battery: 8.387 gSpecific energy: 362 Wh/kg

2340 mAh/g carbon

0.8 mil polymer membrane

Metal mesh 0.7 mm KB carbon electrode

0.5 mm Li foil1 mil separator with binding layer

Cu mesh

+-

Footprint: 4.6 cm x 4.6 cm; thickness = 3.8 mm

1. Development of High Capacity Primary Li-air Batteries

4

Hierarchically Porous Graphene as a Lithium-Air Battery Electrode

a and b, SEM images of as-prepared graphene-based air electrodes c and d, Discharged air electrode using FGS with C/O = 14 and C/O = 100, respectively. e, TEM image of discharged air electrode. f, Selected area electron diffraction pattern (SAED) of the particles: Li2O2.

Jie et al, to be published

5

Graphene as a Lithium-Air Battery Electrode Record Capacity of 15,000 mAh/g

6

Challenges: Stable glass electrolyte at strong acid and alkaline environment

Chemical stability Electrochemical stability Single ion conductivity

Air Electrode

Li+ Glass

Load

Li+

e-

Li Basic-electrolyte

¼ O2

e-

Final product stays in air electrode

Li+ ½ O2 + ½ H2O LiOH

½ H2O

LiOH

Li+ stable coating

Li Anode

In Aqueous Electrolyte

2. Stability of Li-air Batteries Using Aqueous Electrolyte & NASICON Glass

7

Advantages and Challenges in Li-air Batteries

Li/O2 Reaction in Different Electrolyte

Theoretical voltage

Theoretical specific energy based on metal

Theoretical specific energy

based on reactants

(excluding O2)

Theoretical specific

energy based on reaction

products

Specific energy based on full reaction

V Wh/kg Wh/kg Wh/kg Wh/kg With precipitation

(Li/electrolyte/carbon)

Li + ½ O2 ↔ ½ Li2O2 3.10 11972 11972 3622 2790**

Li + 0.25 O2+ 1.5 H2O ↔ LiOH.H2O 3.4-3.86* ~12000 2717 2198 1500

Li + 0.25O2 + HCl ↔LiCl +0.5H2O 3.86-4.44* ~12000 2637 2227 1607

With no precipitation (Li & electrolyte)

Li + 0.25 O2+ 11.14 H2O ↔ LiOH+10.64H2O 3.4-3.86* ~12000 444 428 444

Li + 0.25O2 + HCl +2.29H2O↔LiCl +2.79H2O 3.86-4.44* ~12000 1352 1236 1353

**J. P. Zheng et al, J. Electrochem. Soc., 158 (1), A43- A46 (2011).

* Voltage is a function of pH value: V = 4.268 – 0.059 pH

8

Voltage in Li-Air Batteries Strongly Depends on pH Value of Aqueous Electrolyte

3.0

3.2

3.4

3.6

3.8

4.0

4.2

4.4

4.6

4.8

5.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

pH

Vol

tage

(V

)

Strong base (LiOH)electrolyte (1M, 3.4V)

Strong acid (HCl) electrolyte (1M, 4.21 V)

Neutral electrolyte(3.855 V)

V(x)* = 4.268 − 0.059 pH (V)

*Xia et al, NATURE CHEMISTRY, VOL 2, SEPTEMBER , 760 (2010)

9

LiOH １週間

HCl ３週間

1.8 ㎛×10ｋ未処理

H2O 半年超

LiCl ３週間

酸性

1.02.0

3.0

4.05.0

6.07.0

8.09.0

10.011.0

12.013.0

14.0

塩基性

LiNO3 ３週間

塩基性 ： Li3PO4の結晶析出

酸 性 ： 表面が溶出

basic

acidic

pristine LiCl(aq.)

LiNO3(aq.)

LiOH

HCl

H2O

Yamamoto et al, presentation in Symposium on Energy Storage Beyond Lithium Ion，ORNL, Oct, 2010 https://www.ornl.gov/ccsd_registrations/battery/presentations/Session6-840-Yamamoto.pdf

Stability of LTAP Glass vs. pH

3.00

3.20

3.40

3.60

3.80

4.00

4.20

4.40

4.60

4.80

5.00

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0HCl (M)

Vol

tage

(V

)

Strong acidic (HCl) electrolyte(11. 5 M, 4.33 V)

Li + 0.25O2 + HCl +2.29H2O↔LiCl +2.79H2O

7 0 pH

Neutral electrolyte0 mol HCl

pH = 7V = 3.855 V

1 mol HClpH = 04.268 V

0% 100%8.7%

Soluability limit: 1353 Wh/kg Glass Stability limit:

118 Wh/kg

Capacity (%)

10

Limitation of Li-air Batteries in Acid Electrolyte

To utilize the full capacity of acid electrolyte based Li-air batteries, pH value of the electrolyte has to be limited to be larger than 4.

11

Limitation of Li-air Batteries in Alkaline Electrolyte

3.00

3.20

3.40

3.60

3.80

4.00

4.20

0.0 1.0 2.0 3.0 4.0 5.0LiOH (M)

Vol

tage

(V

)

Neutral electrolyte0 mol LiOH

pH = 7V = 3.855 V

pH147

0.1 mol LiOHpH = 133.501 V

0.5 mol LiOHpH = 13.73.460 V

1 mol LiOHpH = 143.442 V

Capacity (%) 19.2% 100%0%

Li + 0.25 O2+ 11.14 H2O ↔ LiOH+10.64H2O

5.3 LiOH3.4 V

Soluability limit: 444 Wh/kg

Glass Stability limit: 85 Wh/kg

To utilize the full capacity of alkaline electrolyte based Li-air batteries, pH value of the electrolyte has to be limited to be less than 10.

12

Electrochemical Stability of NASICON Glass in Aqueous Electrolytes

0

20

40

60

80

100

120

140

160

180

200

0 0.5 1 1.5 2 2.5E

MF

/ mV

lg (C1/C2)

LiCl EMF

HCl EMF

K+ ion in nonaqueous

Linear (LiCl EMF)

NASICON solid electrolyte cannot distinguish between H+ ions and Li+ ions, at least on the interface.

Both Li + and H+ ion in chamber B participate in the electrochemical process. Increase of pH value in chamber B may indicate the reaction of H+ with glass electrolyte.

See Dr. Fei Ding’s poster for more details

13

3. Stability of Li-air Batteries Using Non-Aqueous Electrolyte

a. 2Li + O2 Li2O2

Air electrode Separator

Anode

Load

2Li+

2e-

2Li Non-aqueous electrolyte

O2-2

O2

2e-

Li2O2 Final product stays in air electrode

2Li + O2 Li2O2 E0 = 3.1V Schematic of reaction processes in metal-air batteries.

In non-aqueous electrolyte

b. 2Li + O2 Li2O2 In non-aqueous electrolyte

Challenges: Stable non-aqueous electrolyte in oxygen rich environment

Investigation on the Reaction Products in Li-O2 Batteries

14

Developed unique characterization tools

O2 in

Teflon container

- +

Mass Spec ---CO2?

GC ---CO2?

Compressed O2, 2 atm

Li/air coin cells filled with 1M LITFSI in PC/EC (low vapor pressure electrolyte )

Coin cells holder

In situ analysis

X-ray FTIR NMR

ex situ analysis

2Li + O2 Li2O2 ?

15

O2 in

Teflon container

- +

Mass Spec ---CO2?

GC ---CO2?

Compressed O2, 2 atm

Li/air coin cells with 1M LITFSI in different electrolytes

Coin cells holder

In situ analysis

X-ray FTIR NMR

ex situ analysis

2Li + O2 Li2O2 ?

3.1. Reaction Products During Discharge Process

16

Li2O2 Was Not Observed in the Discharge Products When Carbonate Solvents Was Used

10 20 30 40 50 60 70 802 Theta (Degree)

Inte

nsity

(a. u

.)2.7 V

2.6 V

2.8 V

2.5 V2.4 V2.2 V

2.0 VLi2CO3

Li2O2

Li2OTeflon

KB carbon

LPDC

LEDC

• For all DODs (from 2.8 V to 2.0 V), nearly no Li2O2 and Li2O were detected. • Lithium alkylcarbonates (LPDC and LEDC) and Li2CO3 are identified to be the

primary discharge products.

Ether Based Electrolytes Can Lead to the Formation of Li2O2 During Discharge

PC-EC

Triglyme

Teflon

Carbon

Li2O2

Li2CO3

Li2O Sebaconitrile

TEPa

DMSO

BDG

Yes. If the right electrolytes are used

2Li + O2 Li2O2

See Dr. Wu Xu’s poster for more details 17

3.2 Reaction Products During Charge Process

18

O2 in

Teflon container

- +

Mass Spec ---CO2?

GC ---CO2?

Compressed O2, 2 atm

Li/air coin cells filled with 1M LITFSI in PC/EC

Coin cells holder

In situ analysis

2Li + O2 Li2O2 ?

X-ray FTIR NMR

ex situ analysis

19

What Can be Charged?

First cycle charge capacities:

a) SP: 4.1 mAh/g b) Li2CO3 : 4.8 mAh/g c) LEDC: 99.6 mAh/g (331

mAh/g, 30.1%) d) LPDC: 153.3 mAh/g (308

mAh/g, 49.8%) e) Li2O2 : 1078.6 mAh/g (1165 mAh/g, 92.5%) a) Li2O: 63.3 mAh/g

0 10 20 30 40 50 60 70 80 90 100

Test time (hour)

Cel

l vol

tage

(V)

(f) Li2O

(e) Li2O2

(d) LPDC

(c) LEDC

(b) Li2CO3

(a) SP

Li2O2: Highly chargeable (>92%) LEDC and LPDC: Partially chargeable (~42-69%) and is responsible for apparent recharge-ability reported before. Li2O, SP, and LiC2O3: Not chargeable

20

What Are the Charge Products? (GC/MS Analysis)

0

0.1

0.2

0.3

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

2

3

4

Volta

ge (V

)

(a) SP

Voltage

Argon

CO/N2

CO2

O2

H2O

0

0.1

0.2

0.3

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

1

2

3

4

Volta

ge (V

)

(b) Li2CO3/SP

Voltage

CO/N2

CO2

Argon

H2O

O2

0

0.1

0.2

0.3

0.4

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

2

3

4

5

Volta

ge (V

)

CO2

Voltage

CO/N2

Argon

(c) LEDC/SP

H2O

O2

0

0.1

0.2

0.3

0.4

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

2

3

4

5

Volta

ge (V

)

Voltage

CO2

CO/N2

Argon

H2O

O2

(d) LPDC/SP

0

0.1

0.2

0.3

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

2

3

4

5

Volta

ge (V

)

O2

Voltage

CO/N2

CO2

H2OArgon

(e) Li2O2/SP

0

0.1

0.2

0.3

0 4 8 12 16 20

Time (hour)

Gas

com

posi

tion

(%)

2

3

4

5

Volta

ge (V

)

Voltage

CO/N2

Argon

CO2

H2O

O2

(f) Li2O/SP

a) SP – mainly CO with minor CO2 b) Li2CO3 – small amount of CO2 and CO c) LEDC – large amount of CO2 d) LPDC – large amount of CO2 e) Li2O2 – mainly O2 with minor CO2 f) Li2O – some CO2 and CO. It cannot be oxidized.

21

Charge Efficiency in Different Non-Aqueous Electrolytes in Oxygen-Rich Environment

2 3 4 5-160

-80

0

80

ORR

Curr

ent /

uA

E / V (Li / Li+)

O2

Ar

OER

DME (0.1M): QOER/QORR = 97.5%

2 3 4 5-60

-40

-20

0

20

Curre

nt /

uA

E / V (Li/Li+)

LiPF6 / EC

bkgrd O2

2 3 4 5-120

-80

-40

0

40

Curre

nt /

uA

E / V (Li/Li+)

LiPF6 / DMC

bkgrd O2

2 3 4 5-90

-60

-30

0

30

Curre

nt /

uA

E / V (Li/Li+)

LiPF6/PC

bkgrd O2

a) b) c)

2 3 4 5-120

-60

0

60

Curr

ent /

uA

E / V (Li / Li+)

LiPF6 / BDG

bkgrd O2

Butyl Diglycol Ether (1.0M) QOER/QORR = 96.7%

No OER peak forLiPF6/PC solution indicates that all the ORR products react with PC which leads to the products

that can not be oxidized/recharged within the

potential range.

Cyclic voltammograms (100 mV/s) on glass carbon in O2 and Ar saturated LiPF6 in different non-aqueous solvent

No OER peak for LiPF6/PC solution indicates that all the ORR products react with PC which leads to the products that can not be oxidized/recharged within the potential range.

Large QOER/QORR ratio for LiPF6/ether solution indicates that most ORR products can be oxidized/recharged within the potential range.

See Dr. Yuyan Shao’s poster for more details

22

Li2CO3 is still found in all electrodes discharged in all different electrolytes. The formation of Li2CO3 during discharge process of a Li-air battery will limit the

cycle life of the batteries. Where is Li2CO3 from – carbon electrode or electrolyte?

The characteristic peak of carbon for Li2CO3 is located at 169 ppm and the peak at 112 ppm is ascribed for Kel-F signal from the two end plugs inside the MAS rotor.

What Are the Source of Limited Efficiency? --NMR Investigation on Discharged Electrodes

23

Li2CO3 Comes From the Electrolyte --13C-MAS NMR Study

The signal of Li2CO3 in the 13C-labeled carbon electrode is not larger, but even less, than the corresponding signal in the natural abundance carbon electrode.

⇒ This result unambiguously reveals that the Li2CO3 is originated not from the carbon electrode but from the electrolyte.

Oxygen and/or superoxide radical anions may oxidize the alkyl groups (CH3 or CH2) or ether bond (C–O) into the carbonate groups that in turn form Li2CO3.

13C-labeled air electrode was prepared and discharged in the same electrolyte.

IfLi2CO3 comes from the carbon electrode, the 13C-MAS NMR signal corresponding to Li2CO3 would increase by nearly 90 folds using 99% 13C labeling since the natural abundance of 13C is only 1.1% without isotope enrichment.

See Dr. Jian Zhi Hu’s poster for more details

24

4. Summary 1. Graphene based air electrode exhibits very high capacity (~15,000 mAh/g) due to its

dual pore structure and defect activity.

2. Specific energy of Li-air batteries with aqueous electrolyte is strongly limited by the operational pH window of NASICON glass. Electrolyte additives and other approaches which can stabilize pH value of electrolyte is critical for long term operation of aqueous based Li-air batteries.

3. In carbonate electrolytes, the majority of the discharge products in Li-air batteries are lithium alkylcarbonates and Li2CO3. Apparent rechargeability is due to oxidation of Lithium alkylcarbonates which release CO2 and CO. This process is not sustainable.

4. Li2O2 can be formed during discharge process when ether based electrolytes were used, Li2CO3 was still observed and need to be minimized.

5. Electrolyte is the key for rechargeable, long term operation of Li-air batteries.

25

Acknowledgments

Technical Team: Wu Xu, Jie Xiao, V. Viswanathan, Jianzhi Hu, Vijayakumar Murugesan, Silas A. Towne, Phillip Koech, Donghai Mei, Fei Ding, Zimin Nie, Yuyan Shao, Jian Zhang, Dehong Hu, Deyu Wang, Jun Liu, and Gordon L. Graff

Financial support: •Defense Advanced Research Program Agency •Laboratory Directed R&D Program (Transformational Materials Science Initiative) of PNNL