Frederick E. Pinkerton - UCSB · 2 content (11.9%) Pinkerton et al., J. Phys. Chem. B 109, 6...

New Li-B-N-H Quaternary Hydrides

Frederick E. PinkertonMaterials and Processes Laboratory

General Motors R&D Center

International Symposium on Materials Issues in Hydrogen Production and Storage

Santa Barbara, CA, 21-25 August 2006

Acknowledgements

• Greg Meisner• Martin Meyer• Mike Balogh• Jan Herbst• Lou Hector• Charlene Hayden• Matt Kundrat• Matt Scullin• Aimee Bailey• Laura Confer• Richard Speer, Jr.

• Yaroslav Filinchuk• Klaus Yvon

• John Vajo• Sky Skeith

• Jason Graetz• Santanu Chaudhuri• Alex Ignatov

Outline

• Hydrogen storage requirements for solid hydrides

• Building high hydrogen capacity storage reactions

• New quaternary Li-B-N-H hydride– Synthesis from LiNH2 and LiBH4

– Hydrogen release properties– Crystal structure

• Metal additives to promote hydrogen release

• Concluding remarks

The hydrogen storage problem…

Dr Arnold Hydrogen Storage / TAA

Global Alternative Propulsion CenterWith thanks to Dr. Gert Arnold

Storage method

15 M

Pa gas

35 M

Pa gas

70 M

Pa gas

Liquid

H2

LaNi5H

6FeT

iH2

MgH2

Mg2NiH

4NaA

lH4

Li-N-H

2LiBH4+

MgH2

Li3BN2H

8Gas

oline

*

Vol

umet

ric d

ensi

ty (k

g H

2/lite

r)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

Solid hydrides offer compact storage…

*energy equivalent

kg of hydrogen per liter of material

Storage method

15 M

Pa gas

35 M

Pa gas

70 M

Pa gas

Liquid

H2

LaNi5H

6FeT

iH2

MgH2

Mg2NiH

4NaA

lH4

Li-N-H

2LiBH4+

MgH2

Li3BN2H

8Gas

oline

*

Vol

umet

ric d

ensi

ty (k

g H

2/lite

r)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

Solid hydrides offer compact storage…

system performance will be substantially reduced

*energy equivalent

Other aspects are more challenging

Equilibrium Temperature for PH2=1 atm (°C)0 100 200 300

Theo

retic

al h

ydro

gen

capa

city

(wei

ght %

)

0

2

4

6

8

10

12

Approximate ∆H (kJ/mol H2) [~T∆SH2]

30 40 50 60 70

MgH2

Mg2NiH4

LaNi5Hx

TiV1.6Ni0.4HxTiFeHx

NaAlH4 (stage 1)

Na3AlH6 (NaAlH4 stage 2)

Other aspects are more challenging

Equilibrium Temperature for PH2=1 atm (°C)0 100 200 300

Theo

retic

al h

ydro

gen

capa

city

(wei

ght %

)

0

2

4

6

8

10

12

Approximate ∆H (kJ/mol H2) [~T∆SH2]

30 40 50 60 70

MgH2

Mg2NiH4

LaNi5Hx

TiV1.6Ni0.4HxTiFeHx

NaAlH4 (stage 1)

Na3AlH6 (NaAlH4 stage 2)

LiNH2+LiH (Chen et al.)Mg(NH2)2+2LiH

(Luo)

LiBH4+½MgH2+TiCl3(hydrogenation)

(Vajo et al.)

Li3BN2H8+0.2NiCl2(dehydrogenation only)

New materials research• Capacity: High specific mass• Thermodynamics: Moderate ∆H (~35 kJ/mol H2)

– Hydrogen release temperature 20-80 C in 2-5 bar H2– Thermal management: insertion and extraction of ∆H

• Fast kinetics– Complex hydrides tend to be kinetically limited, requiring

high temperature even if the thermodynamics are good

GM “Sequel”Hydrogen Fuel Cell prototype

Li-B-N-H: Building high-capacity reactions

NaAlH4 1/3 Na3AlH6 + 2/3 Al + H2 3.7 wt%NaH + Al + 3/2 H2 5.6 wt% (7.5 wt%)

LiNH2 + LiH Li2NH + H2 6.5 wt% (9.8 wt%)Mg(NH2)2 + 2 LiH Li2Mg(NH)2 + 2 H2 5.6 wt% (8.4 wt%)LiBH4 LiH + B + 3/2 H2 13.9 wt% (18.5 wt%)LiBH4 + ½ MgH2 LiH + ½ MgB2 + 2 H2 11.5 wt% (14.4 wt%)

Not all of the hydrogen is released

Strategy: Identify HYDROGEN-FREE compounds involving light elements that could be the DECOMPOSITION PRODUCTS of

reactions between hydrogen-containing materials corresponding to COMPLETE HYDROGEN RELEASE

Hypothetical reaction

• Lithium Boronitride: Li3BN2– Several known polymorphs

• Tetragonal P42212 low temperature phase (<860°C)– a = 4.6435 Å c = 5.2592 Å

• Monoclinic P21/c high temperature phase– a = 5.1502 Å b = 7.0824 Å c = 6.7908 Å β = 112.96º

• High pressure phase (DeVries and Fleischer)

• Hypothetical reaction:2 LiNH2 + LiBH4 → Li3BN2 + 4 H2 11.9 wt% H

• Success! > 11 wt% H2 release– But here’s the twist:

We formed a new quaternary Li-B-N-H phase

Synthesis of 2 LiNH2 + LiBH4: Ball milling

2 LiNH2 + LiBH4

ball milled powders

2θ (degrees)

10 15 20 25 30 35 40 45 50 55

Inte

nsity

(arb

irary

uni

ts)

LiBH4

LiNH2

Li2O

α phase

0

10

2040

960

160

Millingtime(min)

300

Pinkerton et al., J. Phys. Chem. B 109, 6 (2005). Published on line: 17 Dec 2004

α phase

Premill LiNH2 and LiBH4separately for 10 min, then mix and heat

The in situ XRD data at 75ºC is identical to that at room temperature (RT) when the experiment started.

After mixing and storing for 12 days at RT, the mixture spontaneously formed a substantial quantity of the α phase!

Heating above ~95ºC completes the conversion to the α phase.

XRD Results: Mix & Heat

LiBH4 + 2 LiNH2 conversion to α Li-B-N-H by heating

2θ (degrees)

15 20 25 30 35 40 45 50 55

Inte

nsity

(arb

itrar

y un

its)

LiBH4

LiNH2

α phase

75°C

109°C

Pinkerton et al., J. Phys. Chem. B 109, 6 (2005).

In situ XRD: Evolution of α Li-B-N-H at 73 °C

100

25

500

200

50

0

Time(min)

2θ (degrees)LiBH4 LiNH2

αααα α

α

α

αααα

2 LiNH2 + LiBH4mixed powders

Time (min)

0 20 40 60 80

Tem

pera

ture

(°C

)

0

50

100

150

200

250

300

350

400

Wei

ght (

%)

86

88

90

92

94

96

98

100

102

Residual gas analysis

Time (min)

0 20 40 60 80

Par

tial p

ress

ure

(Tor

r)

10-11

10-10

10-9

10-8

10-7

Mass 2 - H2

Mass 17 - NH3

Mass 16 - NH2 radical, CH4

RGA mass spectrometry

13.1 wt%

~2 mole% NH3

11.5 wt% H2, 1.6 wt% NH3

Hydrogen desorption

(TGA)2 LiNH2 + LiBH4mixed powders(“mix and heat”)

13.1 wt% lossexceeds theoretical H2 content (11.9%)


In situ XRD – ball milled 2LiNH2 + LiBH4

α-phase

Liquid

Li3BN2

Melts~190°C

Dehydrogenation> 250°C

Temperature(°C)

Reaction path

2 LiNH2 + LiBH4 → α Li-B-N-H (solid)

→ α Li-B-N-H (liquid) ~195ºC

→ Li3BN2 (solid) + 4 H2 >250ºC

“Destabilized LiBH4” ?

New compound, less stable than LiBH4

α Li-B-N-H: a new quaternary hydride

2θ (degrees)

5 10 15 20 25 30 35 40 45 50 55 60

Inte

nsity

(cou

nts)

0

2000

4000

6000

8000

10000

2 LiNH2 + LiBH4 ball-milled 5 hrs:α Li-B-N-H

Li2O

Plastic film

bcc crystal structurea = 10.78 Å

“Li3BN2H8” α phase in Li-B-N-H

2

α-phase crystal structure

• Single crystals formed by recrystallizing from the melt

• Body-centered cubic space group:

I213 (#199)a = 10.676 Å

• NH2− and nearly

tetrahedral BH4− units

persist in the structure• Equilibrium composition:

Li4BN3H10 !(= 3 LiNH2 + LiBH4)

Filinchuk et al., Inorg. Chem. 45, 1433 (2006).

Li

B

N

H

Lattice shift of ball milled material

2 LiNH2 + LiBH4 ball milled 16 hrs

2θ (degrees)

25 30 35 40 45 50

Inte

nsity

(cou

nts)

0

2000

4000

6000

8000

10000

Experimental patterna = 10.78 ACalculated Li4BN3H10

a = 10.68 A

0

10

20

30

40

50

10 20 30 40 50 60 70Angle (2θ)

Diff

ract

ion

Inte

nsity

(cou

nts/

sec)

x = 1

x = 0.75

x = 0.667

x = 0.655

x = 0.636

x = 0.6

x = 0.556

x = 0.5

x = 0.333

x = 0

LiNH2

α + δ

α

γ +α

β + γ + α

β + γ + α

β + γ

β

LiBH4 + β

LiBH4

X-Ray Diffraction Data for Ball Milled

(LiNH2)x(LiBH4)1-x

*

Meisner et al., J. Phys. Chem. B 110, 4186 (2006).

Li2O

3:1 α + δ2:1 α

1:1 β

} α+β+γ

What about 3 LiNH2 + LiBH4?

If the α phase is Li4BN3H10, then what is the dehydrogenation behavior of samples made

at the 3 LiNH2 + LiBH4 composition?

RGA mass spectrometry

Hydrogen desorption

(TGA)3 LiNH2 + LiBH4Ball milled 5 hrs

18 wt% loss(theoretical content

11.1 wt%)

3 LiNH2 + LiBH4ball milled 5 hrs

Time (min)

0 20 40 60 80 100 120

Tem

pera

ture

(°C

)

0

100

200

300

400

500

Wei

ght (

%)

80

85

90

95

100

Residual gas analysis

Time (min)

0 20 40 60 80 100 120

Parti

al p

ress

ure

(Tor

r)

10-10

10-9

10-8

10-7

Mass 2 - H2

Mass 17 - NH3

Mass 16 (x 1/2) - NH2 radical

18 wt%

~9 mole% NH3

9.6 wt% H2, 8.3 wt% NH3

0

2

4

6

8

10

12

14

16

18

20

22

0.3 0.4 0.5 0.6 0.7 0.8Composition (x)

Des

orpt

ion

(wt%

)

Total hydrogen contentHydrogen desorbed as HAmmonia desorbedTotal hydrogen desorbedTotal desorption

H2 and NH3 release in (LiNH2)x(LiBH4)1-x

Meisner et al., J. Phys. Chem. B 110, 4186 (2006).

2:1

3:1

Li3BN2

LiBH4 LiNH2

What’s so special about 2:1?

• Dehydrogenation from the liquid is not controlled by the starting α-phase, but rather by the product Li3BN2 phase

• For 2 LiNH2 + LiBH4:Li3BN2H8 (liquid) Li3BN2 + 4H2

• For 3 LiNH2 + LiBH4:Li4BN3H10 (liquid) Li3BN2 + ½ Li2NH + 4H2 + ½ NH3

Li3BN2 !

Dehydrogenated material: New Li3BN2 polymorph

Li3BN2 - I41/amd

2θ

10 20 30 40 50 60 70

Inte

nsity

0

1000

2000

3000

4000

5000

6000

DataFitDifferenceLi2O

101

112200

103

202

121004

301

204

105224

215

400233

305 107422

404051,341 244

Body-centered tetragonalcrystal structureI41/amd (#141)a = 6.60 Å, c= 10.35 Å

Pinkerton and Herbst, J. Appl. Phys. 99, 113523 (2006).

Li

B

N

Promoting H2 release with metal additives

• Composition: 2 LiNH2 + LiBH4 + additive

• All samples ball milled for 5 hrs– Fully converted to α phase

• Metals or metal compounds added prior to ball milling– Metal powder– Metal dichlorides– “Pt/Vulcan carbon”: 2 nm diameter Pt nanoparticles

supported on a Vulcan carbon substrate

TiCl3 additive

• 2 mole% TiCl3additive has been shown to be very effective in some hydrogen storage systems– NaAlH4

(Bogdanović et al.)– LiBH4 + ½ MgH2

(Vajo et al.)

• TiCl3 does not significantly improve dehydrogenation of Li-B-N-H

LiB0.33N0.67H2.67 ball-milled 5 hrs

Temperature (°C)150 200 250 300 350 400

Wei

ght (

%)

84

86

88

90

92

94

96

98

100

102

Without additive5 wt% TiCl3


Pt/Vulcan carbon additions

Temperature (°C)150 200 250 300 350 400

Wei

ght (

%)

84

86

88

90

92

94

96

98

100

102

Without additive5 wt% Pt powder6.8 wt% PtCl21 wt% Pt/Vulcan carbon 2 wt% Pt/Vulcan carbon3 wt% Pt/Vulcan carbon5 wt% Pt/Vulcan carbon10 wt% Pt/Vulcan carbon2.4 wt% Vulcan carbon

LiB0.33N0.67H2.67 with Pt

∆T = -90°C


In situ XRD – Pt/Vulcan carbon added

Temperature(°C)

α-phase

Li3BN2 + Li2Pt3BIntermediate solids

In situ XRD - 5 wt% Pt/Vulcan carbon

Quaternary α-phase

Intermediate 1

Li3BN2 + Li2Pt3B

Intermediate 2


20 40 60 80 100 120 140 160 180 200

Intensity(counts)

Accelerated isothermal H2 release

Li3BN2H8

Time (min)0 200 400 600 800

Wei

ght (

%)

86

88

90

92

94

96

98

100

102

Uncatalyzed (210°C)

5 wt% Pt/Vulcan carbon(200°C)

NiCl2 additions

Temperature (°C)150 200 250 300 350 400

Wei

ght (

%)

84

86

88

90

92

94

96

98

100

102

Without additiveNi powder (< 53 µm)Ni flake (< 44 x 0.37 µm thick)Raney Ni 2800Nanosized Ni (<40 nm)NiCl211 wt% NiCl2

LiB0.33N0.67H2.67 with Ni additives(5 wt% unless noted otherwise)

10 wt% Pt/Vulcan carbon

∆T = -112°C

Pinkerton et al., JALCOM, available online.

TEM: LiB0.33N0.67H2.67 + 5 wt% NiCl2

Nanocrystalline Ni3B (≤ 8 nm)

Mass spectrometry gas analysis

• Additive-free Li-B-N-H: – H2 and NH3 release occur

together above 250°C– Evolved gas ~2 mole% NH3

• NiCl2-added:– onset of H2 release is 120°C– Total NH3 release reduced

by an order of magnitude

LiB0.33N0.67H2.67

0 20 40 60 80

Wei

ght (

%)

86

88

90

92

94

96

98

100

102

Tem

pera

ture

(°C

)

0

100

200

300

400

0 20 40 60 80

Parti

al P

ress

ure

(Tor

r)

10-11

10-10

10-9

10-8

10-7

Time (min)0 20 40 60 80

Parti

al P

ress

ure

(Tor

r)

10-10

10-9

10-8

10-7

10-6

LiB0.33N0.67H2.67 + 5 wt% NiCl2

Mass 2 - H2~11.0 wt%

Mass 18 - H2O

Corrected NH3 ~0.2 wt%

Mass 17 - NH3, OH −

Mass 2 - H2~11.5 wt%

Corrected NH3 ~1.6 wt%Mass 17 - NH3, OH −

Mass 18 - H2O

(b) LiB0.33N0.67H2.67

(a)

(c) LiB0.33N0.67H2.67 + 5 wt% NiCl2


5 wt% NiCl2 addition

Quaternary α-phase

Intermediate

Li3BN2

Distinct from both intermediates in Pt/Vulcan carbon!


Comparison of intermediate phases

LiBH4 + 2 LiNH2 + 5 wt% Pt/Vulcan C226°C

Inte

nsity

(cou

nts)

020406080

100120140160

LiBH4 + 2 LiNH2 + 5 wt% Pt/Vulcan C261°C

Inte

nsity

(cou

nts)

0

20

40

60

80

100Li2Pt3B

LiBH4 + 2 LiNH2 + 11 wt% NiCl2

2θ (degrees)

5 10 15 20 25 30 35 40 45

Inte

nsity

(cou

nts)

0

20

40

60

80

100

120

Pt/Vulcan carbonIntermediate 1(226°C)

Pt/Vulcan carbonIntermediate 2(261°C)

NiCl2Intermediate(~210-~290°C)

What are the additives doing?

– Small quantities (~1 mole%) have a large effect

– ∆T½ scales with the specific surface area (m2/g) of the additive particles

– Effect appears to saturate at low addition levels (~2 mole% for NiCl2)

• Likely acting as a dehydrogenation catalyst

Ni additions

Specific surface area (m2/g)0.1 1 10 100 1000

| ∆T 1/

2 | (°

C)

0

20

40

60

80

100

120

Ni powder

Ni flake40 nm Ni nanoparticles

Raney Ni

NiCl2 (< 10 nm Ni3B)

(5 wt% Ni)

What about reversibility?

• Dehydrogenation appears to be exothermic

• Thermodynamically unstable

• => difficult to reverse (off-board regeneration)

• Caveat:It’s not a simple system: H2 release, NH3 release, and Li3BN2 solidification are happening simultaneously Pinkerton et al., J. Phys. Chem. B 110, 7967 (2006).

LiB0.33N0.67H2.67 ball milled 300 min

Temperature (°C)100 200 300 400

Hea

t flo

w (W

/g);

endo

ther

mic

up

-12

-10

-8

-6

-4

-2

0

2

4

No additive5 wt% Pt/Vulcan carbon

α-phasemelting

Desorptiononset

Desorptiononset

β-phasemelting

DSC

Exothermic

DFT estimates of reaction enthalpy

• 2 LiNH2 + LiBH4 Li3BN2 + 4 H2– ∆H ~ 23 kJ/mol H2 - Aoki et al., Appl. Phys. A 80, 1409 (2005)– ∆H = 18-24 kJ/mol H2 - Alapati et al., JPC B 110, 8769 (2006)– Caveats: zero T calculations excluding zero point energy– Both suggest reversibility of Li3BN2 to the two-phase mixture

• Li4BN3H10 Li3BN2 + ½ Li2NH + ½ NH3 + 4H2– ∆H = 24 kJ/mol H2 – Herbst & Hector, APL 88, 231904 (2006)– Includes zero point energies and phonons: 298 K values– Endothermic hydrogen release suggests reversibility– Caveat: does not include α-phase melting or Li3BN2 solidification

• May be too unstable: – 1 bar H2 equilibrium temperature is ~130-240 K

Aside:

• Reaction enthalpy for 3 LiNH2 + LiBH4 Li4BN3H10

– ∆H = -6 kJ/mol – Herbst & Hector, APL 88, 231904 (2006)

• Formation of Li4BN2H10 is slightly exothermic• Consistent with observed conversion

2 LiNH2 + LiBH4 α Li-B-N-H

Effect of H2 pressure

LiB0.33N0.67H2.67 + 5 wt% Pt/Vulcan carbon

Temperature (°C)150 200 250 300 350 400

Wei

ght (

%)

86

88

90

92

94

96

98

100

102

100 kPa Ar130 kPa H2

4200 kPa H2


Attempt to rehydride LiB0.33N0.67H2.67 + 5 wt% Pt/Vulcan carbon

LiB0.33N0.67H2.67 + 5 wt% Pt/Vulcan carbon

Time (min)0 2000 4000 6000

Pre

ssur

e (M

Pa)

0

2

4

6

8

Tem

pera

ture

(°C

)

0

20

40

60

80

100

120

140

160

180

Wei

ght (

%)

90

92

94

96

98

100

102

Switch to H2

8.4 MPaH2

150°C0.13 MPa He150°C

1.3 wt%


Challenges for quaternary Li-B-N-H

• Thermodynamics– Understand reaction enthalpies for dehydrogenation of

α Li-B-N-H at different compositions• Kinetics

– High temperatures required to overcome slow diffusion and strong hydrogen binding and in complex hydrides

• Catalysis– Why does Ni or Ni3B work so well?– Why doesn’t TiCl3 work?

• What if we could reduce the Li-B-N-H hydrogen release temperature below the α phase melting temperature (similar to what was done for NaAlH4)?– Reversibility?– Suppress NH3 release?– More practical? (Or at least, less impractical?)

On-board hydrogen storage materials-

• Compressed gas• Liquid H2• Physisorption at cryogenic temperatures

– Activated carbon, nanotubes, MOFs

• Hydrolysis hydrides– NaBH4 + H2O NaBO3 +

• Organic ring compounds• Solid hydrides

– Metal hydrides (LaNi5, Mg, Mg2Ni, AB, AB2 compounds– Complex hydrides (NaAlH4, LiBH4, LiAlH4, …)

H2

Quaternary hydrides are fertile hunting ground

in the search for new hydrogen storage materials

Summary

• Strategy of looking for hydrogen-free products has been successful

• We have discovered a new quaternary hydride, α Li-B-N-H (Li4BN3H10), that releases all of its hydrogen above 250°C– H2 and NH3 release are strong functions of composition, with

optimum H2 release near (LiNH2)0.67(LiBH4)0.33 [2:1]

• Numerous other Li-B-N-H phases exist– 4 phases along the (LiNH2)x(LiBH4)1-x tie line (2 metastable)– 4 additional phases during Li-B-N-H decomposition with additives– 3 more phases have been found in the Li-B-N-H phase diagram

(Torgersen et al., MRS 2004 Fall Meeting)

• Metal nanoparticle additions reduce the dehydrogenation temperature by up to 112°C (NiCl2)

• Quaternary hydrides are fertile hunting ground in the search for new hydrogen storage materials

Frederick E. Pinkerton - UCSB · 2 content (11.9%) Pinkerton et al., J. Phys. Chem. B 109, 6...

Documents

Transcript of Frederick E. Pinkerton - UCSB · 2 content (11.9%) Pinkerton et al., J. Phys. Chem. B 109, 6...