effects and mechanisms of mechanical activation on ... · leon l. shaw, tippawan markmaitree,...

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

Effects and Mechanisms of Mechanical Effects and Mechanisms of Mechanical Activation on Hydrogen Sorption/Desorption Activation on Hydrogen Sorption/Desorption

of Nanoscale Lithium Nitridesof Nanoscale Lithium Nitrides

Leon L. Shaw, Tippawan Markmaitree, William Osborn, Xuefei Wan, Kyle CrosbyUniversity of Connecticut

Z. Gary Yang, Jianzhi Hu, Ja Hu KwakPacific Northwest National Laboratory

June 9 – 13, 20082008 DOE Hydrogen Program Review

Project ID: ST14

2

OverviewOverview

• Project start date: December 2004• Project end date: December 2009• Percent complete: 55%

• Total project funding:DOE share: $1.6 milContractor share: $0.4 mil

• Funding received in FY07: $200,000 for UConn; $90,000 for PNNL• Funding for FY08: $250,000 for UConn; $80,000 for PNNL

Timeline:

Budget:

Partners/Collaborators:• Pacific Northwest National Laboratory – NMR analysis• Sandia National Laboratory – Information exchange• Universidad de Extremadura, Badajoz, Spain – X-ray analysis• HRL Laboratories – Information exchange• University of Pittsburgh – Information exchange

Barriers

A. System weight and volume: 2 kWh/kg & 1.5 kWh/L

E. Charging/discharging rates: 3 min for 5 kg

3

ObjectivesObjectivesObjective in FY 07:

Identify hydriding/dehydriding reaction mechanisms and rate-limiting steps of (LiNH2 + LiH) systems

Enhance hydriding/dehydriding rates via nano-engineering and mechanical activation

Improve hydriding/dehydriding properties via thermodynamic destabilization

Objective in FY 08:

Further improvement in hydriding/dehydriding properties of (LiNH2 + LiH) systems via nano-engineering, mechanical activation, and thermodynamic destabilization

Establishment of the atomic level understanding of the reaction mechanism and kinetics of mechanically activated, nano-engineered (LiNH2 + LiH) systems

Nano-engineering and mechanical activation of LiBH4-based materialsDemonstration of hydrogen uptake and release of (LiBH4 + MgH2) systems

with a storage capacity of ~ 10 wt% H2 at 2000C

4

MilestonesMilestonesMonth/Year Milestone or Go/No-Go Decision

Nov-07 Milestone: Identify the hydriding/dehydriding mechanisms and rate-limiting steps of (LiNH2 + LiH) systems. Establish the effect of mechanical activation. Enhance the reaction kinetics of mechanically activated, nano-engineered (LiNH2 + LiH) systems.

Dec-07 Go/No-Go Decision: Demonstrate mechanically activated Li3N-based materials with ~ 7.5 wt% reversible hydrogen storage capacity and uptake/release at temperatures below 1300C and a few bars of hydrogen pressure.

Nov-08 Milestone: Establish the atomic level understanding of the reaction mechanism and kinetics of hydriding/dehydriding of mechanically activated, nano-engineered (LiNH2 + LiH) systems. Demonstrate

hydriding and dehydriding reactions of (LiBH4 + MgH2) systems at the solid state, i.e., below the melting point of LiBH4 (Tm = 2800C).

Demonstrate hydrogen uptake and release of (LiBH4 + MgH2) systems with a storage capacity of ~ 10 wt% H2 at 2000C.

Dec-08 Go/No-Go Decision: Demonstrate hydrogen uptake and release of (LiBH4 + MgH2) systems with a storage capacity of ~ 10 wt% H2 at

2000C.

5

ApproachesApproaches

Identifying the mechanism and rate-controlling step of hydriding and dehydriding reactions of (LiNH2 + LiH) systems and LiBH4-based materials.

Developing the understanding of the effect of mechanical activation and nano-engineering on hydriding and dehydriding properties

Applying the understanding of mechanical activation and the established reaction mechanism and rate-controlling step to enhancing hydriding and dehydriding properties and long-term cyclical stabilities of hydrogen storage materials.

6

LiNH2 + LiH Li2NH + H2

Identification of Reaction MechanismIdentification of Reaction Mechanism

Reversible storage capacity = 6.5 wt% H2 [1]

ΔH (dehydriding) = 65.6 kJ/mol H2 [2]

The current operating temperatures for hydriding and dehydriding are typically at ~ 2800C, higher than the thermodynamic prediction [1-3].

Why is hydriding very fast, whereas dehydriding is very slow [4]?

Why is the formation of Li2NH and NH3 from LiNH2 slow [5], whereas the reaction between NH3 and LiH is very fast [6]?

There is NH3 emission associated with this storage system [3,7].

[1] P. Chen, et al., Nature, 420, 302 (2002).[2] Y. Kojima, Y. Kawai, J. Alloys Comp., 395, 236 (2005).[3] L. Shaw, et al., J. Alloys Compd., 448, 263 (2008).[4] L. Shaw, et al., J. Power Sources, 177, 500 (2008).[5] T. Markmaitree, R. Ren, L. Shaw, J. Phys. Chem. B., 110, 20710 (2006).[6] H. Y. Hu, E. Ruckenstein, J. Phys. Chem. A, 107, 9737 (2003).[7] T. Ichikawa, et al., J. Phys. Chem. B, 108, 7887 (2004).

7

Composition analysis of the effluent gas from LiNH2+LiH mixtures during heating

1.00E-09

1.00E-08

1.00E-07

1.00E-06

1.00E-05

0 100 200 300 400 500 600 700 800

Temperature (OC)

Inte

nsity

(Tor

r.) H2

H2

OH*+NH3O2

N2

(a)

0.00E+00

5.00E-11

1.00E-10

1.50E-10

2.00E-10

2.50E-10

3.00E-10

1 2 3 4

Am

mon

ia L

evel

, tor

r/mg

Identification of Reaction MechanismIdentification of Reaction MechanismComparisons in the NH3 emission

from different hydride systems

LiNH2 with 180 min milling

LiNH2+LiH without milling

LiNH2+LiH with 180

min milling

MgH2without milling

The LiNH2–only sample exhibits the highest NH3 emission.LiNH2+LiH mixtures display much less NH3 emission, indicating the “capturing”

of NH3 by LiH and offering the evidence of two elementary steps for dehydrding reaction.

Ball milling reduces the NH3 emission dramatically (by about 36 times); In fact, the NH3 intensity is so low that it is below the detection limit of the mass spectrometer.

-10-50510ppm

NH3surface LiNH2

(a)

(b)

(c)

VT 300MHz in situ 1H MAS NMR Spectra

30°C

50°C

75°C

-50510

NH3

H2

(d)

(e)

(f)

ppm

Identification of Reaction MechanismIdentification of Reaction Mechanism

LiNH2

40-150°C (during 77s)

150-180°C

LiNH2+LiH

25°C

NH3 releases from LiNH2 at temperatures as low as 30oC. The speed of NH3 release increases dramatically at ~ 75oC.

LiH reacts with NH3 at RT; however, the minimum temperature for LiH to capture all the NH3 molecules is higher than the temperature at which a large quantity of NH3 is released from LiNH2.

9

Isothermal hydrogen uptake/release cycles of the LiNH2 + LiH mixture at 2850C*

Identification of RateIdentification of Rate--Limiting StepsLimiting Steps

A rapid hydrogen uptake rate (i.e., approaching the theoretical storage capacity in ~ 5 min)

A slow release rate (i.e., incomplete release of hydrogen in 2.5 h).

Modeling of the release kinetics under a constant temperature and constant hydrogen pressure to investigate the rate-limiting step.

* L. Shaw, W. Osborn, T. Markmaitree, X. Wan, J. Power Sources, 177, 500 (2008).

10

Dehydriding of LiNH2 + LiH proceeds with two elementary reactions:

LiNH2

Li2NH NH3

LiHLiNH2

LiNH2

H2

NH3

LiNH2 = ½ Li2NH + ½ NH3 (1) ½ NH3 + ½ LiH = ½ LiNH2 + ½ H2 (2)


[1] T. Markmaitree, R. Ren, L. Shaw, J. Phys. Chem. B., 110, 20710 (2006).[2] H. Y. Hu, E. Ruckenstein, J. Phys. Chem. A, 107, 9737 (2003).[3] L. Shaw, et al., J. Power Sources, 177, 500 (2008).

Both reactions produce solid products. However, Reaction (1) occurs very slow because of the diffusion-controlled reaction [1], whereas Reaction (2) takes place in microseconds [2] because of cracking and flaking off of the LiNH2 product [3].

The cracking and flaking off of the LiNH2 product is due to the fact that the volume of the solid product from Reaction (2) is substantially larger than that of the solid reactant (2 times larger) [3].

11

Based on the reaction pathway proposed, the hydrogen release of the LiNH2 + LiH system can be described at least with the following 6 steps.

(decomposition at the LiNH2/Li2NH interface) (4)#21

21

322 NHNHLiLiNH +⎯→⎯

*21#

21

33 NHNH ⎯→⎯

*21

21*

21

33 +⎯→⎯ NHNH

Δ⎯→⎯Δ+ 33 21

21

21 NHNH

Δ+⎯→⎯+Δ 223 21

21

21

21 HLiNHLiHNH

Δ+⎯→⎯Δ21

21

21

22 HH

(diffusion of NH3 to a Li2NH surface site, *) (5)

(desorption of NH3 from the Li2NH surface site) (6)

(adsorption of NH3 on a LiH surface site, Δ) (7)

(reaction at the LiH surface) (8)

(desorption of H2 from the LiH surface site) (9)

The rate of dehydriding will be controlled by the slowest step.


12

( ) 2/12/1

3/1 6.51'1 tRDff −=−

Conclusions*: Dehydriding of the LiNH2 + LiH system is controlled by the slowest step of the decomposition of LiNH2 which is a diffusion controlled reaction.

Curve Fitting of the Amount of H2 Released at 2850CIdentification of RateIdentification of Rate--Limiting StepsLimiting Steps

The shrinking-core model with the moving speed of the LiNH2/Li2NH interface controlled by a diffusing specie (e.g., N or NH3)

* L. Shaw, W. Osborn, T. Markmaitree, X. Wan, J. Power Sources, 177, 500 (2008).

13

Challenges for Improving Hydriding Challenges for Improving Hydriding and Dehydriding Propertiesand Dehydriding Properties

1. Nano-engineering to decrease the diffusion distance.

2. Increasing the composition gradient to enhance diffusion via advanced catalysts.

3. Increasing the diffusion coefficient to augment the diffusion rate via mechanical activation and doping.

We have established that dehydrogenation of the LiNH2 + LiH system is a diffusion-controlled process and thus takes place very slowly.

How can we increase the rate of a diffusion-controlled reaction?

Dtx ∝2

dxdCDJ −=

⎟⎠⎞

⎜⎝⎛−=RTQDD exp0

NanoNano--Engineering Engineering and Mechanical Activation and Mechanical Activation through Highthrough High--Energy Energy Ball MillingBall Milling

0

10

20

30

40

50

60

70

1 2 3 4

Spec

ific

surf

ace

area

(m2 /g

)

Crystallite size distribution of LiH particles

Surface area increases with milling time

No milling Milling for 90 min

Milling for 180 min

Milling for 24 hr

LiNH2 + LiH mixtures with particles of 6 to 120 μm before milling

3h milling

Particle sizes reduced to 0.5 to 6 μm after 3h milling

No milling

-10-505

LiNH2 45 minutes mill


FWHM=148Hz

FWHM=156Hz

45 mins

180 mins

10 -10-5ppm

5



2.08ppm

FWHM=148Hz

1.68ppm

FWHM=156Hz

(a) 45-min milling

(b) 180-min milling

10

LiNH2 at RT

-10-50510

300MHz 6Li MAS NMR Spectra*

NanoNano--Engineering Engineering and Mechanical Activation and Mechanical Activation through Highthrough High--Energy Energy Ball MillingBall Milling

(c) 90-min milling

(d) 180-min milling

2.37ppm

2.26ppm

ppm

LiNH2 + LiH at 180oC

Broadening and up-field shifting of 6Li peaks indicate that ball milling has not only resulted in structural refinement, but also led to the alternation in the local electronic state around Li nuclei.

* C. Lu, et al., J. Power Sources, 170, 419 (2007).

16

0

50

100

150

200

250

300

0 200 400 600 800 1000 1200 1400 1600

Milling Time, min

Act

ivat

ion

Ener

gy, k

J/m

ol

The peak temperature for hydrogen release has been reduced by ~ 100oC via ball milling. The weight loss for the sample without milling is substantially larger than the theoretical

prediction (~ 5.5 wt%) because of the NH3 emission. In contrast, the milled sample is in excellent agreement with the theoretical prediction.

The activation energy for dehydrogenation decreases with an increase in the milling time, indicating the effect of mechanical activation.

The activation energy is also affected by the milling and mixing sequence, suggesting that LiH has the catalytic effect on the decomposition of LiNH2.

LiNH2 onlyLiNH2 + LiH (mixing & milling)

LiNH2 + LiH (milling & mixing)

90

92

94

96

98

100

102

104

0 100 200 300 400 500 600

Temperature (0C)

Wei

gth

(%) No milling

Milled for 180 min @ RT

TGA Traces of LiNH2 + LiH

Enhancing Dehydrogenation Enhancing Dehydrogenation of LiNHof LiNH22 + LiH Mixtures + LiH Mixtures via via NanoNano--Engineering & Mechanical Engineering & Mechanical ActivationActivation

17

Enhancing Hydrogenation/Dehydrogenation Enhancing Hydrogenation/Dehydrogenation through through Mechanical Activation at Liquid Mechanical Activation at Liquid

Nitrogen TemperatureNitrogen Temperature

0

1

2

3

4

5

6

7

0 5 10 15 20 25 30 35 40

Time [hr]

Co

mp

osi

tio

n [

wt%

H2]

RT MillLN2 MillCycling Temperature = 2850C

LiNH2+LiH ball milled at liquid nitrogen temperature (LN2) displays much faster hydriding/dehydriding kinetics than that ball milled at RT. As a result, LN2-processed LiNH2+LiH has improved the utilization of the theoretical storage capacity by 22%.

VT 300MHz in situ 1H MAS NMR Spectra

-50-2502550

H2

(a1)

(b1)

(c1)

25-100°C

160-180°C

180°C

LN2-180min mill

-50-2502550

H2

(a2)

(b2)

(c2)

RT-180min mill



The narrow peak at 4.1ppm corresponds to both gaseous and chemical and physical adsorbed H2 molecules, whereas the broad peak beneath the narrow peak is due to the rigid lattice protons.

Gaseous H2 is observed below 100oC for LN2-milled samples, but not for RT-milled samples.At 180oC, the amount of H2 released from LN2-milled samples is 4 times more than that from

RT-milled samples.

ppm

-10-50510

LN2-180 min mill

No mill

RT-45 min mill

RT-180 min mill

2.7

2.7

1.4

2.02.7

2.0

1.5

0.0

2.6

1.80.0

ppm

RT 900MHz 6Li MAS NMR Spectra

0

5

10

15

20

25

Room Temp Dry Ice Liquid N2

Cry

stallit

e S

ize [

nm

]

Milling Temperature

LiNH2

LiH

0

10

20

30

40

50

60

180 min mill at RT 180 min mill at LN2

SSA

(m2 /g

)



Fig. B

LN2-processed LiNH2+LiH has larger crystallite sizes (Fig. A) and a lower specific surface area (Fig. B) than the LiNH2+LiH processed at RT.

Fig. C Fig. A

NMR analysis (Fig. C) reveals an up-field shift of the 6Li peak and a stronger surface defect peak (1.8ppm peak) for the LN2-ball-milled sample.

20

Enhancing the Reaction Enhancing the Reaction Kinetics of Mechanically Kinetics of Mechanically Activated, NanoActivated, Nano--Engineered Engineered LiNHLiNH22+LiH via +LiH via

Advanced CatalystsAdvanced Catalysts

Addition of 1 mol% nano-catalyst particles increases the hydriding/dehydriding kinetics.

Analysis of the dehydriding kinetics indicates that the dehydriding reaction of the catalyzed system is still diffusion-controlled. Thus, it is hypothesized that the added catalyst enhances desorption of NH3 from the Li2NH surface and thus creates a large composition gradient for diffusion within the Li2NH solid.

0

1

2

3

4

5

6

7

0 10 20 30 40

Co

mp

osi

sito

n (

wt%

H2

)

Time (hr)

Pure

Ceria

No catalyst

With catalyst

21

Diffusion-controlled hydriding

The slow hydriding process needs to be solved to make this material a viable storage system.

Improving Hydriding/Dehydriding Properties via Improving Hydriding/Dehydriding Properties via Thermodynamic DestabilizationThermodynamic Destabilization

Li2Mg(NH)2 + 2H2 Mg(NH2)2 + 2LiH

34.18m2/g

Cracking-assisted dehydriding

SSA: 22.99m2/g

58.11m2/g

Reversible storage capacity = 5.5 wt% H2ΔH (dehydriding) = 44 kJ/mol H2

Cycling at 200oC between H2 pressures of 68 and 0.01 bars

22

1) Hydrogenation below the melting temperature of LiBH4 is a strong function of the crystallite size and lattice microstrain.

2) The largest hydrogen uptake of 8.3 wt.% in the solid state is achieved using the particles with the smallest crystallites and largest lattice microstrain.

SolidSolid--State Hydrogenation/Dehydrogenation of State Hydrogenation/Dehydrogenation of LiBHLiBH44+MgH+MgH22 Enabled via NanoEnabled via Nano--Engineering and Engineering and

Mechanical ActivationMechanical ActivationLiBH4 + ½ MgH2 LiH + ½ MgB2 + 2 H2

0 2 4 6 80

50

100

150

200

250

300

Time (hr)

Tem

pera

ture

(o C

)

-2

0

2

4

6

8

d

c

b

Hydrogen (w

t%)

a

e

Curve b

Curve d

Curve e

Crystallite size of LiH

(nm)6.9 7.9 29.0

Crystallite size of

MgB2 (nm)6.1 9.9 18.1

Equivalent particle

diameter (nm)

258 230 217

Lattice microstrain

of MgB2

2.87% 1.94% 1.65%

23

SolidSolid--State Hydrogenation/Dehydrogenation of State Hydrogenation/Dehydrogenation of LiBHLiBH44+MgH+MgH22 Enabled via NanoEnabled via Nano--Engineering and Engineering and

Mechanical ActivationMechanical Activation

1.E-12

1.E-11

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

0.00 100.00 200.00 300.00 400.00 500.00Temperature (oC)

Inte

nsity

(Tor

r)

H2

H+

H2ON2

OH-

O2-

O2

BHx

CO2

The gas released during dehydrogenation of LiBH4 + MgH2 in the solid state is H2. Borane, if present, is below the detection limit of the mass spectrometer.

24

• Remainder of FY 2008:

Develop the atomic level understanding of the reaction mechanism and rate-limiting steps of mechanically activated, nano-engineered (LiBH4 + MgH2) systems (PNNL and UConn)

Further enhance hydriding/dehydriding rates of (LiBH4 + MgH2) systems(UConn)

Improve the hydriding/dehydriding cyclic stability of (LiBH4 + MgH2) systems by integrating mechanical activation, nano-engineering, and thermodynamic destabilization (UConn)

Further improve hydriding/dehydriding properties of (LiNH2 + LiH) systems via nano-engineering, mechanical activation, and thermodynamic destabilization (UConn)

Establish the atomic level understanding of the reaction mechanism and kinetics of mechanically activated, nano-engineered (LiNH2 + LiH) systems (PNNL and UConn)

Demonstration of hydrogen uptake and release of (LiBH4 + MgH2) systems with a storage capacity of ~ 10 wt% H2 at 2000C (UConn)

• FY 2009:

Future Work Future Work

25

Project Summary Project Summary Relevance: Explore fundamental mechanisms related to mechanical

activation and nano-engineering necessary for improving kineticsof reversible hydrogen storage materials.

Approach: Investigate hydriding/dehydriding properties of (LiNH2+LiH) andLiBH4-based materials with different degrees of mechanical activation and nano-engineering; Enhance the storage performance based on the understanding developed.

Technical Accomplishments: (i) Identified the reaction mechanism and rate-limiting step in the dehydriding process of the LiNH2 + LiH systems; (ii) Established the effect of mechanical activation on hydride particlesand their hydriding/dehydriding properties; (iii) Proved low temperature milling can introduce a large amount of defects to nano-particles, which can in turn enhance hydriding and dehydriding reactions, (iv) Demonstrated the improved hydriding/dehydriding rates via nano-engineering, mechanical activation, and advanced catalysts even through the reaction rate is controlled by diffusion; and (v) Demonstrated hydriding and dehydriding reactions of (LiBH4 + MgH2) systems at the solid state, i.e., below the melting point of LiBH4 (Tm = 2800C) with 8.3 wt% H2uptake.

effects and mechanisms of mechanical activation on ... · leon l. shaw, tippawan markmaitree,...

Documents