-1-

Effect of Trace Elements on Long-Term Cycling and Aging Properties

of Complex Metal HydridesDhanesh Chandra

University of Nevada, Reno (UNR)

This presentation does not contain any proprietary or confidential information

June 12, 2008

ST 37

-2-

Project start date – FY05Project end date – FY10Percent complete – ~60%

Total project funding (5yrs.) : $ 1.5 M (Requested)

DOE share (5yrs.) : $ 1.2 M

Contractor share (5yrs.) : $ 301 K

Funding received in FY07 : $ 520 K(Includes funding for major equipment)

Funding received in FY08 : $ 250 K

• SNL – Ewa Rönnebro• GE – Dr. J.C. Zhao (Now at Ohio State)• ESRF, Grenoble , France – Yaroslav Filinchuk• Univ. of Utah, - Dr. Z. Fang

Overview

Barriers Addressed• The effect of trace impurities on materials • Poor mechanistic understanding of some materials• Lack of characterization of material volatilization

Timeline

BudgetPartners

-3-

OverallObjective

UNR’s Focus Areas:I. The primary objective of the UNR Project is to determine the effects of gaseous trace impurities such as O2, CO, H2O, CH4 etc. in H2 on long-term behavior of the complex hydrides/precursors by pressure cycling and/or thermal aging with impure H2.II. Secondary related objectives: (a) Vaporization behavior of hydrides

(b) Crystal Structure studies 2006 Constructed high pressure (up to 100 bar) cycling equipment.

Performed hydrogen cycling studies on amide-imide and mixed alanates.

Vapor pressure behavior of Li3N and Mg(BH4)2 initiated.

HP DSC experiments, in-situ neutron, and x-ray diffraction studies2007(May, 15 2007-April 1, 2008)

Thermodynamic Studies:

A. Extrinsic Hydrogen Charging/Discharging effects: Determined the effects of gaseous impurities in hydrogen on Li2NH-LiNH2 and other systems.

B. Vaporization Thermodynamics: Worked on Mg Borohydride, and identified vapor species at moderate temperatures.

Crystal Structure Studies:

In-situ phase transformation studies on Ca(BH4)2

Objectives

-4-

Significance of Amide-Imide Studies

1. Li2NH-LiNH2 → total capacity of ~10.5 wt.% hydrogen , but currently ~5.6 wt.% is reversible. Further studies may lead to increased capacities

Temperature (oC)

Rev

ersi

ble

stor

age

capa

city

(wt %

H)

Courtesy of Dr. Zhang of GE

Significance of Mg(BH4)2 and CaBH4)2 Studies

1. Mg(BH4)2 → ~15 wt.% H2 capacity with ΔH ~ - 53 kJ/mol

2. Ca(BH4)2 → Potential candidate for hydrogen storage candidate

3. Mg(BH4)2 and Ca(BH4)2 → Vapor Pressures are important to understand vaporization during evacuation cycle of the hydriding/dehydriding.

Current Work

2. LiNH2-Li3AlH6 → Mixed Amide-Alanates are important because of their theoretical ~7 wt.% hydrogen storage capacity

16Mg(BH4)2

650°C

500 600

407°C

> 650°C

ΔH=-96 kJ/mol H2

ΔH=-97 kJ/mol H2

ΔH=-53 kJ/mol H2

LiBH4

NaBH4

320°C380°C

DOE 2010 Goal

0 100 200 300

2

0

4

6

810

1214

NaAlH

MgH2

LiNH2

2

400

DOE 2015 Goal

DOE 2010 Goal

DOE 2010 Goal

0

2

0

4

6

810

1214

LaNi5H6

NaAlHNaAlH4 Mg2NiH4

2LiBH4+MgH2

MgH2+2LiNH2

16

-5-

0 2 4 6 8 101E-4

1E-3

0.01

0.1

1

10

1000 1 2 3 4

P (B

ar)

wt% H

H: Li-N-H

Li4NH*

8.78%

9.10%

10.07 %

Absorption

Desorption

Pressure Cycling Li-N-H

LiNH2 + LiH

Imide

LiN

H

Li2NH

Li2ND – Cubic - Fm-3m - Z = 4a = 5.0476 Å, Vol = 128.602 Å3

LiNH

LiNH2 - BCTI-4 - Z = 8a = 5.0695 Å, c = 10.2599 ÅVol = 263.68 Å3

K. Miwa, N. Ohba, S. Towata, Y. Nakamori, and S. Orimo Physical Review, B71(2005), 195109.

Cubic

Fm3m

Li

N

Li3N

α-Li3N – Hexagonal P6/mmm - Z = 1a = 3.6587 Å, c = 3.876 ÅVol = 44.933 Å3

Cycling Region

2.36% 3.49%Li2NH

Li3N +2H2 ⇔ Li2NH+LiH

*Weidner et al.

*Weidner, E., D.K. Ross, et al. Chemical Physics Letters, 2007. 444(1-3): p. 76-79.

T. Noritake, Orimo, et al. J. Alloys Compds.393 (2005),264-268.

-6-0.01

0.1

1

10

100

0 1 2 3 4 5 6 7wt% H

P (b

ar)

1 Cycle H2100 Cycles CH4163 Cycles H2

3.6

H2

(1 cycle)

100 cycles 100 ppm CH4 in H2

4.2

163 cycles H2

Objective To assess Loss in Hydrogen

Capacity after Pr. cycling Nominally for ~100 cycles

Experiments1. Li2NH ↔ LiNH2 Pressure Cycled

~ 20 atm/vacuum at 225oC. Top–left

2. Absorption/desorption Isotherms (up to ~12 bar) using the Sieverts apparatus

Summary

Effect of 100 ppm CH4 in H2 :About 0.7 wt.% H2 capacity was lost after 100 pressure cycles. There was virtually no change in kinetic behavior

55 60 65 70

0

500

1000

1500

2000

0 100 200

0

500

1000

1500

2000

59 60 61 62 63 640.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

P (k

Pa)

t (hours)

Cycling Data of Imide-Amide with 100 ppm CH4 in UHP Hydrogen

hours

Absorption Cycle Desorption Cycle

Li2NH-LiNH2CyclesPr

essu

re (b

ar)

This

cyc

ling

data

from

CH

4–

H2

hours

-7-

0.1

1

10

100

0 1 2 3 4wt% H

P (b

ar)

501 Cycles H2560 Cycles H2 + H2O560 Cycles H2 + O2

1.5

O2H2O

2.7

Results: Thermodynamics: After ~500 Pressure cycles at 225oC → remaining Hydrogen

capacity is 1.5wt.% (with O2 additions) and 2.7 wt.% out of ~5.6 wt.% with H2O ( total reversible capacity). Cycling of Li2NH ↔ LiNH2 with Industrial hydrogen Water ~32 ppm, O2~10ppm showed ~2.6% hydrogen loss (500 cycles under similar cycling conditions)Loss in Capacity due to formation Li2O, and LiH and LiOHImportance: Presence of water in H2 is expected to have more impact on the loss of hydrogen capacity but it appears that there is greater loss observed when the experiments were conducted with O2 impurity in H2

Major Phases: Li2O, Li2NH and small amount LiOH phase

X-ray Diffraction Pattern - 560 Cycles with 100 ppm H2O in H2 Li2NH: Cubic

(Fm-3m)

LiH: Cubic(Fm-3m)

Li2O (MAJOR): Cubic (Fm-3m)

LiOH (small Fraction): Tetragonal (P4/nmm)

Effect of O2 and H2O in H2 Cycling between Li2NH-LiNH2

-8-

Dual Combined 100 bar H2Cycling and Sievert’s apparatus

Glove Box

Results: Pressure Cycling : Significantly more losses with O2 and H2O impurities. (Desorption cycles not shown for clarity)Thermal aging: 100 ppm CO with H2→verylittle loss in H2capacity for the imide–amide system (results not shown)XRD results show that there is residual Li2NH along with Li2O (major phases) and LiH phase.

SUMMARY PLOTS of Isotherms Before and After Li2NH-LiNH2 Cycling (No catalyst) with Industrial Hydrogen and 100 ppm levels of O2 , H2O,

CH4, in UHP H2

Summary: Weight loss After Cycling with 100 ppm Impurities in H2 gas

T=255oC

0.01

0.1

1

10

100

0 1 2 3 4 5 6 7

wt% H

P (b

ar)

0 0.5 1 1.5 2 2.5H/Li-N-H

1 Cycle H2100 Cycles CH4163 Cycles H2501 Cycles H2560 Cycles H2 + H2O560 Cycles H2 + O2

560O2

501 H2(Ind.)560H2O100CH4

163 H2(Ind.)1 H2(Ind.)

Li2NH-LiNH2Cycling

-9-

Li2NH ⇔ LiH ⇔ LiNH2 Cycle 2 Bar - Industrial Hydrogen @T= 255oC

Cycles k (1/s)

1 0.0075

163 0.0064

502 0.0035

1100 0.00075

*Please note: Note that cycling Li2NH ↔ LiNH2 in Industrial hydrogen Water ~32 ppm, O2~10ppm and others showed ~3.2% (out of ~4.4%) hydrogen loss after 1100 cycles under similar cycling conditions

ln oA

A

Ckt

C=

0 . .in= Concentration at time (t) & C @ 0

o

A

A A A

C Intial Concof Hyd gas reactionC C t

= −= =

The PCI’s were shown before ( as a reference) , and the kinetic data (right) is new

Rate Constant for the Hydriding reaction:0 1 2 3 4 5 6

0.1

1

10

0.0 0.5 1.0 1.5 2.0

Li2NH Starting Material After 56 Cycles After 163 Cycles After 501 Cycles After 1100 Cycles

P(b

ar)

W t% H

H: Li-N-H

~10.25

~0.86

k= 0.00075/s

k= 0.0075/s

Imide/amide Pr. Cycling with Industrial H2containing small levels* of O2 and H2O

y = 0.0035x + 0.0259

y = 0.0064x + 0.0418

y = 0.0007x + 0.014

y = 0.0075x + 0.0873

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 20 40 60 80 100t (seconds)

ln P

0/P

502 Cycles163 Cycles1100 CyclesNo cycles

Kinetic Plot for Cycling Data

Old Results

New Analysis

ktCC

A

A =0ln

0t@CC A0A ==

Kinetic Losses From Cycling in Industrial Hydrogen

-10-

Pressure Measurements

Torsion effusion system, available at UNR used to determine total equilibrium pressure we use the following Eq.: (‘K’=Fiber Constant)

( )( ) ( ) ( )

( ) ( )

22

21 1 1 2 2 2 1/ 2AVG 2

11 1 2 2

22 i

n

ii

dw a f d a f dRT dtM m Ma c a cK

πθ

−

=

⎛ ⎞ ⎡ ⎤+⎜ ⎟ ⎡ ⎤= ⋅ =⎢ ⎥ ⎣ ⎦⎜ ⎟ +⎢ ⎥⎜ ⎟ ⎣ ⎦⎝ ⎠∑

Molecular Weight Measurements of Vapors

Determined by rate of weight loss (TGA)

Disproportionation equations (below) in the vapor phase determined by equating the experimental MAVG to the theoretical Mol. Wt. of the effusing gas species:

( )( ) ( )1 1 1 2 2 2

2T

KP

a f d a f dθ

=+

Vapor Pressure Measurement Aspects of Mg(BH4)2 Using Knudsen Torsion Effusion Method

Typical Pr. Temp. and Sample Size:

• Turbo Pump vacuum (<10-5 Torrs)

• ~ 1 gram

• Temperature capability: -20oC to 600-700oC

. Tungsten Knudsen Cells Used

Pressure Equation

-11-

Knudsen Torsion Effusion apparatus used to measure thermodynamics

V.P and Molecular weight of the effusing species.Mg(BH4)2 (s) →Mg (s) +2B (s) +4H2(g)

VP Eq. : log P (bar) = 2.646- 3871/T)

Mg Metal +B (amorphous)?

Vapor Pressures of Mg(BH4)2, Mg, and MgH2

1. There are two principal reactions:Mg(BH4)2(s) → Mg(BH4)2(g)………..(~ 2%)

Eq. : log P (bar) = 2.216- 4485/T2. Mg(BH4)2(s) → Mg (s) +B(s) +4H2(g)

Eq. : log P (bar) = 2.646- 3871/T)thus Disproportionates………..(~98%)

3. The Average Mol. Wt. of effusing gas is 2.42 g/mol suggests that majority hydrogen is major component in vapor phase.

Portion of Glass tube with the baffles and Knudsen cell housing is shown

XRD Pattern Showing Mg Metal (residual sample)

-12-

Results:

• Vapor pressures of Mg (s) → Mg (g), and decomposition of MgH2 to Mg metal and H2

• Gibbs energy calculation of vaporization are listed in Figure

• At 225oC: PH2 = 8.8x10-6 atm, and PMg(BH4)2 = 2.03x10-7 atm

• No detrimental cations appear to be effusing out –stable

• Above 508 K (233.1oC) the ΔGo becomes negative and vaporization does starts.

( ) ( ) ( )2 6 2B H 2 B 3 Hg so lid g→ +

@ 298.15 K, 1 Bar ΔGRXN= - 87.6 kJ/mol (Possible)

( ) ( ) ( )2 6 2B H 2 B 3 Hg g a s g→ +

@ 298.15 K, 1 Bar ΔGRXN= +1042.4 kJ/mol (unlikely)

Issues related to formation solid or gaseous Boron (have significance in release of Borane gas)

Note: T > 275 oC rapid decomposition and there is vigorous hydrogen release

Gibbs Energies of Vaporization of Mg(BH4)2 and others

-13-

CRYSTAL STRUCTURE AND PHASE TRANSFORMATION STUDIES ON Ca(BH4)2

These studies were performed using High resolution and high temperature synchrotron x-ray diffraction at ESRF, Grenoble to understand the phase stability of Ca(BH4)2

SummaryThe sample contains 87% α-Ca(BH4)2 and 13% β-Ca(BH4)2 phases at room temperature.Phase transition from α → α’ occurred at 222oC (second order transformation) Then α’ → β occurred at 297oC Manuscript prepared: “Crystal Structures and Phase Transformations in Ca(BH4)2 , Y.

Filinchuck, E. Ronnebro and D. Chandra, 2008.

Phases Present : α + β α’ + β β

Synchrotron Data from the Ca(BH4)2 Specimen

-14-

α+βα+β

α αc↑

β α αβ

α

αβ

β α

β

β

ββ

α

β αa↓

β

44oC

222oC

297oC382oC

α+ β

α’+ β

In-situ synchrotron XRD patterns taken between 317- 573K

Synchrotron x-ray data Taken at Grenoble( λ = 0.711385 Å) α β+

Major Phase

Phases

α→ α’ is complete @ 222oC, α’ Bragg Peaks in Red (line) Pattern

The Bragg peaks of the β phase are shown in The XRD pattern (green line) shows complete transformation of α’-to- β phase

Summary of Phase TransitionsThe following Phase transition were observed

'α α β→ →222oC 297oC

α→α’ (2nd order transition complete at 222oC)

β Phase decomposes to unknown phases above 382oC

α’

-15-

Variation of the Unit cell Parameters of α and α’-Ca(BH4)2 as a Function of Temperature

Volume of the Ca(BH4)2 / formula unit in the α, , α’and β-polymorphs as a function of temperature

Summary

No change to lattice parameters of α phase until ~165oC,

Unit cell volume per formula unit of the α and α’phases shows increases as temperature ramps up.

The volume of the unit cell of β phase shows linear increase during heating

Small amount of β phase are always present at the start of the experiment mixed with the α- phase

Latti

ce P

aram

eter

s (A

ngst

rom

)

'222o C

α α→

300 320 340 360 380 400 420 440 460 480 500 520 540 560 5807

8

9

10

11

12

13

21/2a

F2dd I-42d

cb

a

c

Phase−α ' Phaseα −

Temperature (K)

Temp (K) 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580

104

105

106

107

108

109

110

111

112

113

114

P42nm

I-42d

F2dd

Phase−α' Phaseα −

Phase−β

Vol.

/Z (A

3 )

-16-

New Crystal Structures α, α’ and βphases of Ca(BH4)2

α’ Phaseβ Phaseα Phase

Projection onab plane Projection on

ac plane 3D- view

Summary

New phase transitions were observed and crystal structures determined as shown above.

Crystal Structures were determined in collaboration with Dr. Yaroslav Filinchuk (ESRF-Grenoble) and Dr. Ewa Ronnebro (Sandia National Laboratory).

Structure determination in final Stages in Collaboration with ESRF (Grenoble) and Sandia National Laboratory

CaBH

CaBH

CaBH

-17-

1. Continue Work on Effect of Impurities on Specific Contaminantso Pressure Cycling on mixed Mg-Li based complex hydrideso New 8 station combined cycling/Sievert’s hydriding apparatuso Testing of hydrides developed by MHCoE partners

2. In-Situ Neutron and X-ray Diffraction Studies on Hydriding/Dehydriding o Studies on Borohydride using X-ray and neutron diffraction

3. Vapor Pressure Studies on LiBH4 and other Borohydrides o Thermodynamics of vaporization of LiBH4 and others

4. Phase Diagram Determination of Mixed Complex Hydrideso Develop experimental non-equilibrium/equilibrium phase diagramso CALPHAD modeling at UNR

5. High Pressure Differential Scanning Calorimetric Research o Dynamic heating behavior at up to ~ 50 bar hydrogen

6. Hydrogen Lattice Dynamics Studies on Complex Hydrides- Prof. Cantelli, Univ. of Rome - IPHE Proposal

o "Hydrogen Dynamics, Lattice interactions, and Atomic-scale Structure of Complex/Chemical Hydrides"

o Collaboration between Cantelli-Rome, Italy and Chandra-Jensen, USA7. IEA/IPHE Collaborative Studies at Uni. of Geneva and CRNS (France)

o Proposal to study defect structures in the complex hydrides such as Li-Al hydrides , Mg-Li amides, and others

Future Work on Complex Hydrides(FY ‘08 and Beyond)

-18-

Summary of Technical Accomplishments

Imide-Amide (Li2NH-LiNH2) Impurity Effects (UNR Sample)Studies on trace amounts of impurity gases (100 ppm) such as O2, CO, H2O, and CH4, in H2 and industrial hydrogen, up to ~1100 cycles. O2 was most detrimental to the performance of amide-imide hydrides The kinetic analyses showed one order of magnitude change of the rate constants; from cycle 1 at 7.5x10-3/sec to 7.5x10-4/sec. after 1100 cycles.

Vapor Pressure Measurement of Mg(BH4)2 (Sample from GE )No significant vaporization of Mg(BH4)2. Below 233oC not possible to record any data. Above 233oC the ΔGo becomes negative and vaporization starts. ΔGo of Mg(BH4) was determined.Partial Pressures: PH2 = 8.8x10-6 atm, P Mg(BH4)2 = 2.03x10-7 atm, at 225oCMg(BH4)2 (s) →Mg(BH4)2 (g) (ΔH= 93.4 kJ/mol) (only ~2% of the vaporization). The majority of the vaporization was due to disproportionation of Mg(BH4)2 →Mg(s)+B(s)+4H2(g), ΔH= 44.82 kJ/mol (~98%)

Structure and Phase Transformations in Ca(BH4)2 (UNR-SNL Sample)in-situ synchrotron data showed two polymorphs of α-Ca(BH4)2 and a small amount of β-phase formed upon removal of solvent from Ca(BH4)2·2THF.A second order α→ α’ phase transition occurred at 222oC (confirmed by DSC).Another phase transition, α’→ β phase upon heating above 297oC and decomposes at 382oC into unknown products which according to TGA is associated with a weight loss, likely due to release of hydrogen.