Development of high hydrogen capacity complex and chemical hydrides for hydrogen storage

DEVELOPMET OF HIGH HYDROGE CAPACITY COMPLEX AD CHEMICAL HYDRIDES FOR HYDROGE STORAGE CHUA YOG SHE ATIOAL UIVERSITY OF SIGAPORE 2011... DEVELOPMET OF HIGH HYDROGE CAPACITY COMPLEX AD C

DEVELOPMET OF HIGH HYDROGE CAPACITY COMPLEX AD CHEMICAL HYDRIDES FOR

HYDROGE STORAGE

CHUA YOG SHE

ATIOAL UIVERSITY OF SIGAPORE

2011

DEVELOPMET OF HIGH HYDROGE CAPACITY COMPLEX AD CHEMICAL HYDRIDES FOR

HYDROGE STORAGE

CHUA YOG SHE

(B.Sc (Hons.), UTM)

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMET OF CHEMISTRY

ATIOAL UIVERSITY OF SIGAPORE

2011

in laboratory skills and structural analyses

Special thanks also go to Dr S Thomas Autrey, Dr Abhi Karkamkar and Dr Wendy J Shaw from Pacific Northwest National Laboratory (PNNL) for their useful discussion and assistance in high field NMR experiments in the collaboration works

I also want to thank all current and past members in Prof Chen’s group, whose assistance and encouragement composed every step in the past four years’ research course

I would like to thank National University of Singapore for providing the PhD research scholarship Much appreciation also goes to the staffs of the CMMAC laboratories, especially Mdm Han Yan Hui and Mr Wong Chee Ping for NMR support I am also thankful to the assistance that I have received from the laboratory officers of Surface Science Laboratory, Mr Wong How Kwong and Mr

Ho Kok Wen

Finally and most importantly, I would like to thank my family and friends, especially Ng Jeck Fei and New Siu Yee, for their support throughout the years

Table of Contents

Chapter 1 Introduction 1

1.1 DOE target for hydrogen storage 2

1.2 Materials for hydrogen storage 3

1.2.1 Metal hydrides 6

1.2.2 Complex hydrides 9

1.2.2.1 Activation for hydrogen release from complex aluminum hydride 11

1.2.3 Chemical hydrides 12

1.2.3.1 Proposed dehydrogenation mechanism 16

1.2.3.2 Activation for hydrogen release from AB 17

1.3 Thermodynamic alteration on hydrides 20

1.3.1 Destabilization 20

1.3.2 Amide-hydride interaction 23

1.3.3 Cationic substitution 28

1.3.3.1 Interaction of metal hydride and AB 28

1.3.3.1.1 Driving forces and reaction mechanisms 31

1.3.3.2 Interaction of metal amide/imide/nitride with AB 34

1.4 Motivation 35

Chapter 2 Experimental 38

2.1 Sample preparation and synthesis 38

2.1.1 Solid state ball milling 38

2.1.2 Liquid state reaction 39

2.2 Dehydrogenation 39

2.2.1 Open system dehydrogenation 39

2.2.2 Close system dehydrogenation 40

2.3 Characterizations 40

2.4 Quantification of NH3 41

Chapter 3 Investigations on the solid state interactions between lithium alanate with binary and complex amides 42

3.1 Introduction 42

3.2 Experimental section 43

3.2.1 Sample preparations and synthesis 43

3.2.2 Characterizations 45

3.3 Results and discussion 45

3.3.1 Reaction of lithium alanate with sodium amide 45

3.3.1.1 Metathesis between LiAlH4 and NaNH2 45

3.3.1.2 Dehydrogenation 50

3.3.1.3 Proposal of reaction mechanisms 53

3.3.1.4 Thermal decompositions of the post milled products 56

3.3.2 Reaction of lithium alanate with lithium aluminum ternary amide 58

3.3.2.1 Hydrogen release from the interaction of LiAlH4 and LiAl(NH2)4 58

3.3.2.2 Dehydrogenation pathway 64

3.4 Conclusion 68

Chapter 4 Syntheses, structures and dehydrogenations of alkaline earth metal amidoboranes and their ammoniate complexes 70

4.1 Introduction 70

4.2 Experimental section 75

4.2.1 Sample preparations 75

4.2.2 Characterization 77

4.3 Results and discussion 78

4.3.1 Calcium amidoborane (CaAB) 78

4.3.1.1 Synthesis and crystal structure 78

4.3.1.2 Dehydrogenation 82

4.3.2 Calcium amidoborane diammoniate (CaAB·2NH3) 83

4.3.2.1 Optimization 83

4.3.2.2 Crystal structure 85

4.3.3 Calcium amidoborane monoammoniate (CaAB·NH3) 103

4.3.3.1 Tuning NH3 content in CaAB·2NH3 103

Chapter 5 Mechanistic investigation on the formation and

dehydrogenation of calcium amidoborane diammoniate by 15

'H 3 and 'D 3 isotopic substitution

5.3 Results and discussion 136

5.3.1 Mechanistic understanding on the formation of calcium

Summary

The depletion of fossil fuel stimulates tremendous efforts in setting up a sustainable energy system Because of its abundance, high energy output and zero emission hydrogen is recognized as the most prospective energy carrier for the future energy system To utilize H2 as a fuel for transportation, a safe and efficient storage medium is needed

Lithium aluminum hydride (LiAlH4) and ammonia borane (AB) possess high hydrogen content of 10.5 wt% and 19.6 wt%, respectively, and thus, are attractive for hydrogen storage However, direct uses of these chemicals for hydrogen storage are not feasible due to their poor dehydrogenation kinetic and thermodynamic properties Therefore, the aim of this study is to improve the dehydrogenation properties of LiAlH4 and AB by chemically altering (or modifying) their dehydrogenation thermodynamics

Significant improvement in the dehydrogenation properties of LiAlH4 has been achieved by reacting it with NaNH2 or LiAl(NH2)4 As results of chemical alterations, LiAlH4 underwent different dehydrogenation pathways, releasing hydrogen rapidly under ambient temperature In the dehydrogenation process, the large combination potential of Hδ+ and Hδ- to H2 together with the formation of thermodynamic favorable Al-N bond induce a direct interaction of the materials to form H2, giving rise to the formation of Li-Al-N-H product

In the chemical modification of AB by reacting it with calcium amide in THF solvent, high purity calcium amidoborane (CaAB) was yielded; whereas in a solid-state reaction of AB with Ca or Mg amides or imides, a new class of materials, namely metal amidoborane ammoniate, with high hydrogen content was discovered These newly developed materials possess high hydrogen content of > 11 wt% and demonstrate attractive dehydrogenation behaviors In the study, NH3 was found to play a vital role in stabilizing crystal structures of ammoniates, forming strong coordination with metal cation and establishing a complete dihydrogen bond network within the structures As a consequence of dihydrogen bonding, N-H and B-H bonds are weakened, resulting in a significant reduction in dehydrogenation temperature Furthermore, stoichiometric conversion of NH3 to H2 was achieved, allowing more hydrogen to be released in the dehydrogenation

Due to the attractive properties of amidoborane ammoniates, mechanisms of their formation and dehydrogenation were studied by using isotopic labeling It was found that the formation of ammoniate sample was attributed to the combination of one Hδ+ from AB and [NH2]- unit in amide, i.e Ca(NH2)2, to form NH3, and the remaining [NH2BH3]- unit of AB may bond to Ca2+ as a result of electrostatic force

In the investigation of the dehydrogenation mechanism, it was evidenced that the improved dehydrogenation properties in amidoborane ammoniate was resulted from the participation of NH3 molecule in the dehydrogenation of metal amidoborane

Stimulate by the mechanistic study, the dehydrogenation thermodynamics of CaAB was chemically modified by reacting it with LiNH2 and the results showed that an even lower dehydrogenation temperature with a milder reaction enthalpy can be achieved in comparison to that of CaAB

Overall, significant improved dehydrogenation properties have been achieved in the chemically modified materials as compared to the pristine materials (LiAlH4 and AB) A few new materials were synthesized which exhibit remarkable dehydrogenation properties

List of Tables

Table 1.1 DOE targets for onboard hydrogen storage systems for

light-duty vehicles

3

Table 1.2 Comparisons on currently available hydrogen storage methods 6

Table 1.3 Amide-hydride combinations for hydrogen storage

Reproduced with permission from Ref 93

Table 4.4 Calculated structural parameter of Mg(NH2BH3)·NH3 (space

group P21/A, a = 8.8174 Å, b = 9.0794 Å, c = 8.1299 Å and β

Table 4.8 Amount of NH3 coproduced in the dehydrogenation process of

MgAB·NH3

129

List of Figures

Figure 1.1 Representation of physisorption (cryosorption) and

chemisorption processes Reproduced with permission from Ref 7 Reproduced by permission of the PCCP Owner Society

5

Figure 1.2 Low-temperature (orthorhombic) crystal structure of

ammonia borane (AB) N, B and H atoms are depicted in blue, purple, and gray, respectively The dihydrogen bonds are delineated by dashed lines Reproduced with permission from Ref 53

13

Figure 1.3 Energy diagram on hydride destabilization through alloy

formation Reproduced with permission from Ref 13

21

Figure 1.4 Weight variations during hydrogen absorption and

desorption processes over Li3N samples Abs: absorption;

Des: desorption Reproduced with permission from Ref 16

24

Figure 1.5 Schematic diagram of the crystal structure of LiNH2BH3 and

NaNH2BH3 determined from high-resolution X-ray powder diffraction data at room temperature Boron is represented

by orange spheres; nitrogen by green spheres; hydrogen by white spheres and lithium by red spheres Reproduced with permission from Ref 80

29

Figure 1.6 Time dependence of hydrogen desorption from alkali

amidoboranes and post milled BH3NH3 sample at about 91

°C Reproduced with permission from Ref 80

30

Figure 1.7 Schematic energy profiles for the L* (oligomerization) and

L (non-oligomerization) pathways of dehydrogenation in (LiNH2BH3)2 Reproduced with permission from Ref 112

33

Figure 3.1 XRD patterns of self made LiAl(NH2)4 and LiAlH4 44

Figure 3.2 Time dependences of hydrogen evolution from the S-I

(LiAlH4-2NaNH2) and S-II (2LiAlH4-NaNH2) samples in the milling process, respectively

46

Figure 3.3 FTIR spectra of the S-I and S-II samples collected at

different reaction stages

48

Figure 3.4 XRD patterns of the S-I (LiAlH4-2NaNH2) sample collected

at different reaction stages

48

Figure 3.5 XRD patterns of the S-II (2LiAlH4-NaNH2) sample

collected at different reaction stages

49

Figure 3.6 27Al, 23Na and 7Li NMR spectra of samples collected at

different milling stage of the S-I and S-II (a-b) S-I stage 2 and 4, (c-d) S-II stage 2 and (e-f-g) are LiAlH4, LiAl(NH2)4

and LiH included for comparison

53

Figure 3.7 DSC (a) and TPD (b) measurements on the post milled

products of S-I (solid line) and S-II (dash line) samples

57

Figure 3.8 Time dependences hydrogen desorption from ball milling

LiAlH4-LiAl(NH2)4 mixtures at the mole ratios of 1:1 and 3:1 at ambient temperature

59

Figure 3.9 TPD-MS curves of LiAl(NH2)4/LiAlH4(1:3) homogenous

(up) and pure LiAl(NH2)4 (down)

60

Figure 3.10 DSC measurement of LiAl(NH2)4 and homogenous mixing

LiAl(NH2)4-LiAlH4 (1:3)

61

Figure 3.11 Volumetric release of grinded LiAl(NH2)4-LiAlH4 (1:3) 62

Figure 3.12 XRD pattern of LiAlH4/LiAl(NH2)4 (1:3) homogenous

mixing (a), evolved 8 equiv H (b), 10 equiv H (c), 16 equiv

H (d) and post milled LiAlH4/LiAl(NH2)4 (1:1) (e) Peaks correspond to platinum (Pt) are marked with asterisks

66

Figure 3.13 FTIR spectra of self-made LiAl(NH2)4 (a),

LiAl(NH2)4/LiAlH4 (1:3) evolved 8 equiv H (b), 10 equiv H atoms (c) and 16 equiv H atoms (d)

66

Figure 3.14 27Al (A) and 7Li (B) MAS solid state NMR spectra and

post milled LiAl(NH2)4/LiAlH4 (1:3) with 16 H evolved (a),

10 H evolved (b), 8 H evolved (c), homogenous mixing (d), LiAl(NH2)4 (e), LiH (f), AlN(g) and LiAlH4 (h) Asterisks denote the spinning sidebands

67

Figure 4.1 Time dependence of NH3 release from the reaction between

Ca(NH2)2-2AB in THF solution at 25 °C

79

Figure 4.2 Proton coupled (top) and decoupled (bottom) 11B NMR

spectra of calcium amidoborane

80

Figure 4.3 XRD pattern of calcium amidoborane (CaAB) upon THF

removal

80

Figure 4.4 Molecular packing and close contacts in crystal CaAB

Calcium is represented by green spheres, nitrogen by blue spheres, boron by pink spheres and hydrogen by white spheres

81

Figure 4.5 TPD-MS measurement on the CaAB synthesized by

reacting Ca(NH2)2 and 2 equiv of AB in THF

82

Figure 4.6 XRD patterns of the mixture of Ca(NH2)2-AB with the

molar ratios of 1:1, 1:2 and 1:4 collected after 10 hours of ball milling

84

Figure 4.7 XRD patterns of the mixture of Ca(NH2)2-AB with the

molar ratio of 1:2 collected at different ball milling intervals

84

Figure 4.8 (A) XRD patterns of AB, Ca(NH2)2, CaAB (derived by

heating CaAB·2NH3 at 100oC in vacuum) and CaAB·2NH3 (B) Experimental (smooth line) and calculated (+) profiles

of CaAB·2NH3 The differences between the experimental and calculated values were shown below the observed patterns

85

Figure 4.9 Molecular packing and network of N–H···H–B dihydrogen

bonding in crystal CaAB·2NH3 Calcium is represented by green spheres, nitrogen by blue spheres, boron by pink spheres and hydrogen by white spheres

88

Figure 4.10 FTIR spectra of Ca(NH2)2, AB, CaAB·2NH3 and the

post-dehydrogenated CaAB·2NH3 sample

88

Figure 4.11 11B MAS solid state NMR spectra of AB, CaAB,

CaAB·2NH3 and post-dehydrogenated CaAB·2NH3 sample

89

Figure 4.12 DSC and TPD measurements on the decomposition of

CaAB·2NH3

92

Figure 4.13 XRD pattern of the sample collected after ammoniation of

CaAB Asterisks denote the XRD reflections of CaAB·2NH3

92

Figure 4.14 Volumetric release of CaAB and CaAB·2NH3 Temperature

was raised to 300 °C at a ramping rate of 0.5 °Cmin-1

94

Figure 4.15 Volumetric release of CaAB·2NH3 at 80 °C Temperature

was raised to 80 °C at a ramping rate of 0.5 °Cmin-1 and held for 2 hours

95

Figure 4.16 MS measurement of the gaseous product collected after

heating CaAB·2NH3 at 80 °C for 2 hours H2 is the main gaseous product

96

Figure 4.17 Volumetric release of CaAB·2NH3 at different ramping rate

to 150 °C and held for overnight

97

Figure 4.18 MS measurement of the gaseous product collected after

heating CaAB·2NH3 to 150 °C at the rate of 0.5 °Cmin-1 H2

is the main gaseous product

98

Figure 4.19 MS measurement of the gaseous product collected after

heating CaAB·2NH3 to 150 °C at the rate of 2 °Cmin-1 H2 is the main gaseous product

98

Figure 4.20 MS measurement of the gaseous product collected after

heating CaAB·2NH3 to 150 °C at the rate of 5 °Cmin-1 H2 is the main gaseous product

99

Figure 4.21 Volumetric release of CaAB·2NH3 at pre-set temperatures 101

Figure 4.22 DSC and TPD measurements on the CaAB·2NH3 collected

after volumetric release at (i) 90 °C, (ii) 130 °C and (iii) 150

°C, respectively

102

Figure 4.23 C80 measurement of CaAB·2NH3 102

Figure 4.24 XRD patterns of the post milled CaAB·2NH3 + nCaAB

samples, with n = 1, 2 and 3 XRD spectra of CaAB and CaAB·2NH (n = 0) were included for comparison

104

Figure 4.25 Raman spectra of the post milled CaAB·2NH3 + nCaAB

samples with n = 1, 2 and 3 CaAB and CaAB·2NH3 (n = 0) were also included for comparison

105

Figure 4.26 XRD patterns of post milled CaNH-2AB and CaAB·2NH3 +

CaAB samples

106

Figure 4.27 DSC and TPD measurements on the post milled

CaAB·2NH3 + nCaAB samples with n = 1, 2 and 3

107

Figure 4.28 Volumetric release measurements of the post milled

CaAB·2NH3 + nCaAB samples with n = 1, 2 and 3

Temperature was raised to 300 °C at a ramping rate of 0.5

°C/min

109

Figure 4.29 C80 measurements of post milled CaAB·2NH3 + nCaAB

samples with n = 1, 2 and 3 Temperature was raised to 300

°C at a ramping rate of 0.5 °C/min

109

Figure 4.30 FTIR measurements on the post dehydrogenated

CaAB·2NH3 + nCaAB samples with n = 1, 2 and 3

110

Figure 4.31 Reactions of NH3BH3 (AB) and hydrides in THF 112

Figure 4.32 XRD patterns of the post milled Mg(NH2)2-AB in the molar

ratios of 1:1 and 1:2

114

Figure 4.33 11B solid state MAS NMR measurements of the mixture of

Mg(NH2)2-AB in the molar ratio of 1/2 before and after being treated at different temperatures 11B NMR spectra of

AB was included for comparison

114

Figure 4.34 FTIR spectra of AB, MgAB·2NH3 and the

post-dehydrogenated MgAB·2NH3 sample

115

Figure 4.35 DSC and TPD measurements on the decomposition of

MgAB·2NH3 under a flow of argon

116

Figure 4.36 Volumetric measurements on MgAB·2NH3 at different

temperatures with a heating rate of 0.5 °C/min

117

Figure 4.37 XRD patterns of post milled MgNH-2AB and post

deammoniated Mg(NH2)2-2AB samples

120

Figure 4.38 Rietveld fit of the XRD pattern of MgAB·NH3. 121

Figure 4.39 Molecular packing and network of N-H···H-B dihydrogen

bondin MgAB·NH3 (a) and close contacts around Mg2+

center (b) Mg is represented by green spheres, N by blue spheres, B by orange sphere, H by white sphere

121

Figure 4.40 FTIR spectra of AB (a), MgAB·NH3 (b) and

post-dehydrogenated MgAB·NH3 (c)

123

Figure 4.41 Simulated structure of Mg(NH2BH3)2 Magnesium is

represented by green sphere, nitrogen by blue sphere, boron

by orange sphere, and hydrogen by white sphere

124

Figure 4.42 High field (18.8T) 11B solid state NMR spectra of

MgAB·NH3 samples before (a) and after (b) dehydrogenation Asterisks denote spinning side bands

125

Figure 4.43 TPD-MS (a) and volumetric release (b) measurements on

MgAB·NH3 at the heating rates of 2 °C/min (a) and 0.5

Figure 4.45 C80 calorimetric measurement on MgAB·NH3 at the

heating rate of 0.5 °C/min

128

Figure 5.1 TPD measurement on Ca(NH2BH3)2·215NH3 137

Figure 5.2 Mass spectrometer blank test on 15NH3 gas 138

Figure 5.3 TPD of the post milled Ca(15NH2)2-2AB 138

Figure 5.4 Raman characterizations on a) CaAB, b) CaAB preabsorbed

2ND3, c) CaAB preabsorbed 2NH3, d) CaCl2·2ND3 and (e) CaCl2·2NH3 (for comparison)

141

Figure 5.5 TPD-NH3 of CaAB preabsorbed ND3 142

Figure 5.6 Composition of gaseous product obtained at different stage

of close system dehydrogenation

143

Figure 5.7 In-situ 11B NMR characterization of CaAB·2NH3 at 90 °C 145

Figure 5.8 11B MAS solid state NMR spectra of CaAB·2NH3 collected

at a) before heating and b) t = 0 at 90 °C

145

Figure 5.9 The 1H→11B cross-polarization spectrum of the post 90 °C

dehydrogenated CaAB·2NH3 sample

146

Figure 5.10 Volumetric release measurement of CaAB·2NH3 147

Figure 6.1 XRD patterns of CaAB and post milled CaAB-LiNH2

sample

153

Figure 6.2 XRD patterns of the mixture of CaAB-LiNH2 with the mole

ratio of 1:2, 1:1 and 2:1 collected after 5 hours of ball milling

Figure 6.5 11B MAS solid state NMR spectra of CaAB, post milled

[CaAB-LiNH2] sample and post dehydrogenated [CaAB-LiNH2] sample

156

Figure 6.6 TPD measurement on the post milled CaAB-LiNH2 mixture

TPD on CaAB was included for comparison

157

Figure 6.7 C80 calorimetric measurements on the post milled

CaAB-LiNH2 and the pristine CaAB samples

159

Figure 6.8 Volumetric release measurements on the post milled

CaAB-LiNH2 sample Doted line indicates the heating profile

160

Figure 6.9 XRD pattern of the post milled [CaAB-LiNH2] before and

after dehydrogenation

161

List of Schemes

Scheme 1.1 The proposed thermal dehydrogenation mechanism of

ammonia borane showing discrete induction, nucleation and growth steps leading to hydrogen release Reproduced

with permission from Ref 63

17

Scheme 1.2 Schematic diagram of the decomposition of AB based on

the in situ NMR results Reproduced with permission from

Ref 65

17

Scheme 1.3 Mechanism involved in the amide-hydride interaction 24

Scheme 1.4 Mechanism involved in the cation substitution 32

Scheme 3.1 Reaction pathways of S-I and S-II samples in the milling

Chapter 1

Introduction

The soaring global demands on energy together with the increasing awareness of the depleting resources of fossil fuel, have brought about an imperious need to shift to a cleaner and more sustainable energy system The potential of hydrogen to be utilized as an alternative energy carrier has long been suggested due to its cleanlines and ubiquitousness When burnt in an internal combustion engine, hydrogen produces zero emissions; when powering

a fuel cell, its only by-product is water In order to utilize it for vehicular application, hydrogen has to be stored on-board using a safe and energy efficient storage medium To realise hydrogen powered vehicle, hydrogen storage system has to match up with the properties of gasoline Based on the performance of gasoline powered vehicle, the U.S Department of Energy (DOE) has set the targets for onboard hydrogen storage systems for light-duty vehicles, which include high gravimetric and volumetric H2 densities, moderate operating temperature, fast dehydrogenation kinetics, good reversibility, low toxicity, etc.1These targets have been widely used as a benchmark to evaluate the potential of materials for hydrogen storage However, till now none of the viable hydrogen storage systems, even the most well-developed pressurized tank and cryogenic

liquid hydrogen techniques, could meet all the DOE targets Thus, hydrogen storage still remains a challenging technical issue

1.1 DOE target for hydrogen storage

To realize hydrogen powered vehicle, DOE has translated the performance

of a gasoline powered vehicle into the targets for hydrogen storage system1 Table 1.1 shows the latest DOE hydrogen storage targets established in 2009 The 2010

targets would allow some low volume vehicles to achieve a driving range of 300 miles (~480 km) To allow such a driving range, a system capacity with gravimetric and volumetric densities of 4.5 wt% and 2.8 kgH2/L is a prerequisite Even more challenging targets (by year 2015) have been identified to enable the same driving range for mid-sized vehicles It is worthy of noting that these targets are for an integral hydrogen storage system (including auxiliary parts associate with the storage system) Hence, the hydrogen storage capacity required for materials alone should far exceed the stated targets On top of it, hydrogen delivery temperature (to fuel cell) of -45 to 85 °C is also identified based on the operating temperature limit

of the fuel cell A temperature tolerance of up to 85 °C indicates that the waste heat from fuel cell can be utilized for hydrogen generation To have a comparable driving experience with that of gasoline vehicle, hydrogen charging and discharging rates of 1.2 kgH2/min and 1.5 kgH2/min should also be achieved by year 2010 and 2015, respectively Furthermore, a high durability of the storage system of 1000 cycles by 2010 and 1500 cycles or equivalent to 150,000 miles by

2015 has to be achieved

Table 1.1 DOE targets for onboard hydrogen storage systems for light-duty

System volumetric capacity 0.9 kWh/L

0.028 kgH 2 /L system

1.3 kWh/L 0.04 kgH 2 /L system

Durability/operability

Min/max delivery temperature

Operational Cycle life

System fill time (for 5 kg H 2 )

Min full flow rate

4.2 min 0.02 (g/s)/kW

3.3 min 0.02 (g/s)/kW

1.2 Materials for hydrogen storage

Today’s available hydrogen storage technologies can be classified into two major categories, conventional and solid state hydrogen storage systems Conventional hydrogen storage systems are the most well developed hydrogen storage technology today,2, 3 in which hydrogen is compressed or liquefied in a tank Most of today’s prototype fuel cell vehicles use either of them as hydrogen storage system.4, 5 Not surprisingly, the state-of-the-art hydrogen storage tank can store up to 700 bar H2.6 Besides compressed H2, storing hydrogen in liquid form offers higher mass density per volume than that in compressed H2 However, these conventional systems are less energy efficient as large amount

of energy is penalized for compressing or liquefying gaseous H2 Furthermore, safety issues remain a concern for vehicular application Solid-state hydrogen storage, i.e., storing hydrogen in solid material, offers several advantages over the conventional hydrogen storage systems with regards of the H2 storage volumetric and gravimetric densities, energy efficiency and safety In general, solid-state hydrogen storage materials can be classified into two groups, depending on how hydrogen is stored These materials either undergo hydrogen

physisorption (physical storage) or chemisorption (chemical storage)

Comparisons on physical and chemical storage are shown in Figure 1.1

In a physisorption, molecular hydrogen is adsorbed onto the surface of materials as a result of weak Van der Waals forces between substrates and hydrogen molecules The H2-substract surface interaction potential has an attractive interaction originating from long-range forces, and a repulsive force originating from short-range interaction The former is produced by fluctuations

in the charge distribution of the hydrogen molecules and of the atoms on the surface giving rise to dipole-dipole attraction, while the latter is caused by the overlap between the electron cloud of the hydrogen molecule and of the substrate The combination of these two forces results in a small adsorption enthalpy of 1-10 kJ/mol As the force involved in the interaction between H2 gas and substrate is relatively weak, physisorption usually takes place only at low temperatures Generally, hydrogen uptake in these materials is governed by specific surface area, pore sizes and the structure of the materials Representative examples for physorption include those of highly porous solids, comprising mainly metal organic frameworks (MOFs) and carbon materials

In a chemisorption process, an activation energy is needed to break the H–H bond when the adsorption is dissociative and the hydrogen atom is stored

interstitially in the lattice of heavier atoms (Figure 1.1) As hydrogen is

strongly bound to the host as a result of the formation of chemical bond, the enthalpy involved in chemisorption is greater than that for physisorption The host materials are represented by metals or metal alloys, complex hydrides and chemical hydrides

Figure 1.1 Representation of physisorption (cryosorption) and chemisorption

processes.7 Reproduced by permission of the PCCP Owner Society

Table 1.2 summarizes the potential of each hydrogen storage method for

vehicular applications Obviously, both physisorption and chemisorption H2

storage methods show basic fulfilment of DOE primary targets on H2gravimetric and volumetric densities However, for physisorption, a significant amount of H2 adsorption can only be achieved at cryogenic temperature (77K) This prerequisite shades its attractiveness for onboard application For chemisorption, although most metal or complex hydrides demonstrate high thermal stabilities, recent efforts in thermodynamic and kinetic improvements shine a light towards possible vehicular applications Chemical hydride, on the other hand, releases H2 irreversibly, nevertheless, its relatively high H2 capacity make it attractive as a “single-use” fuel, where the spent material has to be removed from the vehicle for off-board regeneration Thus, a focus on chemical storage (chemisorption H2) with emphasis on complex and chemical hydrides will be reviewed in the following section

Table 1.2 Comparisons on currently available hydrogen storage methods

Storage methods Advantages Drawbacks

(MOFs, carbon materials)

High gravimetric and volumetric densities

Low temperature (77K) for H 2 adsorption Interstitial metal hydride

(Chemisorption H 2 )

High volumetric density, suitable thermodynamic properties

Low gravimetric density

1.2.1 Metal hydrides

Metal hydrides have been extensively studied as hydrogen storage materials for a few decades.8, 9 Most of the metal hydrides are stable under ambient condition Due to the exothermic reaction between metal and hydrogen, reversible release of

H2 from hydride is possible under appropriate temperature and pressure.9 Several intermetallic alloys of AB5, AB3, AB2, AB and A2B types have been investigated for hydrogen storage In these alloys, element A is usually a rare earth or an alkaline earth metal and tends to form a stable hydride; whereas element B is usually a transition metal and forms unstable hydride Hydrogen is stored interstitially in the lattice of the heavy atoms of A and B Most of the intermetallic alloys store hydrogen reversibly In general, the formation and decomposition of an interstitial metal hydride can be represented by the equation 1.1:

The reversibility of hydrogen storage in hydride is governed by its thermodynamic property Thermodynamic factors dictate the equilibrium relationship between the operating temperature and pressure and can be described

by the van’t Hoff equation 1.2:

H P

where PH2 is the equilibrium pressure, P0 is a reference pressure (atmospheric

pressure), R is the gas constant, T is the absolute temperature, and ∆H is the enthalpy change and ∆S is the entropy change involved in the formation of the

metal hydride

For example, well known interstitial metal hydrides for hydrogen storage are the AB5-type (LaNi5)hydride and A2B-type (Mg2Ni) hydride LaNi5 has a hydrogen content of ~1.4 wt% and a H2 equilibrium pressure of 2 bars at ambient temperature,10 whereas Mg2Ni has a higher hydrogen content of ~3.6 wt% with H2

equilibrium pressure of 1 bar at 555K11

Due to their appropriate H2 reversibility and cyclability, metal hydrides are well suited for stationary applications However, their applications for vehicular purpose are plagued by high price and low gravimetric density (due to the high atomic weight of alloys) This is also true for AB, AB2 and AB3-type alloys.9Therefore, research efforts have shifted towards the use of lighter and cheaper metal hydrides, for instances, simple binary hydrides.12, 13 LiH and MgH2 are both light metal binary hydrides which attract significant attention owing to their high

hydrogen capacities (~12.5 and 7.7 wt % of H2, respectively) However, these metal hydrides are too stable that 910 °C and 277 °C are required for an equilibrium pressure of 1 bar for LiH and MgH2, respectively.14, 15 Because of its thermal stability, LiH was not considered as a practical hydrogen storage material Until recently, Chen et al.16 reported an interaction of LiNH2 and LiH which utilizes H- on Li and H+ on NH2 for direct dehydrogenation at much reduced temperatures with reversible H2 storage Vajo et al.13, on the other hand, demonstrated an effective way to destabilize LiH by using Si More detailed reviews on the thermodynamic alterations of LiH will be discussed in section 1.3 MgH2, which demonstrates a better thermodynamic property than LiH, has been studied extensively.17-23 In addition to its thermodynamic stability, MgH2 also suffer from slow dehydrogenation kinetic.24, 25 Therefore, research efforts have been focused on both catalytic improvements17-23 and chemical compositional alterations11, 13, 16 of MgH2 Improved dehydrogenation kinetics at low temperatures have been achieved with additions of transition metals, such as Ti, Nb V, Mn, Fe,

Co, Ni, and some of their oxides.17-23 One of the most notable study was the nano-sized Ni-doped MgH2 reported by Hanada and coworkers23 which showed a 6.5 wt% H2 storage capacity at 150-250 °C Despite of significant kinetic improvement in the H2 charging and discharging, the reaction enthalpy involved in the dehydrogenation process remains unchanged Research efforts in modifying dehydrogenation thermodynamic of metal hydrides have been achieved by employing destabilization or amide-hydride interaction strategies In those modifications, dehydrogenation thermodynamic of the precursors (metal hydrides) can be altered by the additions of alien species, which in turn resulted in the

significant changes in the dehydrogenation pathways.11, 13, 16 Detailed reviews will

be discussed in the section 1.3

1.2.2 Complex hydrides

Complex hydrides (M+ [XH]-) are salt-like materials which are made up of cations and anions The cations and anions are ionically bonded to each other The negatively charged anion group consists of hydrides which are covalently bonded

to a metal or metalloid atom, such as boron or aluminum Complex aluminum hydrides have been known for many years and they are widely used in organic synthesis as reducing agents.26-28 In addition, their low molecular weights and high hydrogen contents have drawn research focuses on their hydrogen storage properties.29-32

The researches on complex aluminum hydrides for hydrogen storage are somehow limited until the breakthrough investigation on the Ti-catalyzed NaAlH4

by Bogdanovic et al.33 in 1997 Among complex aluminum hydrides, NaAlH4 is the most widely investigated material in the past 10 years NaAlH4 was first synthesized by reacting NaH with AlBr3 in the presence of dimethyl ether.34 This synthetic approach yielded only 60% of NaAlH4 In 1960’s, a high yield direct synthesis of NaAlH4 was reported by reacting either NaH or metallic Na with Al in THF under elevated H2 pressure (see equations 1.3 and 1.4).35 Lately, due to the exceptionally effective TiCl3 catalyst in the dehydrogenation of NaAlH4, Ti catalyst was introduced directly during the synthesis of NaAlH4.36-38

Na + Al + 2H2 → NaAlH4 (1.4)

NaAlH4 contains 7.4 wt% of H2 and undergoes step-wise dehydrogenation which is accompanied by phase transformations Thermal decomposition of NaAlH4 was extensively studied by Dilts et al.39 and Dymova et al.40 in 1970’s NaAlH4 melts at temperatures between 165-205 °C,39 follows by an endothermic reaction to give ca 3.4 wt% of H2 and Na3AlH6 (equation 1.5) The subsequent release of H2 from Na3AlH6 in the temperature range of 250-300 °C is also an endothermic reaction,39 with approximately 2.1 wt% of H2 detached (equation 1.6) The final dehydrogenation step involves the decomposition of NaH to Na (equation 1.7) which occurs at relatively high temperatures (> 300 °C).39 Therefore, only H2

from the first and second steps (5.4 wt %) are useful for hydrogen storage

The endothermic natures of the first and second reactions suggest that H2 can

be recharged on Na3AlH6 + 2Al or 3NaH + Al to regenerate NaAlH4 From the thermodynamic data, decomposition of NaAlH4 and Na3AlH6 can occur at much lower temperatures than the experimentally observed temperatures The difference

in temperature is resulted from kinetic barrier which can be overcome by catalytic modification or nanosizing

LiAlH4 which has a hydrogen content of 10.6 wt% is more attractive than

NaAlH4 in term of hydrogen capacity Since the discovery of the reversible Ti-doped NaAlH4, LiAlH4 has attracted a flurry of research interest as a potential hydrogen storage material LiAlH4 is thermodynamically unstable although it decomposes in a similar manner to that of NaAlH4 However, due to kinetic hindrance, LiAlH4 is stable at ambient temperature and even under high energy ball milling.31 LiAlH4 decomposes exothermically (∆H = -10 kJ/mol H2) into Li3AlH6

at 150-175 °C to release 5.3 wt.% of H2 (equation 1.8).39 Further increasing in temperature to 180-220 °C would result in the decomposition of Li3AlH6 to LiH and Al (see equation 1.9), releasing 2.6 wt% of H2 endothermically (∆H = 25 kJ/mol H2).39, 41

Due to the presence of severe kinetic barrier, high temperature is required for the dehydrogenation Therefore, research efforts have been focused on the kinetic improvement of LiAlH4 by nanosizing and catalytic modifications

1.2.2.1 Activation for hydrogen release from complex aluminum hydrides

In 1997, Bogdanovic and Schwickardi33 demonstrated a remarkable improvement in dehydrogenation properties of NaAlH4 by doping it with TiCl3 By adding 2% of TiCl333 to NaAlH4, a temperature reduction of ~80 °C can be achieved, giving an equilibrium pressure of 1 bar at ~36 °C for the first dissociation

step (∆H = 37 kJ/mol-H2) and at ~116 °C for the second step (∆H = 47 kJ/mol-H2).37

Similar to that of NaAlH4, introducing Ti-based catalyst into LiAlH4 by ball milling was reported to show kinetic enhancement in the dehydrogenation process

In fact, improved dehydrogenation kinetics can also be achieved in the dehydrogenation of LiAlH4 to Li3AlH6 with increasing ball milling time.42 As reported by Blanchard and coworkers43, adding 5% of TiCl3.1/3(AlCl3) reduced the onset dehydrogenation temperature of LiAlH4 for ~90 °C On the other hand, Balema et al.44 reported a spontaneous decomposition of LiAlH4 to Li3AlH6 by ball milling 3 mol% of TiCl4 with LiAlH4 for 5 minutes A microcrystalline intermetallic Al3Ti, which was formed in the ball milling process, acts as an active catalyst in the dehydrogenation of LiAlH4.45

Besides catalytic improvements, recent efforts have also been focused on tuning the reaction thermodynamic of LiAlH4 to achieve reversible hydrogen storage One of the prospective thermodynamic modification strategies is amide-hydride interaction Various combinations of LiAlH4-amide have been reported and in all cases, H2 is released with ease as a result of direct combination

of Hδ- (hydride) and Hδ+ (amide).46-50 In fact, some combinations demonstrated reversible release of H2 Detailed review on amides-hydrides modification will be discussed later

1.2.3 Chemical hydrides

Chemical hydride discussed here is referring to ammonia borane (denoted as

AB) AB, with a chemical formula of NH3BH3, is an analog of ethane which contains an ammonia group with protic N–H bonds and a borane group with hydridic B–H bonds The NH3 and BH3 groups in AB are bound together with a strong B–N bond (1.56 Å)51 and the molecular AB is stabilized in the crystal structure by establishing strong intermolecular dihydrogen bonding network (with

BHδ-···δ+HN distance of 2.02Å, see Figure 1.2)52, 53, resulting in AB being a solid

at ambient temperature Furthermore, with an exceptionally high gravimetric storage density of ~19.6 wt% H2 and good stability in air and moisture, AB has been regarded as a good candidate for hydrogen storage

Figure 1.2 Low-temperature (orthorhombic) crystal structure of ammonia borane

(AB) N, B and H atoms are depicted in blue, purple, and gray, respectively The dihydrogen bonds are delineated by dashed lines Reproduced with permission from Ref 53

As reported in literature, AB can be prepared by two approaches, i.e., metathesis and direct reaction In the first approach, metathesis occurs between

alkali metal borohydrides and ammonium salts in organic solvents, i.e., ether54 or THF55 (equations 1.10 and 1.11)

LiBH4 + NH4Cl → NH3BH3 + H2 + LiCl (1.10) LiBH4 + (NH4)2SO4 → 2NH3BH3 + 2H2 + Li2SO4 (1.11)

Shore and Parry reported that the synthesis of AB that used LiBH4 as the precursor would yield only 45% of AB (soluble in ether) and the remains appeared

to be insoluble “diammoniate of diborane”.54 Later in 2007, Ramachandran and Gagare conducted a more extensive investigation on the AB synthesis, in which different kinds of solvents paired with precursors were studied In their study, they achieved relatively higher yield and purity by reacting sodium borohydride with ammonium sulfate in THF solution.55

In the second approach, a direct reaction between NH3 with different sources

of boranes yielded AB In the reaction between NH3 with diborane in different solvents, symmetric and asymmetric cleavage of the B2H6 would result in the formation of a mixture of “diammoniate of diborane” (DADB) and AB (the reactions 1.12 and 1.13) Mayer reported that the yield of AB is strongly dependent

on the solvents’ basicity, which is in the sequence of n-hexane < diisopropyl ether <

diethyl ether < monoglyme.56 Other than diborane, BH3·Et2O, BH3·S(CH3)2 and

BH3·THF also react with NH3 to form AB.57, 58

Asymmetric cleavage: B2H6 + 2NH3 → [(NH3)2BH2][BH4] (1.12) Symmetric cleavage: B2H6 + 2NH3 → 2NH3BH3 (1.13)

Decomposition of AB has been studied extensively due to its attractiveness

as potential hydrogen storage material.59-61 AB decomposes in a three-step process with one equivalent of H2 being released in each step.61 In the first step, AB decomposes to release one equiv H2 to give a white amorphous product of (NH2BH2)n, oralso known as polyaminoborane, at temperatures < 130 °C (equation 1.14) In the second step, polyaminoborane further decomposes to release another equivalent of H2 at temperatures above 150 °C, giving an amorphous polymeric product, polyiminoborane (NHBH)n (equation 1.15) Releasing the last equiv of H2, however, can only be achieved at temperatures above 500 °C (equation 1.16), and thus, only the first two steps are considered for producing “usable” hydrogen

1.2.3.1 Proposed dehydrogenation mechanism

To understand the dehydrogenation properties of AB in both solid and liquid states Autrey’s group employed in-situ NMR to monitor and characterize the intermediates formed in the dehydrogenation processes In the solid state dehydrogenation of AB, Autrey and coworkers prescribed the overall dehydrogenation mechanism of AB as induction, nucleation and growth (scheme 1).63 The mechanism lies behind the experimental observations of long induction period61, 64 was identified to involve disruptions of dihydrogen bonding which

results in a mobile phase of AB (AB* as shown in Scheme 1.1) Subsequently, H2

started to evolve which results in the detection of DADB corresponding BH2

species and two other BH2 group of possible linear and cyclic dimers of aminoborane Thus, it was proposed that in the nucleation process, formation of DADB or the dimer species is responsible for prompting the growth process for rapid H2 release

On the other hand, investigation on the thermal decomposition of AB in glyme also found that DADB is the active species that lead to the release of hydrogen65 (Scheme 1.2), similar to those reactions detected in solid-state63 and ionic liquid64 However, DADB was found to be unstable in glyme and rapidly undergo cyclization to release H2 and form clyclodiborazane (CDB) CDB subsequently reacts with the remaining AB to release H2 and give B-(cyclodiborazanyl) aminoborohydride (BCDB) In addition, cyclotriborazane (CTB) which may be converted from CDB or BCBD, was also detected However, there was no concrete evidence on its formation pathway