The Effect of Metal Oxides Additives on the Absorption ...

The Effect of Metal Oxides Additives on the Absorption/Desorption of MgH2 for Hydrogen Storage Applications

Author

Alsabawi, Khadija, Gray, Evan

Published

2019-07-15

Thesis Type

Thesis (PhD Doctorate)

School

School of Environment and Sc

DOI

https://doi.org/10.25904/1912/1452

Copyright Statement

The author owns the copyright in this thesis, unless stated otherwise.

Downloaded from

http://hdl.handle.net/10072/386537

Griffith Research Online

https://research-repository.griffith.edu.au

i

The Effect of Metal Oxides Additives on the

Absorption/Desorption of MgH2 for Hydrogen

Storage Applications

Khadija Alsabawi

BScHons

School of Environment and Science

and

Queensland Micro- and Nanotechnology Centre

Griffith University, Queensland, Australia

Submitted in fulfilment of the requirements of the degree of

Doctor of Philosophy

FEBRUARY 2019

To my parents

My loving Husband Youssef

And my wonderful kids

Naomi & Ameer

STATEMENT

This work has not previously been submitted for a degree or diploma in any university.

To the best of my knowledge and belief, the thesis contains no material previously

published or written by another person except where due reference is made in the thesis

itself.

Khadija Alsabawi

i

ACKNOWLEDGEMENT

This project was completed under the supervision of Dr Jim Webb. I am immensely

grateful for his patience, invaluable guidance and the support he has given me

throughout the project. The lessons he’s taught me have sculpted me into an

independent and critical minded researcher and for this I am very grateful. I feel very

fortunate to have worked with him and believe that I can still learn much from him in

our future projects.

I would like to express my gratitude to Prof Evan Gray for his indispensable assistance

and suggestions throughout the years.

To all the PhD candidates in my group, many thanks for making my time a forever

enjoyable and memorable experience. Special thanks go to Tim, Donya, Lubna,

Nazanin, and Samaneh who shared their PhD experience with me.

In addition, I would like to acknowledge the financial support received from the

Australian government through the Australian Government Research Training Program

Scholarship. Another big thanks to the members of Griffith University Higher Degree

Research department, the school of Environment and Science and the Queensland Micro

and Nanotechnology Centre (QMNC).

Finally, a tremendous thanks to my family for taking this journey with me, without their

support and encouragement this would not have been possible.

ii

ALL PAPERS INCLUDED ARE CO-AUTHORED

Acknowledgement of Papers included in this Thesis

Included in this thesis are papers in Chapters 2, 4, 5 and 6 which are co-authored with

other researchers. My contribution to each co-authored paper is outlined at the front of

the relevant chapter. The bibliographic details for these papers including all authors, are:

Chapter 2:

Alsabawi K, Gray EM and Webb CJ. ‘A review of the effect of metal oxide additives on

the sorption kinetics of MgH2 for hydrogen storage’. International Journal of Hydrogen

Energy. (To be submitted)

Chapter 4:

Alsabawi K, Gray EM and Webb CJ. ‘A Sieverts apparatus with TPD/TDS for accurate

high-pressure hydrogen uptake and kinetics measurements’. International Journal of

Hydrogen Energy. (Submitted)

Chapter 5:

Alsabawi K, Webb TA, Gray EM and Webb CJ. ‘Kinetic enhancement of the sorption

properties of MgH2 with the additive titanium isopropoxide’. International Journal of

Hydrogen Energy. 2017;42(8):5227-34;10.1016/j.ijhydene.2016.12.072.

Chapter 6:

Alsabawi K, Gray EM and Webb CJ. The effect of ball-milling gas environment on the

sorption kinetics of MgH2 with/without additives for hydrogen storage. International

Journal of Hydrogen Energy. 2019;10.1016/j.ijhydene.2018.12.026.

Chapter 7:

Alsabawi K, Gray EM and Webb CJ. ‘The effect of transition metal oxide additives on

the sorption kinetics of MgH2 for hydrogen storage.’ International Journal of Hydrogen

Energy. (Submitted)

iii

Appropriate acknowledgements of those who contributed to the research but did not

qualify as authors are included in each paper.

(Signed) _____________________(Date)__14/02/2019____________

Khadija Alsabawi

(Countersigned) ______________ (Date)__ 14/02/2019________________________

Supervisor: Jim Webb

iv

ABSTRACT

Magnesium is considered as one of the more promising candidates for hydrogen

storage, primarily because of its abundance and the high hydrogen capacity of MgH2,

magnesium dihydride (7.6 wt%). Unfortunately, practical applications of MgH2 are

limited by poor hydrogen sorption kinetics and high thermodynamic stability, resulting

in slow charging and the need for high temperatures to release the hydrogen. Methods to

improve the sorption kinetics of MgH2, through ball-milling, alloying and introducing

small amounts of additives are currently under investigation. The aim of this work was

to investigate the effect of different transition metal oxides, as well as other catalysts, on

the hydrogen sorption kinetics, hydrogen release temperature and cycling stability of

MgH2.

The rate of MgH2 absorption is determined by the physisorption and dissociation of

molecular hydrogen, diffusion through the hydride layer and then nucleation of the

hydride. Whereas desorption is determined by nucleation of metal phase, diffusion of

atomic hydrogen through the metal and hydride, and recombination to form molecular

hydrogen at the surface before dissociation of the molecule form the surface.

The addition of small amounts of additives (< 10 mol%) to MgH2 during milling have

been shown to have a significant effect on the kinetics of absorption and desorption of

hydrogen. In addition, the effect of oxygen as a component of these additives has been

extensively studied. Although a surface layer of MgO is known to slow the diffusion of

hydrogen into metal, niobium oxide, Nb2O5, is one of the best additives for kinetic

improvement of MgH2. It has been suggested that higher valance oxides have a greater

effect on the kinetics; also recent studies have indicated that the formation of

magnesium-niobium ternary oxide compound may be responsible for the enhancement,

however the exact mechanism by which Nb2O5 enhances the kinetics is still unclear.

Various transition metal oxides have proved to be effective additives for hydrogen

absorption/desorption of Mg-based hydrides e.g., Nb2O5 and TiO2. In this work organo-

metallic additives based on transition metal oxides (Ti and V) and halides have been

chosen, some of which are liquid at room temperature. These additives were chosen to

extend the oxide valency to higher values, provide a comparison between liquid and

solid oxide additives, and compare a non-oxide transition metal additive (Ti-based

chloride). Nb2O5 was used as the benchmark for comparisons.

v

A range of different amounts of transition metal (Ti and V) oxides and other additives,

both liquid and solid, were ball milled with pre-milled MgH2 under different gas

environments and the hydrogen sorption behaviour of the ball milled composite samples

was investigated using a TPD/TDS Sieverts apparatus. A key difference from previous

work in this area was the use of organo-metallic compounds, some of which were liquid

at room temperature. Samples were characterized using X-Ray Diffraction, and, where

feasible, Scanning Electron Microscopy and Raman Spectroscopy before and after

recording the hydrogen sorption behaviour.

A Sievert apparatus was improved, then modified to include TPD and TDS capability to

determine the hydrogen uptake and release and to perform all these measurements in

one consistent environment.

With the aim of understanding what effect the ball milling gas environment might have

on the MgH2 sorption kinetics, two different gases, Ar and H2, were used. MgH2, with

and without additives (C60, TTIP (Titanium TetraIsoPropoxide) and Nb2O5) were milled

under the two different gases. In most cases the milling gas had little to no effect on

desorption kinetics, but a noticeable effect in the absorption uptake was observed - by

up to 2 wt% - depending on the gas used.

To understand the role of an organo-metallic oxide liquid additive on MgH2 sorption

kinetics, a systematic survey was performed by milling up to 2 mol% of TTIP with

MgH2 and the results compared to composite samples with Nb2O5 as the additive. TTIP

was found to be equally as effective as Nb2O5 with superior hydrogen capacity, and, just

as for Nb2O5, only a small amount of additive, 0.5 mol% of TTIP was found to be

sufficient for kinetics enhancement. Interestingly, 0.5 mol% TTIP hand-mixed with the

pre-milled MgH2 was also found to be effective for desorption, but not for the

absorption kinetics.

To further investigate the effect of organo-metallic liquid oxides, particularly transition

metal oxides, on the enhancement of the sorption kinetics of MgH2, a range of liquid

oxides (Ti-based and V-based oxides) were milled with MgH2 and compared to powder

oxides. Ti-based oxides were found to have superior desorption enhancement, with the

liquids performing better than the powders. The V-based oxides (all liquids) showed

faster absorption and higher uptake when compared to Ti-based oxides. The Ti-chloride

based organo-metallic additive was investigated to compare a non-oxide transition metal

additive to the oxide additives. It was found that a small amount of 1 mol% of additive

vi

milled for short time (1 h) had the best desorption of all the additives investigated in this

work - but with quite poor absorption.

Overall, this project established that the use of transition metal oxides as additives has a

great impact on improving the sorption kinetics of MgH2. The TTIP additive produced

results at least as good as the benchmark additive Nb2O5, but with the significant

advantage of being able to be mixed with MgH2 without ball-milling. The Ti and V

based oxides additives were also shown to be effective, Ti-based achieving better

desorption enhancement, whereas V-based oxide samples had faster absorption and

higher uptake. It was also determined the ball milling gas environment can have a

significant effect on the sorption kinetics, but the effect depended on the additive. The

difference in the effect of additives on desorption and absorption cycles confirms the

need to study combinations of additives for optimal overall benefit.

CONTENTS

ABSTRACT .................................................................................................................... iv

CONTENTS .......................................................................................................................

CHAPTER ONE ............................................................................................................... 1

1. Introduction ............................................................................................................... 1

1.1. Renewable energy .............................................................................................. 1

1.2. Hydrogen energy system (closed cycle) ............................................................ 1

1.3. Hydrogen properties .......................................................................................... 2

1.4. Hydrogen production ......................................................................................... 4

1.5. Hydrogen storage ............................................................................................... 4

1.5.1. Gaseous storage .......................................................................................... 7

1.5.2. Liquid storage ............................................................................................. 7

1.5.3. Solid state materials .................................................................................... 7

1.5.3.1 Physisorption on carbons: ................................................................... 8

1.5.3.2 Simple and Intermetallic Metal Hydrides: .......................................... 8

1.5.3.3 Complex hydride: .............................................................................. 11

1.6. Magnesium hydride MgH2 ............................................................................... 11

1.6.1 The effect of additives ....................................................................... 12

1.6.1.1 Transition metals and Metal Oxides: .......................................... 13

1.7. Thesis structure ................................................................................................ 14

1.8. List of publications and presentations ............................................................. 15

1.9. References ........................................................................................................ 16

CHAPTER TWO ............................................................................................................ 24

Abstract ........................................................................................................................... 24

2. Introduction ............................................................................................................. 25

2.1. Kinetic processes in magnesium hydride ......................................................... 26

2.2. Mechanical ball milling and additives ............................................................. 27

2.3. Oxide additives ................................................................................................ 28

2.3.1. Magnesium and Metal Oxide additives: ................................................... 28

2.3.2. Zirconium dioxide (ZrO2) ......................................................................... 33

2.3.3. Cerium (IV) oxide (CeO2) ........................................................................ 34

2.3.4. Vanadium oxide ........................................................................................ 35

i

2.3.5. Titanium Oxide based additives ............................................................... 38

2.3.6. Magnesium oxide: .................................................................................... 42

2.3.7. The effect of Cr2O3 on MgH2 ................................................................... 44

2.3.8. Niobium Pentoxide Nb2O5: ...................................................................... 48

2.4. Summary .......................................................................................................... 61

2.5. Reference ......................................................................................................... 63

CHAPTER THREE ........................................................................................................ 74

3. Introduction ............................................................................................................. 74

3.1. Synthesis .......................................................................................................... 74

3.1.1. Ball-milling ............................................................................................... 74

3.1.1.1. Mechanical grinding (MG) ................................................................... 75

3.1.1.2. Mechanical alloying (MA) .................................................................... 75

3.1.1.3. Reactive ball milling (RBM) ................................................................ 75

3.2. Preparation of milled hydride/additive mixtures ............................................. 76

3.3. Characterisation techniques ............................................................................. 78

3.3.1. X-Ray Diffraction (XRD) analysis ........................................................... 78

3.3.1.1. Bragg’s Law .......................................................................................... 78

3.3.2. Scanning Electron Microscopy (SEM) ..................................................... 80

3.3.3. Synchrotron Radiation X-Ray Powder Diffraction (SR-XRPD) .............. 82

3.3.4. Kinetics characterisation........................................................................... 83

3.3.5. Temperature Programmed Desorption and Thermal desorption spectroscopy ............................................................................................................ 84

3.4. References ........................................................................................................ 86

CHAPTER FOUR .......................................................................................................... 89

4. Introduction ............................................................................................................. 89

4.1. References ........................................................................................................ 91

4.2. Paper Sieverts instrument ................................................................................ 91

Abstract ....................................................................................................................... 93

Introduction .................................................................................................................... 94

Uptake measurements ................................................................................................. 95

The Sieverts technique ................................................................................................ 97

Errors and uncertainties using the Sieverts method .................................................... 98

Temperature Programmed Desorption (TPD) ............................................................ 99

Thermal Desorption Spectroscopy (TDS)/ (TDS-MS) ............................................... 99

Hydrogen pressure amplification.................................................................................... 99

ii

Instrument design ......................................................................................................... 102

Temperature control .................................................................................................. 102

Leak management ......................................................................................................... 103

Preparation section .................................................................................................... 104

Reference section .......................................................................................................... 104

Sample section .............................................................................................................. 104

Sample environment ................................................................................................. 105

Experimental procedure ............................................................................................ 105

Volume calibration ................................................................................................... 105

Empty-cell isotherm ..................................................................................................... 106

Divided Volume Calibration ..................................................................................... 109

Instrument validation: ................................................................................................... 112

Standard Sample measurements ............................................................................... 112

TPD ............................................................................................................................... 114

TDS ............................................................................................................................... 115

Conclusion .................................................................................................................... 116

References ................................................................................................................. 118

CHAPTER FIVE .......................................................................................................... 125

5. Introduction ........................................................................................................... 125

5.1. References ...................................................................................................... 126

5.2. TTIP paper: .................................................................................................... 127

CHAPTER SIX ............................................................................................................ 137

6. Introduction ........................................................................................................... 137

6.1. : Milling Environment Paper ......................................................................... 138

CHAPTER SEVEN ...................................................................................................... 144

7. Introduction ........................................................................................................... 144

7.1. References ...................................................................................................... 146

7.2. Paper .............................................................................................................. 146

Abstract ......................................................................................................................... 148

Introduction .................................................................................................................. 149

Experimental ................................................................................................................. 152

Results and discussion:(1 mol%) Ti-based oxide additives milled for 2 h ................... 153

Different amount of Ti-ethoxide milled for 2 h ............................................................. 155

1 mol% V-based oxide additives milled for 2 h ............................................................ 157

iii

BisTiCl2 (Bis(isopropylcyclopentadienyl) titanium dichloride) additive ..................... 159

Comparison of all additives .......................................................................................... 161

Conclusion .................................................................................................................... 162

Acknowledgements ...................................................................................................... 163

References: ................................................................................................................... 164

CHAPTER EIGHT ....................................................................................................... 170

8. Summary and Conclusion ..................................................................................... 170

8.1. Future work .................................................................................................... 174

8.2. Reference ....................................................................................................... 176

iv

List of Figures

2 Fig 1-1:Typical operating systems for various hydrogen storage technologies and their operating temperatures[22] ............................................................................................... 5 Fig 1-2: (a) Pressure-Concentration-Temperature (PCT) curve and (b) van ’t Hoff plot [31]. .................................................................................................................................. 9 Fig 1-3: Table of the binary hydrides and the Allred-Rochow electronegativity [32]. .. 10 Fig 1-4:Lennard-Jones potential of hydrogen approaching a metallic surface [34]. ...... 10 Fig 2-1:Transition metal oxides and their effect on hydrogen desorption kinetics of MgH2 [31] ....................................................................................................................... 28 Fig 2-2: Desorption rates for metal oxides milled with MgH2, from [29]...................... 29 Fig 2-3: Decomposition temperature of MgH2 milled with different metal oxides [32].32 Fig 2-4: (a) A side view (b) top view of (1*3) V2O5 (001) surface. (c) the optimized MgH2 clusters. red (O), grey (V), green (Mg) and white balls (H)[49]. ........................ 36 Fig 2-5: Crystal structure of TiO2 a) anatase b) rutile [57]. ........................................... 38 Fig 2-6: Correlation between desorption temperature achieved upon oxide addition and electronegativity of the oxide additives [68]. ................................................................. 43 Fig 2-7: The selected structural models of MgH2 on (a) Nb(1 0 0), (b) NbO(1 0 0), (c) NbO(1 10), (d) NbO(1 1 1) and (e) Nb2O5(1 0 0). Atomic colour codes: Mg, green; H, white; Nb, blue; O, red. [106] ......................................................................................... 59 Fig 3-1: Bragg diffraction in a 2D lattice ....................................................................... 79 Fig 3-2: A schematic view of a laboratory X-ray Diffractometer (XRD) (image form PANAlytical) [13]. ......................................................................................................... 80 Fig 3-3: Electron beam sample interaction [14]. ............................................................ 81 Fig 3-4: Schematic layout of the 10-BM-1 Powder Diffraction beamline “Australian synchrotron [17]” ............................................................................................................ 82

List of Tables

Table 1-1: Physical properties of molecular hydrogen ................................................................. 3 Table 1-2: Heating values of some common fuels [13] ................................................................ 3 Table 1-3:basic hydrogen storage methods and phenomena, [23] modified from table 1. ........... 6 Table 3-1:Chemicals and materials used in this project .............................................................. 77 Table 3-2: The Australian Synchrotron Parameters. ................................................................... 83

1

CHAPTER ONE

1. Introduction

1.1. Renewable energy

The increasing world population continues to grow energy demand. As of 2017, three

quarters of the world’s energy was provided by fossil fuels [1], which are forecast to be

limited in the near future and when combusted they release greenhouse gases into the

atmosphere posing a threat to the global environment and climate change [2]. Profound

changes are needed regarding the production, distribution and the consumption of

energy. Renewable energy is an emerging solution for a clean, sustainable, affordable

and long-term source of energy for the future. Sources of renewable energy include

solar, wave, tide, wind, geothermal and biomass, however, practical implementation of

most of these sources is constrained by their intermittent and unpredictable nature, such

as wind, for example. Therefore, the use of an energy storage system is necessary to

store energy during excess production and re-supply the energy when needed.

Electricity is an energy carrier which can be produced from different sources and is

technically easy to transport, although this can be quite expensive, especially over long

distances [3]. It has a relatively low environmental impact when produced from

renewable sources. Hydrogen is another efficient and versatile energy carrier, and

together with electricity they may satisfy all the energy requirements, and form a clean

energy system when sourced using renewable energy [4]. Jules Verne in 1874 was the

first to suggest that hydrogen derived from water electrolysis will be used in place of

coal for energy purposes - “water will be the coal of the future”. In the 1930s Rudolf

Erren suggested the use of hydrogen as a transportation fuel, and, after another 30 years,

the concept of a hydrogen economy was introduced as a future energy system to replace

depleting fossil fuels [5].

1.2. Hydrogen energy system (closed cycle)

Hydrogen is the most abundant element accounting for 75% of the mass of the universe

[6]. However, elemental or molecular hydrogen does not occur naturally as a gas on

earth but is found in compounds combined with other elements such as oxygen, carbon

and nitrogen including many organic compounds such as hydrocarbons. Hydrogen has

2

relatively high specific energy content (140 MJ/kg), with liquid hydrogen being used by

NASA since the 1970s to propel the space shuttle and other rockets into orbit. Hydrogen

also has the advantage over hydrocarbons that hydrogen can be generated from water,

and, when combusted or used in a fuel cell to generate electricity, combines with

oxygen to turn back into water [7]. This closed cycle produces little to no pollutants,

whereas hydrocarbons convert mainly to carbon dioxide, a greenhouse gas. In its

molecular form (H2), hydrogen is a gas that burns in oxygen following the simple

reaction:

2 2 2

12

H O H O heat+ → + (1)

Hydrogen is an ideal fuel for the fuel cell which was invented in 1839 [8], although

solid oxide fuel cells for ammonia [9] and methane [10] exist. The fuel cell is an

electrochemical device that converts chemical potential energy directly into electrical

energy and since the ideal fuel cell is reversible, it is not limited by the Carnot

efficiency imposed by the 2nd law of thermodynamics on heat engines. A PEM (Proton

Exchange Membrane) fuel cell can use hydrogen gas and oxygen as fuel, where they are

converted directly into electricity, water and heat [11].

1.3. Hydrogen properties

Hydrogen is the simplest atomic element, consisting of one proton and one electron

only, in its simplest form. Hydrogen has three isotopes, H1 (protium), H2 (deuterium),

both stable isotopes, and H3 (tritium), which has a half-life of 12.32 years. It is a

colourless, odourless gas at ambient temperature and pressure with no toxic effects. It is

the lowest density element and has the lowest melting and boiling points after helium

(see Table 1-1) for physical properties of hydrogen).

3

Table 1-1: Physical properties of molecular hydrogen

Parameter Hydrogen

Molecular mass 2.01588 g/mole

Melting point 13.99 K

Boiling point 20.271 K

Critical point 32.938 k, 12.85 bar

Density (as gas) 0.089 g/L

Density (as liquid) 0.07 g/ cm3

Hydrogen is very reactive element due to its single valence electron and usually

combines two atoms to form the molecule H2. It is found in a wide range of inorganic

and organic chemical compounds, but rarely in its gas form. Combined with carbon,

hydrogen has the ability to form long chains and complex molecules, and this has a

major role in organic life (hydrocarbons, carbohydrate). Hydrogen, as an energy carrier,

plays an important part in the metabolism of plants, animals, and humans [12].

Table 1-2: Heating values of some common fuels [13]

Fuel High Heating value (MJ/kg) Low Heating value (MJ/kg)

Hydrogen 142 120

Gasoline 46 44

Diesel 45 43

Methane 55 50

As seen in Table 1-2, hydrogen has the highest energy to weight ratio when compared

to other common fuels; since the other fuels listed all contain carbon atoms which are

significantly heavier than hydrogen. By weight, hydrogen produces energy during a

reaction of about 2.5 times the heat of combustion of common hydrocarbon fuels.

4

1.4. Hydrogen production

Hydrogen can be produced from fossil fuel based hydrocarbons, such as methanol,

gasoline, propane and natural gases, by applying heat in the presence of a catalyst, in a

process known as ‘reforming’ [14]. Currently 96 % of all the hydrogen produced is

derived from fossil fuels, with an estimated 49 % derived from natural gas, 29 % from

liquid hydrocarbons, 18 % from coal and about 4 % from electrolysis [15]. Electrolysis

of water produces hydrogen when an electrical current is passed through water

separating it into oxygen and hydrogen.

2 2 22 2H O electricity H O+ → + (2)

Other research currently exploring new methods for producing hydrogen from

renewable resources includes the use of sunlight, biomass and biological organisms

(bacteria and algae). Semiconductor or biological organisms can be used to absorb

sunlight, split water and produce hydrogen. Some biological organisms produce

hydrogen naturally through their normal metabolic function, and semiconductors can

generate an electric current that splits water to produce hydrogen. NREL researchers

have developed a device (PEC system) that uses photovoltaic cells immersed in acidic

electrolyte to initiate a water electrolysis reaction. This system converts about 16.2 % of

available sunlight into hydrogen [16].

1.5. Hydrogen storage

Hydrogen gas has a very low volumetric energy density which makes it inefficient to

store in small vessels. The density of hydrogen in a storage container can be increased

by applying work to either compress the gas (high pressure gas cylinder), and/or

decrease the temperature below the critical point (liquid hydrogen) or finally, reduce the

repulsion between hydrogen molecules by utilizing an interaction between hydrogen

and another material (solid state storage). It is important for a hydrogen storage system

to reversibly store and release the hydrogen, and for this to be practical, at temperatures,

pressures and efficiencies able to be employed in power stations, homes and motor

vehicles (Fig 1-1). Commercial hydrogen powered cars such as the Toyota Mirai,

Honda Clarity and Hyundai ix35 fuel cell vehicles are already available [17-19], but due

to its low volumetric energy density, 0.5 kWh/dm3 for 350 bar H2 [12], compared to

gasoline (~7 kWh/dm3), large storage tanks of compressed hydrogen are required in

order to have a sufficient driving range. High pressure storage increases the energy

density, but then requires stronger, more expensive storage tanks [20] to be safe. Liquid

5

hydrogen cryogenic tanks need costly and energy consuming cooling systems [21]. In

addition, any compression system requires parasitic energy use that reduces the capacity

of the total system. A viable alternative is solid state storage.

Fig 1-1:Typical operating systems for various hydrogen storage technologies and

their operating temperatures[22]

Züttel [23] described the possible methods for reversible hydrogen storage with high

volumetric and gravimetric density, (see Table 1-3). Each of the listed methods have

their advantages and disadvantages. ρm is the gravimetric capacity, ρv the volumetric

capacity, T the working temperature and P the pressure. RT: room temperature (25°C).

6

Table 1-3:basic hydrogen storage methods and phenomena, [23] modified from

table 1.

Storage method ρm(mass%) ρV(kg

H2m−3

)

T(°C) p(bar) Phenomena and remarks

High-pressure

gas cylinders

13 <40 RT 800 Compressed gas

(molecular H2) in

lightweight composite

cylinder (tensile strength

of the material is

2000 MPa)

Liquid hydrogen

in cryogenic

tanks

Size

dependent

70.8 −252 1 Liquid hydrogen

(molecular H2),

continuous loss of a few

% per day of hydrogen at

RT

Adsorbed

hydrogen

≈2 20 −80 100 Physisorption (molecular

H2) on materials, e.g.

carbon with a very large

specific surface area,

fully reversible

Absorbed on

interstitial sites

in a host metal

≈2 150 RT 1 Hydrogen (atomic H)

intercalation in host

metals, metallic hydrides

working at RT are fully

reversible

Complex

compounds

<18 150 >100 1 Complex compounds

([AlH4]− or [BH4]−),

desorption at elevated

temperature, adsorption

at high pressures

7

1.5.1. Gaseous storage

Gaseous hydrogen is still the most commonly-used method to store hydrogen for a

range of applications, despite having a low energy density compared to other methods.

The gaseous hydrogen has to be compressed to high pressures, in high-pressure gas

cylinders, to improve its density [24]. However, the very high pressure needed to store

the gas (up to 800 bar, Table 1-3) imposes a safety concern, and together with the

relatively low hydrogen density and energy losses in compressing the gas, makes this an

unsuitable method for storing hydrogen for the future of most applications.

1.5.2. Liquid storage

Storing hydrogen in a liquid form requires a large amount of energy to compress, cool

and liquefy the gas. The liquefaction process requires the following steps: the gas is

compressed then cooled in a heat exchanger, then undergoes an isenthalpic Joule-

Thompson expansion by passing through a throttle valve, producing liquid hydrogen.

This process (named the Joule-Thompson cycle) has been used for space projects and is

considered one of the simplest methods to liquefy gases. The liquefied hydrogen is

stored in well-insulated tanks at very low temperature. However, liquid hydrogen

suffers from boil-off due to heat leaks and para-ortho H2 conversion [25], as well as a

low ignition threshold, which poses safety threats [26].

1.5.3. Solid state materials

Solid state materials such as metal hydrides (including elemental metal hydrides such as

MgH2, and alloys, such as LaNi5) and complex hydrides (e.g. amides, borohydrides and

alanates) [27], as well as physisorption on high surface area materials (e.g. graphite,

graphene, polymers, metal-organic frameworks and carbons), are all capable of storing

hydrogen with varying gravimetric and volumetric densities. The physisorption process

has a relatively low operating pressure, the price of materials involved is affordable, and

the simplicity of the storage system design, are all advantages for using this process.

However, physisorption has drawbacks such as the low amount of hydrogen adsorbed

and the cryogenic temperatures needed to attain a suitable capacity.

Metal hydrides and complex hydrides can store a large amount of hydrogen in a safe

and compact way. Reversible intermetallic hydrides are available that work at ambient

temperature and atmospheric pressure, but their gravimetric hydrogen density is limited

to <3 mass% [23]. Chemical hydride compounds such as the simple metal hydrides and

8

the complex metal hydrides often have high formation energies and therefore require

high temperatures to release the hydrogen, which makes them unsuitable for many

practical applications [28].

1.5.3.1 Physisorption on carbons:

Physisorption is the adsorption a solid surface by a molecular fluid via the same

intermolecular forces as those involved in the condensation of vapours and real gas,

such as van der Waals forces [29]. During absorption and desorption the electronic

orbital patterns of the species involved are not significantly affected by the

intermolecular forces, preserving the chemical nature of the solid/fluid [29]. Hydrogen

can be stored via physisorption in light-weight nanoporous materials, such as carbon,

where the interaction between hydrogen and the material are mainly weak van der

Waals forces. The key to maximising hydrogen storage lies in having a high surface

area, high characteristic binding energy of the hydrogen molecule with the material, and

appropriate pore size [30]. Nanostructured carbon, activated carbon and carbon

nanotubes (CNTs), have large surface areas which make them possible structures for

physisorption of H2.

The operating temperature for a hydrogen storage system based on physisorption on

solid state materials, is determined by the binding energy. The hydrogen binding

energies for pure carbon materials are in the range of 4-15 kJ/mol, where the lower is

typical for graphite and activated carbon, and the upper for internal and interstitial sites

of single walled nanotubes (SWNT) and SWNT bundles. The confined geometry effects

of the SWNT sites are associated with higher binding energy [30]. As seen in Table 1-3,

most carbons need a very low temperature (<–80 ºC) to enable reasonable adsorption

capacity.

1.5.3.2 Simple and Intermetallic Metal Hydrides:

Hydrogen storage in metal hydrides has been extensively researched and explored, due

to their capability to reversibly absorb large amounts of hydrogen. Pressure-

Composition isotherms (PCT) seen in Fig1-2 are used to describe the thermodynamic

properties during the formation of hydride from gaseous hydrogen.

9

Fig 1-2: (a) Pressure-Concentration-Temperature (PCT) curve and (b) van ’t Hoff

plot [31].

Initially some hydrogen will dissolve into the host metal as solid solution (α-phase).

With increasing hydrogen pressure and concentration in the host metal the α-phase will

change to metal hydride (β-phase). The isotherm ideally shows a flat plateau when the

α-phase coexists with the β-phase, where its length determines the amount of H2 that

can be reversibly stored at the equilibrium dissociation pressure. The critical point TC

occurs when the temperature is high enough that there is no difference between the α-

and β-phases. Since the equilibrium pressure ‘plateau’ depends strongly on temperature,

it is related to the changes in the enthalpy, ∆H, and entropy, ∆S, by the van ’t Hoff

equation:

0

ln p H S

p RT R

∆ ∆= −

(3)

Entropy loss occurs in the change of molecular hydrogen gas to dissolved solid

hydrogen, and this enthalpy loss characterises the energy stability of the metal hydrogen

bond. The metal hydride entropy of formation leads to a significant heat evolution ∆Q =

T·∆S during the hydrogen absorption (exothermal reaction) and consequently this heat

must be provided for desorption (endothermal reaction) [23].

10

Fig 1-3: Table of the binary hydrides and the Allred-Rochow electronegativity

[32].

Metal-hydrogen compounds can be formed via reaction of hydrogen gas with a metal,

metal alloy or intermetallic compound. According to the Allred-Rochow

electronegativity (Fig 1-3), all these compounds exist as either ionic, volatile covalent,

polymeric covalent or metallic hydrides. The reaction of hydrogen gas with metal can

be described by a one-dimensional Lennard-Jones potential energy curve (see Fig1-4 )

[33].

Fig 1-4:Lennard-Jones potential of hydrogen approaching a metallic surface [34].

11

The van der Waals force is the first attractive interaction of the hydrogen molecule,

leading to the physisorbed state (EPhys ≈ -5 kJ/molH). When hydrogen approaches the

surface of a metal, it has to overcome an activation barrier for the dissociation of the

molecule before formation of the bonds between the resultant hydrogen atoms and the

surface material. The chemisorbed state is reached when the hydrogen atoms share their

electrons with atoms at the surface of the metal (EChem ≈-50 kJ/molH). Then the

chemisorbed hydrogen atom is able to jump into the subsurface layer and the host metal

lattice enables the atom to diffuse on interstitial sites [34].

1.5.3.3 Complex hydride:

Elements in group 1, 2 and 3 e.g. Li, Mg, B and Al are the building blocks for a large

range of metal hydrogen complexes. Their light weight and the high number of

hydrogen atoms per metal atom makes them interesting for hydrogen storage media

[34]. A large number of complex metal hydrides, such as alanates [AlH4]−, amides

[NH2]−, imides and borohydrides [BH4]−, have been investigated [35]. In such systems

hydrogen is often located in the corners of a tetrahedron with B or Al in the centre. The

alanates and borohydrides may have potential for further investigation due to the very

high capacity, for example, Al(BH4)3 has the highest volumetric density (150kg/m3) and

LiBH2 has the highest gravimetric density at room temperature (18 mass%) [36],

however, these materials are either not reversible or require extremely high temperature

to release the hydrogen.

1.6. Magnesium hydride MgH2

Magnesium hydride (MgH2) is a promising candidate for a solid state hydrogen storage

material. The metal hydride bond, crystal structure and properties of MgH2 make it

unique among metal hydrides. Under normal conditions, α-MgH2 has a tetragonal

crystal structure of rutile type and under high pressure undergoes a polymorphic

transformation to form β-MgH2 with hexagonal structure and γ-MgH2 with

orthorhombic structure [37, 38]. The large hydrogen storage capacity (up to 7.6 wt%)

was first observed in the late nineteenth century [39]. MgH2 is thermodynamically

stable (ΔH=75 kJ.mol-1H2), which corresponds to unfavourable operating temperatures

(of the order of 573–673 K). In addition, the Mg-H system has slow hydrogen

absorption/desorption kinetics [40]. For practical use it is crucial to find a convenient

way to overcome the high stability and slow kinetics.

12

The slow kinetics of MgH2 is due to multiple factors, including a passivation layer

(MgO) covering the surface of Mg particles, slow dissociation of hydrogen at the

magnesium surface, a slow hydrogen diffusion rate in bulk Mg and particularly in the

forming hydride layer, MgH2. The MgO layer is not transparent to H2 and prevents

penetration into the metal; therefore, for Mg to absorb H2 the oxide layer must be

perforated or cracked. Annealing at 673 K in vacuum is known to help crack the oxide

layer [40], as does ball-milling. The slow dissociation rate can be overcome by the

presence of a catalytic metal e.g. Pd or Ni.

Ball milling has been used to reduce the grain size of magnesium metal, however this is

difficult due to the high reactivity and cold welding (re-agglomeration) of the

magnesium nanoparticles [41]. More recently, milling magnesium hydride rather than

Mg metal, or milling Mg under a hydrogen atmosphere, has been more successful at

producing a material with much faster desorption kinetics. This improvement is due to

the small particle size, introduction of defects and the increase of the surface area during

milling, which increases the nucleation site density and reduces the diffusion lengths.

However, a degree of re-agglomeration of the material with cycling can partially

remove these benefits [42].

Changing the enthalpy is much more difficult and is not achieved through defect

enhancement or catalysts. Attempts to lower the enthalpy have included alloying with

other metals, confinement in other materials and reducing the particle size to very low

nano-sizes. Wagemans et al.[43] investigated the thermodynamic stability of MgH2, by

determining the effect of decreasing the crystal grain size of magnesium to nanoscale.

The change of enthalpy during hydriding and dehydriding of magnesium was found to

decrease significantly provided the particle size was less than ~1.3 nm [43].

1.6.1 The effect of additives

Generally, the addition of comparatively small amounts of additives or catalysts has

been shown to be highly effective in enhancing the kinetic performance of the MgH2/

Mg system. Many additive materials have been introduced into MgH2, including various

types of carbons, halides, oxides, metal and non-metal elements, intermetallics, hydrides

and others (carbides, nitrides and borides) [44]. Most of these additives were reported to

improve the hydrogen storage performance [45]. The rate of absorption is controlled by

the following factors: hydrogen dissociation rate at the surface, hydrogen ability to

penetrate from the surface which is typically covered by an oxide layer, and the

13

hydrogen diffusion rate into the bulk metal and through the hydride. The addition of

additives could improve the kinetics by facilitating hydrogen dissociation at the surface,

increasing the available surface area, and by introducing defects that promote hydrogen

diffusion. The additives investigated in this study mainly fall into the category of

transition metal oxides.

1.6.1.1 Transition metals and Metal Oxides:

Over the past few decades, numerous transition metals have been investigated as

additives (i.e. Ti, Nb, V, Co, Mo, Fe, Mn, Ni and Ce) and their compounds, on the kinetics

of MgH2 and found to be effective in reducing the operating temperature and improving the

rate of absorption and desorption [46]. Among the investigated transition metals, niobium,

vanadium and titanium have shown superior results in kinetic enhancement of MgH2 [47].

The addition of transition metals was found to reduce the nucleation barrier and thus

improve the sorption properties of MgH2 [48].

Different types of metal oxides have been used as an additives to improve the sorption

kinetics of MgH2, including Nb2O5 [49-53], MgO [54-56], TiO2 [57-59], V2O5 [60, 61],

Cr2O3 [62-64] and many others. Niobium pentoxide, Nb2O5, is considered as a bench

mark for improving MgH2 kinetics, although the reasons for the kinetic enhancement of

this and other additives are less well-known. There are different explanations in the

literature for the superiority of Nb2O5, including that it acts as a dispersive during

milling, which yields a smaller average particle size of MgH2 [65], that it can act as an

active reservoir that helps with the association and disassociation of H2 in Mg/MgH2

system [66], or that it partially reduces during milling to produce catalytic compounds,

including a ternary Mg-Nb oxide, suggested to facilitate the diffusion through the MgO

layer [67, 68].

It has also been proposed that the metal’s valence state in the oxides plays an important

role in improving the sorption kinetics [69], as well as the high defect density

introduced during milling at the surface of the metal oxide particles [70]. It has been

found that milling MgH2 with a metal oxide yields a smaller average particle size than

milling with metals [65]. However, the exact mechanism of the kinetic enhancement is

still under investigation.

14

1.7. Thesis structure

The overall objective of this project was the enhancement of the absorption/desorption

kinetics of MgH2 using different types of additives with ball-milling methods. The

choice of additive was guided by the literature on high oxygen containing transition

metal compounds. A number of organic compounds, some of which are liquids at room

temperature as well as others containing little to no oxygen or halides, were chosen for

comparison.

The history of the use of additives for kinetic improvement of MgH2 is detailed in the

literature survey in Chapter 2. It provides a critical review for a range of metal-based

oxide additives used to improve the kinetics of MgH2 through ball milling.

Chapter 3, reports on the experimental techniques and materials preparation used

during this work, along with the explanation of the different characterization techniques

used in studying the composite materials. Chapter 4, describes the experimental method

used – in particular the Sieverts apparatus, including the design, performance, operation,

data acquisition and calculation.

Chapters 5, 6 and 7, contain the experimental results of the various additive materials

used to modify the sorption kinetics of MgH2. Chapter 5, is a published paper on the

effect of titanium tetraisopropoxide on the sorption kinetics of MgH2, Chapter 6, is

another published work on the effect of the ball milling environment on the kinetics of

MgH2 with and without additives, and Chapter 7 is a submitted paper showing the effect

of different transition metal oxide additives (powder vs liquid) on the kinetics of MgH2.

Conclusions are presented in Chapter 8 and suggestions for future work are given.

15

1.8. List of publications and presentations

Journal papers

1) Alsabawi K, Webb TA, Gray EM and Webb CJ. The effect of C60 additive on

magnesium hydride for hydrogen storage. International Journal of Hydrogen

Energy. 2015;40(33):10508-15;10.1016/j.ijhydene.2015.06.110.

2) Alsabawi K, Webb TA, Gray EM and Webb CJ. ‘Kinetic enhancement of the

sorption properties of MgH2 with the additive titanium isopropoxide’.

International Journal of Hydrogen Energy. 2017;42(8):5227-

34;10.1016/j.ijhydene.2016.12.072.

3) Alsabawi K, Gray EM and Webb CJ. The effect of ball-milling gas

environment on the sorption kinetics of MgH2 with/without additives for

hydrogen storage. International Journal of Hydrogen Energy.

2019;10.1016/j.ijhydene.2018.12.026.

4) Alsabawi K, Gray EM and Webb CJ. ‘The Effect of Metal Oxides on the

Sorption Kinetics of MgH2 for Hydrogen Storage, a review’. International

Journal of Hydrogen Energy. (To be submitted)

5) Alsabawi K, Gray EM and Webb CJ. ‘A Sieverts apparatus with TPD/TDS

for accurate high-pressure hydrogen uptake and kinetics measurements’.

International Journal of Hydrogen Energy. (submitted)

6) Alsabawi K, Gray EM and Webb CJ. ‘The effect of transition metal oxide

additives on the sorption kinetics of MgH2 for hydrogen storage.’ International

Journal of Hydrogen Energy. (submitted)

Conferences

Alsabawi K, Gray EM and Webb CJ, ‘Kinetic enhancement of the sorption

properties of MgH2 with the additive titanium isopropoxide’. The

16th International Symposium on Metal-Hydrogen Systems (MH2018),

Guangzhou, China, October 28-November 2, 2018. (Oral presentation)

K. Alsabawi, T.A. Webb, E.MacA. Gray and C.J. Webb, ‘Kinetic Enhancement

of the Sorption Properties of MgH2 with The Additive Titanium Isopropoxide’.

2017 International Conference on BioNano Innovation, Brisbane, Australia, Sept

2017. (Poster)

16

1.9. References

1. Global Energy & CO2 Status Report 2017, https://www.iea.org/geco / ; 2018

2. Hughes TP, Kerry JT, Alvarez-Noriega M, Alvarez-Romero JG, Anderson KD,

Baird AH, et al. Global warming and recurrent mass bleaching of corals. Nature.

2017;543(7645):373-7;10.1038/nature21707.

3. Gray EM, Webb CJ, Andrews J, Shabani B, Tsai PJ and Chan SLI. Hydrogen

storage for off-grid power supply. International Journal of Hydrogen Energy.

2011;36(1):654-63;10.1016/j.ijhydene.2010.09.051.

4. Sherif SA, Barbir F and Veziroglu TN. Wind energy and the hydrogen

economy—review of the technology. Solar Energy. 2005;78(5):647-

60;10.1016/j.solener.2005.01.002.

5. Hoffmann P and Dorgan B. Tomorrow's Energy: Hydrogen, Fuel Cells, and the

Prospects for a Cleaner Planet: MIT Press; 2012.

6. Momirlan M and Veziroglu T. The properties of hydrogen as fuel tomorrow in

sustainable energy system for a cleaner planet. International Journal of


7. Solar Bay. Combining Solar Power and Hydrogen Fuel Cells,

http://solarbay.com.au/combining-solar-power-and-hydrogen-fuel-cells/; 2017

8. Blomen LJMJ and Mugerwa MN. Fuel Cell Systems: Springer US; 2013.

9. Afif A, Radenahmad N, Cheok Q, Shams S, Kim JH and Azad AK. Ammonia-

fed fuel cells: a comprehensive review. Renewable and Sustainable Energy

Reviews. 2016;60:822-35;10.1016/j.rser.2016.01.120.

10. Liu J and Barnett SA. Operation of anode-supported solid oxide fuel cells on

methane and natural gas. Solid State Ionics. 2003;158(1-2):11-6;10.1016/S0167-

2738(02)00769-5.

11. Hydrogenics. Fuel Cells, http://www.hydrogenics.com/technology-

resources/hydrogen-technology/fuel-cells/ ; 2018

12. Hirscher M and Hirose K. Handbook of Hydrogen Storage: New Materials for

Future Energy Storage: Wiley; 2010.

13. NIST Chemistry WebBook, NIST Standard Reference Database Number 69,

https://webbook.nist.gov/chemistry/ ; 2019

14. Gattrell M, Gupta N and Co A. Electrochemical reduction of CO2 to

hydrocarbons to store renewable electrical energy and upgrade biogas. Energy

17

Conversion and Management. 2007;48(4):1255-

65;10.1016/j.enconman.2006.09.019.

15. Chemical Economics Handbook, https://ihsmarkit.com/products/hydrogen-

chemical-economics-handbook.html; 2018

16. Young JL, Steiner MA, Döscher H, France RM, Turner JA and Deutsch Todd G.

Direct solar-to-hydrogen conversion via inverted metamorphic multi-junction

semiconductor architectures. Nature Energy.

2017;2(4):17028;10.1038/nenergy.2017.28.

17. Toyota Motor Sales. Hydrogen Fuel Cell Car | Toyota Mirai,

https://ssl.toyota.com/mirai/fcv.html; 2017

18. Hyundai ix35 Fuel Cell, http://www.hyundai.com.au/why-hyundai/design-and-

innovation/fuel-cell ; 2017

19. American Honda Motor. 2017 Clarity fuel cell,

https://automobiles.honda.com/clarity-fuel-cell; 2017

20. Jorgensen SW. Hydrogen storage tanks for vehicles: Recent progress and current

status. Current Opinion in Solid State and Materials Science. 2011;15(2):39-

43;10.1016/j.cossms.2010.09.004.

21. Durbin DJ and Malardier-Jugroot C. Review of hydrogen storage techniques for

on board vehicle applications. International Journal of Hydrogen Energy.

2013;38(34):14595-617;10.1016/j.ijhydene.2013.07.058.

22. von Helmolt R and Eberle U. Fuel cell vehicles: Status 2007. Journal of Power

Sources. 2007;165(2):833-43;10.1016/j.jpowsour.2006.12.073.

23. Zuttel A. Hydrogen storage methods. Naturwissenschaften. 2004;91(4):157-

72;10.1007/s00114-004-0516-x.

24. Rohland B, Eberle K, Ströbel R, Scholta J and Garche J. Electrochemical

hydrogen compressor. Electrochimica Acta. 1998;43(24):3841-6;10.1016/s0013-

4686(98)00144-3.

25. Petitpas G, Aceves SM, Matthews MJ and Smith JR. Para-H2 to ortho-H2

conversion in a full-scale automotive cryogenic pressurized hydrogen storage up

to 345 bar. International Journal of Hydrogen Energy. 2014;39(12):6533-

47;10.1016/j.ijhydene.2014.01.205.

26. Zacharia R and Rather Su. Review of Solid State Hydrogen Storage Methods

Adopting Different Kinds of Novel Materials. Journal of Nanomaterials.

2015;2015:1-18;10.1155/2015/914845.

18

27. Ley MB, Jepsen LH, Lee Y-S, Cho YW, Bellosta von Colbe JM, Dornheim M,

et al. Complex hydrides for hydrogen storage – new perspectives. Materials

Today. 2014;17(3):122-8;10.1016/j.mattod.2014.02.013.

28. Dornheim M. Thermodynamics of metal hydrides: tailoring reaction enthalpies

of hydrogen storage materials. Thermodynamics-Interaction Studies-Solids,

Liquids and Gases: InTech; 2011.

29. Koopal LK. manual of symbols and terminology for physicochemical quantities

and units. Appendix II, Definitions, Terminology and Symbols in Colloid and

Surface Chemistry, Internet Consultation, International Union of Pure and

Applied Chemistry;2001.

30. Bénard P and Chahine R. Storage of hydrogen by physisorption on carbon and

nanostructured materials. Scripta Materialia. 2007;56(10):803-

8;10.1016/j.scriptamat.2007.01.008.

31. National Institute of Standards and Technology. Pressure-Composition-

Temperature (PCT) Measurements,

https://www.ctcms.nist.gov/hydrogen_storage/research_pct.html ; 2008

32. Züttel A, Borgschulte A and Schlapbach L. Hydrogen as a future energy carrier: John Wiley & Sons; 2011.

33. Lennard-Jones JE. Processes of adsorption and diffusion on solid surfaces.

Transactions of the Faraday Society. 1932;28:333;10.1039/tf9322800333.

34. Züttel A. Materials for hydrogen storage. Materials Today. 2003;6(9):24-

33;10.1016/s1369-7021(03)00922-2.

35. Orimo S, Nakamori Y, Eliseo JR, Zuttel A and Jensen CM. Complex hydrides

for hydrogen storage. Chem Rev. 2007;107(10):4111-32;10.1021/cr0501846.

36. Zhou L. Progress and problems in hydrogen storage methods. Renewable and

Sustainable Energy Reviews. 2005;9(4):395-408;10.1016/j.rser.2004.05.005.

37. Bortz M, Bertheville B, Böttger G and Yvon K. Structure of the high pressure

phase γ-MgH2 by neutron powder diffraction. Journal of Alloys and

Compounds. 1999;287(1-2):L4-L6;10.1016/s0925-8388(99)00028-6.

38. Gennari FC, Castro FJ and Urretavizcaya G. Hydrogen desorption behavior

from magnesium hydrides synthesized by reactive mechanical alloying. Journal

of Alloys and Compounds. 2001;321(1):46-53;10.1016/s0925-8388(00)01460-2.

39. Ellinger FH, Holley CE, McInteer BB, Pavone D, Potter RM, Staritzky E, et al.

The Preparation and Some Properties of Magnesium Hydride1. Journal of the

American Chemical Society. 1955;77(9):2647-8;10.1021/ja01614a094.

19

40. Zaluska A, Zaluski L and Ström–Olsen JO. Nanocrystalline magnesium for

hydrogen storage. Journal of Alloys and Compounds. 1999;288(1-2):217-

25;10.1016/s0925-8388(99)00073-0.

41. Zhou S, Chen H, Ding C, Niu H, Zhang T, Wang N, et al. Effectiveness of

crystallitic carbon from coal as milling aid and for hydrogen storage during

milling with magnesium. Fuel. 2013;109:68-75;10.1016/j.fuel.2012.09.002.

42. Huot J, Liang G, Boily S, Van Neste A and Schulz R. Structural study and

hydrogen sorption kinetics of ball-milled magnesium hydride. Journal of Alloys

and Compounds. 1999;293-295(0):495-500;10.1016/s0925-8388(99)00474-0.

43. Wagemans RW, van Lenthe JH, de Jongh PE, van Dillen AJ and de Jong KP.

Hydrogen storage in magnesium clusters: quantum chemical study. J Am Chem

Soc. 2005;127(47):16675-80;10.1021/ja054569h.

44. Crivello JC, Dam B, Denys RV, Dornheim M, Grant DM, Huot J, et al. Review

of magnesium hydride-based materials: development and optimisation. Applied

Physics A. 2016;122(2):1-20;10.1007/s00339-016-9602-0.

45. Wang Y and Wang Y. Recent advances in additive-enhanced magnesium

hydride for hydrogen storage. Progress in Natural Science: Materials

International. 2017;27(1):41-9;10.1016/j.pnsc.2016.12.016.

46. Liu Y, Du H, Zhang X, Yang Y, Gao M and Pan H. Superior catalytic activity

derived from a two-dimensional Ti3C2 precursor towards the hydrogen storage

reaction of magnesium hydride. Chem Commun (Camb). 2016;52(4):705-

8;10.1039/c5cc08801a.

47. Charbonnier J, de Rango P, Fruchart D, Miraglia S, Pontonnier L, Rivoirard S,

et al. Hydrogenation of transition element additives (Ti, V) during ball milling

of magnesium hydride. Journal of Alloys and Compounds. 2004;383(1-2):205-

8;10.1016/j.jallcom.2004.04.059.

48. Liang G, Huot J, Boily S, Van Neste A and Schulz R. Catalytic effect of

transition metals on hydrogen sorption in nanocrystalline ball milled MgH2–Tm

(Tm=Ti, V, Mn, Fe and Ni) systems. Journal of Alloys and Compounds.

1999;292(1-2):247-52;10.1016/s0925-8388(99)00442-9.

49. Barkhordarian G, Klassen T and Bormann R. Fast hydrogen sorption kinetics of

nanocrystalline Mg using Nb2O5 as catalyst. Scripta Materialia.

2003;49(3):213-7;10.1016/s1359-6462(03)00259-8.

20

50. Barkhordarian G, Klassen T and Bormann R. Catalytic mechanism of transition-

metal compounds on Mg hydrogen sorption reaction. The journal of physical

chemistry B. 2006;110(22):11020-4;10.1021/jp0541563.

51. Barkhordarian G, Klassen T and Bormann R. Effect of Nb2O5 content on

hydrogen reaction kinetics of Mg. Journal of Alloys and Compounds.

2004;364(1-2):242-6;10.1016/s0925-8388(03)00530-9.

52. Fátay D, Révész Á and Spassov T. Particle size and catalytic effect on the

dehydriding of MgH2. Journal of Alloys and Compounds. 2005;399(1-2):237-

41;10.1016/j.jallcom.2005.02.043.

53. Friedrichs O, Klassen T, Sanchezlopez J, Bormann R and Fernandez A.

Hydrogen sorption improvement of nanocrystalline MgH2 by Nb2O5

nanoparticles. Scripta Materialia. 2006;54(7):1293-

7;10.1016/j.scriptamat.2005.12.011.

54. Aguey-Zinsou KF, Nicolaisen T, Ares Fernandez JR, Klassen T and Bormann R.

Effect of nanosized oxides on MgH2 (de)hydriding kinetics. Journal of Alloys

and Compounds. 2007;434-435(0):738-42;10.1016/j.jallcom.2006.08.137.

55. Ares-Fernández J-R and Aguey-Zinsou K-F. Superior MgH2 Kinetics with MgO

Addition: A Tribological Effect. Catalysts. 2012;2(3):330-

43;10.3390/catal2030330.

56. Aguey-Zinsou KF, Ares Fernandez JR, Klassen T and Bormann R. Using MgO

to improve the (de)hydriding properties of magnesium. Materials Research

Bulletin. 2006;41(6):1118-26;10.1016/j.materresbull.2005.11.011.

57. Hanada N, Ichikawa T, Isobe S, Nakagawa T, Tokoyoda K, Honma T, et al. X-

ray Absorption Spectroscopic Study on Valence State and Local Atomic

Structure of Transition Metal Oxides Doped in MgH2. The Journal of Physical

Chemistry C. 2009;113(30):13450-5;10.1021/jp901859f.

58. Croston DL, Grant DM and Walker GS. The catalytic effect of titanium oxide

based additives on the dehydrogenation and hydrogenation of milled MgH2.

Journal of Alloys and Compounds. 2010;492(1-2):251-

8;10.1016/j.jallcom.2009.10.199.

59. Pitt MP, Paskevicius M, Webb CJ, Sheppard DA, Buckley CE and Gray EM.

The synthesis of nanoscopic Ti based alloys and their effects on the MgH2

system compared with the MgH2 + 0.01Nb2O5 benchmark. International Journal

of Hydrogen Energy. 2012;37(5):4227-37;10.1016/j.ijhydene.2011.11.114.

21

60. Schimmel HG, Huot J, Chapon LC, Tichelaar FD and Mulder FM. Hydrogen

cycling of niobium and vanadium catalyzed nanostructured magnesium. J Am

Chem Soc. 2005;127(41):14348-54;10.1021/ja051508a.

61. Milošević S, Kurko S, Pasquini L, Matović L, Vujasin R, Novaković N, et al.

Fast hydrogen sorption from MgH 2 –VO 2 (B) composite materials. Journal of

Power Sources. 2016;307:481-8;10.1016/j.jpowsour.2015.12.108.

62. Song M, Bobet J-L and Darriet B. Improvement in hydrogen sorption properties

of Mg by reactive mechanical grinding with Cr2O3, Al2O3 and CeO2. Journal

of Alloys and Compounds. 2002;340(1-2):256-62;10.1016/s0925-

8388(02)00019-1.

63. Bobet JL, Desmoulins-Krawiec S, Grigorova E, Cansell F and Chevalier B.

Addition of nanosized Cr2O3 to magnesium for improvement of the hydrogen

sorption properties. Journal of Alloys and Compounds. 2003;351(1-2):217-

21;10.1016/s0925-8388(02)01030-7.

64. Polanski M, Bystrzycki J, Varin RA, Plocinski T and Pisarek M. The effect of

chromium (III) oxide (Cr2O3) nanopowder on the microstructure and cyclic

hydrogen storage behavior of magnesium hydride (MgH2). Journal of Alloys and

Compounds. 2011;509(5):2386-91;10.1016/j.jallcom.2010.11.026.

65. Agueyzinsou K, Aresfernandez J, Klassen T and Bormann R. Effect of Nb2O5

on MgH2 properties during mechanical milling. International Journal of


66. Dolci F, Chio MD, Baricco M and Giamello E. Niobium pentoxide as promoter

in the mixed MgH2/Nb2O5 system for hydrogen storage: a multitechnique

investigation of the H2 uptake. Journal of Materials Science. 2007;42(17):7180-

5;10.1007/s10853-007-1567-0.

67. Friedrichs O, Agueyzinsou F, Fernandez J, Sanchezlopez J, Justo A, Klassen T,

et al. MgH with NbO as additive, for hydrogen storage: Chemical, structural and

kinetic behavior with heating. Acta Materialia. 2006;54(1):105-

10;10.1016/j.actamat.2005.08.024.

68. Friedrichs O, Sanchez-Lopez JC, Lopez-Cartes C, Klassen T, Bormann R and

Fernandez A. Nb2O5 "pathway effect" on hydrogen sorption in Mg. The journal

of physical chemistry B. 2006;110(15):7845-50;10.1021/jp0574495.

69. Pukazhselvan D, Nasani N, Correia P, Carbó-Argibay E, Otero-Irurueta G,

Stroppa DG, et al. Evolution of reduced Ti containing phase(s) in MgH 2 /TiO 2

22

system and its effect on the hydrogen storage behavior of MgH 2. Journal of


70. Walker G. Solid-state hydrogen storage: materials and chemistry: Elsevier;

2008.

23

STATEMENT OF CONTRIBUTION TO CO-AUTHORED PUBLISHED PAPER

This chapter includes a co-authored paper. The bibliographic details of the

co-authored paper, including all authors, are:

Alsabawi K, Gray EM and Webb CJ.

‘The Effect of Metal Oxides on the Sorption Kinetics of MgH2 for Hydrogen Storage,

A review’. International Journal of Hydrogen Energy. (To be submitted)

My contribution to the paper involved:

1) Literature review

2) Collection of information

3) Manuscript preparation

(Signed) _________________________________ (Date)_12/02/2019___________

Khadija Alsabawi

(Countersigned) __________________________(Date)12/02/2019______________

Corresponding author of paper: Khadija Alsabawi

(Countersigned) ___________________________ (Date)_12/02/19_____________

Supervisor: Colin Jim Webb

24

CHAPTER TWO

A review of the effect of metal oxide additives on the sorption kinetics of MgH2 for

hydrogen storage

K. Alsabawi*, E.MacA. Gray, C.J. Webb

Queensland Micro- and Nanotechnology Centre, Griffith University, Nathan, 4111

Brisbane Australia

Abstract

Magnesium hydride is a most promising hydrogen storage material, due to its high

hydrogen capacity and low cost of production. However, slow kinetics and a high

enthalpy precludes most practical applications. One strategy to improve the hydrogen

sorption kinetics of the Mg–H system that has met with considerable success is ball-

milling with small amounts of additives, which facilitate faster absorption, desorption or

both. In particular, the metal oxide additives, notably niobium pentoxide, but also many

others, have been shown to be particularly successful. This review details the progress

in the improvement of the sorption properties of MgH2 by the addition of different

metal oxides and in terms of their effect on the activation energy, kinetics and

thermodynamic properties.

Keywords: Magnesium hydride; Oxides; Additives; Review; Surface properties.

* Corresponding author: Tel.: +61 7 37353625 E-mail address: [email protected]

25

2. Introduction

Hydrogen is considered to be a viable energy carrier that, when produced from

renewable primary energy sources, can be used to generate clean, affordable energy [1].

Although not a primary energy source, as it is naturally found in compounds with other

elements, it can readily be produced from water through electrolysis or other means and

converted back to heat or electricity when required. When used in a fuel cell to generate

electricity, or combusted in oxygen or air, the emission is predominantly water. With

the increase in power generation from renewable energy sources, the requirement to

store large amounts of energy, to buffer the differences between supply and demand, is

increasing. This is due to the intrinsically intermittent nature of some renewable

sources, such as solar photovoltaic arrays and wind turbines. Hydrogen is an enabling

energy vector with a minimal environmental impact.

Although there are established methods for producing and regenerating energy from

hydrogen, storing hydrogen still presents a major challenge, because of its low boiling

point (20.4 K at 1 atm) and low density in the gaseous state (90 g/m3) [2]. Compressed

hydrogen gas, as currently used in commercial fuel cell vehicles [3] still has a

significantly lower volumetric energy density than gasoline, and presents higher

containment costs and safety issues. Storage as a cryogenic liquid improves the

volumetric density but introduces parasitic energy and economic costs to maintain the

temperature as well as additional safety requirements. Storage in a solid state material

has the advantages of high energy density (higher than liquid hydrogen density), lower

containment pressures than compressed gas, and long term stability compare to batteries

[4, 5].

In solid state material the hydrogen can be stored through physical or chemical

absorption. Materials such as zeolites, carbons, porous polymers and metal organic

frameworks adsorb molecular hydrogen on the surface and in pores through

physisorption. The weak van der Waals forces between the molecular hydrogen and the

surface of the material necessitate low temperatures to achieve sufficient capacity.

Metal and complex hydrides store ionic or atomic hydrogen through chemisorption,

often with covalent or ionic chemical bonds with the opposite problem – the strong

interaction due to the high enthalpy of formation requires high temperatures to release

the hydrogen. The addition of a non-fuel storage material to the actual energy vector

increases the overall weight, and consequently light materials are preferable as solid

state materials for hydrogen storage [2, 6]. The light metal hydrides are low weight,

26

simple hydrogen containing materials with a relative high hydrogen capacity.

Unfortunately, some of the light metal hydrides have high enthalpies, such as LiH

(−90.65 kJ/mol) making it difficult to retrieve the hydrogen. In addition, most metal

hydrides have poor sorption kinetics, making the hydriding and de-hydriding processes

impractically slow. While it is more difficult to overcome thermodynamic limitations,

there has been considerable success in improving the kinetics of light metal hydrogen

sorption, through high energy ball milling and the use of suitable additives that facilitate

the absorption/desorption processes. In particular, Mg-based hydrides have been shown

to have dramatically improved kinetics after ball-milling with small amounts of

additives, and consequently MgH2, with high volumetric and gravimetric capacities of

~0.1g/cm3 and 7.6 wt%, respectively, remains a potential candidate for hydrogen

storage applications [7-9].

A vast range of additives have been tested for kinetic enhancement of MgH2 [10]. While

there is still no clear understanding of what role the additive plays in improving the

absorption and desorption, it is clear that transition metal-based additives are

particularly successful [11-13] and, in particular, the oxides of transition metals. While

transition metals are known to facilitate the dissociation of molecular hydrogen, which

could increase the rate of absorption provided the dissociation step is a rate-limiting

step, the presence of oxygen in Mg is largely regarded as impeding absorption, because

of the very low diffusivity of H in MgO.

2.1. Kinetic processes in magnesium hydride

In order to phase transition from solid magnesium to solid magnesium dihydride in the

presence of molecular hydrogen gas, a number of sequential steps must occur. A

hydrogen molecule must first adsorb via physisorption to the surface of the material.

The molecule must then dissociate into hydrogen atoms which then diffuse through the

bulk material before nucleating the new phase. Desorption starts with nucleation of the

metal phase, followed by diffusion of hydrogen atoms to the surface where they can

associate to form molecular hydrogen, and finally leave the surface.

Slow kinetics is expected to be due to one or more of these sequential steps occurring at

a particularly slow rate. For hydrogenation: (i) the ready formation of a oxide or

hydroxide layer on the surface can prevent or slow the dissociation of molecular H2 as

well as the diffusion of atomic H into the bulk; (ii) the low diffusion coefficient of

MgH2 compared to Mg, means it is much slower for the H to diffuse through a MgH2

27

layer formed on the surface of the Mg [14, 15]; and (iii) the rate of nucleation may

depend on the presence or otherwise of different phases. Slow dehydrogenation may be

cause by: (i) the strong Mg-H bond; (ii) low diffusion rate of H in MgH2; (iii)

nucleation of Mg on the surface of MgH2 requiring a high energy; and (iv) a low

recombination rate of hydrogen atoms at the surface [16, 17].

Extensive research has been devoted to exploring ways to decrease the desorption

temperature and enhance the kinetics and cycle life of MgH2. Possible methods include

changing the microstructure of the hydride by ball milling, and the addition of materials

that reduce the stability of the hydride or catalyse the sorption reactions [18-22].

Zaluska et al. [23] assumed that the crystallite size of ball milled samples is one

important factor in improving the hydrogen sorption kinetic, however other reports

disagree and relate the improvement to the particle size. Dornheim et al. [2] stated that

the effect of crystallite size on sorption kinetics is clear when the particle size is larger

than 700 nm and Dehouche et al. [24] showed that when the crystallite size is smaller

than 80 nm, the influence of the crystallite size is not substantial.

2.2. Mechanical ball milling and additives

Mechanical ball milling of metal powders under reactive gas is widely used to synthesis

metal hydrides [2,25]. The resulting hydrides are usually nanocrystalline materials,

which have a fast hydrogen sorption kinetics, due to the introduction of defects and

fresh surface areas as well as the reduction of particle size, due to the milling [25, 26].

Ball milling pure Mg is often associated with cold welding and reagglomeration of the

magnesium, resulting in large particles and crystallite size [27]. However, ball-milling

MgH2, which is a brittle material, successfully produces quite small particle and gain

sizes. Huot et al. [28] indicated that the specific surface area of the milled β- MgH2 had

increased of about 10 fold compared to the un-milled. The milling destroys the oxide

layer on the surface of Mg by repeatedly breaking the particles and producing fresh

metal surfaces for interaction with hydrogen accelerating dissociation of the molecules.

This also reduces the diffusion lengths for hydrogen in the bulk material [14, 28].

Nanocrystalline ball-milled MgH2 has reasonable absorption kinetics at 300 ºC,

however, this is still too high for practical applications [29]. The ability of metallic Mg

to dissociate hydrogen molecule is poor, due to the low probability of adsorption of the

hydrogen molecule to the Mg-surface of 10-6 [30]. Transition metal oxides are widely

used in small mole fractions as catalysts, as transition metals are known to facilitate

28

hydrogen dissociation, and have proven successful in improving the sorption kinetics of

MgH2 [29, 31, 32]. Furthermore, oxides are typically cheap and only small amounts are

required. Among all the transition metal oxides, see Fig 2-1, niobium pentoxide (Nb2O5)

is still considered one of the most effective additives.

Fig 2-1:Transition metal oxides and their effect on hydrogen desorption kinetics of

MgH2 [31]

During the complex process of milling MgH2 with metal oxides, the metal oxides acts

as a refinement agent for MgH2 powder which in turns facilitate the formation of

nanocrystalline magnesium hydride, that exhibit faster hydrogen sorption behaviour [32,

33]. Oxides with small particle size and a higher initial surface area have been found to

have better dispersion in MgH2 during milling and subsequently provide a greater

number of active sites for enhancing dehydrogenation [34].

2.3. Oxide additives

2.3.1. Magnesium and Metal Oxide additives:

The influence of (Sc2O3, TiO2, V2O5, Cr2O3, Mn2O3,Fe3O4,CuO,Al2O3 and SiO2) metal

oxides on the sorption kinetics of nanocrystalline Mg-based systems was investigated

by Oelerich et al. [29] in 2001. MgH2 was milled for 20 h, and then it was further milled

with the addition of 5 mol% metal oxides for 100 h. The comparison of the results of

MgH2 with and without the additives showed a noticeable enhancement in the

absorption and desorption kinetics for the composite samples. The ball-milled MgH2

composite samples with additives V2O5 and Fe3O4 had the fastest desorption kinetics

29

(Fig 2-2), whereas SiO2 impeded the kinetics even in comparison to pure

nanocrystalline MgH2. Cr2O3 had the highest rate for hydrogen absorption.

Fig 2-2: Desorption rates for metal oxides milled with MgH2, from [29].

As well as 5 mol%, different amounts of additive (0.2, 1 and 5 mol%) were investigated

using the MgH2/Cr2O3 system as an example, since Cr2O3 gave the fastest hydrogen

absorption. The results indicated that the absorption kinetics was independent of the

catalyst amount, for the amounts tested. They showed that only transition metal oxides

catalysts lead to a significant enhancement of hydrogen sorption kinetics and proposed

that the reason for the enhancement by transition metals was due to their different

valence states. This could indicate that the ability of the metal atom to take different

electronic states is a factor in the improvement of the kinetics of the Mg-H2 reaction.

The catalytic effect of Nb2O5 on the sorption kinetics of MgH2 was first studied in 2003,

by Barkhordarian et al. [35] and compared it to pure Nb metal catalysts [36, 37] and

metal oxides. They suggested that the oxides of metals with multiple valence states have

greater catalytic effect, through the electronic exchange reactions with hydrogen

molecules, and accelerating the gas-solid reaction.

In 2004, Castro et al.[38] mechanically milled 10 wt% tungsten oxide (WO3) with

MgH2 under hydrogen pressure for a range of times (1, 2, 4, 8, 12 and 16 h) to study the

30

hydrogen sorption properties of the composite samples. Tungsten (W) was chosen due

to its multivalence state and it was expected that the addition of WO3 would improve

the hydrogen sorption of MgH2. After milling the composite sample for 16 h, the XRD

still showed the reflections of WO3, indicating that oxide did not completely reduce

during milling. With increasing milling time to 8 h, the crystallite size of the MgH2

decreased to 12 nm and a uniform distribution of Mg and WO3 could be seen, little

improved was observed for longer milling times. Absorption studies at 300 ºC and

desorption studies at 330 ºC were performed and it was found that the addition of WO3

was substantially accelerated – at least double the absorption and desorption rates of

pure Mg. For the sample milled for 8 h, it was shown that there was no difference

between the absorption and desorption for first and the 15th cycle. However, the authors

were unable to explain how WO3 improved the sorption kinetics of Mg.

The effect of milling 10 wt% of Fe2O3 in Mg was investigated in 2005 by Kwon et al.

[33]. Samples were milled under hydrogen pressure at various revolution speeds and

times. It was concluded that the post-milling hydriding-dehydriding cycles played a role

in increasing the hydrogen sorption rates, by facilitating nucleation and introducing

cracks on the surface of Mg particles and reducing the particle size that shorten the

diffusion paths of hydrogen atoms, similar to milling. Their optimal milling parameters

were 250 rpm for 24 h.

In 2006, Jung et al .[39] synthesized composites of MgH2 and a range of metal oxides

(V2O5, Cr2O3, Fe3O4, and Al2O3) by ball milling, to investigate their hydriding

properties. This work was similar to the previous work by Oelerich et al. [29] but used

shorter milling times (1 and 2 h). They found that at 300 ºC the composite sample

MgH2/Al2O3 had the highest hydrogen absorption capacity of 4.09 wt%. At lower

temperatures (250-200 ºC) the MgH2/V2O5 composite sample showed even faster

kinetics with absorption capacities up to 3.2 and 2.25 wt%, which was in agreement

with the previous work [29].

Barkhordarian et al.[31] also investigated the mechanism of the catalytic behaviour of

non-metallic and metallic transition-metal doped samples. They argued that electronic

structure of transition metal ions had an important role in improving their catalytic

effect on MgH2. They discussed four factors that could have a role in changing the

electronic properties of the transition metal. The first was the introduction of structural

defects and increasing fresh surface area by high energy ball milling. The second was

31

the thermodynamic stability of the transition metal compound, which should be low

enough to allow a chemical interaction with magnesium but high enough to avoid a

complete reduction of the transition metal compound. From their results they showed

that oxides with low stability had the highest catalytic activity, but when the stability is

too low as in the Re2O7 the compound reacted completely with magnesium and lost its

catalytic ability. The third factor was the nature of the transition metal compound,

where they proposed that it should have a high valence state. This was established when

they compared the catalytic efficiency of different niobium oxides (NbO, NbO2 and

Nb2O5) and found that the catalytic activity increased with the valence state of the

niobium. Lastly, the fourth factor was a high interaction energy of the transition metal

compound with hydrogen.

In 2007, Aguey-Zinsou et al. [32] studied the effect of nanosized transition and non-

transition metal oxides on hydriding/dehydriding kinetics of MgH2. They classified the

non-transition metal oxides as additives with no catalytic potential and the transition

metal oxides as having potential catalytic effects on chemisorption, confirming previous

studies [29]. However, they found that the only non-TMO that exhibit similar kinetics to

TMO was Al2O3 (Fig 2-3). They suggested TMO partial or total reduction to the

elemental metal during mechanical milling could have a catalytic effect on hydrogen

absorption/desorption and the presence of MgO also contributed in improving the

sorption kinetics of MgH2. However, this could not be proven since XRD studies

showed that the TMOs were not reduced during milling. They argued that the oxides

with more ionic than covalent metal-oxygen bonds build up charge on the surface of the

sample during milling as well as balance the surface forces generated by electrostatic

charges on MgH2 particles. They suggested this could be the key parameter to stabilize

the size of MgH2 nanoparticles and prevent cold welding and agglomerations.

32

Fig 2-3: Decomposition temperature of MgH2 milled with different metal oxides

[32].

In 2007, Varin et al. [40] studied the hydrogen desorption properties of MgH2 milled

with 5 wt% of (Al2O3 and Y2O3). They found that doping with Al2O3 had no beneficial

effect on hydrogen desorption kinetics contradicting previous studies [32,39].

Conversely, Y2O3 had a limited effect on the hydrogen desorption kinetics. They were

able to overcome this problem by doping the nanocomposites with specialty Ni which

reduced the desorption temperature and increased the desorption kinetics.

Polanski et al. [41] in 2009 studied the influence of different nano-sized metal oxides

(TiO2, Cr2O3, In2O3, Fe3O4, Fe2O3 and ZnO) on the hydrogen sorption of MgH2. They

employed a simplified geometrical model to calculate the amount of metal oxide needed

to cover the surface of micro-sized particles, considering their geometrical parameters

and densities. They found that as the catalyst particle radius decreased, a smaller

amount was needed to cover the same area on the surface of the sample. A significant

difference between the oxide covering ratios was observed, depending on the oxide

density. Denser oxides had a lower covering ratio and number of active centres on the

surface of MgH2. For a catalyst with nanoparticle size less than 20 nm diameter, they

observed a sharp increase in the number of particles and active centres on the MgH2

surface. Absorption/desorption studies showed that the addition of Cr2O3 and TiO2 had

a significant improvement on the hydrogen sorption properties. Both Fe3O4 and Fe2O3

showed a slight improvement in the desorption process with almost no influence on the

33

absorption process, whereas both In2O3 and ZnO acted as inhibitors rather than catalysts

in the hydrogenation process.

Bellemare et al. [42], in 2012 compared the kinetics after ball milling and cold rolling

MgH2 with transition metal oxides. Samples were milled for 30 min under argon or cold

rolled for five rolling passes to obtain a nanocrystalline size. 2 at.% of OX was mixed

with MgH2 where OX= (Co3O4, Cr2O3, Cu2O, Fe2O3, MnO, Nb2O5, NiO, TiO2, and

V2O5). They found that the sorption kinetics of the samples prepared by the cold rolling

method were slightly slower than that of the ball milled samples. The NiO and Nb2O5

additives gave the fastest desorption kinetics irrespective of the mixing technique,

whereas ball milled V2O5 composite sample showed faster kinetics than the cold rolled

sample. As per the authors recommendation, this comparison is not entirely valid, since

the cold rolling could have been performed under an inert atmosphere rather than in air,

and for a longer time.

In 2017, the catalytic effect of the iron (Fe, Fe2O3 and FeF3) and niobium group (Nb,

Nb2O5, NbF5) additives on MgH2 were examined by Floriano et al. [43]. 2 mol% of

each additive was cryomilled with MgH2 at 77 K for 3 h under Ar, then the hydrogen

desorption properties were investigated by DSC and in a manometric PCT apparatus.

SEM and XRD analysis were performed to determine the microstructure and

morphology of the samples. They found that the fluoride composite samples had the

smallest crystallite size and a very heterogeneous particle size distribution followed by

the oxide composites. These results translated to the absorption/desorption kinetics

where the fluoride containing samples had the fastest absorption and the best

desorption. It was also noticed that the niobium group additives were superior to the

iron additive samples.

2.3.2. Zirconium dioxide (ZrO2)

In 2012 Zeng et al. [44], studied the effect of ZrO2 on the dehydrogenation kinetics of

MgH2 through milling. Different amounts of MgH2 (100 mg- 300 mg) were milled with

ZrO2 ball-mill balls for 20 h, and the kinetics were compared to MgH2 + 1 mol% ZrO2

milled with steel balls. They found the desorption peak for the sample with the smallest

amount of MgH2, milled with ZrO2 balls, to be at around 240 ºC, which was similar to

the sample with 1 mol% ZrO2 milled with the same amount of MgH2. These results

indicated contamination with ZrO2 occurred during milling, which was also confirmed

by the ICP-AES analysis revealing Zr in the samples milled with ZrO2 balls. They have

34

concluded that the presence of ZrO2 helped improving the kinetics hydrogenation and

dehydrogenation behaviours of MgH2.

In 2016, Chen et al. [45] studied the catalytic effect of milling (1 and 5 mol%) ZrO2

with. TEM images showed the grain size of the 5 mol% ZrO2 composite sample was

smaller than the composite with 1 mol%. The 5 mol% sample absorbed 6.73 wt% in 100

s at 150 ºC. At room temperature this sample was able to absorb 4 w% whereas the

sample with 1 mol% was able to absorb 3 wt%. These results indicated that milling

ZrO2 with MgH2 helped to in refining the grain of the composite which was beneficial

in enhancing the hydrogen absorption/desorption capacity of the system.

2.3.3. Cerium (IV) oxide (CeO2)

Singh et al. [46] 2013, prepared different sizes of CeO2 nanoparticles by ball milling

and then studied their effect as additives on the absorption kinetics and decomposition

temperature of MgH2. The synthesis of the nanoparticle was done by varying the ball

milling time (5- 50 h) where 30 h was the optimal time for obtaining nano-crystalline

CeO2. For their studies they used the synthesised CeO2 that was milled for 30 h and 50

h. Different amounts (0.5- 5 wt%) were milled with MgH2 and it was found that the 2

wt% for the 30 h ball milled CeO2 had a 45 ºC decrease in the desorption temperature

when compared to MgH2 and in the first 10 min it absorbed 3.8 wt%, compared to 2.5

wt% for MgH2.

The catalytic effect of CeO2 nano-powder on MgH2 was investigated by Mustafa et al.

[47] 2017, where 5 wt% CeO2 was ball milled with MgH2 for 1 h under Ar, and a

Sievert apparatus used to perform the sorption measurements. Isothermal

dehydrogenation and rehydriding measurements were performed on as-received and as-

milled MgH2 and the composite sample at different temperatures 300 and 320 ºC where

the higher temperature had better kinetics. They indicated that the addition of 5 wt%

nano-powder CeO2 improved the kinetics of MgH2, during dehydrogenation, with the

composite releasing 3.6 wt% hydrogen after 30 min and absorbed 3.95 wt% after 5 min

at 320 ºC. The activation energy for MgH2 of 133.62 kJ/mol also decreased to 108.65

kJ/mol after the addition of CeO2. The XRD patterns showed formation of new phases

of CeH2, and the authors believed the enhancement in the kinetics was due to the

combination of CeO2 and the formation of CeH2.

35

2.3.4. Vanadium oxide

The catalytic effect of high-purity V and different vanadium containing compounds on

the sorption kinetics of MgH2 were investigated by Oelerich et al. [29] in 2001. A low

content (1 mol%) of each of V2O5, VN and VC was milled with MgH2. Negligible

enhancement in the kinetics of MgH2 was found when using high-purity V, whereas

(V2O5, VN, VC) showed a significant enhancement. They found that V2O5 was the most

effective for desorption – at 8 times faster than MgH2. To further extend their

understanding of the kinetic improvement, they exposed the (MgH2)0.99(V2O5)0.01 to air

for 17 h at room temperature, which improved the sorption kinetics. Oelerich et al.

argued that the enhancement of the V2O5, despite its low content, would be due to the

milling behaviour of metal oxides which produce a finer particle distribution and finer

microstructure. In addition, they suggest a stepwise exposure to small amounts of air

also enhances the kinetics. This added to the evidence that transition metal oxides have

a favourable catalytic effect on the sorption kinetics of MgH2 [48].

Du et al. 2008 [49] theoretically investigated the effect of V2O5 on the hydrogenation

and dehydrogenation of MgH2, by performing ab initio density functional theory (DFT)

calculations. They characterized the V2O5 (001) surface as having single, two- and

three-coordinated oxygen sites (O1, O2 and O3) and two types of fourfold hollow sites

(Hole I, Hole II).

36

Fig 2-4: (a) A side view (b) top view of (1*3) V2O5 (001) surface. (c) the optimized

MgH2 clusters. red (O), grey (V), green (Mg) and white balls (H)[49].

The Mg-H bond length in the MgH2 clusters positioned on the V2O5 (001) surface, were

found to be elongated and weakened and the hydrogen molecule was formed at the hole

site on the V2O5 (001) surface (Fig 2-4). The activation barrier for the hydrogen

dissociation on the top vanadium was found to be 1/5 of that of pure Mg (001). They

established that oxygen and hole sites on the V2O5 (001) surface are catalytically active,

which could help accelerate the dissociation of H2.

Grigorova et al. [50] 2013, studied the hydrogen sorption properties of MgH2 + 10 wt%

V2O5 composite samples using the manometric method. At 300 ºC, an absorption

capacity of 6.3 wt% H2 was achieved, but the desorption was slow, after 1 h less than

0.5 wt% was desorbed. By increasing the temperature to 350 ºC desorption finished in

25 min with ~5.4 wt% H2 released. TEM studies showed that the particle size of the

composite decreased further after cycling. They have observed that after 10 cycles the

V2O5 did not completely reduce to V.

In 2015, the kinetics of MgH2 milled with transition metal hydrides, chlorides and

oxides was investigated by Korablov et al. [51]. The oxide sample was 5 mol% (V2O5),

milled with MgH2 under argon. Diffraction peaks for V2O5 were not observed in the in

situ SR-PXD, suggesting a redox reaction during milling. V2O5 was found to readily

reduce with hydrogen or during milling with MgH2, as also observed for Nb2O5 [39,52].

After heating the sample to 450 ºC and cycling twice, XRD patterns for V2H and V

were clearly observed. This suggested that V2H/V system acts as an active catalyst [53],

37

since it does not form an alloy or intermetallic compound with Mg. They observed that

the MgH2 /V2O5 sample had the fastest hydrogen desorption and absorption kinetics at

320 ºC. A drawback of using V2O5 is the formation of considerable amount of MgO

during its reduction in the MgH2 -V2O5 system.

2016, Milosevic et al. [54] studied the hydrogen sorption kinetics of MgH2 milled with

(5 and 15 wt%) VO2 which had been hydrothermally synthesized from V2O5. XRD

showed that after cycling, the composite MgH2-VO2 sample had ~46% of MgO, as also

reported by other authors [51]. The oxidation could be due to high temperature reaction

with oxygen impurities and/or from the reactions:

2 2MgH Mg H→ + (1)

22 2Mg VO MgO V+ → + (2)

H2 and V interaction could lead to formation of an interstitial solid solution, VH~0.8, or

VH2 depending on the pressure and temperatures. XRD analysis confirmed the

formation of a VH2 phase in both samples after cycling. The XRD also showed that in

the 5 wt% VO2 sample, most of the VO2 had been reduced to VH2, whereas the sample

with 15 wt% was only partially reduced, and this result was confirmed with Raman

spectroscopy. The 5 wt% composite was able to desorb 4.9 wt% in 120 s, while the 15

wt% composite desorbed 4.3 wt% after 85 s at 350 ºC, - in agreement with Grigorova

et al. [50].

The hydrogen storage properties of MgH2 doped with (5, 10 and 15 wt%) as-

synthesized VNbO5 was investigated in 2018, by Valentoni et al. [55]. VNbO5 was

prepared by a prolonged annealing of a mixture of V2O5 and Nb2O5 in a molar ratio 1:1.

TPD experiments using a Sievert instrument and Pressure-Composition-Isotherm (PCI)

measurements, as well as calorimetric measurements, were performed to study the

sorption properties of the samples, and structural and microstructure characterization

was performed with XRD. The desorption temperature of the MgH2 was reduced to 280

ºC due to the addition of the VNbO5 catalyst. The composite with 15 wt% VNbO5 was

found to be the best performing absorbing 5.5 wt% hydrogen in 10 min at 160 ºC under

20 bar of hydrogen and full reversibility at 275 ºC for more than 50 cycles. XRD

patterns showed that the catalyst was dispersed in the whole system. Activation energies

were lowered to 99 kJ/mol from the 115 kJ/mol values obtained for the undoped

sample.

38

2.3.5. Titanium Oxide based additives

Wang et al. [56], 2000, employed the reaction ball milling (RBM) method to prepare

Mg-TiO2 composite, using rutile TiO2, in order to study the hydriding properties at

moderate temperatures. XRD studies showed that the high intensity milling resulted in

nanostructured TiO2 and nanocrystalline MgH2 and γ-MgH2. The composite sample

showed superior kinetics at 350 ºC under 20 bar H2 and high hydrogen capacity. There

was no evidence of oxide reduction and they considered the existence of the n-TiO2 was

the reason for the favourable performance. They also mentioned that the distribution

state of the catalyst played an important role. By performing RBM, a brittle Mg hydride

phase was formed which enabled the TiO2 to be highly dispersed. This helped the

hydrogen diffusion through the TiO2 to the Mg-TiO2 interface and into the Mg bulk,

improving the sorption kinetics.

Fig 2-5: Crystal structure of TiO2 a) anatase b) rutile [57].

High energy milling of MgH2 with various types of TiO2 (rutile, anatase (Fig 2-5) and

commercial P25) was investigated by Jung et al. [58] in 2007, to determine the

hydrogen storage capacity and hydriding kinetics of Mg. Different concentrations of the

rutile (3, 5, 7 and 10 mol%) were milled with MgH2, and by performing Selected Area

Diffraction (SAD) patterns they found that by increasing the concentration of the rutile,

MgH2 became increasingly single-crystal-like and the diffraction intensity of the rutile

ring decreased. The formation of nanocomposite MgH2-(rutile)0.05 was observed. It was

indicated that a uniform distribution of the nanocomposite only occurred when the

concentration of the rutile TiO2 was below 5 mol%. The nanocomposite MgH2-

39

(rutile)0.05 exhibited the best hydrogen absorption kinetics, absorbing 4.4 wt% hydrogen

at 300 ºC which was 10 times faster than that of ball milled MgH2.

In 2009, the valence state and the local structure of the metal oxides (V2O5, Nb2O5 and

TiO2nano) ball milled with MgH2 was investigated by Hanada et al. [59] to determine the

precise structures of the catalysts. MgH2 was ball milled for 20 h and then further milled

with 1 mol% (V2O5, Nb2O5 and TiO2nano) for 20 h. Below 250 ºC, all composite samples

desorbed more than 6 mass% of hydrogen, and after rehydrogenation the desorbed

hydrogen in all samples decreased by less than 1 mass%. XRD profiles for the

composite samples showed a growth of the MgO phase in addition to Mg phase after

rehydrogenation at 450 ºC. Based on the Nb, V and Ti K-edge XANES spectra they

were able to confirm that the samples were reduced during ball milling to NbO, VO and

Ti2O3. The K-edge EXAFS spectra for the samples were performed after ball milling,

dehydrogenation and re-hydrogenation with references of Nb and NbO for niobium and

V, VO for vanadium and Ti, TiH2 and Ti2O3 for titanium. The authors analysed the

Fourier transformation of the EXAFS spectra and found that two peaks corresponded to

metal-metal and metal-oxygen bonds. They concluded that the catalytic effect of the

metal oxides on the hydrogen sorption of MgH2 occurred due to the lower oxidation

state and the disarranged local structure.

The catalytic effect of titanium oxide based additives was further investigated by

Croston et al. [34] in 2010. The additives used for this study had been prepared from

alkoxide precursors using a sol-gel route and calcined at temperatures from 70 ºC to 800

ºC. The amorphous oxide structures were calcined at 70 ºC and 120 ºC and the anatase

and rutile structures of the sample were calcined at 400 ºC and 800 ºC respectively. The

addition of titanium oxide based additives to the ball milled MgH2 showed a significant

reduction in the dehydrogenation onset temperature. The dehydrogenation of MgH2-

Ti70 started at 257 ºC, MgH2-Ti120 at 291 ºC and for MgH2-Ti400 started at 263ºC,

however for the MgH2-Ti800 sample, the hydrogenation started at the higher

temperature of 310 ºC. These authors were the first to report the influence of oxide

calcination temperature on the dehydrogenation temperature of MgH2. They also

emphasized that the anatase was more effective than the rutile at lowering the

dehydrogenation temperature of MgH2. They suggested that the Ti+4 in TiO2 samples

were reduced to metallic Ti0 during milling and further reduced after dehydrogenation.

Both anatase and rutile TiO2 were reduced at similar rate, but the higher surface area of

40

the anatase resulted in higher dispersion of active sites throughout the MgH2, compared

to the lower surface area rutile.

Pitt et al. [60] in 2012, studied the effect of synthesized nanoscopic Ti based

compounds on MgH2 and compared it to the effect of Nb2O5. By ball milling

MgH2+0.02 TiO2 for 30 min, they noticed that it was reduced by ca. ¾ of the original

TiO2 forming MgO. Since they were not able to detect residual Ti from the TiO2

reduction in the XRD diffraction pattern, it was assumed that the Ti was

inhomogenously dispersed and only weakly covered large regions of the MgH2 with

distinctly concentrated particles on the outer MgH2 surface. For the reduced additives

such asNb2O5 and TiO2, the authors noticed that a Mg/transition-metal exchange

occurred forming a tertiary structure. These tertiary structures e.g., MgxNb1-xO are

considered to be hard phases that could act as effective grinding agent to reduce MgH2

external powder dimensions.

In 2015, Jardim et al. [61] compared the enhancement of the hydrogen uptake of MgH2

milled with 5 wt% of two different one dimensional TiO2 based additives. The samples

were titanate nanotubes (TTNT-Low) which were synthesized by an alkaline

hydrothermal method and annealed at 100 ºC. This sample was further heated to 550 ºC

to produce the second sample, TiO2 nanorods (TTNT-550 C). The TTNT-550 C was

able to desorb 5.2 wt% at 300 ºC after 10 min compared to the TTNT-Low sample that

only desorbed 2.5 wt%. The TTNT-550, also showed better absorption kinetics

attaining 5.5 wt% of H2 at 350 ºC while the TTNT-Low sample only absorbed 1.7 wt%.

TEM and STEM were used to study the effect of milling on the composite sample,

where they found that the crystal structure and the morphology of TTNT-550 C were

preserved, whereas the TTNT-Low was partially destroyed. They concluded that it is

essential to preserve the morphology of the 1 D TiO2 based nanomaterials in order to

improve the kinetics of the MgH2.

MgH2 was milled with rutile and anatase TiO2 by Vujasin et al. [57], 2016, to study the

desorption properties. XRD, particle size (PSD) analysis and SEM were used to

investigate the role of the additive in improving the kinetics of MgH2. XRD patterns

showed four phases: tetragonal β-MgH2, anatase and rutile TiO2 as well as metallic

magnesium. The PSD results showed that around 10% of particles have average size

around 0.5 mm while around 90% are sized of ~20 μm. Isoconversional kinetic analysis

of the differential thermal analysis (DTA) spectra together with TDS were employed to

41

investigate the mechanism of the hydrogen desorption of the composite samples. The

hydrogen desorption for the composite samples was found to proceed in three steps,

where they suggest that at low temperature H is released and OH is formed, at

intermediate temperature peaks were formed which relates to the catalytic effect of the

composite and at high temperature H2 was desorbed from non-catalysed MgH2. They

found that the desorption activation energy was decreased when using rutile TiO2

additives but the anatase TiO2 had negligible effect. They also found that the Avrami–

Erofeev model for pure MgH2 was changed from n=3 to n=4 for the composite material.

This increase in the Avarmi parameter, n, refers to an increase in nucleation and/or

multidimensional growth of nuclei, reducing the effect of diffusion such that it is no

longer the rate-limiting step.

In 2017, titanium tetraisopropoxide (TTIP), a Ti-based organo-metallic material which

is liquid at room temperature, was used for the first time as an additive to improve the

sorption kinetics of MgH2 alone (Alsabawi et al. [62]). Different amounts of TTIP (0.5-

2 mol%) were ball milled with MgH2 for various times. The effect of TTIP on the MgH2

was compared with the bench mark Nb2O5, and it was found that both have a similar

effect on the absorption and desorption kinetics. For both samples, 0.5 mol% of additive

was sufficient. A hand mixed version of the composite sample also showed excellent

desorption kinetics, but poorer absorption compared to the ball milled MgH2.

The catalytic reaction mechanism of TiO2 additive with MgH2 was investigated by

Pukazhselvan et al. [63] in 2017. They analysed the different compositions of TiO2

milled with MgH2 using XRD, XPS, DSC, HRTEM and chemical mapping. The XPS

study confirmed the existence of Ti3+and Ti2+ valance states and the DSC study showed

a reduction of the Ti oxide phase. The reaction of TiO2 additives with MgH2 was found

to produce a mono-oxide rock salt phase. XRD, HRTEM/SAED and chemical mapping

studies suggested that this phase contain Ti and metal dissolved MgO. They also found

the mono-oxide rock salt phase to be coexisting with Ti2+ cations, which the authors

suggested plays a critical step in the catalysis of the TiO2/MgH2 hydrogen storage

system.

Pukazhselvan et al. [64] 2017, investigated the chemical interaction between MgH2 and

TiO2 by performing a quantitative XRD studies and to explore the formation of a new

phase. MgH2 was ball milled with xTiO2 (x=0.25, 0.33, 0.5 and 1) for 30 h under N2

atmosphere. XRD patterns showed that all composite samples contain a rock salt phase

42

(MgxTiyOx+y), believed to be the product of the chemical interaction between the TiO2

and MgH2, which may promote the dehydrogenation of MgH2. The authors also studied

the cycling life of the composite samples, where they speculated that MgxTiyOx+y salt

phase varies during cycling with a negative effect on the desorption kinetics after ten

cycles. They concluded that the titania additives are not suitable for long term MgH2

cycling.

Following their previous studies [64] Pukazhselvan et al. [65] 2018, chose to investigate

the effect of two synthesized ternary oxides (Mg2TiO4 and MgTiO3) on MgH2. 3 wt% of

each of these oxides was milled with MgH2 for 5 h and DSC profiles were recorded. It

was found that the Mg2TiO4 composite sample released hydrogen at a slightly lower

temperature than the corresponding Mg2TiO3 sample. An XRD study of 10 wt%

additive samples found that during the interaction with MgH2 both additives reduced

and formed an active in-situ MgxTiyOx+y, however, this had slower kinetics than using

titania additives. This work confirmed their previous work [63] and showed that the

reduction and the formation of in-situ MgxTiyOx+y is the most important step in order for

the titania based additive to improve the MgH2.

2.3.6. Magnesium oxide:

In 2006, Aguey-Zinsou et al [66] used MgO as an additive to improve the (de)hydriding

properties of Mg. MgH2 was milled with 10 wt% MgO for milling times from 20 to 100

h, then analyses of the MgH2 particles morphology, crystalline structure,

thermodynamic and H-sorption properties were conducted. The hydrogen kinetics were

improved due to the decrease in the particle size of MgH2. MgO has lubricant and

dispersing properties which reduce the agglomeration and cold welding of MgH2 during

milling. However, according to the XRD patterns there were no changes to the

structural properties. They also found that MgO does not modify the thermodynamic

properties of MgH2.

Usually surface metal oxide layers are not transparent to hydrogen molecules, so a MgO

layer prevents or shows hydrogen penetration [23, 67]. In 2007, Aguey-Zinsou et al.

[32], studied the effect of milling microsized MgO (MgOmicro) and nanosized MgO

(MgOnano) with MgH2 on the sorption kinetics of MgH2. A decrease of the

decomposition temperature was observed for the MgOmicro while the decomposition

temperature of the MgOnano milled for 50 h was found to be higher than pure MgH2

milled for 200 h. They also analysed the morphology of MgOnano and MgOmicro

43

additives and found that smaller particles of MgH2 are formed for MgOmicro than for

MgOnano, although agglomeration occurred for both additives.

Ares-Fernandez et al. [68] presented a detailed analysis in 2012 on the kinetic effects of

modifying MgH2 with MgO. They observed a significant decrease in the hydrogen

desorption temperature for the milled MgH2+10 wt% MgO sample, as the desorption

peak shifted from 336 ºC for milled MgH2 to 262 ºC for the composite sample. It was

observed that hydrogen absorption or desorption could be achieved in less than 100 s at

300 ºC. They also observed that full absorption was completed at 250 ºC in less than

200 s and desorption in less than 800 s. They indicated that the absorption process was

limited by diffusion of hydrogen into MgH2 in a one dimensional process. However, for

desorption the rate limiting process was shifted from nucleation and growth, to be

controlled by the Mg/MgH2 interface as the MgH2 core is shrinking. The rate limiting

steps for MgH2 milled with MgO were comparable to those observed with Nb2O5

addition [69]. However, since MgO was unable to catalyse the reaction of hydrogen

with magnesium, the authors considered the effect of MgO was to reduce the size of the

MgH2 during the milling.

Fig 2-6: Correlation between desorption temperature achieved upon oxide addition

and electronegativity of the oxide additives [68].

Fig 2-6 shows the correlation between the potency of a metal oxide in improving the

hydrogen kinetics and the electronegativity. The higher the oxide electronegativity, such

as MgO, the lower the hydrogen desorption temperature.

44

2.3.7. The effect of Cr2O3 on MgH2

In 2001, Oelerich et al. [29] investigated the influence of metal oxides, such as Cr2O3,

on the sorption kinetics of nanocrystalline Mg-based systems. XRD studies for the

composite sample MgH2/(Cr2O3)0.05 showed that the sample composition did not change

and both Cr2O3 and MgH2 were present. They also found that Cr2O3 did not reduce

during milling, which could be due to using MgH2 instead of pure Mg during the

milling process. Different amounts of Cr2O3 (0.2, 1 and 5 mol%) were used to

determine their effect on the hydrogen sorption property. They found that for the lowest

content of Cr2O3 the hydrogen capacity reached 6.7 wt% and for the highest it was 4.7

wt% within 2 min at 300 ºC.

Song et al. [70] studied the effect of reactive ball-milling Mg with 10 wt% Cr2O3, Al2O3

and CeO2 under H2 on the hydrogen sorption properties in 2002. The sample with Cr2O5

had the best hydriding rate at 300 ºC, with 5.87 wt% of hydrogen absorbed in 60 min;

whereas Al2O3 and CeO2 absorbed 5.66 and 3.43 wt% respectively. The Cr2O5 also had

the highest dehydriding rate at 300 ºC, desorbing 4.44 wt% of hydrogen over 60 min;

whereas Al2O3 and CeO2 desorbed 0.58 and 0.88 wt% respectively. All samples showed

a decrease in the amount of hydrogen desorbed and absorbed during 60 min after the 5th

cycle. XRD studies showed that the Cr2O5 in the Mg + 10 wt% Cr2O5 powder was

reduced during cycling, concluding that it was due to the higher chemical affinity of Mg

than Cr for oxygen.

In 2002, Dehouche et al. [24] investigated the thermal and cycling stability (up to 1000

cycles), of nanostructured MgH2 milled with 0.2 mol% Cr2O3. Based on the PCT

measurements and dynamic hydrogen absorption and desorption curves, they observed

an increase in the reversible hydrogen storage capacity from 5.9 to 6.4 wt%, between

the first and the 500 or 1000 cycle at 300 ºC. Conversely, the desorption kinetics was

reduced by about a factor of four which was attributed to crystallite growth and

structural relaxations. They also reported that the change of sorption properties during

cycling was due to the reduction of Cr2O3 to Cr metal.

A mixture of crystalline Mg and 5 wt% nanocrystalline Cr2O3, synthesized by the

supercritical fluid process was prepared by, Bobet et al. [71] in 2003 by reactive ball-

milling, to compare their morphology, crystallinity, chemical composition and hydrogen

sorption properties with Mg + crystalline Cr2O3 They hypothesized that differences

between the mechanical properties and the particle size of the nano and micro-sized

45

additives may lead to the formation of finer particle size when using the crystalline

Cr2O3. After 2 min the hydrogen sorption capacity was higher for the FSC Cr2O3. Zeta

potential measurement was performed on both samples milled for different times; for

the nanocrystalline Cr2O3 mixture a decrease in zeta potential was observed indicating

that some Cr3+ may have been reduced to Cr metal. They concluded that the presence of

Cr atom clusters or atoms may have an important role in improving the sorption

properties.

In 2004, Bobet et al. [72], following their previous work [71], investigated the effect of

milling parameters on the chemical composition, morphology and crystallinity of Mg +

5 wt% Cr2O3. The XRD patterns obtained after different milling duration, for vibratory

and planetary mill, and for different rotation speeds indicated that the higher rotation

speed allowed for a higher formation rate of MgH2. They concluded that the ball to

powder weight ratio parameter did not affect the efficiency of the milling, since the

efficiency is linked to the total energy of milling and the shock probability. A

subsequent work, Bobet et al. [73] in 2005 focused only on the effect of reactive ball-

milling on hydrogen sorption properties of Mg + Cr2O3 mixture. They showed that by

increasing the injected power of the milling, the hydriding rates increased. To achieve

high hydrogen sorption properties, the authors noted that one must use a planetary ball

mill with high rotation speed and high numbers of balls in the milling vials.

In 2006, Vijay et al. [74] ball-milled nanocrystalline Mg-x wt% Cr2O3 (x=5-20)

composites to study the absorption/desorption kinetics at 100, 200 and 300 ºC as well as

the Cr2O3 distribution and grain size. The absorption rate increased with increasing

Cr2O3 content at each temperature but decreased with decreasing temperature. The

reason for the near-instantaneous absorption was linked to the saturation of hydrogen

atoms on surface and grain boundary of Mg by adsorption and surface/grain boundary

diffusion. At 100 ºC the composite Mg-15 wt% Cr2O3 showed the highest absorption

capacity, suggested to be due to a smaller interparticle distance of Cr2O3 on Mg

surface/grain boundaries. For 200 and 300 ºC the Mg-5 wt% Cr2O3 composite showed

the highest absorption. For desorption, it was observed that it was incomplete in all the

composites at 300 ºC and marginal at 200 and 100 ºC. The desorption rates increased

with increasing the content of Cr2O3 and a similar pattern was observed for the

desorption capacity till 15 wt% then decreased with further increase of Cr2O3 at 300 ºC.

46

In 2006 Pranzas et al. [75] investigated the structural changes of high-energy ball-

milled MgHx and MgH2/10 wt% Cr2O3 with small and ultra small-angle neutron

scattering (SANS/USANS). Different milling parameters were investigated, such as

milling time, and vial and ball material (ZrO2 balls and Al2O3 vials), to gain an insight

into the hydrogen sorption mechanisms for the MgH2/Cr2O3 composite. Crystallite

coarsening was found to take place during cycling at 300 ºC, in agreement with the

literature [24]. They also found that during cycling, particles with radii larger than 10

μm began to break up. They compared the scattering curves of MgH2 and samples

containing Cr2O3 using SANS/USANS, and found that Cr2O3 lead to much smaller

Mg/MgH2 particle sizes, indicating Cr2O3 acts as an agent to reduce particle size during

the milling process [76].

In 2007, Aguey-Zinsou et al. [32] studied the effect of milling micro- and nanosized

Cr2O3 (Cr2O3micro and Cr2O3

nano) with MgH2 on the sorption kinetics. Their results

showed a decrease in the decomposition temperature for both nano and micro

Cr2O3 samples with only slight differences, and there was greater improvement with

increased milling time. SEM and XRD studies showed that the nano and micro

Cr2O3/MgH2 samples had similar particle size and XRD patterns. However, the time for

Cr2O3nano to desorb hydrogen was about half of the time taking the Cr2O3

micro.

In 2008, Polanski et al. [77] investigated microstructural and catalytic activity of nano-

Cr2O3 particles on the hydrogen sorption properties of nanocrystalline MgH2. After

milling MgH2 with 10 wt% Cr2O3 for 20 h, after only 1 min 6 wt% of hydrogen was

absorbed at 300 ºC and 10 bar, and the material was completely desorbed within 5 min

at 325 ºC under 1 bar, in agreement with previous work [32]. The SEM analysis

revealed that Cr2O3 particles were embedded in the MgH2 matrix forming a

nanocomposite structure and were distributed heterogeneously and closely spaced,

forming brittle chains. STEM studies confirmed that the nano-Cr2O3 particles were

present both on the surface and within the micrometric particles of MgH2. They

performed BET specific surface area on the MgH2 powder after milling with nanosized

Cr2O3 and found that it decreased from 43.9 m2/g without catalyst to 9.0 m2/g. Based on

this, they concluded that the distribution of nano-Cr2O3 and the morphology of MgH2

were important but not crucial factors in enhancing the hydrogen sorption properties.

Patah et al. [78] in 2009, milled MgH2 + Nb2O5 + Cr2O3 to investigate the hydrogen

sorption kinetics, activation energy for hydrogen desorption and the morphology

47

changes of the hydride. They prepared a composite sample with 1 mol% Cr2O3 which

desorbed 1.5 wt% of hydrogen in 20 min at 300 ºC, whereas the composite sample with

1 mol% Nb2O5 desorbed 3.6 wt%, indicating the superiority of the Nb2O5 additive.

They then compared MgH2+1 mol% Nb2O5+0.2 mol% Cr2O3 and MgH2+1 mol%

Nb2O5+1 mol% Cr2O3 samples. They found that the small addition of Cr2O3 improved

the desorption kinetics up to 5.1 wt% in 20 min and absorbed 6 wt% in 5 min, but

increasing the amount of Cr2O3 had a detrimental effect on the hydrogen capacity,

desorbing 4.3 wt% and absorbing 5.3 wt%. This suggests that the type of the oxide is

not the only factor in improving the hydrogen capacity, but the amount of the oxide is

also important. The activation energy of MgH2 decreased from 206 kJ mol-1 to 136 kJ

mol-1 after the addition of both oxides. XRD showed that the addition of Cr2O3 together

with Nb2O5 increased the formation of the ternary Mg-Nb-oxide, thereby decreasing the

formation of MgO. The formation of the ternary oxide is linked to improving the

hydrogen desorption kinetics and hydrogen capability of MgH2.

In 2011, Polanski et al. [79] studied the cyclic behaviour (desorption/absorption) and

the microstructure of ball milled MgH2 + 10 wt% Cr2O3. Scanning transmission electron

microscopy (STEM) showed that Cr2O3 was uniformly distributed in the MgH2. The

composite sample was cycled 150 times at 325ºC where a systematic loss of hydrogen

capacity was observed. In the first cycle the sample hydrogen capacity was ~6.7 wt%,

reducing to ~ 4.6 wt% after 75cycles. For the following 75 cycles the hydrogen capacity

did not change, suggesting that most of the microstructural changes responsible for the

catalytic effect occurred during the first 75 cycles. XRD, STEM and XPS studies

showed that the major microstructural changes upon cycling were sintering of

individual powder particles into agglomerates, Cr2O3 particle segregation at the

interfaces between the sintered Mg particles, grain/crystallite growth to larger than 100

nm in the agglomerated particles, and the reduction of Cr2O3 to form Cr and MgO.

Song et al. [80] 2011, used spray conversion to prepare nano-structured Cr2O3 and used

it as an oxide additive to MgH2. The effect of ball milling Mg with 10 wt% of both the

synthesized and as-purchased Cr2O3, for 2 h under 10 bar of hydrogen, on the hydrogen

absorption/desorption kinetics of MgH2 was investigated. SEM images showed that the

addition of Cr2O3 during milling facilitated pulverization of Mg, created multiple

defects and reduced the Mg particle size. The nano-structured synthesized Cr2O3 had a

superior effect on the hydrogen absorption/desorption kinetics when compared to the

48

other additives - it absorbed 4.57 wt% H for 5 min at 593 K and desorbed 0.55 wt% H

over 10 min at 603 K.

2.3.8. Niobium Pentoxide Nb2O5:

Niobium pentoxide (Nb2O5) is known as a promoting agent and an acid catalyst that

enhances the catalytic activity of other catalysts such as molybdenum used for

dehydrogenation of alkanes [81-83]. Nb2O5 is considered one of the best additives for

enhancing the kinetics of MgH2 present [10]. However, the mechanism of hydrogen

sorption enhancement by Nb2O5 is still unclear. When milling MgH2 and Nb2O5

together, Nb2O5 is increasingly embedded in the MgH2 matrix covered by MgH2 and an

outer surface of oxide layer forms during the milling process. This layer prevents further

oxidation of the sample, but it hinders hydrogen diffusion, as does the MgH2 phase [84].

It has been reported that MgH2 milled with Nb2O5 improves the kinetics of the system

due to formation of a ternary Mg-Nb-O phases during the hydrogen sorption cycles

[78,85].

In 2003, Huot et al. [37] investigated the dehydrogenation mechanism of MgH2-Nb

nanocomposites, and its effect on the crystal structure. They performed in-situ XRD

under 1 bar of hydrogen pressure, to gain insight into the structural changes in the NbH

phase. They concluded that the change in the lattice parameters of NbH phase was due

to the hydrogenation/dehydrogenation of the magnesium and not to the change in

temperature. This demonstrated that niobium hydride was transformed into a metastable

NbH0.6 phase before dissociating to form Nb. The formation of an ordered pattern of H

vacancies was found in the hydrogen sub-lattice of the β phase. This indicated that

hydrogen diffusion out of magnesium hydride has to pass through niobium [36,37].

In 2003, Barkhordarian et al. [35] used Nb2O5 as a catalyst for the first time, and tried

to understand its catalytic effect. They compared the use of pure Nb metal catalysts

[36,37] and metal oxides, and they found considerable differences with respect to the

amount of catalysts required for similar kinetics. They suggested that the oxides of

metals with multiple valence states have greater catalytic effect, which in turns improve

the electronic exchange reactions with hydrogen molecules, thus accelerating the gas-

solid reaction. They assumed that Nb2O5 reduced the kinetic barriers and suggested that

the thermodynamic driving force dominates the reaction speed and compensates for

changes in the kinetic barriers. They also found that the absorption kinetics were

negatively influenced by decreasing temperature and that this decreased the

49

thermodynamic driving force, whereas increasing temperature had an opposite effect on

the desorption kinetics, increasing the rate of desorption.

In 2004, Barkhordarian et al. [69] compared different amount of Nb2O5 ranging from

0.05 to 1 mol%, to study the kinetics of the nanocomposite. Adding 0.5 mol % Nb2O5 to

MgH2 and milling for 100 h, the sample desorbed 7 wt% of hydrogen gas within 90 s

under vacuum at 300 ºC and then absorbed ~7 mass% hydrogen within 60 s. The

absorption kinetics were largely independent of the amount of the catalyst used,

suggesting that 0.5 mol% was sufficient, and the activation energy decreased

exponentially with the addition of the catalysts.

In 2005, Fatay et al. [86] studied the effect of milling time of nanocrystalline MgH2

with Nb2O5 on the H-desorption kinetics and desorption temperature. They performed

XRD studies on the milled samples and suggested that the mixing between the hydride

and the catalyst particles occurs on the nanoscale, without any observable reaction

between the hydride and the catalyst. In addition, they found that when the milled

hydride was subsequently milled with a catalyst, the grain size refinement in the second

milling was negligible. The addition of Nb2O5 decreases the activation energy for short

milling times, however at longer milling times subsequent increase in activation energy

was found. They suggested that the difference in the activation energy could be due to

penetration of the catalytic particles into the MgH2 rather than the surface where it was

more active reducing the activation energy.

In 2005, Huhn et al. [87] studied the thermal stability of nanocrystalline MgH2 with

and without Nb2O5. The samples were annealed for 5 h at temperatures between 300

and 400 ºC. The sorption properties for the 20 h milled MgH2 samples annealed at 300

and 310 ºC showed a distinct deterioration; whereas at 310 ºC up to 370 ºC the samples

were quite similar. The absorption and desorption kinetics showed a slight deterioration

for samples annealed at 380 and higher temperatures for 5 h. For the catalysed samples

the kinetics were unchanged at 300 ºC up to the annealing temperature of 380 ºC.

However, the reaction kinetics slowed down and the storage capacity decreased for

temperatures above 380 ºC. Based on their results they indicated that Nb2O5 is stable up

to temperatures below 380 ºC, they also found that the catalysed sample annealed at 400

ºC was worse than the uncatalyzed sample. They concluded that at high temperatures,

380 ºC, Mg reduces the Nb2O5 to form Nb and MgO.

50

In 2005, Friedrichs et al. [85] studied the chemical, structural and kinetic behaviour of

MgH2 and Nb2O5 composite samples with heating. Two mol% of Nb2O5 was milled

with MgH2 for different times. The samples then underwent a series of characterization

tests: structural analysis using XRD with in-situ heating to analyse the cycled samples

and kinetics sorption measurement using a volumetric Sieverts apparatus. Based on

their XRD characterization they hypothesized the existence of a ternary magnesium-

niobium oxide (MgNb2O3.67) and thus a reduction of the Nb2O5 by the magnesium. As a

result of the Nb2O5 reduction, a large amount of MgO is formed in the sample which

exceeds the oxidation level of the ‘pure’ MgH2 sample.

During their in-situ heating experiment, the diffraction patterns showed that Mg reduces

Nb2O5 to form MgO and a ternary oxide with niobium, as well as metallic niobium. An

XPS experiment was performed to study the surface chemical composition of the

catalysed samples, however they were not able to detect any catalyst on the surface of

the milled sample. The cycled sample showed metallic niobium and niobium oxide

phases, which were initially dispersed through the sample and then emerged on the

surface after the first cycle. Their kinetics sorption measurement showed that the initial

capacity of the catalysed MgH2 was 6.5 wt% compared to 7.5 wt% for pure MgH2, due

to the non-absorbing Nb2O5. However, the kinetics of the catalysed samples both in

absorption and desorption of hydrogen was much faster than the pure MgH2 sample

[35,69] which is in agreement with previous studies.

In 2006, Friedrichs et al. [88] studied the effect of Al2O3 and Nb2O5 nanoparticles as

milling additives for MgH2. 10 wt% of additives were milled with MgH2 for different

times, the kinetics sorption measurements were carried out using volumetric Sievert

apparatus and the crystalline characterization was done using XRD. For long milling

times, both additives had improved kinetics when compared to pure MgH2. However,

when samples were milled for shorter times, only Nb2O5 additives had better kinetics,

suggesting that Al2O3 effect on the MgH2 is only due to the reduction in the particle size

rather than a catalytic one. They also compared the kinetic differences between Nb2O5

nanoparticles and microparticles. They found that the MgH2/Nb2O5-nano kinetics

behaviour for shorter milling time is about 10 times faster than the MgH2/Nb2O5-micro

sample which enabled them to analyse the sorption kinetics without long term milling.

This confirmed the catalytic effect of Nb2O5 .

51

The ternary oxide found previously was confirmed [85]. They detected Mg-Nb oxide

with an increased magnesium oxide signal in the diffraction pattern of the cycled MgH2/

Nb2O5 sample. Nb2O5 reduced completely during the hydrogen cycling which was

confirmed by XRD and X-ray photoelectron spectroscopy (XPS) studies for the

MgH2/Nb2O5-micro sample. Therefore, they concluded that the Nb2O5 by itself had no

catalytic effect on the samples. They suggested different niobium phases with lower

oxidation states or the non-stoichiometric magnesium-niobium oxides were responsible

for the catalytic effect on the samples.

In 2006, Friedrichs et al. [84] proposed a pathway model that facilitates hydrogen

transportation through a sample, which was based on their characterization of the

MgH2/ Nb2O5 nano-powder system. By taking into account their previous finding [88]

regarding the reduction of the milling time, they were able to exclude long term-milling

effects when analysing the hydrogen sorption kinetics. When they compared a

physically mixed sample and a sample ball milled for short time, the results indicated

that milling the sample for short time improved the sorption kinetics of the sample.

However, the XRD patterns for both samples showed identical results, where the Nb2O5

peaks disappeared after cycling and a new peak was observed. The new peak could be

due to magnesium oxide phase with larger lattice spacing or by the formation of a

ternary magnesium oxides [85,88]. The ternary magnesium oxide phase was confirmed

by using electron diffraction (ED) where the d spacing values were close to the d

spacing of the ternary oxide found by the XRD. After the hydrogen was desorbed, the

Nb2O5 reacted with the liberated Mg and their product penetrated the outer oxide

surface and formed a network of pathways inside the MgH2 phase.

In 2007, Friedrichs et al. [89] performed an in-situ energy-dispersive XAS and XRD

study on MgH2/Nb2O5. They demonstrated that Nb2O5 partially reduced when milled

with MgH2 then during the first desorption a fast decrease of energy and further

reduction of the sample was observed. Since Nb2O5 has a very low redox potential, the

freed Mg led to faster reduction of the niobium oxide at the initial desorption, after

which it slowed and the oxidation state of the Nb2O5 approached a minimum l. Based

on this work and their previous works [84, 85, 88], they concluded the following; Nb2O5

reduced during milling and was further reduced during heating in hydrogen and the first

absorption process. The crystalline content of Nb2O5 decreased and an apparent MgO

peak growth shifted to lower diffraction angles, indicating the formation of ternary

oxide MgxNbxO.

52

In 2006, Bhat et al. [90] compared the effect of, Nb2O5, with its counterpart halide

NbCl2 and a non-transition metal halide CaF2, in order to understand the role of the

oxygen content in the oxides in improving the hydrogen desorption in MgH2. They

showed that both transition and non-transition metal halides were able to desorb 5.2

wt% and 5.8 wt% of hydrogen in 50 min respectively, whereas Nb2O5 desorbed nearly

5.2 wt% in 80 min. They had concluded that all three samples had similar catalytic

properties. They had prepared Nb2O5-based catalyst using the precipitation technique to

increase the surface area to determine the relation between surface area and the

improvement in desorption kinetics. When they compared the desorption properties of

the commercial and the prepared Nb2O5, they found that the higher surface area of the

prepared sample played a strong role in improving the kinetics even at low temperature

(200 ºC). Then they went on to determine the role of the transition metal cations, Nb+5

and the anion O-2, by performing hydrogen sorption of MgH2 in the presence of the

three catalysts. It was found that the catalytic activity for NbCl2 and CaF2 were higher

than that for Nb2O5, which indicate that neither transition metal cation nor oxide ion

were crucial for hydrogen sorption.

In 2006, Hanada et al. [91] studied the kinetics of hydrogen sorption on the MgH2

composite doped with Nb2O5 by ball milling. They examined the chemical bonding

state of milled MgH2/Nb2O5 sample by XAFS measurement and found that Nb2O5 was

partially reduced by MgH2 during milling, which is in agreement with previous studies

[89]. They considered the H-absorption kinetics had been controlled by a thermally

activated process like the Arrenius process. They reported that 1 mol% Nb2O5+ MgH2

milled for 20 h, was able to absorb ~ 4.5 wt% hydrogen after full desorption under 1.0

MPa within 15 s at ambient temperature. A Kissinger plot was used to determine the

activation energy of H-desorption for the catalysed MgH2, which was deduced from the

first order reaction. The plots indicated that the activation energy was sufficiently

decreased by the catalytic effect of Nb2O5. In the same year, Hanada et al. [92]

confirmed their previous finding [91] by examining the catalytic effect of Nb2O5 and

found that the Nb2O5 was reduced by MgH2 to other Nb compounds, with a valence

state of less than 5+ for the Nb atom.

In 2007, Hanada et al. [93] examined the hydrogen absorption kinetics of MgH2

catalysed with Nb2O5 under various pressure and temperature conditions. For 1 mol%

Nb2O5 milled with MgH2, they found that the absorption reaction rate decreased with

increasing temperature above 150 ºC under any pressure, whereas the hydrogen

53

absorption amount increased significantly with increasing pressure at lower

temperatures. They argued that the reason for this behaviour could be due to the

variation of surface coverage of hydrogen atoms on the surface of Mg with reaction

temperature. However, this assumption does not explain the behaviour of the sample at

high pressure. Therefore, they had to consider the heat generation effect of Mg caused

during the absorption reaction. They proposed that the high reaction speed at low

temperature under high pressure in the short time range was initiated in the temperature

increase by heat generation due to hydrogenation. Also, they found that the high

pressure leads to high reaction speed due to increasing the temperature of the sample

from low to equilibrium temperature (400 ºC).

In 2007, Aguey-Zinsou et al. [83] mechanically milled 17 wt% of Nb2O5 with MgH2 to

investigate the effects on crystalline structure, particle morphology, thermodynamic and

hydrogen sorption properties. Based on their analysis of the crystalline structure they

observed that Nb2O5 was not reduced and the γ-MgH2 was not detected. Also, they

found that the amount of MgO and Nb2O5 stayed constant during the first cycle at 300

ºC, indicating no significant decomposition of Nb2O5. When they performed SEM on

the composite sample, they found that the average particle size of MgH2 milled with

Nb2O5 appeared to be significantly smaller than pure MgH2. Based on their sorption

measurements they found that the addition of the Nb2O5 reduced the decomposition

temperature of MgH2. However, when the composite sample was cycled, they found

that Nb2O5 did not change the thermodynamic properties of MgH2. As a result of their

study they demonstrated that Nb2O5 which is considered as a lubricant, dispersing and

cracking agent during milling, helped in reducing the MgH2 particle size.

In 2007, Dolci et al. [94] investigated the hydrogen uptake for the MgH2/Nb2O5 system.

They used different techniques to explain the features of the interaction of hydrogen

with Nb2O5 system, including: XRD, Electron Paramagnetic Resonance (EPR), Diffuse

Reflectance UV-Vis Spectroscopy, Thermal Desorption Spectroscopy –Mass

Spectrometry (TDS-MS), Differential Scanning Calorimetry and Thermal Programmed

Desorption (TPD). They performed three different types of experiments: treatment in

molecular hydrogen at 673 K, interaction with atomic hydrogen and interaction with

‘nascent’ hydrogen generated by metal-acid reaction in solution. The resulting EPR

spectrum suggested that due to a chemical action of hydrogen, reduction of the material

had occurred where the released electrons were either localized by Nb+5 ions, which

reduced to Nb+4 or delocalized in the conduction band of the solid. The effect of

54

hydrogen treatment resulted in a wide absorption band, spanning form the UV to the

Near Infrared region. They found that the oxide treated with ‘nascent’ hydrogen in

hydrochloric acid solution was the most effective oxide in desorbing hydrogen based on

their observation of the (TDS-MS), confirming that Nb2O5 can act as an ‘active-

reservoir’ of hydrogen and might have a role in the kinetic activation of hydrogen

uptake in mixed system.

Dolci et al. [95] in 2008, investigated the hydrogen uptake in Mg-Nb-O ternary phases

following on the from work of Friedrichs et al. [84, 88] that proposed a ‘reactive

pathway’ mechanism caused by the presence of the ternary oxide (Mg3Nb6O11) which

improved the hydrogen sorption kinetics. The purpose of their study was to understand

the interaction of hydrogen with several mixed Mg-Nb-O compounds. They were able

to obtain three individual ternary phases MgNb2O6, Mg4Nb2O9 and Mg3Nb6O11. Based

on thermal desorption mass spectrometry (TDMS) mixtures with the first two oxides

had similar results, where they underwent irreversible interaction with H2, a slow

reduction of the oxides occurred when contacted with H2 molecule forming water.

However, the third mixture with Mg3Nb6O11 showed a reversible interaction with

molecular hydrogen. They hypothesized that the reason for this reactivity was due to the

occurrence of niobium octahedral clusters inside the oxidic lattice. A similar structure

was observed for NbO, which has similar reactivity with molecular hydrogen.

Therefore, they concluded that both Mg3Nb6O11 and Nb2O5 were active in absorption

and desorption of molecular hydrogen.

In 2009, Dolci et al. [96] compared the interaction of hydrogen with different oxidic

promoters (Nb2O5, WO3 and a ternary Mg/Nb/O oxide), in order to understand the role

of oxides as promoters for hydrogen storage in MgH2. To test the reactivity of the

oxides they used either “nascent” hydrogen or molecular hydrogen. TDMS experiments

were performed on the oxides and it was found that water was the only product

desorbed by WO3 after contact with hydrogen. In comparison Nb2O5 released molecular

hydrogen after reacting with nascent hydrogen and the Mg/Nb/O multiphase mixture

showed reversible uptake of nascent hydrogen, but even better uptake when reacted

with molecular hydrogen.

55

In 2008, Yang et al. [97] applied the Chou model to understand the effects of particle

size and the ball milling time on dehydriding kinetic mechanism of MgH2- Nb2O5 1. The

results of using these calculations indicated that the rate-limiting step is the surface

penetration of hydrogen atoms. Based on their linear regression method they found that

the dehydriding reaction rate of the ball-milled nano particle composite was 1-9 times

faster than that for the micro particle.

Porcu et al. [98], 2008 used TEM to study the microstructure of a MgH2-Nb2O5

nanocomposite formed by ball milling. They compared the particle size of MgH2 milled

for 30 h with MgH2-Nb2O5 milled for the same amount of time. They found only a

slight decrease of particle size in the mixed sample, which is attributed to the presence

of oxides that facilitate the cracking of the MgH2 particles. They indicated the

possibility of an interaction between MgH2 and Nb2O5 during ball milling due to the

iconic character, which promotes the diffusion of hydrogen through the Mg matrix.

They suggested that the defective and highly dispersed structure of Nb2O5 may provide

easy pathways for hydrogen atoms to diffuse through the Mg particle core and the

surface area of the catalyst was increased due to ball milling to provide a wider area for

the Mg/MgH2 reaction to occur. Also, they indicated that Nb2O5 partially reduced to

Nb2O forming MgO during desorption process. Then when Nb2O5 inter-diffused with

MgO, a new mixed oxide MgNb2O3.67 was formed which enhanced the kinetics.

In 2009, Aurora et al. [99] performed kinetic studies and microstructural investigation

to distinguish the difference between the catalytic effect of nanometric and micrometric

Nb2O5 on MgH2. A morphological refinement of the micro Nb2O5-MgH2 composite was

shown as well as a progressive reduction of the dimensional distribution. Isothermal

hydrogen absorption/desorption measurements were carried out and MgH2 milled with

nano Nb2O5 showed an improved kinetic performance when compared to the micro

Nb2O5. They showed that using nano Nb2O5 reduced the milling time requirement to

one third of the milling time and halved the desorption time, under the same

experimental condition. XRD confirmed the formation of ternary oxides (MgNbxOy) in

the cycled MgH2/ Nb2O5 milled sample, which was also found in previous studies,

1 ‘Chou model involves a series of analytic equations for various kinds of rate-controlling steps. These

formulae express the reacted fraction ξ of hydriding/dehydriding reaction as an explicit function of time

t, temperature T, hydrogen partial pressure p(��) and particle size ��. Among them, the reacted

fraction (or transformation fraction) ξ is a key parameter to describe the kinetic behavior’98. Liu Y,

Li Q and Chou K-c. Dehydriding reaction kinetic mechanism of MgH2-Nb2O5 by Chou model. Trans

Nonferrous Met Soc China. 2008;18(Copyright (C) 2015 American Chemical Society (ACS). All Rights

Reserved.):s235-s41;10.1016/s1003-6326(10)60209-9..

56

however its role in the sorption process is still unclear [84,89]. Based on electron

microscopy observations, they found that after milling, the catalyst particle size for the

nano Nb2O5 was lower than the micro Nb2O5, this corresponded to a higher volume of

MgH2 phase in contact with the catalyst. Hence, this increased the surface of the

interface between the catalyst and MgH2, which in turns increased the number of

nucleation sites of the reacted phase (Mg). The authors found that per unit time there

was a higher number of nucleation sites of the reacted phase, which explains the faster

kinetics of nano Nb2O5 respect to micro Nb2O5 sample.

In 2011, Rahman et al. [100] following the steps of Dolci et al.[95] investigated the

hydrogen sorption and desorption in MgH2 ball milled with 1 mol% MgNb2O6,

Mg4Nb2O9 and Mg3Nb6O11. These three distinct phases were formed during ball milling

MgH2 with Nb2O5. Using XRD to characterize the crystal structure after milling, they

found that the relative amounts of additives were unchanged and hence there was no

interaction with MgH2. They suggested that the occurrence of nanostructured phase

after milling together with thermodynamic effects had an important role on the reaction

kinetics by reducing the release time of hydrogen under isothermal conditions. During

the desorption reaction, a significant decrease in the amount of the additives was

observed, for the MgNb2O6 mixture, whereas the reduction was strong for Nb2O5 and

limited for Mg4Nb2O9 mixture. For the Mg3Nb6O11 mixture they were able to detect a

progressive decrease in the amount of additive as function of desorption time, which

may suggest a chemical interaction with the Mg/MgH2 matrix or an interaction of the

oxide phase with MgO. Based on these observations, they confirmed that the additives

play an active role in improving the kinetics which is in agreement with suggested

“pathway effects” for hydrogen sorption in Mg [84]. The presence of the octahedral

niobium clusters in the structure of Mg3Nb6O11 additive, act as an insertion site for

hydrogen uptake and release and had a role in the reversible interaction of hydrogen and

improved kinetics.

In 2011, Rahman et al. [101] wanted to understand the synthetic route for the

preparation of MgNb2O6, Mg4Nb2O9 and Mg3Nb6O11 phases, by facilitating solid state

reactions between MgO, Nb2O5 and/or metallic Nb precursors. The synthesis of these

ternary compounds were reported previously by Pagola et al. [102] using a conventional

solid state reaction based on a simple annealing of commercial precursor materials.

However, the formation mechanisms of theses phases were unclear and further

investigations were needed. Rahman et al. characterized the pure ternary phases in their

57

reactivity with molecular hydrogen. The samples were prepared at different

temperatures of 600 ºC, 800 ºC and 1000 ºC, then the samples at the highest temperature

were tested in hydrogen absorption/desorption experiments. A negligible amount of

water was desorbed by the three phases, but for the Mg3Nb6O11 sample it was observed

that the H2 desorption drastically changed. They concluded that both MgNb2O6 and

Mg4Nb2O9 were totally inert, whereas Mg3Nb6O11 had an active reversible interaction

with molecular hydrogen occurring at around 395 ºC.

In 2011, Ma et al. [103] studied the hydrogen desorption properties, microstructure,

kinetics and chemical bonding states of catalyst surfaces for MgH2 and 1 mol% Nb2O5

composites milled for different times under 10 bar H2 atmosphere. They found during

ball milling the catalytic effect of Nb2O5 was gradually activated and improved with

increasing the milling time. The desorption activation energy (Ea) was calculated by the

Kissinger method and Ea decreased with increasing ball milling times. They suggested

that the better distribution and refined size of Nb2O5 was responsible for improving the

desorption kinetics especially for the sample milled for 20 h.

In 2012, Nielsen et al. [52] investigated the MgH2-Nb2O5 composite by in situ

synchrotron X-ray diffraction in order to further understand the effect of Nb2O5 for

hydrogen release and uptake in Mg/MgH2 system. They also were trying to explore the

chemical reactions in the Mg-MgH2-Nb2O5 system. Based on their SR-PXD data and

Rietveld refinement analysis, they proposed that a reaction occurred during ball milling

between MgH2 and Nb2O5, and at elevated temperatures Nb2O5 reacted with Mg to

produce the ternary oxide MgxNb1-xO. They also showed that the formation of the

ternary oxide MgxNb1-xO lead to a decrease in the gravimetric hydrogen storage

capacity after ball milling and cycling the sample at elevated temperature (>300 ºC). In

this work they suggested that the ideal ternary composition was Mg0.6Nb0.4O which

considered the active material for the prolific kinetic properties for the hydrogen release

and uptake in the system.

In 2013, da Conceicao et al. [104] compared three Nb materials, commercial oxide (c-

Nb2O5), high surface area synthesized oxide (s-Nb2O5) and metallic Nb in order to study

their catalytic capability on hydrogen sorption of MgH2. The hydride phase stability for

the three samples were examined by using a Differential Scanning Calorimeter (DSC)

and based on their results they found that the hydride decomposition temperature

decreased when adding higher amounts of Nb2O5. They compared the absorption

58

kinetics of the composite samples and found that all samples absorbed the same amount

of 5.2 wt% hydrogen. However, the s- Nb2O5 was 6 times faster than c-Nb2O5 sample

and the metallic Nb sample showed the slowest kinetics. Also, they found that

increasing the temperature from 300 to 350 ºC had a negative effect on the absorption

kinetics as indicated in previous study [35]. They had suggested that increasing the

amount of Nb2O5 improved the desorption kinetics however the metallic Nb showed

slower kinetics compared to the other samples. They concluded that the high surface

area improved the kinetics as well as other factor such as the amount of the catalyst and

the experimental conditions.

In 2013, Ma et al. [105] investigated the Nb-contained catalysts (Nb, NbO and Nb2O5)

in MgH2 nanocomposites during full cycling. They found that the Nb2O5 reduces during

milling with MgH2 to form NbH2 and MgO, with the reaction equation written as:

�� +�

�� = �� +

�� +

�

↑ (3)

Nb also reacted with MgH2 during milling to form NbH

�� + �� = �� + �� (4)

However, NbO did not react during milling which was confirmed by the XRD profiles.

After one cycle for the Nb-doped and Nb2O5-doped samples the compositions were the

same, except for the presence of MgO in the Nb2O5-doped sample. They then

investigated the role of MgO by comparing Nb-doped sample and Nb-MgO-doped

sample and found that both exhibited the same dehydrogenation peak-temperature,

indicating that the presence of MgO provided no contribution to the desorption

property. They concluded that the catalytic effect in MgH2- Nb2O5 and MgH2-Nb

nanocomposites were attributed to Nb, which is based on the Nb-gateway model, where

Nb facilitates the hydrogen transportation from MgH2 to the outside. They compared

the size of the samples using the Scherrer equation:

� =��

�� (5)

and their morphology by SEM and found that the Nb2O5 doped sample grain size was

distinctly smaller than the other samples. Since Nb2O5 is considered as a ceramic, it

exhibits a hard and brittle character which helps refine the composite sample during

milling, while Nb metal and NbO, which have metallic properties are ductile and

difficult to mill. They suggested that the smaller size had a partial effect on the

59

dehydrogenation property. The Nb-gateway model was validated in MgH2- Nb2O5

system, when Nb and NbH2 with 10-20 nm in size were observed in the 1 mol% Nb2O5-

doped sample. Finally, they concluded that Nb crystals which were highly dispersed in

the composites act as the essential catalyst and followed the Nb-gateway model in the

enhancement of desorption properties.

Fig 2-7: The selected structural models of MgH2 on (a) Nb(1 0 0), (b) NbO(1 0 0),

(c) NbO(1 10), (d) NbO(1 1 1) and (e) Nb2O5(1 0 0). Atomic colour codes: Mg,

green; H, white; Nb, blue; O, red. [106]

A density functional theory study was conducted by Takahashi et al.[106] in 2013, to

investigate the catalytic effect of Nb, NbO and Nb2O5 with different surface planes (Fig

2-7) on the dehydrogenation of MgH2. They showed that the adsorption site of Mg and

H are different depending on the surface planes and found that among all the surface

planes NbO (1 1 1) had the best kinetics. NbO (1 1 1) electron pairing was responsible

for providing very high adsorption energy instead of the charge transfer, also they

indicated that there was a strong interaction between the s-state of H and d-state of Nb

based on the electronic structure of NbO (1 1 1).

60

Pukazhselvan et al.[107] 2015, studied the formation of Mg-Nb combined active rock

salt (RS) nanoparticle MgxNbyOx+y. They tried to clarify the chemical interaction

between MgH2 and Nb2O5, by monitoring ball milled MgH2+ n Nb2O5 (n=0.083-1.5) for

different time intervals (1, 10 and 30 h). By performing DSC studies, they confirmed

that the in-situ formed RS structures were able to enhance the hydrogen storage

properties of MgH2 system. Based on the XRD patterns they suggested that the initially

formed RS had a high concentration of Nb, but with increasing the milling time to more

than 10 h, the Mg concentration increased. An interesting outcome of milling MgH2+

0.5 Nb2O5 for 10 h and 30 h was the coexistence of two different Mg-Nb-O phases,

Mg4Nb2O9 and MgxNbyOx+y Rs, which they relate to the characteristic of the specific

stoichiometry. The crystallographic structure of the RS was found to be similar to the

MgO structure but they possessed higher lattice parameters, at least 1-4% [108], and

cell volume than MgO.

In 2016, Pukazhselvan et al.[109] and following their previous study [107] they have

noticed that during milling a considerable amount of Fe impurities were present as Nb

dissolved in the RS to form MgxNbyOx+y. since Fe was used before as an additive, they

had to investigate the Fe impurities by using different milling medium, such as steel and

zirconia. 6MgH2+Nb2O5 was milled for 30 h using a steel milling medium and a

zirconia milling medium. The steel milled sample had ~2.5 mol% Fe, and monophasic

MgxNbyOx+y, which suggested to be a Mg0.75Nb0.25O composite based on some

occupational analysis. Whereas the zirconia milled sample had ~4.5 mol% zirconia and

more than one Mg-Nb-O rock salt phase, where one phase was closer to NbO and the

other was MgxNbyOx+y.

Then in 2016, Pukazhselvan et al. [108] investigated this further, expanding their

research , by choosing one concentration of Nb2O5 (n = 0.167) and ball-milled MgH2+

0.167 Nb2O5 for different time intervals (2, 5, 15, 30 and 45 min, as well as 1, 2, 5, 10,

15, 20, 25 and 30 h). They confirmed the formation of catalytically active Nb-doped

MgO nanoparticles, by transforming the MgH2+0.167 Nb2O5 composite through an

intermediate phase typified by Mg m-xNb 2n-yO 5n-(x+y). The Mg0.75Nb0.25O composite

exhibited an outstanding catalytic effect on the hydrogen storage performance of MgH2.

They noted that the Nb-dissolved MgO significantly enhanced the sorption kinetics of

MgH2, whereas MgO by itself provide an undesirable hydrogen diffusion barrier to the

MgH2 system.

61

2.4. Summary

• Additives such as metal oxides shown to help in reducing the activation energy of

the MgH2 by providing a dissociation and recombination sites for hydrogen

absorption and desorption.

• During ball milling Mg or MgH2 powder particles break down providing fresh un-

oxidized surfaces which in turns reduces the activation energy. Ball milling with

hard/brittle additives, increases the surface area to the volume ratio of the material.

• Higher surface area increases the rate of hydrogen contacts which leads to a

reduction in the required pressure drive. The increased surface/volume ratios lead to

faster diffusion of the hydrogen through the bulk material due to the shorter path.

• The most effective oxide additives were the ones with multivalent metals, which

appeared to enhance the kinetics compared those with a single oxidation state that

had no effect, The structure of the oxides was found to be important in providing

additional sites for the dissociation of the hydrogen [24, 71].

• The brittle nature of the oxides together with ball milling causes structural

refinement that help in breaking down the hydride into smaller particles [77, 105]

• The effect of metal oxides catalysts might be due to many factors other than its

electronic structure. Other mechanisms that may affect this are: catalyst acting as

milling aid, high defect density, surface area and size effect, crystal structure and

availability of sites for OH groups [110].

• Currently, considering all the additives studied, no materials have been found to

have the required combination of fast reaction kinetics, ability to be cycled and high

H2 capacity for practical applications.

• Nb2O5: has been repeatedly reported as the best metal oxide additives in improving

the hydrogen sorption of MgH2 [31, 35, 69, 86, 88]

• Barkhordarian et al. and Fatay et al. suggested that Nb2O5 acts as a catalyst for

hydrogen desorption and found no evidence for any structural transformation of

Nb2O5 phase.

• Using XPS Friedrichs et al. showed a reduction in the Nb species after cycling

and observed a new phase MgNb2O3.67 using XRD. Therefore, they suggested

that the Nb2O5 is not the active catalyst and proposed a reaction mechanism.

• These two studies, and others, contradicted each other. For most studies, the

difficulty determining the nature of the oxide phase,

62

• TiO2: it was found that the TiO2 reduces during ball milling to Ti2O3 which has a

lower oxidation state, which has been linked to improving the sorption kinetics of

MgH2 [34,59,60]. A uniform distribution of TiO2 on the surface of MgH2 occurs

when the concentration of TiO2 is below 5 mol%, and this nanocomposite was able

to absorb hydrogen 10 times faster than that of ball milled MgH2 [56,58]. By using

the calcination technique it was found to improve the dehydrogenation temperature

and the anatase form of TiO2 was more effective than the rutile in improving the

dehydrogenation of MgH2[34].

• ZrO2: was found to enhance the sorption kinetics by reducing the grain size of the

composite during the milling [45].

• CeO2: was found that after the addition of CeO2 the activation energy of MgH2

decreased, which help enhancing the kinetics of MgH2, and it was observed that the

formation of a new phase CeH2 together with CeO2, also improved the kinetics [47].

• MgO: was found to decrease the desorption temperature, due to its high

electronegativity as well as the ability to decrease the MgH2 particle size when

milled together [32,66,68].

• Cr2O3: was observed that Cr2O3 reduces after cycling to form Cr atom or clusters

which have an important role in speeding the sorption kinetics of MgH2 [70, 71, 79].

The reversible H2 storage capacity increased after cycling for 500 to 1000 cycles but

due to the reduction of the additive, the desorption kinetics slowed [24]. At

temperature as low as 100 ºChigh H2 absorption was observed which was ascribed

to the small interparticle distance of Cr2O3 on Mg surface boundary [74]. When

comparing the micro and nano Cr2O3, it was found that they have similar particle

size and both are able to decrease the decomposition temperatures; however, the

nano sample desorbed hydrogen in half time taken for the micro Cr2O3[32].

• V2O5: was found to be readily reduced in content with hydrogen or milled with

MgH2. TheV2H/V system act as an active catalyst [53] [54], since it does not form

alloy or intermetallic compounds with Mg. A drawback with using V2O5 is the

formation of considerable amount of MgO during its reduction [51]. The

enhancement with V2O5, is attributed to the milling behaviour of metal oxides which

produce a finer particle distribution and finer microstructure [29].

63

2.5. Reference

1. Yu X, Tang Z, Sun D, Ouyang L and Zhu M. Recent advances and remaining

challenges of nanostructured materials for hydrogen storage applications.

Progress in Materials Science. 2017;88:1-48;10.1016/j.pmatsci.2017.03.001.

2. Dornheim M, Eigen N, Barkhordarian G, Klassen T and Bormann R. Tailoring

Hydrogen Storage Materials Towards Application. Advanced Engineering

Materials. 2006;8(5):377-85;10.1002/adem.200600018.

3. Staffell I, Scamman D, Velazquez Abad A, Balcombe P, Dodds PE, Ekins P, et

al. The role of hydrogen and fuel cells in the global energy system. Energy &

Environmental Science. 2019;12(2):463-91;10.1039/c8ee01157e.

4. Züttel A, Borgschulte A and Schlapbach L. Hydrogen as a future energy carrier:

John Wiley & Sons; 2011.

5. Gray EM, Webb CJ, Andrews J, Shabani B, Tsai PJ and Chan SLI. Hydrogen

storage for off-grid power supply. International Journal of Hydrogen Energy.

2011;36(1):654-63;10.1016/j.ijhydene.2010.09.051.

6. Reilly JJ and Sandrock GD. Hydrogen storage in metal hydrides. Sci

Am;(United States). 1980;242(2):118-29.

7. Züttel A. Materials for hydrogen storage. Materials Today. 2003;6(9):24-

33;10.1016/s1369-7021(03)00922-2.

8. Grochala W and Edwards PP. Thermal decomposition of the non-interstitial

hydrides for the storage and production of hydrogen. Chemical Reviews.

2004;104(3):1283-316;10.1021/cr030691s




10. Webb CJ. A review of catalyst-enhanced magnesium hydride as a hydrogen

storage material. Journal of Physics and Chemistry of Solids. 2015;84:96-

106;10.1016/j.jpcs.2014.06.014.

11. Jin S-A, Shim J-H, Cho YW and Yi K-W. Dehydrogenation and hydrogenation

characteristics of MgH2 with transition metal fluorides. Journal of Power


12. Bobet J. Study of Mg-M (M=Co, Ni and Fe) mixture elaborated by reactive

mechanical alloying — hydrogen sorption properties. International Journal of

Hydrogen Energy. 2000;25(10):987-96;10.1016/s0360-3199(00)00002-1.

64

13. Varin RA, Li S, Chiu C, Guo L, Morozova O, Khomenko T, et al.

Nanocrystalline and non-crystalline hydrides synthesized by controlled reactive

mechanical alloying/milling of Mg and Mg–X (X = Fe, Co, Mn, B) systems.

Journal of Alloys and Compounds. 2005;404-406:494-

8;10.1016/j.jallcom.2004.12.176.

14. Zhu M, Lu Y, Ouyang L and Wang H. Thermodynamic Tuning of Mg-Based

Hydrogen Storage Alloys: A Review. Materials (Basel). 2013;6(10):4654-

74;10.3390/ma6104654.

15. Aguey-Zinsou K-F and Ares-Fernández J-R. Hydrogen in magnesium: new

perspectives toward functional stores. Energy & Environmental Science.

2010;3(5):526-43;10.1039/B921645F.

16. Vajeeston P, Ravindran P, Hauback BC, Fjellvåg H, Kjekshus A, Furuseth S, et

al. Structural stability and pressure-induced phase transitions in MgH2. Physical

Review B. 2006;73(22):224102;10.1103/PhysRevB.73.224102.

17. Tanniru M, Tien HY and Ebrahimi F. Study of the dehydrogenation behavior of

magnesium hydride. Scripta Materialia. 2010;63(1):58-

60;10.1016/j.scriptamat.2010.03.019.

18. Sakintuna B, Lamaridarkrim F and Hirscher M. Metal hydride materials for

solid hydrogen storage: A review☆. International Journal of Hydrogen Energy.

2007;32(9):1121-40;10.1016/j.ijhydene.2006.11.022.

19. Reule H, Hirscher M, Weißhardt A and Kronmüller H. Hydrogen desorption

properties of mechanically alloyed MgH2 composite materials. Journal of

Alloys and Compounds. 2000;305(1-2):246-52;10.1016/s0925-8388(00)00710-

6.

20. Jain IP, Lal C and Jain A. Hydrogen storage in Mg: A most promising material.


44;10.1016/j.ijhydene.2009.08.088.



Physics A. 2016;122(2):1-20;10.1007/s00339-016-9602-0.

22. Crivello JC, Denys RV, Dornheim M, Felderhoff M, Grant DM, Huot J, et al.

Mg-based compounds for hydrogen and energy storage. Applied Physics A.

2016;122(2):1-17;10.1007/s00339-016-9601-1.

65



25;10.1016/s0925-8388(99)00073-0.

24. Dehouche Z, Klassen T, Oelerich W, Goyette J, Bose TK and Schulz R. Cycling

and thermal stability of nanostructured MgH2–Cr2O3 composite for hydrogen

storage. Journal of Alloys and Compounds. 2002;347(1-2):319-

23;10.1016/s0925-8388(02)00784-3.

25. Huot J, Ravnsbæk DB, Zhang J, Cuevas F, Latroche M and Jensen TR.

Mechanochemical synthesis of hydrogen storage materials. Progress in Materials

Science. 2013;58(1):30-75;10.1016/j.pmatsci.2012.07.001.

26. Skripnyuk VM, Rabkin E, Estrin Y and Lapovok R. The effect of ball milling

and equal channel angular pressing on the hydrogen absorption/desorption

properties of Mg–4.95 wt% Zn–0.71 wt% Zr (ZK60) alloy. Acta Materialia.

2004;52(2):405-14;10.1016/j.actamat.2003.09.025.

27. David E. Nanocrystalline magnesium and its properties of hydrogen sorption.

Journal of Achievements in Materials and Manufacturing Engineering.

2007;20(1-2):87-90.

28. Huot J, Liang G and Schulz R. Mechanically alloyed metal hydride systems.

Applied Physics A Materials Science & Processing. 2001;72(2):187-

95;10.1007/s003390100772.

29. Oelerich W, Klassen T and Bormann R. Metal oxides as catalysts for improved

hydrogen sorption in nanocrystalline Mg-based materials. Journal of Alloys and

Compounds. 2001;315(1-2):237-42;10.1016/s0925-8388(00)01284-6.

30. Reimann AL. LVII.The clean-up of hydrogen by magnesium. The London,

Edinburgh, and Dublin Philosophical Magazine and Journal of Science.

2009;16(106):673-86;10.1080/14786443309462320.

31. Barkhordarian G, Klassen T and Bormann R. Catalytic mechanism of transition-

metal compounds on Mg hydrogen sorption reaction. The journal of physical

chemistry B. 2006;110(22):11020-4;10.1021/jp0541563.




33. Kwon I-H, Bobet J-L, Bae J-S and Song M-Y. Improvement of hydrogen-

storage properties of Mg by reactive mechanical grinding with Fe2O3. Journal

of Alloys and Compounds. 2005;396(1-2):264-8;10.1016/j.jallcom.2004.12.036.

66




8;10.1016/j.jallcom.2009.10.199.


nanocrystalline Mg using Nb2O5 as catalyst. Scripta Materialia. 2003;49(3):213-

7;10.1016/s1359-6462(03)00259-8.

36. Pelletier J, Huot J, Sutton M, Schulz R, Sandy A, Lurio L, et al. Hydrogen

desorption mechanism in MgH2-Nb nanocomposites. Physical Review B.

2001;63(5):052103;10.1103/PhysRevB.63.052103.

37. Huot J, Pelletier JF, Lurio LB, Sutton M and Schulz R. Investigation of

dehydrogenation mechanism of MgH2–Nb nanocomposites. Journal of Alloys

and Compounds. 2003;348(1-2):319-24;10.1016/s0925-8388(02)00839-3.

38. Castro FJ and Bobet J-L. Hydrogen sorption properties of an Mg + WO3 mixture

made by reactive mechanical alloying. Journal of Alloys and Compounds.

2004;366(1-2):303-8;10.1016/s0925-8388(03)00747-3.

39. Jung KS, Lee EY and Lee KS. Catalytic effects of metal oxide on hydrogen

absorption of magnesium metal hydride. Journal of Alloys and Compounds.

2006;421(1-2):179-84;10.1016/j.jallcom.2005.09.085.

40. Varin RA, Czujko T, Wasmund EB and Wronski ZS. Hydrogen desorption

properties of MgH2 nanocomposites with nano-oxides and Inco micrometric-

and nanometric-Ni. Journal of Alloys and Compounds. 2007;446-447(0):63-

6;10.1016/j.jallcom.2006.10.134.

41. Polanski M and Bystrzycki J. Comparative studies of the influence of different

nano-sized metal oxides on the hydrogen sorption properties of magnesium

hydride. Journal of Alloys and Compounds. 2009;486(1-2):697-

701;10.1016/j.jallcom.2009.07.042.

42. Bellemare J and Huot J. Hydrogen storage properties of cold rolled magnesium

hydrides with oxides catalysts. Journal of Alloys and Compounds.

2012;512(1):33-8;10.1016/j.jallcom.2011.08.085.

43. Floriano R, Deledda S, Hauback BC, Leiva DR and Botta WJ. Iron and niobium

based additives in magnesium hydride: Microstructure and hydrogen storage

properties. International Journal of Hydrogen Energy. 2017;42(10):6810-

9;10.1016/j.ijhydene.2016.11.117.

67

44. Zeng L, Kimura T, Hino S, Miyaoka H, Ichikawa T and Kojima Y, editors.

Improvement of Hydrogenation and Dehydrogenation Kinetics on MgH2 by the

Catalytic Effect of ZrO2. Applied Mechanics and Materials; 2012: Trans Tech

Publ.

45. Chen B-H, Chuang Y-S and Chen Co-K. Improving the hydrogenation

properties of MgH2 at room temperature by doping with nano-size ZrO2

catalyst. Journal of Alloys and Compounds. 2016;655:21-

7;10.1016/j.jallcom.2015.09.163.

46. Singh RK, Sadhasivam T, Sheeja GI, Singh P and Srivastava ON. Effect of

different sized CeO2 nano particles on decomposition and hydrogen absorption

kinetics of magnesium hydride. International Journal of Hydrogen Energy.

2013;38(14):6221-5;10.1016/j.ijhydene.2012.12.060.

47. Mustafa NS and Ismail M. Hydrogen sorption improvement of MgH2 catalyzed

by CeO2 nanopowder. Journal of Alloys and Compounds. 2017;695:2532-

8;10.1016/j.jallcom.2016.11.158.

48. Dornheim M, Doppiu S, Barkhordarian G, Boesenberg U, Klassen T, Gutfleisch

O, et al. Hydrogen storage in magnesium-based hydrides and hydride

composites. Scripta Materialia. 2007;56(10):841-

6;10.1016/j.scriptamat.2007.01.003.

49. Du AJ, Smith SC, Yao XD, Sun CH, Li L and Lu GQ. The role of V2O5 on the

dehydrogenation and hydrogenation in magnesium hydride: An ab initio study.

Applied Physics Letters. 2008;92(16):163106;10.1063/1.2916828.

50. Grigorova E, Khristov M, Peshev P, Nihtianova D, Veliehkova N and Atanasova

G. Hydrogen sorption properties of a MgH2 - V2O5 composite prepared by ball

milling. Bulgarian Chemical Communications. 2013;45(3):280-7.

51. Korablov D, Nielsen TK, Besenbacher F and Jensen TR. Mechanism and

kinetics of early transition metal hydrides, oxides, and chlorides to enhance

hydrogen release and uptake properties of MgH2. Powder Diffraction.

2015;30(S1):S9-S15;10.1017/s0885715615000056.

52. Nielsen TK and Jensen TR. MgH2-Nb2O5 investigated by in situ synchrotron X-

ray diffraction. International Journal of Hydrogen Energy. 2012;37(18):13409-

16;10.1016/j.ijhydene.2012.06.082.


cycling of niobium and vanadium catalyzed nanostructured magnesium. J Am

Chem Soc. 2005;127(41):14348-54;10.1021/ja051508a.

68


Fast hydrogen sorption from MgH2 –VO2 (B) composite materials. Journal of


55. Valentoni A, Mulas G, Enzo S and Garroni S. Remarkable hydrogen storage

properties of MgH2 doped with VNbO5. Physical chemistry chemical physics :

PCCP. 2018;20(6):4100-8;10.1039/c7cp07157d.

56. Wang P, Wang AM, Zhang HF, Ding BZ and Hu ZQ. Hydrogenation

characteristics of Mg–TiO2 (rutile) composite. Journal of Alloys and

Compounds. 2000;313(1-2):218-23;10.1016/s0925-8388(00)01188-9.

57. Vujasin R, Mraković A, Kurko S, Novaković N, Matović L, Novaković JG, et

al. Catalytic activity of titania polymorphs towards desorption reaction of MgH2.


11;10.1016/j.ijhydene.2016.01.095.

58. Jung KS, Kim DH, Lee EY and Lee KS. Hydrogen sorption of magnesium

hydride doped with nano-sized TiO2. Catalysis Today. 2007;120(3-4):270-

5;10.1016/j.cattod.2006.09.028.


ray Absorption Spectroscopic Study on Valence State and Local Atomic

Structure of Transition Metal Oxides Doped in MgH2. The Journal of Physical

Chemistry C. 2009;113(30):13450-5;10.1021/jp901859f.





61. Jardim PM, da Conceição MOT, Brum MC and dos Santos DS. Hydrogen

sorption kinetics of ball-milled MgH2 –TiO2 based 1D nanomaterials with

different morphologies. International Journal of Hydrogen Energy.

2015;40(47):17110-7;10.1016/j.ijhydene.2015.06.172.

62. Alsabawi K, Webb T, Gray EM and Webb C. Kinetic enhancement of the

sorption properties of MgH2 with the additive titanium isopropoxide.

International Journal of Hydrogen Energy. 2017.


Stroppa DG, et al. Evolution of reduced Ti containing phase(s) in MgH 2 /TiO 2

system and its effect on the hydrogen storage behavior of MgH 2. Journal of


69

64. Pukazhselvan D, Nasani N, Sandhya KS, Singh B, Bdikin I, Koga N, et al. Role

of chemical interaction between MgH 2 and TiO 2 additive on the hydrogen

storage behavior of MgH 2. Applied Surface Science. 2017;420:740-

5;10.1016/j.apsusc.2017.05.182.

65. Pukazhselvan D, Nasani N, Yang T, Ramasamy D, Shaula A and Fagg DP.

Chemically transformed additive phases in Mg2TiO4 and MgTiO3 loaded

hydrogen storage system MgH2. Applied Surface Science. 2019;472:99-

104;10.1016/j.apsusc.2018.04.052.




67. Manchester F and Khatamian D, editors. Mechanisms for activation of

intermetallic hydrogen absorbers. Materials Science Forum; 1988: Trans Tech

Publ.



43;10.3390/catal2030330.



2004;364(1-2):242-6;10.1016/s0925-8388(03)00530-9.

70. Song M, Bobet J-L and Darriet B. Improvement in hydrogen sorption properties

of Mg by reactive mechanical grinding with Cr2O3, Al2O3 and CeO2. Journal

of Alloys and Compounds. 2002;340(1-2):256-62;10.1016/s0925-

8388(02)00019-1.

71. Bobet JL, Desmoulins-Krawiec S, Grigorova E, Cansell F and Chevalier B.

Addition of nanosized Cr2O3 to magnesium for improvement of the hydrogen

sorption properties. Journal of Alloys and Compounds. 2003;351(1-2):217-

21;10.1016/s0925-8388(02)01030-7.

72. Bobet JL, Castro FJ and Chevalier B. Effects of reactive mechanical milling

conditions on the physico-chemical properties of Mg+Cr2O3 mixtures. Journal

of Alloys and Compounds. 2004;376(1-2):205-

10;10.1016/j.jallcom.2003.12.014.

73. Bobet JL, Castro FJ and Chevalier B. Effects of RMG conditions on the

hydrogen sorption properties of Mg+Cr2O3 mixtures. Scripta Materialia.

2005;52(1):33-7;10.1016/j.scriptamat.2004.09.001.

70

74. Vijay R, Sundaresan R, Maiya MP and Murthy SS. Hydrogen storage properties

of Mg–Cr2O3 nanocomposites: The role of catalyst distribution and grain size.


93;10.1016/j.jallcom.2005.11.090.

75. Pranzas PK, Dornheim M, Bellmann D, Aguey-Zinsou KF, Klassen T and

Schreyer A. SANS/USANS investigations of nanocrystalline MgH2 for

reversible storage of hydrogen. Physica B: Condensed Matter. 2006;385-

386:630-2;10.1016/j.physb.2006.06.119.

76. Pranzas PK, Dornheim M, Bösenberg U, Ares Fernandez JR, Goerigk G, Roth

SV, et al. Small-angle scattering investigations of magnesium hydride used as a

hydrogen storage material. Journal of Applied Crystallography.

2007;40(s1):s383-s7;10.1107/s0021889807008023.

77. Polanski M, Bystrzycki J and Plocinski T. The effect of milling conditions on

microstructure and hydrogen absorption/desorption properties of magnesium

hydride (MgH2) without and with Cr2O3 nanoparticles. International Journal of


78. Patah A, Takasaki A and Szmyd JS. Influence of multiple oxide (Cr2O3/Nb2O5)

addition on the sorption kinetics of MgH2. International Journal of Hydrogen


79. Polanski M, Bystrzycki J, Varin RA, Plocinski T and Pisarek M. The effect of

chromium (III) oxide (Cr2O3) nanopowder on the microstructure and cyclic

hydrogen storage behavior of magnesium hydride (MgH2). Journal of Alloys and

Compounds. 2011;509(5):2386-91;10.1016/j.jallcom.2010.11.026.

80. Song MY, Kwon SN, Park HR and Mumm DR. Effects of fine Cr2O3 addition

on Mg's hydrogen-storage performance. Journal of Industrial and Engineering

Chemistry. 2011;17(2):167-9;10.1016/j.jiec.2011.02.021.

81. Tanabe K. Catalytic application of niobium compounds. Catalysis Today.

2003;78(1-4):65-77;10.1016/s0920-5861(02)00343-7.

82. Ziolek M. Niobium-containing catalysts—the state of the art. Catalysis Today.

2003;78(1-4):47-64;10.1016/s0920-5861(02)00340-1.

83. Agueyzinsou K, Aresfernandez J, Klassen T and Bormann R. Effect of Nb2O5

on MgH2 properties during mechanical milling. International Journal of


71

84. Friedrichs O, Sanchez-Lopez JC, Lopez-Cartes C, Klassen T, Bormann R and

Fernandez A. Nb2O5 "pathway effect" on hydrogen sorption in Mg. The journal

of physical chemistry B. 2006;110(15):7845-50;10.1021/jp0574495.




10;10.1016/j.actamat.2005.08.024.



41;10.1016/j.jallcom.2005.02.043.

87. Huhn PA, Dornheim M, Klassen T and Bormann R. Thermal stability of

nanocrystalline magnesium for hydrogen storage. Journal of Alloys and

Compounds. 2005;404-406(Copyright (C) 2015 American Chemical Society

(ACS). All Rights Reserved.):499-502;10.1016/j.jallcom.2004.10.087.

88. Friedrichs O, Klassen T, Sanchezlopez J, Bormann R and Fernandez A.



7;10.1016/j.scriptamat.2005.12.011.

89. Friedrichs O, Martínez-Martínez D, Guilera G, Sánchez López JC and

Fernández A. In Situ Energy-Dispersive XAS and XRD Study of the Superior

Hydrogen Storage System MgH2/Nb2O5. The Journal of Physical Chemistry C.

2007;111(28):10700-6;10.1021/jp0675835.

90. Bhat VV, Rougier A, Aymard L, Darok X, Nazri G and Tarascon JM. Catalytic

activity of oxides and halides on hydrogen storage of MgH2. Journal of Power


91. Hanada N, Ichikawa T, Hino S and Fujii H. Remarkable improvement of

hydrogen sorption kinetics in magnesium catalyzed with Nb2O5. Journal of

Alloys and Compounds. 2006;420(1-2):46-9;10.1016/j.jallcom.2005.08.084.

92. Hanada N, Ichikawa T and Fujii H. Catalytic effect of niobium oxide on

hydrogen storage properties of mechanically ball milled MgH2. Physica B:

Condensed Matter. 2006;383(1):49-50;10.1016/j.physb.2006.03.051.

93. Hanada N, Ichikawa T and Fujii H. Hydrogen absorption kinetics of the

catalyzed MgH2 by niobium oxide. Journal of Alloys and Compounds.

2007;446-447(Copyright (C) 2015 American Chemical Society (ACS). All

Rights Reserved.):67-71;10.1016/j.jallcom.2006.11.182.

72

94. Dolci F, Chio MD, Baricco M and Giamello E. Niobium pentoxide as promoter

in the mixed MgH2/Nb2O5 system for hydrogen storage: a multitechnique

investigation of the H2 uptake. Journal of Materials Science. 2007;42(17):7180-

5;10.1007/s10853-007-1567-0.

95. Dolci F, Baricco M, Edwards PP and Giamello E. An investigation of the H2

uptake in Mg–Nb–O ternary phases. International Journal of Hydrogen Energy.

2008;33(12):3085-90;10.1016/j.ijhydene.2008.01.034.

96. Dolci F, Di Chio M, Baricco M and Giamello E. The interaction of hydrogen

with oxidic promoters of hydrogen storage in magnesium hydride. Materials

Research Bulletin. 2009;44(1):194-7;10.1016/j.materresbull.2008.03.006.

97. Liu Y, Li Q and Chou K-c. Dehydriding reaction kinetic mechanism of MgH2-

Nb2O5 by Chou model. Transactions of Nonferrous Metals Society of China.

2008;18(Copyright (C) 2015 American Chemical Society (ACS). All Rights

Reserved.):s235-s41;10.1016/s1003-6326(10)60209-9.

98. Porcu M, Petford-Long AK and Sykes JM. TEM studies of Nb2O5 catalyst in

ball-milled MgH2 for hydrogen storage. Journal of Alloys and Compounds.

2008;453(1-2):341-6;10.1016/j.jallcom.2006.11.147.

99. Aurora A, Mancini MR, Gattia DM, Montone A, Pilloni L, Todini E, et al.

Microstructural and Kinetic Investigation of Hydrogen Sorption Reaction of

MgH2/Nb2O5Nanopowders. Materials and Manufacturing Processes.

2009;24(10-11):1058-63;10.1080/10426910903022361.

100. Rahman MW, Castellero A, Enzo S, Livraghi S, Giamello E and Baricco M.

Effect of Mg–Nb oxides addition on hydrogen sorption in MgH2. Journal of

Alloys and Compounds. 2011;509(0):S438-S43;10.1016/j.jallcom.2011.02.064.

101. Rahman MW, Livraghi S, Dolci F, Baricco M and Giamello E. Hydrogen

sorption properties of Ternary Mg–Nb–O phases synthesized by solid–state

reaction. International Journal of Hydrogen Energy. 2011;36(13):7932-

6;10.1016/j.ijhydene.2011.01.053.

102. Pagola S, Carbonio RE, Alonso JA and Fernández-Dı az MT. Crystal Structure

Refinement of MgNb2O6Columbite from Neutron Powder Diffraction Data and

Study of the Ternary System MgO–Nb2O5–NbO, with Evidence of Formation

of New Reduced Pseudobrookite Mg5−xNb4+xO15−δ(1.14≤x≤1.60) Phases.

Journal of Solid State Chemistry. 1997;134(1):76-84;10.1006/jssc.1997.7538.

103. Ma T, Isobe S, Morita E, Wang Y, Hashimoto N, Ohnuki S, et al. Correlation

between kinetics and chemical bonding state of catalyst surface in catalyzed

73

magnesium hydride. International Journal of Hydrogen Energy.

2011;36(19):12319-23;10.1016/j.ijhydene.2011.07.011.

104. da Conceição MOT, Brum MC, dos Santos DS and Dias ML. Hydrogen sorption

enhancement by Nb2O5 and Nb catalysts combined with MgH2. Journal of

Alloys and Compounds. 2013;550(Copyright (C) 2015 American Chemical

Society (ACS). All Rights Reserved.):179-84;10.1016/j.jallcom.2012.09.094.

105. Ma T, Isobe S, Wang Y, Hashimoto N and Ohnuki S. Nb-Gateway for Hydrogen

Desorption in Nb2O5 Catalyzed MgH2 Nanocomposite. The Journal of Physical

Chemistry C. 2013;117(20):10302-7;10.1021/jp4021883.

106. Takahashi K, Isobe S and Ohnuki S. The catalytic effect of Nb, NbO and Nb2O5

with different surface planes on dehydrogenation in MgH2: Density functional

theory study. Journal of Alloys and Compounds. 2013;580(Copyright (C) 2015

American Chemical Society (ACS). All Rights Reserved.):S25-

S8;10.1016/j.jallcom.2013.02.044.

107. Pukazhselvan D, Bdikin I, Perez J, Carbó-Argibay E, Antunes I, Stroppa DG, et

al. Formation of Mg–Nb–O rock salt structures in a series of

mechanochemically activated MgH 2 + nNb 2 O 5 (n = 0.083–1.50) mixtures.


88;10.1016/j.ijhydene.2015.12.077.

108. Pukazhselvan D, Perez J, Nasani N, Bdikin I, Kovalevsky AV and Fagg DP.

Formation of Mgx Nby Ox+y through the Mechanochemical Reaction of MgH2

and Nb2O5, and Its Effect on the Hydrogen-Storage Behavior of MgH2.

Chemphyschem. 2016;17(1):178-83;10.1002/cphc.201500620.

109. Pukazhselvan D, Otero-Irurueta G, Pérez J, Singh B, Bdikin I, Singh MK, et al.

Crystal structure, phase stoichiometry and chemical environment of

MgxNbyOx+y nanoparticles and their impact on hydrogen storage in MgH2.


15;10.1016/j.ijhydene.2016.04.029.

110. Walker G. Solid-state hydrogen storage: materials and chemistry: Elsevier;

2008.

74

CHAPTER THREE

Synthesis and Characterisation

3. Introduction

This chapter describes the synthesis of the materials investigated in this work as well as

the methods used to characterise the materials produced and their hydrogen sorption

kinetics. Different ball-milling techniques used to prepare commercial and research

materials are described and the synthesis methods used in this work are detailed. X-Ray

Diffraction (XRD) was used to determine the structure, size and composition of the

composite samples using Rietveld refinement for the analysis. The measurement

methods used to investigate the sorption kinetics and maximum uptake of the modified

magnesium hydride samples are detailed. In addition, the handling and the preparation

techniques for the samples are discussed.

3.1. Synthesis

Composite samples, consisting of MgH2 with small amounts of additive (up to 10

mol%) were synthesised by ball-milling. The different additives included Ti-based and

V-based transition metal oxides in liquid and powder forms as well as a Ti-based

chloride. The benchmark Nb2O5 was used for comparisons, see Table 3-1 for the full list

of additives. The purpose of the range of additives was to explore the role of transition

metal oxides in enhancing the sorption kinetics of MgH2. The effect of different

additives, different amounts of additives and the role of the ball-milling process were

investigated. Ball-milling was performed for different amounts of time and under

different gases to provide a systematic evaluation of these parameters.

3.1.1. Ball-milling

Ball-milling is a mechanical process that is widely used to combine materials as well as

reduce the particle and crystallite size of most metal hydride materials create fresh

surfaces with desired properties. In addition, it serves to homogeneously distribute

different materials throughout the sample. Preparing Mg-based materials by ball

milling, at low temperature and short times, has been widely developed over the past

twenty years [1]. Milling parameters such as mill type (vibratory, planetary, attritor,

high energy, shakers/mixers), milling medium (steel, ceramic, zirconia, etc.) [2], and

process variables (milling time and speed, ball-to-powder weight ratio, temperature,

75

atmosphere) [3], can be combined and varied to achieve desirable milling outputs [4].

According to the targeted microstructure and composition of the desired product, ball-

milling may be classified into three types mechanical grinding, mechanical alloying and

reactive ball milling [5].

3.1.1.1. Mechanical grinding (MG)

In MG the material is nanostructured without any change to its initial compositional

phase. In this method the particle size of the material and other phases can be reduced,

due to the balls impact on the powder, to micrometric size but it is difficult to reduce

the size further by milling for longer, while the crystallites size can reach nanometer

scale [5]. For Mg-based materials this technique can be used to incorporate catalyst

additives during grinding, where the additives are homogeneously distributed at the

nanometer scale.

3.1.1.2. Mechanical alloying (MA)

Alloys can be synthesized from combinations of metallic elements by ball-milling

which is usually referred to as mechanical alloying. In this process, an alloy is formed

by violently deforming mixtures of different powders. The material is nanostructured

with a change in the composition of the starting material and the synthesized samples.

For Mg powder, metals in the desired ratios are milled together and nanostructured

alloys are rapidly obtained at room temperature [4, 6]. The sorption kinetics of samples

synthesized by the MA method is better than kinetics of the samples obtained by

conventional metallurgy [7]. This improvement is due to the presence of defects

introduced during plastic deformation and the high density of the active surface

available for hydrogen absorption.

3.1.1.3. Reactive ball milling (RBM)

In the RBM method the starting materials are milled under reactive atmosphere, i.e.

hydrogen or deuterium gas, instead of an inert atmosphere, such as argon. This

technique is used to synthesize hydrides of metals and alloys in a single step process at

low temperature and under hydrogen pressures from 1 to 150 bar [1]. The fresh metal

surfaces created during the milling process provide reaction sites for hydrogen, as it

breaks the particles’ oxide surface layer which often obstructs or slows the formation of

hydrides. Modern milling vials can be equipped with sensors to record the temperature

76

and pressure during the synthesis process, which allows for in situ monitoring of

hydrogen absorption [8, 9].

All the ball milling techniques above can produce nanosized materials with defects and

increased grain boundaries that provide pathways for hydrogen diffusion and this can

improve the kinetics of absorption and desorption [1]. This improvement is due to the

reduction of the hydrogen diffusion path, increased surface area and formation of

defects into the crystal lattice which favours nucleation of the hydride phase.

3.2. Preparation of milled hydride/additive mixtures

MgH2 can react in the air due to the presence of water (humidity) and oxygen typically

forming Mg(OH)2 preferentially, Because of this, exposure to the atmosphere must be

avoided to prevent sample contamination. All sample handling, including loading the

ball mill vial, transferring to sample cells (for kinetics measurements) or capillaries (for

XRD) as well as subsequent storage were carried out in an Ar glove box (MBRAUN

Unilab), with O2 and H2O levels maintained at less than 0.1 ppm. Milling vials were

initially loaded and sealed inside the Ar glove box, then, depending on the milling

environment, were either transferred directly to the ball-mill or connected to a gas-

handling apparatus to evacuate the Ar and load with 10 bar of H2 before sealing and

transferring to the ball-mill.

All sample mixtures used in this project were produced using a PQ-N04 planetary mill

(Across International) to mill the MgH2 (ABCR) and MgH2/additive mixtures. The

milling vial and balls used were zirconia to reduce the possibility of contaminating the

samples with ball-mill materials. Contamination from ball-mill is a problem that has

affected sample synthesis and distorted metal hydride measurements in a number of

instances [1]. This hard ceramic is less likely to deteriorate with the constant high

energy impact of the balls during use, compared to steel milling equipment, for

example. In addition, any zirconium in the sample could be readily detected using the

lab source XRD with a silver X-ray generating tube, as zirconium fluoresces at this

wavelength (0.68901 Å) [10].

MgH2 was initially ball-milled for 20 h using 10 zirconia balls 10 mm diameter with a

ball to powder ratio of 2:1. This pre-milled MgH2 was prepared in 5 g batches and

usually milled under 10 bar of hydrogen pressure or 1 bar of Argon. The mill was

running at speed of 600 rpm for 30 min followed by a rest of 6 min this cycle was

77

repeated to give a total of 24 h of milling time. The pre-milled MgH2 was further milled

with different amounts of the desired additives (0.5 - 10 mol%), see Table 3-1 for a list

of additives, for time durations from 0.5 -20 h. Mixtures were milled using six zirconia

balls at a speed of 600 rpm with a ball to powder ratio 10:1, under different gas

environments. After milling the vial was then transferred to the glove box to store the

samples for further investigation.

Table 3-1: Chemicals and materials used in this project

Materials/ chemicals Formula Purity

(%)

State Supplier

Magnesium hydride MgH2 98 Powder ABCR

Niobium pentoxide Nb2O5 (20 nm) _ powder NOVacentrix

Titanium isopropoxide Ti(OCH(CH3)2)4 97 Liquid Sigma-Aldrich

Titanium(IV) methoxide Ti(OCH3)4 95 powder Sigma-Aldrich

Titanium(IV) ethoxide Ti(OC2H5)4 97 Liquid Sigma-Aldrich

Vanadium(V) oxytriisopropoxide OV(OCH(CH3)2)3 _ Liquid Sigma-Aldrich

Vanadium(V) oxytripropoxide OV(OC3H7)3 98 Liquid Sigma-Aldrich

Vanadium(V) oxytriethoxide OV(OC2H5)3 95 Liquid Sigma-Aldrich

Bis(isopropylcyclopentadienyl)

titanium dichloride

C16H22Cl2Ti 95 powder Sigma-Aldrich

Activated Carbon FiltraSorb-400 powder

Fullerene C60 99.9 powder SES

Hydrogen H2 99.999 Gas BOC (purified by

MH reservoir)

Argon Ar 99.997 Gas BOC

Helium He 99.999 Gas BOC

78

3.3. Characterisation techniques

3.3.1. X-Ray Diffraction (XRD) analysis

XRD is an analytical technique used to investigate material properties such as crystal

structure, phase composition, and texture of powder, solids and some liquid samples.

Typically, XRD is a non-destructive characterisation technique, which makes it suitable

for in situ studies. The XRD diffracted intensity is proportional to the element electron

number, therefore the higher the atomic number the greater the intensity. Most

crystalline materials have unique diffraction patterns, and databases of these can be used

to identify a variety of chemical compounds. The purity of the sample and the

composition of any impurities present may also be determined based on diffraction

patterns for crystalline materials.

3.3.1.1. Bragg’s Law

X-rays are partially scattered by atoms when they strike the surface of a crystal. The

formation of diffraction patterns occurred when the wavelength of the radiation is of the

same scale as the atomic spacing in the crystal, such that the diffraction gives rise to a

set of well-defined beams arranged with a characteristic geometry. XRD patterns are the

result of the relative intensities for each reflection within the set of planes in the crystal,

known as the miller indices (h, k, l), accompanied by the corresponding scattering angle

for that reflection. Fig 3-1 shows Bragg diffraction where a peak is observed when a

constructive interference occurs from scattering of X-rays by the atomic planes in a

crystal [11]. The positions and intensities of the diffracted rays are a function of the

arrangements of the atoms in space and other atomic properties. By recording position

and intensity of the diffracted beam, the nature and arrangement of the atoms in the

crystal can be deduced.

79

Fig 3-1: Bragg diffraction in a 2D lattice

The condition for constructive interference from planes is given by Bragg’s law:

nλ= 2dhklsinθ (1)

where

n = an integer, for peak order,

� = wavelength of the radiation (X-ray),

� =Bragg angle or the incidence of the X-ray angle and

!�" =interplanar spacing of the (h k l) planes in the crystal lattice.

A schematic of a laboratory X-ray Diffractometer (XRD) is shown in Fig 3-2. An

electron gun in the X-ray tube is used as a source of electrons fired at a target ‘sample

holder’. The core electrons of the targeted sample are ejected from the atoms by the

high energy incident electrons. The atoms’ electrons in higher energy orbitals drop back

into the low-lying orbitals in the target, which results in emission of X-ray photons. The

emitted radiation is a characteristic wavelength for the element [12]. The XRD

apparatus used in the project, a PANalytical Empyrean, was fitted with a silver tube to

produce X-rays with an average energy of 22 keV.

80

Fig 3-2: A schematic view of a laboratory X-ray Diffractometer (XRD) (image

form Malvern Panalytical) [13].

XRD patterns were collected before and after sample synthesis as well as after

desorption/absorption measurements. Samples were loaded into glass capillaries of 0.5

or 0.8 mm internal diameter in the glove box and sealed with super glue gel to avoid

oxidation during XRD measurements.

3.3.2. Scanning Electron Microscopy (SEM)

A scanning electron microscope (SEM) can be used to obtain quantitative and

qualitative information about the composition and morphology of a sample. This

technique provides a reproducible characterisation of image features such as particle

size, external morphology (texture) and shape of a solid structure. In SEM a high energy

electron beam (from 0.3 to 30 keV) is focused into a fine probe, in vacuum, which is

rastered over the sample surface. When the electrons penetrate the surface, multiple

interactions occur which result in the emission of electrons or photons from or through

the surface, which are then detected. The principal images produced in an SEM are of

three types: backscattered electron images, secondary electron images and elemental X-

ray maps. The most common imaging technique is the one produced by secondary

electrons (SE). SE result from inelastic interactions between the sample surface atoms

and the primary electron beam. They are used to observe the morphology and

topography of a surface. Due to their low energy; only those emitted within few

nanometres from the sample surface are detected. Backscattered electrons (BSE) are

emitted from elastic interactions between the sample atom and the incident electron

beam. Large atoms have a large cross sectional area and therefore have high probability

of generating a collision with primary electrons. In consequence, the backscattering

81

intensity is proportional to the atomic number Z of the sample. Thus, the signal due to

backscattering and the image brightness increase with increasing atomic number. This

method can probe a sample to a depth of about 1 μm (check Fig 3-3).

Fig 3-3: Electron beam sample interaction [14].

The Energy dispersive X-ray (EDX) technique is a complementary tool to SEM that

uses the unique X-ray emission characteristics of each element to determine the relative

atomic composition of the sample. The inelastic collisions of the incident electron beam

with electrons in the inner atomic orbits excite those electrons to higher levels. The

excited atom decays to the ground state emitting an Auger electron or a characteristic X-

ray photon. The photons for a given element have a fixed wavelength, which is related

to the difference in energy levels of electrons in different shells, enabling different

elements and their relative abundance to be determined.

A JEOL-JSM 6510 SEM was used to examine surface topography and morphology of

the samples. MgH2 samples were placed on a carbon tape and then coated with

palladium, which was done outside the glove box. However, this arrangement was not

ideal and the absence of a preparatory stage inert environment, accelerated the oxidation

of the samples. In addition, MgH2 was unstable under the electron beam and visible

degassing was observed, resulting in the conclusion that these results were unreliable.

82

3.3.3. Synchrotron Radiation X-Ray Powder Diffraction (SR-XRPD)

Synchrotron Radiation X-Ray Powder Diffraction (SR-XRPD) is a characterisation

technique used to investigate the microstructure of materials using X-rays generated by

a synchrotron. SR-XRD are accessed in large facilities, where electrons are produced in

electron gun and accelerated to near light speed to reach multi-GeV level energy. The

electrons are then injected into an ultra-high vacuum storage ring up to several

kilometres in circumference where they travel at relativistic speeds while in the storage

ring. At each instrument, the electron direction is changed by either an undulator or a

bending magnet, and the resulting acceleration causes electromagnetic radiation to be

emitted in a narrow cone. Typically the collimating mirror is a toroid used to improve

the energy resolution by collimating the beam in the vertical direction into parallel light

[15]. The monochromators are typically Si<111> which is used to create a tunable

monochromatic X-ray beam focused in the horizontal direction [16].

Fig 3-4: Schematic layout of the 10-BM-1 Powder Diffraction beamline

“Australian synchrotron, Diagram courtesy of ANSTO” [17].

Fig 3-4, shows a schematic illustration of the Australian SR-beamline, where the main

components are the collimating mirror, monochromators, and focusing mirror.

83

Table 3-2: The Australian Synchrotron Parameters.

Parameter Values

Source AS Bending Magnet

Energy range 3-20 keV

Monochromator Type Si <111> flat and Si<311> sagittal focus

Resolution dE/E 1.4 × 10-4

Flux (photons/sec) 1 × 1011 at 10 keV

Beam size focused (H×V) 0.5 × 0.5 mm

Beam size unfocused (H× V) 5 × 2 mm

The Australian Synchrotron is a relatively small synchrotron with a storage ring

circumference of 216 m. The powder diffraction instrument is mounted on a bending

magnet and the beamline parameters are presented in Table 3-2.

This instrument was used to investigate ball-milled MgH2, mixed with 10 mol% TTIP

additive, where the higher additive content was needed to increase the diffracted

intensity from any crystalline products formed during mixing. Ball milling reduces the

particle size to ~20 nm, and the small particle size limits the ability to resolve small

peaks from any new compounds formed due to peak broadening, so relatively long

collection times were needed to provide sufficient signal. Composite samples were

prepared by both hand-mixing and ball-milling, and then samples were mounted in

quartz capillaries and diffraction scans were taken first at room temperature and then in

situ while heating to 400°C under vacuum to desorb the hydrogen. Finally, a pressure of

30 bar H2 was applied at 250°C and scans taken as the sample absorbed. The energy

range used for the investigated hydrogen storage materials was 16 keV.

3.3.4. Kinetics characterisation

The Sieverts or ‘manometric’ technique and the gravimetric technique are the two main

methods to measure hydrogen uptake by a solid state host. Pressure measurements,

together with temperature and volume values are used to determine the uptake in the

84

Sieverts technique, whereas the gravimetric directly measures the change in the weight

of the sample during absorption or desorption. In this work, the Sieverts method was

used to load the samples with a single shot of hydrogen. The hydrogen sorption for the

sample can be calculated from the non-ideal gas law,

( ),PV

nZ T P RT

= (2)

where n is the number of moles of gas in the volume V, P is the pressure, R is the gas

constant, T the temperature and Z is the compressibility of the gas. Z is a function of

both temperature and pressure (calculated using the NIST REFPROP database2) and

represents the variation from an ideal gas (for which Z=1).

A custom manometric Sieverts apparatus constructed from commercial Sitec and VCR

components was used to perform the hydrogen desorption and absorption kinetics

measurements. This instrument was improved during this project with the addition of

active temperature control, a MH hydrogen reservoir as well the development of TPD

and TDS capabilities. Full details on the equipment information can be found in [18]

which is provided here as chapter 4.

3.3.5. Temperature Programmed Desorption and Thermal desorption

spectroscopy

The kinetics of a hydrogen storage material - how fast the material can absorb and

desorb hydrogen - can be characterized using Temperature Programmed Desorption

(TPD) [19, 20]. TPD involves heating the sample at a constant linear temperature rate

while monitoring the pressure of the H2 released from the desorbing sample. Thermal

desorption spectroscopy, enables the determination of the gases released during

temperature desorption. The addition of TDS complements the TPD technique; while

the sample is being heated at a known rate any decomposition products are carried to a

mass spectrometer to identify the evolved atomic or molecular species [21], and this

might help to understand the decomposition of the sample or additive.

For this project TPD studies were performed on all the samples with the same

temperature program. A Platinum Resistance Thermometer (PRT) was used to record

the internal temperature of the sample cell during the experiment. TPD studies was

2 (NIST) National Institute of Standards and Technology (REFPROP) Reference Fluid Thermodynamic and

Transport Properties Database, https://www.nist.gov/srd/refprop

85

initiated at room temperature, and then sample was heated by a furnace controlled by a

Eurotherm 2404 temperature controller, which is used to set the temperature ramping

rate, under a starting vacuum of typically 3 ×10-6 mbar. A MicroPirani 925 vacuum

gauge was used to record to the flow of the hydrogen pressure released by the sample

while heating. All data needed for calculating the hydrogen sorption kinetics

(temperature and pressure of various parts of the instrument) were recorded using a

laptop computer running the control and logging software Kheiron. TDS measurements

were conducted simultaneously with TPD studies, to account for all the gases released

with respect to temperature. A quadrupole mass spectrometer (HIDEN Isochema

DSMS) was connected to the instrument via a (ID= 5 mm) PTFE pipe, where all the

data was recorded using MASsoft software.

86

3.4. References








15;10.1016/j.ijhydene.2016.04.029.





4. Suryanarayana C. Mechanical alloying and milling. Progress in Materials

Science. 2001;46(1-2):1-184;10.1016/s0079-6425(99)00010-9.

5. Shao H, Xin G, Zheng J, Li X and Akiba E. Nanotechnology in Mg-based

materials for hydrogen storage. Nano Energy. 2012;1(4):590-

601;10.1016/j.nanoen.2012.05.005.

6. Koch CC, Scattergood RO, Youssef KM, Chan E and Zhu YT. Nanostructured

materials by mechanical alloying: new results on property enhancement. Journal

of Materials Science. 2010;45(17):4725-32;10.1007/s10853-010-4252-7.

7. Prasad Yadav T, Manohar Yadav R and Pratap Singh D. Mechanical Milling: a

Top Down Approach for the Synthesis of Nanomaterials and Nanocomposites.

Nanoscience and Nanotechnology. 2012;2(3):22-48;10.5923/j.nn.20120203.01.

8. Oghenevweta JE, Wexler D and Calka A. Understanding reaction sequences and

mechanisms during synthesis of nanocrystalline Ti2N and TiN via magnetically

controlled ball milling of Ti in nitrogen. Journal of Materials Science.

2017;53(4):3064-77;10.1007/s10853-017-1734-x.

9. Doppiu S, Schultz L and Gutfleisch O. In situ pressure and temperature

monitoring during the conversion of Mg into MgH2 by high-pressure reactive

ball milling. Journal of Alloys and Compounds. 2007;427(1-2):204-

8;10.1016/j.jallcom.2006.02.045.

10. Bearden JA. X-ray Wavelength and X-ray Atomic Energy Levels: National

Bureau of Standards, for sale by the Superintendent of Documents …; 1967.

87

11. Encyclopedia of Materials Characterization: Surfaces, Interfaces, Thin Films.

Brundle CR, Charles A. Evans J, Wihon S, editors. Greenwich: Manning

Publication Company; 1992.

12. Wang Q, Sun Q, Jena P and Kawazoe Y. Theoretical Study of Hydrogen Storage

in Ca-Coated Fullerenes. J Chem Theory Comput. 2009;5(2):374-

9;10.1021/ct800373g.

13. PANalytical. Heavy Mineral Sand Analysis for Industrial Process Control,

http://www.usedscreeners.com/news/heavymineralsandanalysisforindustrialproc

esscontrol?tmpl=%2Fsystem%2Fapp%2Ftemplates%2Fprint%2F&showPrintDi

alog=1 ; 2012

14. Secondary Ion TESCAN Detector, https://www.tescan.com/en-

us/technology/detectors/sitd; 2019

15. Strocov VN, Schmitt T, Flechsig U, Schmidt T, Imhof A, Chen Q, et al. High-

resolution soft X-ray beamline ADRESS at the Swiss Light Source for resonant

inelastic X-ray scattering and angle-resolved photoelectron spectroscopies. J

Synchrotron Radiat. 2010;17(5):631-43;10.1107/S0909049510019862.

16. Isshiki M, Ohishi Y, Goto S, Takeshita K and Ishikawa T. High-energy X-ray

diffraction beamline: BL04B2 at SPring-8. Nuclear Instruments and Methods in

Physics Research Section A: Accelerators, Spectrometers, Detectors and

Associated Equipment. 2001;467-468:663-6;10.1016/s0168-9002(01)00441-7.

17. Technical information (Powder), https://www.ansto.gov.au/user-

access/instruments/australian-synchrotron-beamlines/powder-

diffraction/technical-information ; 2019

18. Alsabawi K, Gray EM and Webb CJ. A Sieverts apparatus with TPD/TDS for

accurate, high-pressure hydrogen uptake and kinetics measurements.

International Journal of Hydrogen Energy. 2019;Submitted.

19. Kowalczyk P, Kaneko K, Terzyk AP, Tanaka H, Kanoh H and Gauden PA. New

approach to determination of surface heterogeneity of adsorbents and catalysts

from the temperature programmed desorption (TPD) technique: one step beyond

the condensation approximation (CA) method. J Colloid Interface Sci.

2005;291(2):334-44;10.1016/j.jcis.2005.05.029.

20. Bhattacharya M, Devi WG and Mazumdar PS. Determination of kinetic

parameters from temperature programmed desorption curves. Applied Surface

Science. 2003;218(1-4):1-6;10.1016/s0169-4332(03)00541-5.

88

21. Broom DP. Hydrogen storage materials: the characterisation of their storage

properties: Springer Science & Business Media; 2011.

89

CHAPTER FOUR

A reliable, high accuracy Sieverts instrument with TPD and TDS

(As a Journal paper submitted to International Journal of Hydrogen Energy)

4. Introduction

Accurate measurement of hydrogen uptake and release for investigation of future

hydrogen storage materials is vital to determine suitability for practical use, as well as

understand the properties and nature of the materials in order to assist with the

development of new materials. This chapter describes the design and implementation of

a Sieverts type apparatus for characterising hydrogen sorption kinetics; with pressure

capability to 340 bar and sample environment temperatures from 77-773 K. This

instrument, which existed at the start of the project, was improved and then further

developed during the course of this project to enhance the temperature stability,

improve the ratio of reference and cell volumes and establish a single environment to

perform sorption kinetics measurements. The sorption kinetics are measured by the

addition of temperature programmed desorption (TPD) and temperature desorption

spectroscopy (TDS).

During the TPD measurement process, the sample is heated at a constant rate using a

temperature controller to achieve the required temperature (maximum of 773 K), while

monitoring the pressure or flow of the released gases. The addition of a mass

spectrometer to provide TDS capability enabled the determination of the atomic masses

of the gases released during the TPD processes. A number of experiments to validate

the operation and accuracy of this instrument were conducted by measuring hydrogen

uptake and release, kinetics and decomposition for different, representative samples.

An important facility in this instrument is the use of a Metal-Hydride (MH) compressor

as a reservoir for local hydrogen storage. This enabled recycling of hydrogen desorbed

from the sample, a higher hydrogen purity and higher pressure capability than most

commercial cylinders. Commercial/ industrial MH compressors often use more than one

stage/material to achieve sufficient compression of a large quantity of hydrogen,

suitable for gas cylinder replacement or a hydrogen fuel station, for example. In such

cases, combinations of metal hydrides with different enthalpies are used to increase the

final compression ratio while maximizing the absorption pressures of each stage [2].

90

However, for the simpler application of providing hydrogen to a single experimental

apparatus, a small single stage MH compressor was used, housed in a 150 cc Swagelok

stainless steel volume with a rated pressure of 340 bar, incorporating approximately 400

g of the AB5 intermetallic MmNi4:2Co0:8 (Japan Metals and Chemicals Co.), where

Mm represents mischmetal, a combination of naturally occurring rare earth metals. This

reservoir material was capable of generating hydrogen pressure of at least 340 bar when

heated manually using a heat gun and could absorb hydrogen back into the reservoir

with cooling from liquid nitrogen.

The Sieverts technique is vulnerable to errors and uncertainties when calculating the

amount of hydrogen taken up by the sample. The measurement accuracy of the pressure,

temperature and volumes involved is critical in the determination of the amount of gas

in a volume, which is the basis of the Sieverts technique. The choice of sensors is

discussed as well as the volume calibrations for room temperature calibration and non-

ambient sample temperatures. A new volume calibration procedure, using isotherm

measurements with no sample is described which enables a more accurate determination

of the sample cell volume. A number of validation measurements for the instrument are

shown to demonstrate the uptake, kinetics and decomposition capabilities with different

materials including LaNi5, MgH2, and a low density carbon material (FiltraSorb-400).

Full details of the instrument design and validation are in the accompanying paper.

91

4.1. References

1. Reilly Jr JJ and Wiswall Jr RH. Method of storing hydrogen. Google Patents;

1970.

2. Hopkins RR and Kim KJ. Hydrogen compression characteristics of a dual stage

thermal compressor system utilizing LaNi5 and Ca0.6Mm0.4Ni5 as the working

metal hydrides. International Journal of Hydrogen Energy. 2010;35(11):5693-

702;10.1016/j.ijhydene.2010.03.065.

3. Cieslik J, Kula P and Filipek S. Research on compressor utilizing hydrogen

storage materials for application in heat treatment facilities. Journal of Alloys

and Compounds. 2009;480(2):612-6.

4. Rajalakshmi N, Dhathathreyan KS and Ramaprabhu S. Investigation of a Novel

Metal Hydride Electrode for Ni-Mh Batteries. In: Grégoire Padró CE, Lau F,

editors. Advances in Hydrogen Energy. Boston, MA: Springer US; 2002. p. 121-

30.

4.2. Paper Sieverts instrument

92





A Sieverts apparatus with TPD/TDS for accurate high-pressure hydrogen uptake and

kinetics measurements’. International Journal of Hydrogen Energy (Submitted).


1) Re-Building, and help in the modification of the instrument

2) Perform experiment

3) Calculations and analysis of the results


(Signed) _________________________________ (Date)__12/02/2019___________

Khadija Alsabawi

(Countersigned) _________________________(Date) 12/02/2019______________


(Countersigned) _________________________(Date)____ 12/02/2019__________


93

A Sieverts apparatus with TPD/TDS for high-pressure hydrogen uptake and

kinetics measurements



Brisbane Australia

Abstract

A Sieverts (manometric gas absorption) apparatus has been designed,

constructed and used to measure hydrogen absorption and desorption,

pressure-composition-temperature (PCT) isotherms at pressures up to 340

bar and temperatures to 773 K. The hydrogen pressure is provided using a

metal-hydride reservoir which removes the requirement for a high

pressure cylinder and also enables the hydrogen to be recycled. The

addition of temperature programmed desorption (TPD) and temperature

desorption spectroscopy (TDS) using a mass spectrometer enhances the

capabilities of this instrument and allows measurements of the absorption

and desorption kinetics of potential hydrogen storage materials.

Key words:

Sieverts apparatus, Hydrogen storage, MH compressors, PCI curve, TDS,

TPD.

* Corresponding author: Tel.: +61 7 37353625. E-mail address: [email protected]

94

Introduction

Hydrogen has the highest energy density of all common fuels by weight and can be used

as an energy carrier or storage system. Combustion of hydrogen in air to produce heat,

or in a fuel cell to produce electricity, produces little to no pollutants, with the primary

product being water [1]. Where the production of hydrogen is achieved using a

renewable source of energy such as wind or solar energy and the source of hydrogen is

water, this forms a closed energy cycle without any detrimental environmental

consequences [2, 3]. As a mechanism for storing energy generated by inherently

intermittent renewable energy sources, hydrogen offers a way of maintaining a balance

between energy supply and demand [4].

Commercial hydrogen powered cars such as the Toyota Mirai, Honda Clarity and

Hyundai ix35 fuel cell vehicles are already available [5-7]. However, hydrogen gas has

a low volumetric energy density, 0.5 kWh/dm3 for 350 bar H2 [8], compared to gasoline

(~7 kWh/dm3) which requires large storage tanks of compressed hydrogen in order to

have a sufficient driving range. High pressure storage increases the energy density, but

requires stronger, more expensive storage tanks [9] to be safe. Liquid hydrogen

cryogenic tanks need costly and energy consuming cooling systems [10]. A viable

alternative is solid state storage.

Solid state materials such as metal hydrides (including elemental metal hydrides such as

MgH2, and alloys, such as LaNi5) and complex hydrides (e.g. amides, borohydrides and

alanates) [11], and physisorption on high surface area materials (e.g. graphite, graphene,

metal-organic frame works and aerogels) are all capable of storing hydrogen with

varying gravimetric and volumetric densities. Physisorption has a relatively low

operating pressure, the price of component materials is affordable and the storage

system design can be relatively simple. However, the drawbacks include low capacity

and a requirement for low temperatures.

Metal hydrides and complex hydrides can store large amounts of hydrogen in a safe and

compact way. Many intermetallic alloys reversibly store hydrogen and operate at

ambient temperature and pressure, but the gravimetric hydrogen capacity is <3 wt%

[12]. Chemical hydride compounds typically often have high formation energies and

therefore require high temperatures to release the hydrogen which make them unsuitable

for many practical applications [13].

95

The search for new hydrogen storage materials requires the accurate measurement of

hydrogen uptake [14-16] as well as other characterisations such as the sorption kinetics,

activation energy and the changing sample composition during the sorption processes.

Reversible hydrogen capacity as well as enthalpy and entropy changes for potential

storage materials are commonly measured using the manometric technique (also called

Sieverts or volumetric) or the gravimetric method [17]. Temperature programed

desorption (TPD) [18-20] is used to determine the sorption kinetics and activation

energies and temperature desorption spectroscopy (TDS) [21] can be employed to

investigate gases released during the desorption process.

Uptake measurements

Pressure-Composition-Temperature isotherms (PCT) are typically used to characterise

the thermodynamics of metal hydrides and adsorption materials [17]. PCT isotherms

describe the dependence between the hydrogen equilibrium pressure and the hydrogen

concentration in the material at fixed temperatures [22]. To establish an isotherm using

a manometric instrument, aliquots of hydrogen gas are added or removed from a system

with known volume. For many binary metals and intermetallic alloys, hydrogen at low

pressure first forms a solid solution where the uptake rises quickly with the pressure, as

shown in Fig 1, plot A, as equilibrium pressure vs uptake. At first hydrogen forms a

dilute solid solution in the metal, typically called alpha (α) phase at increasing pressure,

ideally following Sieverts' Law. At a critical concentration, the concentrated hydride

phase (typically called beta (β)) nucleates. Ideally, the equilibrium between the three

phases present (H2, α, β) has only one degree of freedom according to the Gibbs Phase

Rule, so that the concentration becomes independent of pressure while the alpha and

beta phases coexist, leading to a plateau (ideally horizontal) in the pressure-composition

isotherm. Once the alpha phase is fully converted to beta phase, hydrogen dissolves in

the beta phase at increasing pressure.

The equilibrium pressure logarithm derived from PCT isotherms conducted over a range

of temperatures, is used to construct a van ’t Hoff plot, from which thermodynamic

parameters such as the enthalpy, ∆H, and entropy, ∆S, of the hydride can be obtained

[22]. For physisorption using porous materials, the PCT isotherm can be classified as

IUPAC type I – VI isotherm, depending on the porosity [23]. An example of a type I

absolute isotherm (uptake vs equilibrium pressure) is shown in Fig 1, plot B.

96

Fig 1: (A) PCT absorption isotherm, (B) absolute adsorption isotherm, (C) excess

adsorption isotherm.

An absolute adsorption isotherm is difficult to determine experimentally as it requires

an assumption about the density of the adsorbate [24]. As gas adsorbs onto the material,

the increasing layer of adsorbate occludes more of the volume occupied by the gas.

Since the manometric method requires knowledge of the volume available to the gas in

order to calculate the uptake, this leads to an under-estimation of the amount of uptake,

which increases with increasing pressure – an excess isotherm. To convert the measured

excess isotherm to an absolute isotherm, the volume of the adsorbate must be known,

however, there is no accurate experimental method to determine this [25]. It is often

assumed that the adsorbate is of uniform density at the bulk liquid density [26]. Since

this assumption cannot be verified, the measured excess isotherm is frequently

published (Fig 1, plot C).

In the gravimetric technique the hydrogen uptake for an isotherm is determined by the

weight change of the sample following a step change in the hydrogen pressure. In the

manometric technique, hydrogen sorption is determined from measurements of the

hydrogen pressure in a system of fixed known volume, Both Sieverts and gravimetric

techniques have advantages and disadvantages [27], as they both rely on accurate

knowledge of the volume of the sample and the adsorbed phase to measure the absolute

adsorption or capacity [24, 28, 29]. In the manometric case, the volume of the sample is

needed to determine the void volume in the sample cell in order to calculate the amount

97

of hydrogen in the gas phase. For gravimetric measurements, the sample volume is

required for the buoyancy correction [25].

The Sieverts technique

Fig 2: Minimal Sieverts apparatus for determining the uptake of hydrogen by a

sample.

A simplified Sieverts apparatus is shown in Fig 2, which is used to determine the

hydrogen uptake by a solid sample. The reference volume (Vref) and the sample cell

volume (Vcell), separated by a valve (Vs), are used to progressively increase the pressure

on the sample by introducing known increments of gas into the sample cell volume. For

accurate results, both volumes must be carefully calibrated.

At each pressure step in the construction of a PCT isotherm, hydrogen is introduced into

the reference volume and the amount of gas in this volume is calculated using the non-

ideal gas law,

( ),PV

nZ T P RT

= (1)

where n is the number of moles of gas in the volume V, P is the pressure, R is the gas

constant, T the temperature and Z is the compressibility of the gas. Z is a function of

both temperature and pressure and represents the variation from an ideal gas (for which

Z=1).

98

The valve Vs is then opened to allow the hydrogen to react with the sample in the

sample cell volume Vcell. After a suitable time to allow the sample to reach equilibrium

with the hydrogen pressure and temperature and for any temperature excursions to

equilibrate, the pressure throughout the hydrogenator and the temperatures in each

volume are recorded. The amount of gas sorbed (desorbed, adsorbed or absorbed) by the

sample is determined from any difference in the calculated total gas in the system before

and after opening the valve Vs [17]. The process is then repeated, introducing more gas

into the reference volume and increasing the sample pressure in a stepwise fashion.

Following the procedure in [17], for any step k in the procedure, the absolute mass of

hydrogen sorbed by the sample in this step can be calculated.

, ,

1, ,=

( )

k b k a

ref ref refk k

abs absk a k a k

cell cell cell x

Vm M n

V V

ρ ρ

ρ ρ−

− ∆ ∆ =

+ − − (2)

Where

,,

, ,

k a

refk a

ref k a k a

ref ref

MP

Z RTρ = is the gas density in the reference volume at step k after the

valve has been opened and similarly for the other densities. k

xV is the volume of the

sample plus adsorbate at step k, M is the molecular mass of the gas used and the

superscripts b and a refer to before and after the valve Vs is opened, respectively.

The total amount sorbed after n steps is therefore:

, ,

1, ,1 1

= ( )

k b k an n

ref ref refn k

abs absk a k a k

k kcell cell cell x

Vm M n

V V

ρ ρ

ρ ρ−= =

− ∆ =

+ − − (3)

Errors and uncertainties using the Sieverts method

The Sieverts technique is vulnerable to errors and uncertainties when calculating the

amount of hydrogen taken up by the sample [15, 16, 24, 27, 29]. The hydrogen uptake is

calculated by subtracting two relatively large numbers, the amount of hydrogen in the

system before and after the valve is opened. For the same sample, lower amounts of gas

before and after reduces the relative uncertainty – hence, where practical, smaller

volumes for the reference and sample cell can moderate this problem [29].

99

To achieve accurate results when using a manometric instrument, volume calibration of

each component is required. The strong temperature dependence of any pressure reading

necessitates careful control of the temperature of the valve manifold and the reference

volumes. Depending on the nature of an experiment, the sample temperature may differ

from the rest of the system. Therefore, extra care is needed when modelling the

temperature of each volume of the system [17].

Temperature Programmed Desorption (TPD)

TPD can be used to determine the kinetics of the desorption processes or decomposition

reactions. The TPD process involves heating the sample at a constant rate while

monitoring the pressure of the released gases [30]. For hydrogen desorption studies this

can determine the maximum amount of hydrogen released as well as indicating the

temperature range over which the desorption takes place. Where the desorption process

occurs in discrete reaction steps or involves phase changes, this can also be observed.

TPD allows the calculation of the activation energies of the desorption process by

measuring the desorption spectra for different temperature ramp rates [31].

Thermal Desorption Spectroscopy (TDS)/ (TDS-MS)

TDS adds the capability of determining the nature of the gases released during

temperature desorption. As the sample is heated any decomposition products are carried

to a mass spectrometer or other device which can determine the atomic or molecular

species [27]. This technique can be used to monitor the hydrogen signal to gain more

accurate results for the characterization of small samples [32]. Similarly to TPD, TDS

can be used to determine the activation energies using the same method [33].

Hydrogen pressure amplification

If the experimental conditions require a hydrogen pressure greater than that supplied

from the available compressed gas cylinders (typically 100-200 bar depending on local

regulations), amplification of the pressure is needed. Pressure amplification of hydrogen

can be performed using mechanical compression, metal hydride (MH) compression,

thermal compression and electrochemical compression [34-36]. Mechanical

compression includes syringe pumps [37], reciprocating compressors, rotary

compressors and ionic compressors [38]. A syringe pump reduces the volume

containing the gas in order to increase the pressure. Thermal compression uses the

temperature dependence of a gas to put more gas into a fixed volume at a low

100

temperature, which increases the pressure when the temperature is raised.

Electrochemical compressors use proton exchange membranes (PEM) to dissociate pure

hydrogen at the anode and then recombine as highly pressurised hydrogen at the

cathode [35]. MH compression uses the thermodynamic properties of an intermetallic

metal alloy which can absorb hydrogen at relatively low temperature and release at a

higher pressure when the hydride is heated.

Advantages of MH compressors

Non-mechanical compressors, including metal hydride compressors, have advantages

over mechanical compressors such as no moving parts, low maintenance, little to no

noise and minimal energy use [36], although diaphragm compressors are twice as

efficient [39] and syringe pumps are more easily automated. Metal hydride compressors

release high purity hydrogen [39-41] since typical contaminants such as water and

oxygen are adsorbed by the MH and not desorbed until the material is heated to a much

higher temperature than required for hydrogen release. MH compressors for research

applications are typically compact, which reduces the requirements and simplifies the

compressor system [36].

MH compressor operation

The first periodically operated MH compressor was patented in 1970 by Wiswall and

Reilly [42]. At low temperature and low pressure, hydrogen is absorbed into the metal,

due to the low equilibrium pressure, during which heat generated by the exothermic

reaction must be removed. Subsequent heating of the MH increases the thermodynamic

equilibrium pressure, releasing hydrogen at a higher pressure.

To achieve sufficient compression of a large quantity of hydrogen, suitable for gas

cylinder replacement or a hydrogen fuel station, for example, industrial MH

compressors often use more than one stage/material, where combinations of different

metal hydrides are used to increase the final compression ratio while maximizing the

absorption pressures of each stage [43]. For example, using ZrNi alloy, hydrogen can be

compressed at very low pressure (0.00013 MPa) [44]. Subsequent stages can use higher

pressure AB5 materials, which can be tailored by substitution of the A or B element

depending on the desirable properties [45]. In order for the alloys to reach and maintain

temperatures that correspond to the required equilibrium pressure, it is essential to cool

and heat the alloys during absorption and desorption cycles respectively.

101

For the relatively small quantity of hydrogen required for a laboratory instrument,

amplifying the pressure of hydrogen from several bar to several hundred bar can readily

be performed with a manually operated single stage reservoir of suitable MH material.

The performance of single stage MH compressors can be optimised by finding a MH

material that has a maximum reversible sorption capacity, in order to use less material

and to reduce heat loss due to thermal swing. Hysteresis, where, for the same

temperature, absorption occurs at a higher equilibrium pressure than desorption, reduces

the efficiency of the compression cycle. Low hysteresis and low plateau slope optimise

the compressor capacity and efficiency [34].

The thermodynamics of the metal-hydrogen interaction determines both the heat

generated/needed during absorption/desorption and the pressure amplification for fixed

absorption and desorption temperatures. The low temperature condition required for

absorption and the maximum temperature required for maximum desorption are usually

determined by limitations of the laboratory equipment and ease of use. For example, the

maximum desorption temperature used for the laboratory MH compressor described

here is restricted to 473 K due to the use of Teflon in the seals and this is easily supplied

manually using a hot air gun. Liquid nitrogen is readily available to cool the compressor

to attain a very low absorption temperature. Any MH material which absorbs down to

less than 1 bar at the absorption temperature and desorbs to the desired maximum

pressure at 473 K would be suitable.

In this paper we present an experimental apparatus for the study of the hydrogen

sorption kinetics of hydrogen storage material based on the manometric technique and

TPD/TDS. The instrument uses a metal hydride compressor for hydrogen pressure

amplification and allows the determination of the pressure-composition-temperature

isotherms (PCT), activation energies and kinetics of solid–hydrogen systems in the

temperature range from 77-773 K and hydrogen pressures up to 340 bar.

102

Instrument design

Fig 3: Instrument schematic showing the main components for the preparation,

reference and sample sections.

A schematic of this instrument is shown in Fig 3. Conceptually, the instrument is

divided into three main sections. The preparation section is physically separated for

thermal isolation from the reference section and has connections for vacuum and

hydrogen gas supply as well as any external attachments such as a volume calibration

apparatus and the mass spectrometer. The reference section includes an empty volume

with an internal platinum thermometer, and the pressure transducer. This section is

isolated from the preparation section by valve Vp and from the sample section by valve

Vs. The sample section includes the sample cell, thermocouple and any sample

environments (e.g. furnace, liquid nitrogen bath).

Temperature control

For accurate measurements, the temperature in all parts of the apparatus must be known

[16, 29]. Where the gas temperature is measured at one point in a particular volume, it is

essential to ensure that the rest of the volume is at the same temperature, such that all

103

measured temperatures are in isothermal volumes. In addition, the temperature should

not vary significantly with time, requiring a temperature controlled environment.

Whenever gas is released from one volume to another, temperature changes occur due

to the adiabatic gas expansion and the temperature must be allowed to equilibrate before

measurements can be made. The use of stainless steel, with relatively low thermal

conductivity, for valves, pipe and fittings can make temperature equilibration time

consuming.

In this instrument, temperature is measured separately in the main gas volume of the

reference section and inside the sample cell, where the sample is contained. To maintain

a constant temperature and to remove or re-distribute heat generated in the gas handling

processes, the laboratory is controlled at 293 ± 0.1 K and constant temperature water

circulating through the reference volume backplane is used for active thermal

management. The physical layout of the instrument helps to reduce the effect of heating

or cooling the MH compressor on the reference and cell volumes.

Leak management

Leak integrity in manometric instruments is vital to the accuracy of resulting

measurements. For example, any leak out of the sample cell not only affects the current

measurement, but continues while the reference volume is being prepared again and will

affect the next measurement. A drop in pressure in the sample cell may cause desorption

– which for a sample exhibiting hysteresis will make the measurements non-

equilibrium. In addition, the successive accumulation of both data and errors in the

Sieverts technique amplifies any errors due to leaks in a single measurement [29].

The use of components with metal seals tightened to the correct specifications reduces

the possibility of leaks. Leak testing prior to the commencement of a measurement is

typically performed by filling the system with hydrogen gas to ~50 bar, and using a

hydrogen leak detector (Hydrogen Leak Locator H1000, AS90100, with a resolution of

0.00001 cc/s) to determine the existence of any leak to the atmosphere. In addition, each

section is individually checked by applying a hydrogen gas pressure close to the

maximum pressure to be used and monitoring the pressure over time duration similar to

the experiment to be performed. A pressure drop of less than the accuracy of the

pressure transducer (ie no experimentally observable leak) over that period ensures no

additional measurement error is introduced. This ensures that any leak is too small to

affect the experiment.

104

Preparation section

The preparation section includes the MH compressor, connections for gas input and

vacuum and the requirements for TPD and TDS.

Vacuum is provided by a Pfeiffer turbo molecular pump (model TMH 071 P). The flow

of the gas to the vacuum system is regulated by a needle valve (VN), which is used to

reduce vacuum capacity when a small back-pressure is desirable. A vacuum gauge, 925

MicroPirani Transducer, is mounted in the preparation section to monitor the release of

the hydrogen gas out of the sample.

A MH compressor employing an AB5 alloy in a 150cc Swagelok stainless steel volume

provides hydrogen during absorption measurements and recycles hydrogen during

desorption (except for TPD/TDS experiments). The volume has a rated pressure of 340

bar and contains approximately 400 g of MmNi4.35Co0.8Al0.05 (Japan Metals and

Chemicals Co.). Hydrogen is absorbed into the MH compressor alloy by cooling the

reservoir in a liquid nitrogen bath, reducing the pressure below 2 bar and is released by

heating the volume with a heat gun. Hydrogen pressures of 340 bar are readily

achievable.

Reference section

A custom made chrome-plated CuBe volume forms the primary volume in the reference

section. Gas temperature in this volume is measured using a platinum resistance

thermometer (PRT) silver soldered into a plug in the base. An NPT to Sitec adapter

connects the volume to the backplane. A Quartzdyne pressure gauge (model DSB301-

05-C85) with accuracy 0.015% FSR (340 bar) provides accurate pressure measurements

in this volume.

Sample section

The sample cell consists of a male-male VCR4 union. A VCR4 cap with a silver-

soldered thermocouple seals the bottom of the cell using a silver plated stainless steel

gasket to withstand high temperatures (to 773 K). The cell connects to the sample stick

using a silver plated stainless steel filter gasket (500 nm filter) to contain the sample

during pressure changes and avoid contaminating the system.

The components for all sections of the instrument, excluding the MH reservoir, are rated

to a pressure of at least 1000 bar. The pressure gauge is chosen to be the lowest pressure

105

range that satisfies the requirements of the experiment in order to maintain the highest

accuracy [25] and this determines the pressure limit of the current instrument.

Sample environment

For different sample temperatures above ambient, the sample cell can be immersed in a

custom furnace with a maximum temperature of 773 K. The furnace temperature is

maintained by a programmable PID controller (Eurotherm 2404), and has a uniform

temperature zone of 140 mm and a stability of ± 0.5 K. For measurements at a

cryogenic temperature, a liquid nitrogen bath is available.

Experimental procedure

For loading and unloading the sample, the sample cell is detached above the valve Vstick

(see Fig 3) to isolate the sample from the atmosphere during transport to and from the

glovebox. Prior to an experiment the sample (and whole instrument) is evacuated to

~10-6 bar, and any sample preparation, including outgassing, is performed. The sample

cell is then immersed in the required sample environment (furnace or liquid nitrogen).

Temperature and pressure measurements are continuously recorded using a PC-based

custom designed data acquisition program with a graphical interface [37].

Volume calibration

The volumes of the reference volume, Vref, and cell volume, Vcell, must be known

accurately to calculate reliable hydrogen sorption isotherms [17, 46, 47]. To calibrate

the reference volume, the gas expansion method was used [17]. An external, known

calibration volume with a high precision Paroscientific 31K-101 pressure transducer

(accuracy of 0.01 % FSR 70 bar) was connected via the preparation section. The

calibration volume is then loaded with ultra-high purity helium gas and the pressure and

temperature in the calibrated volume is recorded. The mass of gas in this volume is:

i

ca l ca lM n Vρ= (4)

where, Vcal is the calibration volume, ρi is the initial gas density in the calibration

volume and M is the molecular mass of the gas used. The valve is then opened to the

previously evacuated target volume and sufficient time (typically 15 minutes) is allowed

for thermal equilibrium. With the assumption that there is no adsorption on internal

walls of the system components and no leaks, the total amount of free gas is constant

and the unknown volume can be calculated from:

106

i f f

cal cal cal cal T TM n V V Vρ ρ ρ= = + (5)

where the superscript ‘f’’ refers to after the valve is opened and subscript ‘T’ specifies

to the target (unknown) volume. Rearranging Eqn 5 for the preparation volume gives

[ ]i f

cal calT

f f

cal T T

V

V

ρ ρ

ρ ρ= − (6)

from which the unknown volume, VT, can be determined. The procedure is repeated

multiple times to ascertain the uncertainty in the measurement. The effect of any

internal adsorption can be checked by performing the calibration at different pressures.

Subsequent unknown volumes can also be determined by allowing the calibration gas to

expand into each volume in turn, although the uncertainty increases with each additional

volume. Applying the calibration procedure to this instrument using helium gas

determined the values for the preparation and reference volumes shown in Table 1. The

cell volume was also determined in this fashion, but greater accuracy is achieved by

using the empty-cell isotherm method.

Empty-cell isotherm

Although the process of volume calibration is very similar to the process of determining

an isotherm, the differences in the expansion gas (helium) and the pressure transducer

used may result in slightly different measurements. If the same pressure transducer as

used in the actual experiments is used with hydrogen to determine the cell volume, these

differences can be avoided.

Table 1: Ambient temperature volume calibration

Volume Measured (cm3)

preparation 9.45 ± 0.03

Reference 12.33 ± 0.06

Sample cell 2.61 ± 0.08

107

By performing the equivalent of a PCT isotherm at room temperature, but with no

sample present, it is possible to determine both the volume of the empty cell and the

volume occupied by the valve, Vs. Eqn 3 describes the uptake after n pressure steps in

an isotherm calculation. In this equation, the value of volume Vref is determined from

the previous helium volume calibration and is necessarily the volume with the valve Vs

closed. The volume Vcell is also defined as the volume of the sample section with the

valve Vs closed.

For an empty cell, and in the absence of leaks, the mass uptake of hydrogen should be

zero, and Eqn 3 becomes:

, ,

1, ,1 1

0

k b k an n

ref ref refn k

abs absk a k a

k kcell cell cell

Vm M n

V

ρ ρ

ρ ρ−= =

− = ∆ = =

+ − (7)

Rearranging Eqn 7 gives:

( ) ( ),

1

, 1, ,

1

k b k a k a k a cellref ref cell cell

re

n n

k fk

V

Vρ ρ ρ ρ

= =

−= −−− (8)

Plotting ( ),

1

,k b k a

ref ref

n

k

ρ ρ=

− as a function of ( )1

1, ,k a k a

cell c

k

ell

n

ρ ρ=

− − yields a straight line with

intercept zero and slope cell

ref

V

V− , from which Vcell can be calculated.

In order to use values for Vref and Vcell which assume the valve Vs is closed, it is

necessary for the measurements above to have the valve closed at the end of each

isotherm step before recording the pressure. When closing the valve the gas currently

occupying the valve volume is displaced to either side changing the pressure on both

sides and, requiring a short wait for pressure and temperature stabilisation. For this

reason the valve should be closed as slowly as possible allowing the gas pressure to

equalise as much as possible before passage through the valve is stopped.

If the pressure is also measured before the valve is closed, it is possible in principle to

determine the volume of the valve between the reference and cell volumes, as the total

volume occupied by the gas must be the reference and cell volumes plus the valve

volume.

108

( )

( )

( )

( )

, , 1, ,

, ,

1 1

1 1

k b k o k a k on n

k k

ref ref cell cell

cell v

k o k o ref refv v

n n

k k

V V

V V

ρ ρ ρ ρ

ρ ρ

=

=

−

=

=

= −

− −

(9)

Where the superscript o designates the valve is open, such that ,k o

cellρ represents the

density of hydrogen in the cell at the end of the kth isotherm step while the valve is still

open, and ,k b

refρ is the density of gas in the reference volume before the valve has been

opened. For constant cell, valve and reference volumes, plotting ( )

( )

,

1

,

,

1

n

k

k b k o

ref ref

k on

k

v

ρ ρ

ρ

=

=

−

as a

function of ( )

( )

1, ,

1

1

,

k a k on

ce

k

n

k

ll cell

k o

v

ρ ρ

ρ

−

=

=

−

yields a straight line with slope cell

ref

V

V and intercept v

ref

V

V− .

An empty-cell isotherm was performed and the results of plotting Eqn 9 are shown in

Fig 4 The straightness of the line (R2 = 0.999997) confirms the validity of the technique

and the slope and intercept have been used to determine the cell and valve volume

shown in Table 2. The value for the valve volume Vv, has a high uncertainty due to the

uncertainty in the intercept and the small volume, and may not provide useful

information, but the slope, which yields the cell volume, is more accurate.

Table 2: Ambient temperature empty cell isotherm


Vcell 2.606 ± 0.004

Vv 0.014 ± 0.045

109

Fig 4: a straight line generated from the isotherm for hydrogen adsorption of an

empty cell @296 K and to 180 bar.

Divided Volume Calibration

The calibration techniques described above provide ambient temperature values for the

two critical volumes, Vref and Vcell. Temperature uniformity in Vref is maintained at

ambient by the room temperature control and the temperature controlled backplane. The

sample cell is often used at a variety of non-ambient temperatures, which can result in a

temperature gradient in the sample cell. As this is difficult to accurately measure, the

divided volume calibration technique [17] is used.

110

Fig 5: Divided-volume schematic of temperature distribution in the sample cell

volume.

Fig 5, shows the sample cell volume conceptually divided into two volumes, Vamb at

temperature Tamb (normally the room temperature) and Vsam at Tsam (the sample working

temperature as provided by a furnace or other heating or cooling facility) such that:

cell amb samV V V= + (10)

The divided volume calibration procedure starts with a known amount of hydrogen gas

at ambient temperature in the reference and cell volumes. The mass of hydrogen in the

system is given by:

( )i

H ref cellMn V Vρ= +

(11)

The temperature of the sample cell is then increased in steps from ambient to the highest

temperature to be used (in this case 773 K) and after the temperature has equilibrated

the pressure is measured. As the total amount of gas does not change, Eqn 11 becomes:

( )i

H ref cell ref ref sam sam amb ambMn V V V V Vρ ρ ρ ρ= + = + +(12)

111

Rearranging Eqn 12 yields the equation for the ratio of the sample volume to the

ambient volume:

i

sam cell amb

amb sam amb sam

V V

V V

ρ ρ

ρ ρ

= −

(13)

Using Eqns 10 and 13 together with the calibrated ambient temperature value for the

cell volume (Vcell) found from the empty cell calibration, Vsam and Vamb can be

determined. This procedure is repeated several times to ensure a reliable result and to

determine any uncertainty. This technique relies on the accuracy of the assumption that

the temperature gradient in the sample cell can be modelled as two discrete volumes,

one at ambient temperature and one at the sample temperature and that the boundary

between them occurs at the same place, independent of the temperature. For best results,

that part of the volume which is physically subject to the thermal gradient is best kept as

small as possible – such as a pipe with a small internal diameter. If the equipment is

such that the model is invalid, this will show in a poor fit of a straight line to the

measurements at different temperatures. The divided volume calibration was performed

for this instrument for temperatures from 300 K to 773 K, and the results for the

volumes are shown Table 3. The coefficient of determination (R2) was 0.99977.

The divided volume technique works well, provided the identical arrangement and

circumstances apply for both the volume calibration and the isotherm experiments, so it

is important that the same conditions (such as furnace type and positions, liquid

nitrogen bath shape and nitrogen level) are maintained, especially for liquid nitrogen

baths [16].

Table 3: Divided volume calibration in furnace environment


Vamb 1.027 ± 0.008

Vsam 1.580± 0.007

Vcell 2.608± 0.014

112

Instrument validation:

Standard Sample measurements

One method to validate the instrument and the experimental procedure, is to run a

standard material [25] such as LaNi5, to check that the isotherm, is in agreement with

the literature [48].

Fig 6: Absorption (triangles) and desorption (circles) isotherms for LaNi5 at

ambient temperature (296 K).

The LaNi5 isotherm (Fig 6) illustrates the classic phase transformation in AB5

intermetallic compounds, starting with the solid solution α phase followed by a plateau

region during which β hydride phase is formed, and finally a sharp increase in pressure

once the sample is all β-phase [49].

LaNi5 has a fast and reversible sorption of hydrogen at room temperature with a

moderate equilibrium pressure of a few bars and exhibits hysteresis. It has been

extensively studied [50, 51], and modelled [52-54] and provides a suitable material for

initial validation isotherms. However, the high density prevents any check of the

sensitivity of the instrument to inaccurate or uncertain sample density and the fast

kinetics of reversible H2 absorption and desorption of comparatively large amount of H2

limits its ability to validate the instrument [16, 55]. High uptake also fails to adequately

check the sensitivity of the instrument to low-capacity samples. It can, however,

indicate the absence of leaks, the accuracy of the volume calibrations and the suitability

113

of the data processing methods. Low-capacity and low-density materials such as the

light metals [56], graphene [57] and other carbons are needed to further validate the

instrument.

Porous materials such as polymers [58] and the porous carbons are potential low density

hydrogen storage materials [24] and have previously been involved in erroneous

measurements of hydrogen uptake [59]. As such, these provide suitable materials to

perform more challenging validation tests of an instrument.

A guide to the amount of sample required for accurate measurements can be obtained

from the Figure of Merit, η [55].

kS

Pη

δ= (14)

where Sk is the slope of the isochore, which is determined by the volumes and the

quantity of sample, divided by the resolution of the pressure transducer ( Pδ ) . Using

the volumes above and the pressure transducer specification, the LaNi5 sample above

gives a Figure of Merit of 145.11 which is substantially above the value of 100,

suggested by the authors as a rule of thumb for accurate isotherm measurements [55].

The hydrogen adsorption excess isotherm for a commercial activated carbon FiltraSorb-

400 with a density of 1.96 g/cm3 [37], was measured (Fig 7) at ambient temperature

(296 K) to 300 bar. Using the Figure of Merit in Eqn 14, a sample quantity of 0.145 g is

needed for an accurate measurement.

114

Fig 7: H2 adsorption isotherm for a commercial activated carbon(FiltraSorb-400)

obtained at ambient temperature (296 K) and pressures up to 300 bar.

TPD

The apparatus was also tested by running a TPD experiment for MgH2 milled for 20 h

under (10 bar) of H2 pressure (Fig 8). Mg is a light metal which has been investigated

for hydrogen storage applications [60-64]. The TPD studies were carried out following

the procedure in [65].

115

Fig 8: Desorption curve for MgH2 milled 20 h under H2

The MgH2 sample started to desorb at 493K, where the desorption peak occurred at 598

K, which is a typical desorption curve for MgH2 milled for 20 h [18].

TDS

TDS offers the additional advantages of monitoring the release of the gases with respect

to temperature. To demonstrate the use of this system for TDS measurements, a

Dynamic Sampling Mass Spectrometer (DSMS) supplied by HIDEN Analytical was

connected to the instrument, below the vacuum valve seen in Fig 3, to perform a TDS

experiment where a sample of MgH2 + (1 mol%) titanium isopropoxide (TTIP) milled

for 2 h under Ar, was heated to 673 K. This TDS experiment helped to evaluate the

quantities of hydrogen and other gases desorbed from this sample with respect to

temperature.

116

Fig 9: TDS spectra for MgH2 + (1 mol%) TTIP. A, hydrogen (2 amu) , B, oxygen

(16 amu), C, water (18 amu).

Fig 9, shows the TDS spectra of this sample where H2 gas (A) started to desorb at

around 423 K, with a desorption peak at 553 K, confirming earlier desorption than the

MgH2 sample without additives shown in Fig 8 [18]. Other gases such as O2 (16 amu)

(B) had a minimal presence and desorbed in the temperature range (473-573 K). A gas

(C) with mass 18 amu (assumed to be water vapour) was also released in a similar

temperature range of H2. The presence of H2 is expected due to the presence of H2 in

both MgH2 and TTIP (C12H28O4Ti), as well as O2, with other combinations of C, H and

O also possible.

Conclusion

The construction of a Sieverts (manometric gas sorption) apparatus that can measure

PCT isotherms to 340 bar at temperatures from 77 K to 773 K has been reported.

Control of the apparatus opening and closing valves to either evacuate or to introduce

gas to each part of the system is achieved manually and a computer program has been

used to monitor and record data, as well as control the sample environment temperature,

as experiments are performed. Absorption and desorption kinetics measurements of

potential hydrogen storage materials were performed using this apparatus. A

vulnerability of Sieverts technique for errors was mitigated by performing multiple

volume calibrations. A full volume calibration accounted for the different physical

volumes of the apparatus, an empty cell isotherm was performed at ambient temperature

to more accurately confirm the sample cell calibration, and a divided-temperature

117

volume calibration accounted for the temperature changes between the components of

the apparatus in the sample environment.

TPD experiments were performed using this Sievert apparatus, using a temperature

controller to ramp the sample temperature to 773 K. TDS experiments were also

performed by linking the Mass Spectrometer while a TPD experiment is running to

evaluate the desorbed gases from the samples. These validation experiments

demonstrate the applicability of this instrument to measurements of hydrogen uptake

and release, capacity, kinetics and decomposition.

Acknowledgements

KA acknowledges receipt of an Australian Government Research Training Program

Scholarship.

118

References

1. Johnston B, Mayo MC and Khare A. Hydrogen: the energy source for the 21st

century. Technovation. 2005;25(6):569-85;10.1016/j.technovation.2003.11.005.

2. Acar C and Dincer I. Comparative assessment of hydrogen production methods

from renewable and non-renewable sources. International Journal of Hydrogen




4. Moriarty P and Honnery D. Intermittent renewable energy: The only future

source of hydrogen? International Journal of Hydrogen Energy.

2007;32(12):1616-24;10.1016/j.ijhydene.2006.12.008.

5. Toyota Motor Sales. Hydrogen Fuel Cell Car | Toyota Mirai,

https://ssl.toyota.com/mirai/fcv.html; 2017

6. Hyundai ix35 Fuel Cell, http://www.hyundai.com.au/why-hyundai/design-and-

innovation/fuel-cell; 2017

7. American Honda Motor. 2017 Clarity fuel cell,

https://automobiles.honda.com/clarity-fuel-cell; 2017

8. Hirscher M and Hirose K. Handbook of Hydrogen Storage: New Materials for

Future Energy Storage: Wiley; 2010.

9. Jorgensen SW. Hydrogen storage tanks for vehicles: Recent progress and current

status. Current Opinion in Solid State and Materials Science. 2011;15(2):39-

43;10.1016/j.cossms.2010.09.004.

10. Durbin DJ and Malardier-Jugroot C. Review of hydrogen storage techniques

for on board vehicle applications. International Journal of Hydrogen Energy.

2013;38(34):14595-617;10.1016/j.ijhydene.2013.07.058.

11. Ley MB, Jepsen LH, Lee Y-S, Cho YW, Bellosta von Colbe JM, Dornheim M,

et al. Complex hydrides for hydrogen storage – new perspectives. Materials

Today. 2014;17(3):122-8;10.1016/j.mattod.2014.02.013.

12. Zuttel A. Hydrogen storage methods. Naturwissenschaften. 2004;91(4):157-

72;10.1007/s00114-004-0516-x.

13. Dornheim M. Thermodynamics of metal hydrides: tailoring reaction enthalpies

of hydrogen storage materials. Thermodynamics-Interaction Studies-Solids,

Liquids and Gases: InTech; 2011.

119

14. Cho WC, Han SS, Kang KS, Bae KK, Kim CH, Jeong SU, et al. Improvement

of hydrogen sorption measurement using a Sieverts apparatus with consideration

of thermal transpiration. International Journal of Hydrogen Energy.

2013;38(14):6241-8;10.1016/j.ijhydene.2012.12.057.

15. Broom D. The accuracy of hydrogen sorption measurements on potential storage

materials. International Journal of Hydrogen Energy. 2007;32(18):4871-

88;10.1016/j.ijhydene.2007.07.056.

16. Broom D and Webb C. Pitfalls in the characterisation of the hydrogen sorption

properties of materials. International Journal of Hydrogen Energy.

2017;42(49):29320-43; 10.1016/j.ijhydene.2017.10.028.

17. Gray EM. 7 - Reliably measuring hydrogen uptake in storage materials A2 -

Walker, Gavin. Solid-State Hydrogen Storage: Woodhead Publishing; 2008. p.

174-204; 10.1533/9781845694944.2.174.

18. Alsabawi K, Webb T, Gray EM and Webb C. Kinetic enhancement of the


International Journal of Hydrogen Energy. 2017;42 (8) ;5227-

5234;10.1016/j.ijhydene.2016.12.072.

19. Saldan I, Frommen C, Llamas-Jansa I, Kalantzopoulos GN, Hino S, Arstad B, et

al. Hydrogen storage properties of γ–Mg(BH4)2 modified by MoO3 and TiO2.


93;10.1016/j.ijhydene.2015.07.082.




11;10.1016/j.ijhydene.2016.01.095.

21. Jain IP, Lal C and Jain A. Hydrogen storage in Mg: A most promising material.


44;10.1016/j.ijhydene.2009.08.088.

22. Bliznakov S, Lefterova E and Dimitrov N. Electrochemical PCT isotherm study

of hydrogen absorption/desorption in AB5 type intermetallic compounds.


94;10.1016/j.ijhydene.2008.07.054.

23. Thommes M, Kaneko K, Neimark AV, Olivier JP, Rodriguez-Reinoso F,

Rouquerol J, et al. Physisorption of gases, with special reference to the

120

evaluation of surface area and pore size distribution (IUPAC Technical Report).

Pure and Applied Chemistry. 2015;87(9-10):1051-69;10.1515/pac-2014-1117.

24. Broom DP, Webb CJ, Hurst KE, Parilla PA, Gennett T, Brown CM, et al.

Outlook and challenges for hydrogen storage in nanoporous materials. Applied

Physics A. 2016;122(3):1-21;10.1007/s00339-016-9651-4.

25. Broom DP and Webb CJ. Pitfalls in the characterisation of the hydrogen

sorption properties of materials. International Journal of Hydrogen Energy.

2017;42(49):29320-43;10.1016/j.ijhydene.2017.10.028.

26. Broom DP and Thomas KM. Gas adsorption by nanoporous materials: Future

applications and experimental challenges. Mrs Bulletin. 2013;38(5):412-

21;10.1557/mrs.2013.105.

27. Broom DP. Hydrogen storage materials: the characterisation of their storage

properties: Springer Science & Business Media; 2011.

28. Webb CJ and Gray EM. Analysis of uncertainties in gas uptake measurements

using the gravimetric method. International Journal of Hydrogen Energy.

2014;39(13):7158-64;10.1016/j.ijhydene.2014.02.154.

29. Webb CJ and Gray EM. Analysis of the uncertainties in gas uptake

measurements using the Sieverts method. International Journal of Hydrogen


30. Elliott JAW and Ward CA. Temperature programmed desorption: A statistical

rate theory approach. The Journal of Chemical Physics. 1997;106(13):5677-

84;10.1063/1.473588.

31. Chen P, Xiong Z, Luo J, Lin J and Tan KL. Interaction between Lithium Amide

and Lithium Hydride. The Journal of Physical Chemistry B.

2003;107(39):10967-70;10.1021/jp034149j.

32. von Zeppelin F, Haluška M and Hirscher M. Thermal desorption spectroscopy

as a quantitative tool to determine the hydrogen content in solids.

Thermochimica Acta. 2003;404(1-2):251-8;10.1016/s0040-6031(03)00183-7.

33. Zlotea C, Sahlberg M, Özbilen S, Moretto P and Andersson Y. Hydrogen

desorption studies of the Mg24Y5–H system: Formation of Mg tubes, kinetics

and cycling effects. Acta Materialia. 2008;56(11):2421-

8;10.1016/j.actamat.2008.01.029.

34. Lototskyy MV, Yartys VA, Pollet BG and Bowman RC. Metal hydride

hydrogen compressors: A review. International Journal of Hydrogen Energy.

2014;39(11):5818-51;10.1016/j.ijhydene.2014.01.158.

121

35. Grigoriev SA, Shtatniy IG, Millet P, Porembsky VI and Fateev VN. Description

and characterization of an electrochemical hydrogen compressor/concentrator

based on solid polymer electrolyte technology. International Journal of


36. Dehouche Z, Grimard N, Laurencelle F, Goyette J and Bose TK. Hydride alloys

properties investigations for hydrogen sorption compressor. Journal of Alloys

and Compounds. 2005;399(1-2):224-36;10.1016/j.jallcom.2005.01.029.

37. Pyle DS, Gray EM and Webb CJ. A sieverts apparatus for measuring high-

pressure hydrogen isotherms on porous materials. International Journal of


38. Green D and Perry R. Perry's Chemical Engineers' Handbook, Eighth Edition:

McGraw-Hill Education; 2007.

39. Laurencelle F, Dehouche Z, Morin F and Goyette J. Experimental study on a

metal hydride based hydrogen compressor. Journal of Alloys and Compounds.

2009;475(1-2):810-6;10.1016/j.jallcom.2008.08.007.

40. Muthukumar P, Maiya AP and Murthy SS. Experiments on a metal hydride

based hydrogen compressor. International Journal of Hydrogen Energy.

2005;30(8):879-92;10.1016/j.ijhydene.2004.09.003.

41. Sandrock G. Applications of hydrides. NATO ASI Series E Applied Sciences-

Advanced Study Institute. 1995;295:253-80.

42. Reilly Jr JJ and Wiswall Jr RH. Method of storing hydrogen. Google Patents;

1970.

43. Hopkins RR and Kim KJ. Hydrogen compression characteristics of a dual stage

thermal compressor system utilizing LaNi5 and Ca0.6Mm0.4Ni5 as the working

metal hydrides. International Journal of Hydrogen Energy. 2010;35(11):5693-

702;10.1016/j.ijhydene.2010.03.065.

44. Cieslik J, Kula P and Filipek SM. Research on compressor utilizing hydrogen

storage materials for application in heat treatment facilities. Journal of Alloys

and Compounds. 2009;480(2):612-6;10.1016/j.jallcom.2009.02.007.

45. Rajalakshmi N, Dhathathreyan KS and Ramaprabhu S. Investigation of a Novel

Metal Hydride Electrode for Ni-Mh Batteries. In: Grégoire Padró CE, Lau F,

editors. Advances in Hydrogen Energy. Boston, MA: Springer US; 2002. p. 121-

30.

122

46. Webb CJ and Gray EM. The effect of inaccurate volume calibrations on

hydrogen uptake measured by the Sieverts method. International Journal of


47. Webb CJ and Gray EMA. Misconceptions in the application of the Sieverts

technique. International Journal of Hydrogen Energy. 2013;38(33):14281-

3;10.1016/j.ijhydene.2013.08.064.

48. Schlapbach L and Zuttel A. Hydrogen-storage materials for mobile applications.

Nature. 2001;414(6861):353-8;10.1038/35104634.

49. Van Vucht J, Kuijpers F and Bruning H. Reversible room-temperature

absorption of large quantities of hydrogen by intermetallic compounds. Philips

Res Rep 25: 133-40 (Apr 1970). 1970.

50. Boer F, Boom R, Mattens W, Miedema A and Niessen A. Cohesion in Metals:

Transition Metal Alloys North Holland. Amsterdam; 1988.

51. Mohammadshahi SS, Webb TA, Gray EM and Webb CJ. Experimental and

theoretical study of compositional inhomogeneities in LaNi5Dx owing to

temperature gradients and pressure hysteresis, investigated using spatially

resolved in-situ neutron diffraction. International Journal of Hydrogen Energy.

2017;42(10):6793-800;10.1016/j.ijhydene.2016.12.061.

52. Mohammadshahi SS, Gould T, Gray EM and Webb CJ. An improved model for

metal-hydrogen storage tanks – Part 1: Model development. International

Journal of Hydrogen Energy. 2016;41(5):3537-

50;10.1016/j.ijhydene.2015.12.050.

53. Mohammadshahi SS, Gould T, Gray EM and Webb CJ. An improved model for

metal-hydrogen storage tanks – Part 2: Model results. International Journal of


54. Dantzer P. Properties of intermetallic compounds suitable for hydrogen storage

applications. Materials Science and Engineering: A. 2002;329-331:313-

20;10.1016/s0921-5093(01)01590-8.

55. Blach TP and Gray EM. Sieverts apparatus and methodology for accurate

determination of hydrogen uptake by light-atom hosts. Journal of Alloys and

Compounds. 2007;446-447:692-7;10.1016/j.jallcom.2006.12.061.



2007;32(9):1121-40;10.1016/j.ijhydene.2006.11.022.

123

57. Cunning BV, Pyle DS, Merritt CR, Brown CL, Webb CJ and Gray EMA.

Hydrogen adsorption characteristics of magnesium combustion derived

graphene at 77 and 293 K. International Journal of Hydrogen Energy.

2014;39(12):6783-8;10.1016/j.ijhydene.2014.02.054.

58. Ramimoghadam D, Gray EM and Webb CJ. Review of polymers of intrinsic

microporosity for hydrogen storage applications. International Journal of


59. Broom D and Hirscher M. Irreproducibility in hydrogen storage material

research. Energy & Environmental Science. 2016;9(11):3368-80;

10.1039/C6EE01435F



106;10.1016/j.jpcs.2014.06.014.



Physics A. 2016;122(2):1-20;10.1007/s00339-016-9602-0.

62. Crivello JC, Denys RV, Dornheim M, Felderhoff M, Grant DM, Huot J, et al.

Mg-based compounds for hydrogen and energy storage. Applied Physics A.

2016;122(2):1-17;10.1007/s00339-016-9601-1.

63. Webb TA, Webb CJ, Dahle AK and Gray EM. In-situ neutron powder

diffraction study of Mg–Zn alloys during hydrogen cycling. International


9;10.1016/j.ijhydene.2015.03.170.



system compared with the MgH2 + 0.01Nb2O5 benchmark. International


37;10.1016/j.ijhydene.2011.11.114.

65. Alsabawi K, Webb TA, Gray EM and Webb CJ. The effect of C 60 additive on



124




Alsabawi K, Webb TA, Gray EM and Webb CJ.

‘Kinetic enhancement of the sorption properties of MgH2 with the additive titanium

isopropoxide’. International Journal of Hydrogen Energy. 2017;42(8):5227-

34;10.1016/j.ijhydene.2016.12.072.


1) Perform the full experiment



(Signed) ____________________(Date)_ 12/02/2019________________________

Khadija Alsabawi

(Countersigned) _____________ (Date) 12/02/2019_________________________


(Countersigned) ______________ (Date) 12/02/2019_________________________


125

CHAPTER FIVE

The effect of TTIP as an additive on the hydrogen sorption properties of MgH2

(Published as a Journal paper in the International Journal of Hydrogen Energy)

5. Introduction

The practical application of MgH2 as a hydrogen storage material for renewable energy

demand/supply smoothing, or for automotive or remote operation, is limited by its

relatively high thermodynamic stability and very slow kinetics. However, the sorption

kinetics of MgH2 can be significantly improved by ball-milling or other severe

deformation techniques and by mixing with additives such as metal oxides and other

materials. In particular, the use of transition metal oxides has been shown to enhance

the kinetics of the Mg-H reaction, by facilitating the dissociation of the hydrogen

molecule into atomic hydrogen, thus increasing the absorption desorption rates. Nb2O5

is the classic oxide additive, which demonstrably enhances both absorption and

desorption as well as reducing the re-agglomeration of Mg/MgH2 with repeated cycling.

The use of a small amount of Nb2O5 – 0.5 mol% has been shown to be sufficient -

milled for short time has been extensively studied and the kinetic enhancement

confirmed [1, 2]. It has been reported that the beneficial effect of Nb2O5 into MgH2 was

due to the formation of new ternary oxide phase MgNbxOy [3, 4], although other

additives, including TiN, TiC [5] , vanadium oxides [6, 7] , and chlorides [8, 9] and

fluorides [10, 11], and many more [12], have also been effective, which are not

expected to form ternary oxides.

In this chapter, a titanium-based oxide liquid precursor, titanium isopropoxide (TTIP)

as an additive to MgH2 was investigated. TTIP is similar to Nb2O5, in that both are

transition metal-based and contain a large number of oxygen atoms per transition metal

atom (O/Nb ratio of 2.5, cf O/Ti ratio of 4). The differences – a different transition

metal with a different interaction with magnesium as well as the liquid property –

provide an opportunity to study the properties which contribute to the kinetic

enhancement, and may provide a commercial advantage if the additive is employed in a

practical application. TTIP had not been used previously for the purpose of enhancing

the kinetics of MgH2. A study of the kinetic effects of using the organo-metallic oxide

126

liquid, TTIP, as a catalytic additive on the sorption kinetics of MgH2 was made.

Different amounts of TTIP up to 2 mol% milled with MgH2 for different ball milling

times (0.5- 10 h), as well as hand mixing, were investigated and compared to the

benchmark Nb2O5. A Sievert apparatus (described in chapter 4) was used to perform the

kinetics measurements. The desorption kinetics were measured by performing a TPD

experiment and this was followed by a single pressure aliquot of hydrogen to measure

the absorption kinetics. It was found the TTIP was just as effective as Nb2O5. The

question of how this material assists the kinetic processes in both absorption and

desorption of MgH2 remains. Full details of the experiment and the results and

conclusions are given in the attached paper.

5.1. References



7;10.1016/s1359-6462(03)00259-8.

2. Friedrichs O, Klassen T, Sanchez-Lopez JC, Bormann R and Fernandez A.



7;10.1016/j.scriptamat.2005.12.011.




10;10.1016/j.actamat.2005.08.024.

4. Patah A, Takasaki A and Szmyd JS. Influence of multiple oxide (Cr2O3/Nb2O5)

addition on the sorption kinetics of MgH2. International Journal of Hydrogen








Compounds. 2001;315(1-2):237-42;10.1016/s0925-8388(00)01284-6.

127




8. Ivanov E, Konstanchuk I, Bokhonov B and Boldyrev V. Hydrogen interaction

with mechanically alloyed magnesium–salt composite materials. Journal of

alloys and compounds. 2003;359(1-2):320-5.




2015;30(S1):S9-S15.

10. Malka I, Bystrzycki J, Płociński T and Czujko T. Microstructure and hydrogen

storage capacity of magnesium hydride with zirconium and niobium fluoride

additives after cyclic loading. Journal of Alloys and Compounds.

2011;509:S616-S20.






106;10.1016/j.jpcs.2014.06.014.

5.2. TTIP paper:

Alsabawi K, Webb TA, Gray EM, Webb CJ. Kinetic enhancement of the sorption

properties of MgH2 with the additive titanium isopropoxide. International Journal of

Hydrogen Energy, 2017; 42(8): 5227-34)

Kinetic enhancement of the sorption properties ofMgH2 with the additive titanium isopropoxide

K. Alsabawi, T.A. Webb, E.MacA. Gray, C.J. Webb*

Queensland Micro- and Nanotechnology Centre, Griffith University, Nathan 4111, Brisbane, Australia

a r t i c l e i n f o

Article history:

Received 7 September 2016

Received in revised form

12 December 2016

Accepted 15 December 2016

Available online 12 January 2017

Keywords:

Magnesium hydride

Hydrogen storage

Catalyst

Sorption kinetics

a b s t r a c t

The effect of titanium isopropoxide (TTIP) on the kinetics of absorption and desorption of

magnesium hydride was compared to the benchmark additive niobium oxide (Nb2O5). A

systematic survey was performed for ball-milled MgH2 mixed with up to 2 mol% TTIP or

Nb2O5 additive for different ball-milling times. It was found that TTIP was just as effective

as Nb2O5 for the enhancement of the kinetics of absorption and desorption and that 0.5 mol

% was sufficient. The TTIP composite sample had superior hydrogen capacity compared to

the Nb2O5 composite sample. As TTIP is a liquid at room temperature, 30 min ball-milling

with the magnesium hydride was sufficient time for mixing. Hand mixing the TTIP with

MgH2 demonstrated excellent desorption kinetics but the absorption kinetics were poorer

compared to ball-milling.

Crown Copyright © 2016 Published by Elsevier Ltd on behalf of Hydrogen Energy Publica-

tions LLC. All rights reserved.

Introduction

Magnesium is a potential hydrogen storage material, primar-

ily because of a moderate H-storage capacity of 7.6 wt% and

110 g H/l. The abundance of magnesium in the form of salts in

the earth's crust and oceans enables future high volume

implementation of magnesium-based technologies [1]. Un-

fortunately, the practical use of Mg is limited by poor

hydrogen sorption kinetics and the high dissociation tem-

perature of MgH2 [2]. Attempts have been made to overcome

these problems through ball-milling, alloying and the use of

small amounts of additives [3e5].

Both the absorption of hydrogen by magnesium and

desorption of magnesium hydride occur through a number of

sequential steps, where the slowest process determines the

overall kinetic rate. The desorption of magnesium hydride

starts with nucleation of the metal phase, diffusion of

hydrogen atoms through the material, and subsequent

recombination at the surface to form molecular hydrogen

[6,7]. Finally, the hydrogen molecule must desorb from the

surface of the material back into the gas phase.

Similarly, the steps for absorption of hydrogen by magne-

sium are physisorption of hydrogen gas on to the surface of

the material, dissociation of the hydrogen molecule, diffusion

of atomic hydrogen into the bulk material and ultimately

nucleation of the hydride phase. As the absorption proceeds,

hydrogen atoms must diffuse through more of the hydride

layer to reach the magnesium. Diffusion of hydrogen through

magnesium hydride is known to be slow (�2.5 � 10�9 cm2/s at

350 �C) [8], approximately three orders of magnitude slower

than hydrogen diffusion through magnesium [4,9] and diffu-

sion is often considered to be the rate-limiting step for both

desorption and absorption [2,10,11].

Mechanical milling and severe plastic deformation tech-

niques have been shown to improve the kinetics of the

* Corresponding author. Fax: þ61 7 3735 7656.E-mail address: [email protected] (C.J. Webb).

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier .com/locate/he

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 2 ( 2 0 1 7 ) 5 2 2 7e5 2 3 4

http://dx.doi.org/10.1016/j.ijhydene.2016.12.072

0360-3199/Crown Copyright © 2016 Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. All rights reserved.

mailto:[email protected]

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijhydene.2016.12.072&domain=pdf

www.sciencedirect.com/science/journal/03603199

www.elsevier.com/locate/he




absorption and desorption reactions through reduction of

particle size, mixing of the surface oxide layer, and introduc-

tion of defects [4,12]. The reduction in external particle size as

well as crystallite size reduces diffusion path lengths through

the bulk material. Reducing the surface oxide layer and the

introduction of defects and grain boundaries allow faster

diffusion paths for hydrogen [2,5].

The hydrogen sorption kinetics of the MgeH system have

been shown to significantly improve further with the combi-

nation of small amounts of additives, usually <10 mol%. Of

these, compounds of transition metals, particularly oxides,

but also halides have been shown to be effective [2,7]. In

particular Nb2O5 [13e15] is generally recognised as a bench-

mark material for the best enhancement of absorption and

desorption kinetics. Barkhordarian et al. were the first to

report the superior catalytic effect of Nb2O5 in improving the

kinetics of MgH2, compared to other metal-oxide additives

[15,16]. It has been reported that the Nb2O5 additive was

partially reduced during milling with MgH2 and then further

reduced after cycling [17e19]. The partially reduced Nb2O5

forms MgO and other Nb compounds, and Nb2O5 reacts with

MgO during cycling, forming new ternary oxides [20e22]. It

was suggested that the reduced Nb-containing compounds

decreased the activation energy for hydrogen desorption of

the catalysed MgH2 and that the activation energy decreased

with increasing milling time [18,23,24]. The mechanism by

which the kinetic enhancement takes place is not well un-

derstood and knowledge of this mechanism would guide the

development of new additives.

Barkhordarian et al. [25] compared the effect of adding

catalyst (Nb2O5) to MgH2 on the rate-limiting step of the ab-

sorption/desorption reaction by fitting the respective kinetic

equations.They found that the rate-limiting step for desorption

changedwith increasingNb2O5contentormilling time, fromre-

association to interface velocity of the transformed phase.

In subsequent desorption studies, the effects of nano- and

micro-sized Nb2O5 were compared and it was found that for

short milling times the desorption kinetic behaviour of the

nanoNb2O5 catalysedMgH2was about 10 times faster than the

micro [19]. The smaller nano-sized Nb2O5 particles increased

the specific surface area of the interface between the additive

and MgH2, which increased the number of nucleation sites for

the reacted Mg phase [26]. Ball milling also helped to create

defects and disperse Nb2O5 through the Mg matrix, providing

easy pathways for hydrogen atoms to diffuse through the Mg

particle core [20]. Dolci et al. [27] compared three different

MgeNbeO phases and found that the Mg3Nb6O11 phase

showed a reversible interaction with molecular hydrogen.

They then hypothesized that this was due to the occurrence of

octahedral niobium clusters which were also observed for

NbO. Following this work, Rahman et al. [28] confirmed the

presence of an octahedral structure and suggested that it acts

as an insertion site for hydrogen uptake and release. In a later

study they showed that Mg3Nb6O11 had an active reversible

interaction with molecular hydrogen occurring at around

395 �C [29]. A more recent study [30] has suggested that the

presence of MgO after milling Nb2O5eMgH2 had no effect in

improving the sorption kinetics of the sample. The improve-

ment was attributed to Nb, supporting the Nb-gateway model

[31], where Nb clusters facilitate hydrogen transportation out

of the MgH2 particles.

Titanium based additives have also been successfully used

to improveMgH2 hydrogen sorption kinetics, exhibiting a high

affinity towards hydrogen at moderate temperatures. The

desorption of composite samples synthesised from mechan-

icallymilled Ti andMgH2 showed a rapid hydrogen desorption

at temperatures above 250 �C and absorption at temperatures

as low as 29 �C, when compared to other additives such asMn,

Fe, and Ni [32]. A comparison between the addition of Ti and

NieTi catalysts toMgH2was conducted by Lu et al. [33], finding

that Ti significantly decreased the desorption temperatures

but that the Ni/Ti-doped MgH2 composites had the lowest

desorption temperature (ca 200 �C).Wang et al. [34] used ball milling to produce a MgeTiO2

composite which showed superior kinetics at 350 �C under

20 bar of H2 and a high hydrogen capacity attributed to

effective dispersion of the catalyst in the Mg matrix as well as

the existence of n-type TiO2 rutile. Different forms of TiO2

(rutile, anatase and commercial P25) were investigated by Jung

et al. [35] to determine the hydriding kinetics of Mg as well as

the hydrogen storage capacity. They found that, of the TiO2

samples studied, the nano-composite MgH2-(rutile)0.05showed the greatest kinetic improvement [36] absorbing

4.4 wt% hydrogen at 300 �C which was 10 times faster than

that of ball milled MgH2 without additives.

Dobrovolsky et al. [37] mechanically milled (50 wt%) TiB2

with (50 wt%) MgH2 and found that the addition of TiB2

decreased the dissociation temperature of the MgH2 by about

50 �C. This decrease in the temperature was ascribed to the

catalytic effect of TiB2 surface particles on hydrogen desorp-

tion occurring on the surface of the MgH2 particles.

Pitt et al. [38] investigated the effect of nanoscopic Ti-based

additives (Ti, TiC, TiN, TiB2 and TiO2) on MgH2 sorption ki-

netics and hydrogen release temperature. They concluded

that all the Ti-based additives studied produced a strong effect

on hydrogenation kinetics, with fast (ca 20 s) absorption to 90%

of capacity in all samples. The effects of various titanium

compounds (TiF3, TiCl3, TiO2, TiN and TiH2) on the hydrogen

sorption kinetics of MgH2 were examined by Ma et al. [39].

They found that among these compounds the addition of TiF3had the most pronounced improvement on both absorption

and desorption rates. The kinetics of the TiF3 doped sample

was almost double that of TiCl3, TiO2 and TiN samples and

over 10 times that of adding TiH2.

Hence, small amounts (~1 mol%) of both Nb2O5 and Ti

compounds are known to contribute significantly to the

enhancement of the kinetics of the MgeH reaction. Transition

metals facilitate the dissociation of the hydrogen molecule

into atomic hydrogen promoting higher rates for absorption

and desorption, however dissociation and recombination are

not usually the rate-limiting steps [13].

The oxide layer on magnesium is expected to retard the

kinetics of absorption due to the slow diffusion of hydrogen

through surfaceMgO layers [40]. One of the advantages of ball-

milling is to provide fresh non-oxided surfaces for hydrogen

reaction [41]. Despite this, oxides such as Nb2O5 and TiO2 are

among the best additives for improvement ofMgH2 absorption

and desorption kinetics. MgOwas used as an additive by Ares-

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 2 ( 2 0 1 7 ) 5 2 2 7e5 2 3 45228



Fernandez et al. [42] and this showed improvement of the

desorption kinetics but not absorption.

It has been suggested that metal oxides with multiple

valence states have a greater catalytic effect due to improve-

ment of the electronic exchange reaction with hydrogen

molecules [15], however when Nb2O5 was added to MgH2 that

had been ball-milled to grain sizes less than 1.5 nm, no

enhancement was found [43]. These seemingly contradictory

findings indicate that the underlying mechanisms by which

the kinetics are improved are not yet understood.

Titanium isopropoxide, also called titanium tetraisoprop-

oxide (TTIP) is an organic alkoxide of titanium containing Ti

and oxygen as well as carbon and hydrogen (Ti[OCH(CH3)2]4)

which is a liquid at room temperature. For each Ti atom, there

are four oxygen atoms. This material has been used as an

additive to improve the hydrogen sorption kinetics of the

LiBH4eMgH2 reactive hydride composite [44e46], but has not

been used previously with MgH2. TTIP is unlikely to form a

ternary oxide (as suggested for Nb2O5) as Ti and Mg are

immiscible elements in thermodynamic equilibrium due to

their positive enthalpy of mixing, and do not form any ther-

modynamically stable alloy or compound. The enthalpy of

mixing for Mg0.5Ti0.05 mixture is DHmix z 20 kJ/g-atom [47,48].

This work compares the effects of titanium isopropoxide

and Nb2O5 on the sorption kinetic properties of MgH2. A sys-

tematic study of the amount of additive and the milling time

was employed to determine the existence of any link between

these parameters and the desorption properties. The choice of

an additive with some similar properties to Nb2O5 (transition

metal and oxygen) but with marked differences (liquid) may

enable comparisons which facilitate a better understanding of

the underlying mechanisms.

Experimental

Magnesium hydride (MgH2) powder (98% purity, AB109434)

was purchased from ABCR GmbH & Co. KG. Titanium iso-

propoxide (99.999% purity) from ALDRICH and 20 nm Nb2O5

from NovaCentrix were used for the additives. The MgH2 was

pre-milled in a PQ-N04 planetary mill (Across International)

for 20 h using a 50 ml zirconia vial and 10 zirconia balls

(10 mm) with a ball to powder ratio (bpr) of 21:1 and a vial

filling factor [49] of 17.4%. Samples were prepared in 5 g

batches under 1 bar of Ar. Different amounts of TTIP from

0.5 mol% to 2 mol% were subsequently milled with the pre-

milled MgH2 for 2 h. In addition, the 1 mol% TTIP was milled

with the pre-milledMgH2 for different times from 0.5 h to 10 h,

(1 g sample, 6 ZrO2 balls, 600 rpm, 21:1 bpr, vial filling factor

7.7%).

Similar compositions andmilling times were employed for

the Nb2O5 composite samples for comparison purposes. All

materials were handled in a dry Ar atmosphere MBraun glo-

vebox with O2 and H2O levels below 1 ppm.

Hydrogen cycling was performed on a modified Sieverts

apparatus constructed from commercial VCR and Sitec com-

ponents. Samples were loaded under Ar into a stainless steel

sample cell, sealed and transferred to the gas apparatus. A K-

type thermocouple internal to the sample cell was used to

monitor the temperature of the sample. A MicroPirani 925

vacuum gauge was used to record the pressure of the released

hydrogen. Temperature Programmed Desorption (TPD)

studies were carried out on the as-prepared sample under

dynamic vacuum from room temperature to 400 �C with a

temperature ramp-rate of 2 �C/min. Baseline starting vacuum

was typically 3 � 10�6 mbar. The sample remained at 400 �Cfor 2 h to ensure complete desorption. Absorption kinetics

data were recorded at 230 �C, after a single hydrogen pressure

aliquot to reach an initial pressure of 50 bar on the sample. A

Quartzdyne DL series pressure sensor with an overall accu-

racy of 0.015% FSR (340 bar) was used to measure the subse-

quent pressure reduction during absorption.

A PANalytical Empyrean X-ray diffractometer was used to

characterize the samples before and after cycling. Samples

were loaded into glass capillaries to collect powder X-ray

diffraction (XRD) data. A silver source with a wavelength of

0.5594�Awas used, with the diffractometer in DebyeeScherrer

geometry. Topas Academic v5 fitting software was used to

analyse the diffraction data. The crystallite size and strain

analysis was determined using Topas Academic's built in

models, which are based on the Scherrer equation and strain

models from Balzar [50].

Results and discussion

The prepared samples were first desorbed in the TPD appa-

ratus under vacuum and subsequently absorbed hydrogen

under 50 bar at 230 �C. Samples with 1 mol% additive were

prepared with a range of ball-milling times, and samples with

a range of additive amounts were ball-milled for 2 h in order to

investigate the effect of ball-milling duration and additive

amount. Previous studies have shown that maxima in the

enhancement of kinetics of MgH2 desorption can be observed

for both of these parameters [51].

Effect of amount of additive

The desorption kinetics of MgH2with varying amounts of TTIP

additive ball-milled for 2 h are shown in Fig. 1 together with

data for the Nb2O5 composite samples. It can be seen that all

the additives produce a substantial improvement (~80 �C) indesorption kinetics compared to the MgH2 ball milled for 20 h

without additives, although the desorption is spread over a

greater range of temperatures. For both TTIP and Nb2O5

composite samples, 0.5 mol% is sufficient to produce the

improvement in desorption kinetics. In addition, within the

experimental uncertainties, the results for TTIP and Nb2O5 are

essentially the same. Some gas release can be seen for the

TTIP samples as low as 130 �C. This was only observed for the

first cycle of TTIP composite samples and is assumed to be

decomposition of the TTIP.

Effect of ball-milling time

In order to investigate the effect of milling time on the

desorption kinetics, 1 mol% of TTIP was milled with MgH2 for

times from0.5 h to 10 h. Nb2O5 samplesweremilled for 2 h and

10 h to provide a comparison.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 2 ( 2 0 1 7 ) 5 2 2 7e5 2 3 4 5229



Fig. 2 shows the desorption kinetics for these samples.

Again the addition of 1 mol% of additive to the pre-milled

MgH2 produces a substantial improvement in the desorption

kinetics as shown by the earlier onset of the desorption

compared to theMgH2 without additives. There is only a small

improvement with additional milling for both additives and

thus 2 h is sufficientmilling time for bothNb2O5 and TTIP. This

agrees with previous findings for Nb2O5 [52]. The desorption

kinetics for both TTIP and Nb2O5 at both 2 h and 10 h milling

are effectively the same.

Also shown in Fig. 2 is a samplewhere theTTIP additivewas

hand mixed with ball-milled MgH2 (TTIP HM). The desorption

result for this sample is at least asgoodasanyof theball-milled

samples. This sample was made by adding the TTIP liquid to

the pre milled MgH2 in a mortar and pestle and mixing for

about 10 min, when the composite sample appeared uniform.

Fig. 1 e Desorption kinetics of pre-milled MgH2 ball-milled for 2 h with varying amounts of TTIP and Nb2O5.

Fig. 2 e Desorption kinetics of pre-milled MgH2 with 1 mol% additive milled for 0.5e10 h. Results for Nb2O5 are shown for

comparison.




Absorption

Fig. 3 shows the effect on the absorption kinetics of the

different amounts of additives milled as well as different ball-

milling times, along with the pre milled MgH2 without addi-

tives. The pre-milled MgH2 without additives has the highest

hydrogen capacity but the slowest kinetic rate. All the sam-

ples with additives have significantly improved kinetics but

with hydrogen capacity lower than the pure MgH2 sample,

showing that the additives used (TTIP or Nb2O5) make a sig-

nificant enhancement to the absorption kinetics. The reduc-

tion in the gravimetric hydrogen capacity for the composite

samples is partly due to the addition of material which does

not contribute to the hydrogen absorption. All of the com-

posite samples with ball-milled additives produce similar re-

sults, other than themaximum capacity and, in particular, the

Fig. 3 e Absorption kinetics of all the composite samples (pre-milled MgH2 with TTIP or Nb2O5). Ball-milling times varied

from 0 to 2 h and the amount of additive from 0.5 mol% to 1.5 mol%.

Fig. 4 e Desorption kinetics of cycled pre-milled MgH2 with 1 mol% TTIP milled for 0.5 h compared to 1 mol% Nb2O5 milled

for 2 h and pre-milled MgH2 without additives over five desorption/absorption cycles.




uptake slopes for the first minute of the TTIP and Nb2O5 ad-

ditive samples are not distinguishable, indicating that the

TTIP additive performs equally well to the Nb2O5 additive for

absorption. In addition, the absorption kinetics are not

improved with greater quantities of the additive e for both

cases, 0.5 mol% of additive is sufficient to produce the kinetic

enhancement. Interestingly, the hand-mixed TTIP sample

showed a poorer result than the ball-milled additives, despite

improvement in the desorption kinetics.

The effect of cycling

Two TTIP composite samples with one of the best combina-

tions of milling and quantity of TTIP (1 mol% TTIP milled for

0.5 h and 1 mol% TTIP hand mixed) were cycled a total of five

times and the results of the desorption kinetics were

compared to the 1 mol% Nb2O5 composite sample.

The cycling desorption data for MgH2þ 1 mol% TTIP com-

posite sample milled for 0.5 h compared to MgH2þ 1 mol%

Nb2O5 are shown in Fig. 4. Pre-milled MgH2 without additives

is also shown for comparison. While the desorption kinetics

formagnesiumhydridewithout additivesworsenwith cycling

(primarily on the first cycle), both Nb2O5 and TTIP show at

worst only a small deterioration. This has been previously

observed for Nb2O5 [4,53].

Without additives, MgH2 desorption kinetics deteriorate

significantly due to re-crystallisation, increasing the grain size

as shown in Table 1 fromX-ray diffraction data ofMgH2milled

under hydrogen for varying times and then cycled.

The cycling desorption data for handmixed TTIP compared

to (1 mol%) Nb2O5 are shown in Fig. 5. There is a small dete-

rioration for both additives and little difference between the

two additives after the first two cycles. The absorption for the

hand-mixed sample, which was poorer than the ball-milled

samples, also showed little change after cycling.

Conclusion

The effect of ball milling of small amounts of TTIP (0.5e2 mol

%), with 20 h pre-milled MgH2 was systematically surveyed by

milling the samples for different times from 0 to 10 h under

1 bar of Ar pressure. To investigate the kinetics of absorption

and desorption of the samples, thermal desorption under

vacuum followed by a single absorption pressure step at 50 bar

of H2 at 230 �C was employed. The results were compared to

the benchmark Nb2O5 additive. It was found that TTIP had a

positive effect on the absorption as well as the desorption

kinetics.

The use of Nb2O5 additives improved the sorption kinetics

of the MgH2, whichmatches the body of literature, confirming

that Nb2O5 is one of the best additives for MgH2. Equivalent

results were observed when comparing the effect of TTIP on

MgH2 with the effect of Nb2O5. As for Nb2O5, the addition of

only 0.5 mol% of TTIP is found to be sufficient to enhance the

kinetics of MgH2. TTIP liquid mixes readily with the MgH2 and

the handmixed sample produced the best desorption kinetics,

Table 1 e Crystallite size of milled MgH2 and before andafter cycling: determined from X-ray diffraction.

Timeh

Samples gMgH2

wt%bMgH2

wt%Scherrer size

nm

20 MgH2 12.4 85.7 13.4

22 MgH2 12.1 85.7 13.1

25 MgH2 12.2 86.2 12.1

30 MgH2 11.7 85.6 11.4

40 MgH2 11.1 86.8 11.4

20 MgH2 cycled 5 times e 89.03 102

MgH2 “as received” 225.5

Fig. 5 e Desorption kinetics of cycled pre-milled MgH2þ 1mol% TTIP handmixed (HM) compared to 1 mol% Nb2O5 milled for

2 h over five desorption/absorption cycles.




but the absorption kinetics were poorer compared to ball-

milling. TTIP had higher capacities than the Nb2O5 due to

the lower amount of additive required and the lower weight of

Ti compared to Nb.

While this study has shown that TTIP produces equally

good results as Nb2O5 when these compounds are used as

additives for the enhancement of kinetics in MgH2, it is hoped

that the real value of such a study is in increasing the infor-

mation available which may facilitate a future understanding

of the mechanism by which the kinetics are enhanced.

The underlying reason for enhancement of the absorption

and desorption kinetics remains unclear. Studies of the effect

of particle size reduction on Nb2O5 additive suggest that

dispersion throughout the MgH2 is important and TTIP, as a

liquid above 17 �C, may disperse more readily than most

powders. However, different ternary oxides such as

Mg3Nb6O11 or other phases that facilitate H2 pathways or

reduce activation energies are less likely in the case of the

TTIP additive. Further studies with different additives as well

as in-situ studies are needed to help establish the mecha-

nisms by which small amounts of additive significantly

improve the kinetics of absorption and desorption.

Acknowledgements

KA acknowledges receipt of an Australian Postgraduate

Award scholarship.

r e f e r e n c e s

[1] Ley MB, Jepsen LH, Lee Y-S, Cho YW, Bellosta von Colbe JM,Dornheim M, et al. Complex hydrides for hydrogen storage e

new perspectives. Mater Today 2014;17(3):122e8.[2] Webb CJ. A review of catalyst-enhanced magnesium hydride

as a hydrogen storage material. J Phys Chem Solids2015;84:96e106.

[3] Jain I, Lal C, Jain A. Hydrogen storage in Mg: a most promisingmaterial. Int J Hydrogen Energy 2010;35(10):5133e44.

[4] Crivello J-C, Dam B, Denys R, Dornheim M, Grant D, Huot J,et al. Review of magnesium hydride-based materials:development and optimisation. Appl Phys A2016;122(2):1e20.

[5] Crivello J-C, Denys R, Dornheim M, Felderhoff M, Grant D,Huot J, et al. Mg-based compounds for hydrogen and energystorage. Appl Phys A 2016;122(2):1e17.

[6] Evard E, Gabis I, Yartys VA. Kinetics of hydrogen evolutionfrom MgH2: experimental studies, mechanism andmodelling. Int J Hydrogen Energy 2010;35(17):9060e9.

[7] Dornheim M, Doppiu S, Barkhordarian G, Boesenberg U,Klassen T, Gutfleisch O, et al. Hydrogen storage inmagnesium-based hydrides and hydride composites. ScrMater 2007;56(10):841e6.

[8] T€opler J, Buchner H, S€aufferer H, Knorr K, Prandl W.Measurements of the diffusion of hydrogen atoms inmagnesium and Mg2Ni by neutron scattering. J Less-Common Metals 1982;88(2):397e404.

[9] Borgschulte A, B€osenberg U, Barkhordarian G, Dornheim M,Bormann R. Enhanced hydrogen sorption kinetics ofmagnesium by destabilized MgH2�d. Catal Today2007;120(3):262e9.

[10] Stander CM. Kinetics of decomposition of magnesiumhydride. J Inorg Nucl Chem 1977;39(2):221e3.

[11] Fern�andez JF, S�anchez CR. Rate determining step in theabsorption and desorption of hydrogen by magnesium. JAlloys Compd 2002;340(1e2):189e98.

[12] Skripnyuk VM, Rabkin E, Estrin Y, Lapovok R. The effect ofball milling and equal channel angular pressing on thehydrogen absorption/desorption properties of Mge4.95 wt%Zne0.71 wt% Zr (ZK60) alloy. Acta Mater 2004;52(2):405e14.

[13] Barkhordarian G, Klassen T, Bormann R. Catalyticmechanism of transition-metal compounds on Mg hydrogensorption reaction. J Phys Chem B 2006;110:11020e4(Copyright (C) 2015 American Chemical Society (ACS). AllRights Reserved.).

[14] Oelerich W, Klassen T, Bormann R. Metal oxides as catalystsfor improved hydrogen sorption in nanocrystalline Mg-basedmaterials. J Alloys Compd 2001;315(1e2):237e42.

[15] Barkhordarian G, Klassen T, Bormann R. Fast hydrogensorption kinetics of nanocrystalline Mg using Nb2O5 ascatalyst. Scr Mater 2003;49(3):213e7.

[16] Barkhordarian G, Klassen T, Bormann R. Effect of Nb2O5

content on hydrogen reaction kinetics of Mg. J Alloys Compd2004;364(1):242e6.

[17] Friedrichs O, Martınez-Martınez D, Guilera G, S�anchezL�opez JC, Fern�andez A. In situ energy-dispersive XAS andXRD study of the superior hydrogen storage system MgH2/Nb2O5. J Phys Chem C 2007;111(28):10700e6.

[18] Hanada N, Ichikawa T, Hino S, Fujii H. Remarkableimprovement of hydrogen sorption kinetics inmagnesium catalyzed with Nb2O5. J Alloys Compd2006;420(1e2):46e9.

[19] Friedrichs O, Klassen T, S�anchez-L�opez JC, Bormann R,Fern�andez A. Hydrogen sorption improvement ofnanocrystalline MgH2 by Nb2O5 nanoparticles. Scr Mater2006;54(7):1293e7.

[20] Porcu M, Petford-Long AK, Sykes JM. TEM studies of Nb2O5

catalyst in ball-milled MgH2 for hydrogen storage. J AlloysCompd 2008;453:341e6 (Copyright (C) 2015 AmericanChemical Society (ACS). All Rights Reserved.).

[21] Hanada N, Ichikawa T, Fujii H. Catalytic effect of niobiumoxide on hydrogen storage properties of mechanically ballmilled MgH2. Phys B (Amsterdam, Neth) 2006;383:49e50(Copyright (C) 2015 American Chemical Society (ACS). AllRights Reserved.).

[22] Friedrichs O, Aguey-Zinsou F, Fern�andez JRA, S�anchez-L�opez JC, Justo A, Klassen T, et al. MgH2 with Nb2O5 asadditive, for hydrogen storage: chemical, structural andkinetic behavior with heating. Acta Mater 2006;54(1):105e10.

[23] F�atay D, R�ev�esz �A, Spassov T. Particle size and catalytic effecton the dehydriding of MgH2. J Alloys Compd2005;399(1e2):237e41.

[24] Ma T, Isobe S, Morita E, Wang Y, Hashimoto N, Ohnuki S,et al. Correlation between kinetics and chemical bondingstate of catalyst surface in catalyzed magnesium hydride. IntJ Hydrogen Energy 2011;36(19):12319e23.

[25] Barkhordarian G, Klassen T, Bormann R. Kineticinvestigation of the effect of milling time on the hydrogensorption reaction of magnesium catalyzed with differentNb2O5 contents. J Alloys Compd 2006;407(1e2):249e55.

[26] Aurora A, Mancini MR, Gattia DM, Montone A, Pilloni L,Todini E, et al. Microstructural and kinetic investigation ofhydrogen sorption reaction of MgH2/Nb2O5 nanopowders.Mater Manuf Process 2009;24:1058e63 (Copyright (C) 2015American Chemical Society (ACS). All Rights Reserved.).

[27] Dolci F, Baricco M, Edwards PP, Giamello E. An investigationof the H2 uptake in Mg-Nb-O ternary phases. Int J HydrogenEnergy 2008;33:3085e90 (Copyright (C) 2015 AmericanChemical Society (ACS). All Rights Reserved.).


http://refhub.elsevier.com/S0360-3199(16)33619-9/sref1




















































































































































































[28] Rahman MW, Castellero A, Enzo S, Livraghi S, Giamello E,Baricco M. Effect of MgeNb oxides addition on hydrogensorption in MgH2. J Alloys Compd 2011;509(Supplement1(0)):S438e43.

[29] Rahman MW, Livraghi S, Dolci F, Baricco M, Giamello E.Hydrogen sorption properties of ternary MgeNbeO phasessynthesized by solidestate reaction. Int J Hydrogen Energy2011;36(13):7932e6.

[30] Ma T, Isobe S, Wang Y, Hashimoto N, Ohnuki S. Nb-gatewayfor hydrogen desorption in Nb2O5 catalyzed MgH2

nanocomposite. J Phys Chem C 2013;117(20):10302e7.[31] Pelletier J, Huot J, Sutton M, Schulz R, Sandy A, Lurio L, et al.

Hydrogen desorption mechanism in MgH2-Nbnanocomposites. Phys Rev B 2001;63(5):052103.

[32] Liang G, Huot J, Boily S, Van Neste A, Schulz R. Catalyticeffect of transition metals on hydrogen sorption innanocrystalline ball milled MgH2eTm (Tm ¼ Ti, V, Mn, Feand Ni) systems. J Alloys Compd 1999;292(1e2):247e52.

[33] Lu HB, Poh CK, Zhang LC, Guo ZP, Yu XB, Liu HK.Dehydrogenation characteristics of Ti- and Ni/Ti-catalyzedMg hydrides. J Alloys Compd 2009;481(1e2):152e5.

[34] Wang P, Wang AM, Zhang HF, Ding BZ, Hu ZQ.Hydrogenation characteristics of MgeTiO2 (rutile) composite.J Alloys Compd 2000;313(1e2):218e23.

[35] Jung KS, Kim DH, Lee EY, Lee KS. Hydrogen sorption ofmagnesium hydride doped with nano-sized TiO2. CatalToday 2007;120(3e4):270e5.

[36] Croston DL, Grant DM, Walker GS. The catalytic effect oftitanium oxide based additives on the dehydrogenation andhydrogenation of milled MgH2. J Alloys Compd2010;492(1e2):251e8.

[37] Dobrovolsky VD, Ershova OG, Solonin YM, Khyzhun OY,Paul-Boncour V. Influence of TiB2 addition upon thermalstability and decomposition temperature of the MgH2

hydride of a Mg-based mechanical alloy. J Alloys Compd2008;465(1e2):177e82.

[38] Pitt MP, Paskevicius M, Webb C, Sheppard D, Buckley C,Gray EM. The synthesis of nanoscopic Ti based alloys andtheir effects on the MgH2 system compared with theMgH2 þ 0.01 Nb2O5 benchmark. Int J Hydrogen Energy2012;37(5):4227e37.

[39] Ma L-P, Wang P, Cheng H-M. Hydrogen sorption kinetics ofMgH2 catalyzed with titanium compounds. Int J HydrogenEnergy 2010;35(7):3046e50.

[40] Zhu M, Lu Y, Ouyang L, Wang H. Thermodynamic tuning ofMg-based hydrogen storage alloys: a review. Materials2013;6(10):4654e74.

[41] Huot J, Ravnsbæk DB, Zhang J, Cuevas F, Latroche M,Jensen TR. Mechanochemical synthesis of hydrogen storagematerials. Prog Mater Sci 2013;58(1):30e75.

[42] Ares-Fernandez J-R, Aguey-Zinsou K-F. Superior MgH2

kinetics with MgO addition: a tribological effect. Catalysts2012;2:330e43 (Copyright (C) 2015 American ChemicalSociety (ACS). All Rights Reserved.).

[43] Aguey-Zinsou KF, Ares Fernandez JR, Klassen T, Bormann R.Effect of Nb2O5 on MgH2 properties during mechanicalmilling. Int J Hydrogen Energy 2007;32(13):2400e7.

[44] Fern�andez A, Deprez E, Friedrichs O. A comparative study ofthe role of additive in the MgH2 vs. the LiBH4eMgH2

hydrogen storage system. Int J Hydrogen Energy2011;36(6):3932e40.

[45] Deprez E, Justo A, Rojas TC, L�opez-Cart�es C, BonattoMinella C, B€osenberg U, et al. Microstructural study of theLiBH4eMgH2 reactive hydride composite with and withoutTi-isopropoxide additive. Acta Mater 2010;58(17):5683e94.

[46] Deprez E, Mu~noz-M�arquez MA, Rold�an MA, Prestipino C,Palomares FJ, Minella CB, et al. Oxidation state and localstructure of Ti-based additives in the reactive hydridecomposite 2LiBH4 þ MgH2. J Phys Chem C2010;114(7):3309e17.

[47] Boer F, Boom R, Mattens W, Miedema A, Niessen A. Cohesionin metals: transition metal alloys. Amsterdam: NorthHolland; 1988.

[48] Baldi A, Gremaud R, Borsa DM, Bald�e CP, van der Eerden AMJ,Kruijtzer GL, et al. Nanoscale composition modulations inMgyTi1�yHx thin film alloys for hydrogen storage. Int JHydrogen Energy 2009;34(3):1450e7.

[49] Kuziora P, Wyszy�nska M, Polanski M, Bystrzycki J . Why theball to powder ratio (BPR) is insufficient for describing themechanical ball milling process. Int J Hydrogen Energy2014;39(18):9883e7.

[50] Balzar D. Voigt-function model in diffraction line-broadeninganalysis. Int Union Crystallogr Monogr Crystallogr1999;10:94e126.

[51] Alsabawi K, Webb T, Gray EM, Webb C. The effect of C60

additive on magnesium hydride for hydrogen storage. Int JHydrogen Energy 2015;40(33):10508e15.

[52] Ma T, Isobe S, Wang Y, Hashimoto N, Ohnuki S. Catalyticeffect of Nb2O5 in MgH2-Nb2O5 ball-milled composites.Catalysts 2012;2(3):344e51.

[53] Nielsen TK, Jensen TR. MgH2-Nb2O5 investigated by in situsynchrotron X-ray diffraction. Int J Hydrogen Energy2012;37(18):13409e16.




















































































































































































136





The effect of ball-milling gas environment on the sorption kinetics of MgH2

with/without additives for hydrogen storage. International Journal of Hydrogen Energy.

2019;10.1016/j.ijhydene.2018.12.026.





(Signed) _______________________________(Date)_ 12/02/2019_____________

Khadija Alsabawi

(Countersigned) ________________________(Date)_ 12/02/2019_____________


(Countersigned) ________________________(Date) 12/02/2019______________


137

CHAPTER SIX

The effect of the milling environment gas on the hydrogen sorption properties of

MgH2+additives

Published as a Journal paper in the International Journal of Hydrogen Energy,

6. Introduction

The ball milling of metal hydride materials is known to have a significant effect on their

sorption kinetics, through the introduction of fresh surfaces, defects and reduced grain

size, which aids in faster diffusion of hydrogen into the bulk material. However, there

are many aspects to ball-milling, such as the type of mill, gas environment, vial and ball

material, ball to powder ratio, pressure and temperature in the vial, sample volume ratio,

rotation speed and milling/rest time ratios, all of which can have an effect on the final

milled sample. However, authors don’t usually provide all the necessary information

with regard to the ball-milling parameters that would enable a direct comparison.

This chapter reports on an investigation as to whether the gas environment in the ball-

mill vial during milling has an effect on the kinetic enhancement of magnesium hydride

with different additives. To do this, kinetic experiments were conducted on composite

materials ball-milled under argon (1 bar) and hydrogen (10 bar) atmospheres.

Conceptually, argon, being an inert gas, might be expected to play no role during the

repeated high energy and high pressure ball-ball and ball-vial impacts with intervening

particles of the composite sample. Molecular hydrogen might be expected to react with

exposed metal surfaces, or even oxygen if available, but the original materials are

already stoichiometric compounds, and the magnesium metal is already hydrided as

MgH2, so some degree of dissociation of these materials would be required before the

hydrogen gas environment would have a possible reaction path.

In this work, the ball milling process was performed in two stages. Firstly, the MgH2

was ball-milled without additive for 20 h to introduce defects and grain boundaries and

reduce the average crystallite size. This pre-milled MgH2 was then further ball-milled

with 1-2 mol% of the additives involved in this study (titanium isopropoxide, niobium

oxide and C60, carbon buckyballs). In both ball-milling stages, the gas environment

could be either hydrogen or Ar. TPD experiments were carried out to determine the

138

desorption kinetics, followed by a single pressure aliquot of hydrogen for absorption to

determine the sorption kinetics. It was found that the milling environment had a

negligible effect on desorption kinetics of most composites, however, depending the on

the gas used, the absorption uptake of some composites was changed. However, the

changes were different for different composite samples, with the same gases leading to

an improvement on one sample, but not others and vice versa. To understand the effect

of the ball-milling gas environment and the relationship to the additives used, requires

further investigation. The following paper provides the full details on the materials,

methods and results.

6.1. : Milling Environment Paper

Alsabawi K, Gray EM and Webb CJ. The effect of ball-milling gas environment on

the sorption kinetics of MgH2 with/without additives for hydrogen storage.


80;10.1016/j.ijhydene.2018.12.026.

The effect of ball-milling gas environment on thesorption kinetics of MgH2 with/without additivesfor hydrogen storage


Queensland Micro- and Nanotechnology Centre, Griffith University, Nathan, 4111, Brisbane, Australia

a r t i c l e i n f o

Article history:

Received 8 October 2018

Received in revised form

4 December 2018

Accepted 5 December 2018

Available online 6 January 2019

Keywords:

Magnesium hydride

Hydrogen storage

Catalyst

Milling gas

Sorption kinetics

a b s t r a c t

A study of the effect of the ball-milling gas environment on the kinetic enhancement of

MgH2 with different additives was conducted using argon and hydrogen. The as-sourced

MgH2 was milled for 20 h and then milled for a further 2 h after adding 1e2 mol% of one

of the additives titanium isopropoxide, niobium oxide or carbon buckyballs, varying the

gas environment for both ball-milling stages. The milling environment had little or no

effect on the desorption kinetics in most cases. However, in some cases, the absorption

uptake differed by up to 2 wt%, depending on the gas used. This effect was not consistent

among the composite samples surveyed, demonstrating the importance of reporting all

information about the ball-milling processes used, including the gas environment.

© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Introduction

Magnesiumhydride (MgH2) has gravimetric and volumetric H-

storage capacities of 7.6 wt% and 110 g H/l respectively, which

makes it a hydrogen storagematerial of interest. However, the

high operating temperature, due to the thermodynamic sta-

bility of the hydride, limits its application [1]. Ball-milling

MgH2 enhances the sorption kinetics at 300 �C compared to

the non-milled material [2].

Ball-milling metal hydride materials helps create fresh

metal surfaces, increases the surface area, forms micro/

nanostructures and introduces defects (such as dislocations,

stacking faults, vacancies, extra grain boundaries, depending

on the material used) on both the surface and in the interior

of the sample [3,4]. The exposed metal surface aids the

dissociation of hydrogen molecules and increased grain

boundaries promote faster diffusion of hydrogen into the

bulk material [4].

Milling parameters such as milling speed, time, tempera-

ture, atmosphere, ball-to-powder ratio [2], milling medium

(ceramic, steel, zirconia, etc.) [5] and ball-mill type (planetary,

high energy, shakers/mixers), all potentially change the final

material. Some of these parameters are interdependent; for

example, optimum milling time is dependent on the speed

and temperature of themill, size of container and type of mill.

Ball-milling may also affect the sample through the intro-

duction of impurities, the generation of higher temperatures

and the increased handling [3].

* Corresponding author.E-mail address: [email protected] (K. Alsabawi).

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier .com/locate/he


https://doi.org/10.1016/j.ijhydene.2018.12.0260360-3199/© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

mailto:[email protected]

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijhydene.2018.12.026&domain=pdf

www.sciencedirect.com/science/journal/03603199

www.elsevier.com/locate/he

https://doi.org/10.1016/j.ijhydene.2018.12.026



Ball-milling MgH2 with small amounts (up to 10 mol%) of

additives such as carbons, metal oxides, metal halides and

many others, has been shown to significantly enhance the

MgH2 sorption kinetics [6,7]. Previous studies have shown that

the amount of additives used could significantly affect the

hydrogen sorption kinetics of MgH2 [8e11]. Ball-milling addi-

tives with MgH2 helps to create a homogeneous distribution

throughout the composite material, improving the contact

between the metal hydride and additive.

MgH2 milled with additives has been studied by several

authors using different milling parameters [7,12], including

the type ofmill, milling time [1], number and size of balls, ball-

to-powder ratio, rotational speed, milling atmosphere and

hydrogen pressure [1]. Many authors do not provide all the

necessary information with regard to the ball-milling to

enable repeat experiments or to make comparisons. Also,

while changes to the additive, such as decomposition and

interactionwith the Mg or MgH2 may be detailed, there is little

understanding of the role of additives in improving the ki-

netics of MgH2. Previous studies suggested that the gas envi-

ronment used during ball-milling can make a difference

[13e15], with the most commonly used gases being Ar

[16e18] and H2 [19,20] and in some cases under N [21], He [22]

and even under vacuum [23].

Varin et al. [24] studied the effect of ball-milling MgH2

under hydrogen and argon on the desorption rates. They

found that activation energy for desorption was lower for

hydrolysed MgH2 powder, i.e. covered with a layer of Mg(OH)2,

that was milled under H2 when compared to an argon milled

sample. However, they concluded that the amount of

hydrogen desorbed was related to the milling time and was

independent of the milling atmosphere. Asselli et al. [14]

showed that milling (2 Mg þ Fe) and (2MgH2 þ Fe) under

different environments did not affect the kinetics of absorp-

tion and desorption, but milling under hydrogen resulted in

decreasing the particles sizes and doubled the amount of

Mg2FeH6 formation.

Spassov et al. [13], produced nanocrystalline

Mg87Ni3Al3Mm7 hydrides by reactive mechanical milling.

They studied the effect of the gas atmosphere on the micro-

structure of the hydrides by milling one sample under H2 gas

only, and another sample first under Ar and then under H2.

The amount of MgH2 phase was found to be slightly larger for

the sample milled under H2 only. The H-absorption rate was

found to be higher for the sample milled under Ar than H2.

Ershova et al. [15], investigated the effect of milling under

different gases on the kinetics and thermodynamics of MgH2þ10 wt% Al mixtures, where they found that milling under

different environments influenced the crystalline phases.

Their first samplewasmilled under 12 bar of H2 (MA1) for 10 h,

second sample (MA2) under Ar for 17 h then H2 for 10 h, third

sample (MA3) under Ar for 20 h and the last sample (MA4) was

pure MgH2 milled under the same condition as MA1 for com-

parison. XRD patterns were obtained for all the samples after

milling. For MA1, three phases were found to be present,

metallic Mg, MgH2 and crystalline phase Mg17Al12. For MA2

there was only MgH2 andmetallic Mg, whereas MA3 consisted

of metallic Mg and Mg17Al12. Their results showed that

regardless of themilling atmosphere, MgH2 phase was formed

from which the aluminum was absent or present in a minute

content. Their results were in agreement with [25], in which

the Al was found to be rejected from Mg(Al) solid solution

during hydriding.

As an additive for MgH2, Nb2O5 has been studied exten-

sively to determine its catalytical effect on the hydrogen

sorption kinetics of MgH2 [6]. Nb2O5 has been milled with

MgH2 under Ar [26,27], H2 [28,29] or N [5,30] atmospheres, with

the majority of users using Ar. There has been no direct

comparison for Nb2O5 milled under different ball-milling

atmospheres.

Francke et al. [31], studied themodification of the structure

and specific surface area of graphite by high-energy milling

under argon and hydrogen atmospheres. TEM studies showed

that after milling, graphite “stacking packages” of particles

appeared. These “stacking packages” disappeared after 10 h of

milling under argon but stayed even after 15 h ofmilling under

hydrogen. The authors suggested that these packages influ-

ence the specific surface area rather than the structure. The

hydrogen-milled samples hadmore small particles (�15 nm in

length and �3 nm in height) than argon-milled samples,

which indicated that the hydrogen atmosphere prevented

agglomeration. Ong et al. [32], milled natural graphite under

inert and oxygen atmospheres. They found that the oxygen

preserved the crystal structure and suppressed the specific

surface area growth, but the sample was found to be highly

anisotropic with large stacking faults.

Rietsch et al. [33] used a planetary ball mill to evaluate the

influence of the gas atmosphere on the evolution of carbon

properties. They found that the atmosphere inside themilling

vial affects themillingmodes (sliding or shockmode)which in

turn affect the evolution of the graphite composition. The

sliding mode is considered to be gentle on the graphite

structure whereas the shock mode degrades the graphite

structure. The lubricating nature of graphite prevents the

shock mode from occurring, but only in the presence of oxy-

gen which annihilates the fresh reactive sites by forming

functional groups. However, under an inert atmosphere, or

when the oxygen is consumed, the lubricating layer is

destroyed resulting in shock mode. This establishes a link

between graphite structure evolution and the milling

atmosphere.

Xiao et al. [34] milled Ti-doped NaAlH4 complex hydrides

under atmospheres of H2 and Ar, and found thatmilling under

H2 improved the reversible hydrogen storage properties of the

hydrides, resulting in a 0.28 wt% higher hydrogen capacity

with a faster hydrogen desorption rate.

Recently, the kinetics of composite samples of MgH2 and

additive titanium isopropoxide, also called titanium tetraiso-

propoxide (TTIP) were compared to Nb2O5, finding that TTIP

produced equivalent results for both desorption and absorp-

tion [35].

The studies above indicate that the gas environment while

ball-milling can potentially affect both the magnesium hy-

dride and the additive. This work investigated the sorption

kinetics of MgH2 milled with different additives under argon

and hydrogen atmospheres, where theMgH2was initially ball-

milled without additive and then further ball-milled with

additive. Temperature Programmed Desorption (TPD) fol-

lowed by a single pressure aliquot of hydrogen for absorption

was carried out to determine the sorption kinetics of the




milled samples. Care was taken to control the milling condi-

tions so that comparisons could be made between samples,

altering only one aspect at a time. The two best additives for

MgH2 enhancement, Nb2O5 and TTIP were chosen as test

cases, together with C60.

Experimental

Magnesium hydride powder, 98% purity, (AB109434) was

purchased from ABCR GmbH & Co. 99.999% purity TTIP from

SigmaeAldrich, 20 nm Nb2O5 from NovaCentrix and 99.9%

fullerene C60 (SES Research) were used for the additives. The

as-supplied MgH2 was pre-milled in a PQ-N04 planetary mill

(Across International) for 20 h using a 50 ml zirconia vial and

10� 10 mm zirconia balls with a ball-to-powder ratio (bpr) of

21:1 and a vial filling factor [2] of 17.4%. Pre-milled MgH2

samples (preMgH2) were prepared in 5 g batches under either

1 bar of Ar or 10 bar of H2. Additives were subsequently milled

with the preMgH2 for 2 h (1 g sample, 6 ZrO2 balls, 600 rpm, 21:1

bpr, vial filling factor 7.7%) also under both Ar and H2 envi-

ronments. All materials were handled in a dry Ar atmosphere

MBraun glovebox with O2 and H2O levels below 1 ppm.

Hydrogen cycling was performed on a modified custom

Sieverts apparatus. TPD studies were carried out on the as-

prepared samples under dynamic vacuum from room tem-

perature to 400 �C with a temperature ramp rate of 2 �C/min

[8,35]. Baseline starting vacuum was typically 3� 10�6 mbar.

The sample remained at 400 �C for 2 h to ensure complete

desorption. Absorption kinetics data were recorded at

230 �C, after a single hydrogen pressure aliquot to reach an

initial pressure of 50 bar on the sample. XRD was performed

on each 5 g batch of pre-milled MgH2 as well as the 1 g

composite samples to ensure uniformity of the starting

material. No variation was observed in the crystalline

structure between Ar and H2 milled samples. All desorption/

absorption experiments were undertaken at least twice to

ensure repeatability.

No additives: MgH2 milled for 20 h under Ar andH2

Absorption/desorption data for MgH2 milledwithout additives

for 20 h under Ar or H2 are shown in Fig. 1. It can be seen that

the desorption kinetics are similar within experimental error.

The sample milled under H2 has faster absorption initially

than the Ar-milled sample, but does not reach the same ca-

pacity by 400 �C.

PreMgH2 milled with 1 mol% Nb2O5 for 2 h

Two batches of PreMgH2 were prepared by ball-milling for 20 h,

one under 10 bar of H2 and the other under 1 bar of Ar. 1 mol%

of Nb2O5 additive was added to each batch and thenmilled for

2 h under either 10 bar H2 or 1 bar of Ar, resulting in 4 samples.

The desorption kinetics for the composite samples are shown

in Fig. 2. The sample milled under H2 both times finished

Fig. 1 e Absorption/desorption of MgH2 milled for 20 h

under H2 (dashed line) and under Ar (unbroken line).

Fig. 2 e Desorption kinetics of pre-milled MgH2 ball-milled

for 2 h with (1mol%) Nb2O5 under different gas

environments.

Fig. 3 e Absorption kinetics of pre-milled MgH2 ball-milled

for 2 h with (1mol%) Nb2O5 under different environment.




desorbing at a slightly lower temperature than the sample

where the additive was milled under Ar. The onset of

desorption for all of the samples is roughly the same. Samples

that were milled fully or partially under Ar show a greater

spread of desorption with temperature, often with a more

obvious second peak.

The absorption kinetics for the four samples are shown in

Fig. 3. The composite samples where the final milling envi-

ronment was Ar have very similar kinetics, which suggests

that the pre milling environment of MgH2 had no effect on the

absorption kinetics. However, the composite sample that was

milled under H2 for both MgH2 and additives had the slowest

and the least uptake between the four samples, whereas the

sample where MgH2 was milled under Ar then the composite

was milled under H2 had the highest hydrogen capacity and

the fastest kinetics. Although the differences are small, these

results suggest that milling under Ar for either pre-milling

alone or for the total milling procedure is preferable, con-

trary to the desorption results. The TPD experiments were

repeated for these four samples, confirming the same results

within experimental uncertainties.

PreMgH2 milled with 1 or 2 mol% TTIP for 2 h

1 mol% of TTIP was added to the H2-milled MgH2 and the

compositewasmilled underH2 environment and another 1mol

% TTIP was added to Ar-milled MgH2 and thenmilled under Ar.

Fig. 4 shows both the absorption and desorption kinetics of the

composite samples. It can be seen that the desorption kinetics

for the samples are similar, but the Ar-milled sample desorbs

the majority of hydrogen at a slightly lower temperature than

the H2-milled sample. Conversely, the H2 milled sample has

superior absorption. The same results were also found for a

2 mol% TTIP composite sample (not shown).

PreMgH2 milled with 1 mol% C60 for 2 h

Fig. 5 shows the absorption/desorption kinetics of thepremilledMgH2 with 1mol% C60 carbon buckyballs, either both

milled under H2 or both under Ar. While there is little to no

change in the desorption kinetics between the Ar and H2

milled samples, the Ar milled sample had significantly better

absorption kinetics and absorbed ~2 wt% more H2 in 20 min

than the sample milled under H2

Conclusion

The effect of ball-milling MgH2 with and without additives

under different gases (Ar, H2) was surveyed. MgH2, pre-milled

for 20 h under either 10 bar of hydrogen or 1 bar of argon was

milled for an extra 2 h with 1 mol% TTIP, Nb2O5 and C60, also

under either H2 or Ar. The kinetics of absorption and desorp-

tion of the composite samples was investigated by thermal

desorption under vacuum, followed by a single absorption

pressure step of 50 bar of H2 at 230 �C.The milling environment had little to no effect on the

desorption kinetics inmost cases, with only the TTIP composite

sample showing marginal improvement. However, the ab-

sorption results demonstrate a significant difference between

Ar and H2 milling with the additives. Unfortunately, the dif-

ference reverses between the TTIP and other samples, with

the TTIP composite sample showing better absorption for H2

milling whereas the Nb2O5 and C60 composite samples had

superior absorption with Ar milling.

While this study has shown there is an effect of the milling

environment on the absorption kinetics of MgH2, it does not

suggest a simple consistent answer as to which gas produces

the best results as it varies according to additive used.

Acknowledgements

K.A acknowledges receipt of an Australian Government

Research Training Program Scholarship.

Fig. 4 e Absorption/desorption of H2-milled preMgH2 milled

with (1 mol%) TTIP for 2 h under H2 (dashed line) and under

Ar (unbroken line).

Fig. 5 e Absorption/desorption of preMgH2 milled with

1 mol% C60 for 2 h under H2 (dashed line) and under Ar

(unbroken line).




r e f e r e n c e s

[1] Crivello J-C, Dam B, Denys R, Dornheim M, Grant D, Huot J,et al. Review of magnesium hydride-based materials:development and optimisation. Appl Phys A2016;122(2):1e20.

[2] Polanski M, Bystrzycki J, Plocinski T. The effect of millingconditions on microstructure and hydrogen absorption/desorption properties of magnesium hydride (MgH2) withoutand with Cr2O3 nanoparticles. Int J Hydrogen Energy2008;33(7):1859e67.

[3] Suryanarayana C. Mechanical alloying and milling. ProgMater Sci 2001;46(1):1e184.

[4] Sakintuna B, Lamari-Darkrim F, Hirscher M. Metal hydridematerials for solid hydrogen storage: a review. Int J HydrogenEnergy 2007;32(9):1121e40.

[5] Pukazhselvan D, Otero-Irurueta G, P�erez J, Singh B, Bdikin I,Singh MK, et al. Crystal structure, phase stoichiometry andchemical environment of MgxNbyOxþy nanoparticles andtheir impact on hydrogen storage in MgH2. Int J HydrogenEnergy 2016;41(27):11709e15.

[6] Webb CJ. A review of catalyst-enhanced magnesium hydrideas a hydrogen storage material. J Phys Chem Solid2015;84:96e106.

[7] Wang Y, Wang Y. Recent advances in additive-enhancedmagnesium hydride for hydrogen storage. Prog Nat Sci:Materials International 2017;27(1):41e9.

[8] Alsabawi K, Webb T, Gray EM, Webb C. The effect of C60

additive on magnesium hydride for hydrogen storage. Int JHydrogen Energy 2015;40(33):10508e15.

[9] Zhang B, Li W, Lv Y, Yan Y, Wu Y. Hydrogen storageproperties of the mixtures MgH2eLi3N with different molarratios. J Alloy Comp 2015;645:S464e7.

[10] Zhang Q, Wang Y, Zang L, Chang X, Jiao L, Yuan H, et al.Core-shell Ni3N@ nitrogen-doped carbon: synthesis andapplication in MgH2. J Alloy Comp 2017;703:381e8.

[11] da Conceic~ao MOT, Brum MC, dos Santos DS, Dias ML.Hydrogen sorption enhancement by Nb2O5 and Nb catalystscombined with MgH2. J Alloy Comp 2013;550:179e84.

[12] Crivello J-C, Denys R, Dornheim M, Felderhoff M, Grant D,Huot J, et al. Mg-based compounds for hydrogen and energystorage. Appl Phys A 2016;122(2):1e17.

[13] Spassov T, Petkov V, Solsona P. Hydriding/dehydriding ofMg87Ni3Al3Mm7 (Mm¼La, Ce-rich mischmetal) alloyproduced by mechanical milling. J Alloy Comp2005;403(1):363e7.

[14] Asselli A, Huot J. Investigation of effect of millingatmosphere and starting composition on Mg2FeH6formation. Metals 2014;4(3):388.

[15] Ershova OG, Dobrovolsky VD, Solonin YM, Khyzhun OY,Koval AY. The effect of Al on thermal stability and kinetics ofdecomposition of MgH2 prepared by mechanochemicalreaction at different conditions. Mater Chem Phys2015;162:408e16.

[16] Cova F, Larochette PA, Gennari F. Hydrogen sorption inMgH2-based composites: the role of Ni and LiBH4 additives.Int J Hydrogen Energy 2012;37(20):15210e9.

[17] Han SS, Goo NH, Lee KS. Effects of sintering on compositemetal hydride alloy of Mg2Ni and TiNi synthesized bymechanical alloying. J Alloy Comp 2003;360(1e2):243e9.

[18] Narayanan DL, Lueking AD. Mechanically milled coal andmagnesium composites for hydrogen storage. Carbon2007;45(4):805e20.

[19] Lototsky MV, Denys RV, Yartys VA. Combustion-typehydrogenation of nanostructured Mg-based composites forhydrogen storage. Int J Energy Res 2009;33(13):1114e25.

[20] Xiao X, Liu G, Peng S, Yu K, Li S, Chen C, et al. Microstructureand hydrogen storage characteristics of nanocrystalline Mgþx wt% LaMg2Ni (x¼ 0e30) composites. Int J Hydrogen Energy2010;35(7):2786e90.

[21] Imamura H, Sakasai N. Hydriding characteristics of Mg-based composites prepared using a ball mill. J Alloy Comp1995;231(1e2):810e4.

[22] Varin R, Czujko T, Mizera J. The effect of MgNi2 intermetalliccompound on nanostructurization and amorphization ofMgeNi alloys processed by controlled mechanical milling. JAlloy Comp 2003;354(1e2):281e95.

[23] Rud A, Lakhnik A, Ivanchenko V, Uvarov V, Shkola A,Dekhtyarenko V, et al. Hydrogen storage of the MgeCcomposites. Int J Hydrogen Energy 2008;33(4):1310e6.

[24] Varin RA, Jang M, Czujko T, Wronski ZS. The effect of ballmilling under hydrogen and argon on the desorptionproperties of MgH2 covered with a layer of Mg(OH)2. J AlloyComp 2010;493(1e2):L29e32.

[25] Tanniru M, Slattery DK, Ebrahimi F. A study of stability ofMgH2 in Mge8at%Al alloy powder. Int J Hydrogen Energy2010;35(8):3555e64.

[26] Ma T, Isobe S, Wang Y, Hashimoto N, Ohnuki S. Nb-gatewayfor hydrogen desorption in Nb2O5 catalyzed MgH2

nanocomposite. J Phys Chem C 2013;117(20):10302e7.[27] Pitt MP, Paskevicius M, Webb C, Sheppard D, Buckley C,

Gray E. The synthesis of nanoscopic Ti based alloys andtheir effects on the MgH2 system compared with the MgH2þ0.01 Nb2O5 benchmark. Int J Hydrogen Energy2012;37(5):4227e37.

[28] Isobe S, Umeda A, Wakasugi T, Ma T, Yamagami R, Hino S,et al. Microscopic study on hydrogenation mechanism ofMgH2 catalyzed by Nb2O5. Mater Trans 2014;55(8):1175e8.

[29] da Conceicao MOT, Brum MC, dos Santos DS, Dias ML.Hydrogen sorption enhancement by Nb2O5 and Nb catalystscombined with MgH2. J Alloy Comp 2013;550:179e84.Copyright (C) 2015 American Chemical Society (ACS). AllRights Reserved.

[30] Porcu M, Petford-Long AK, Sykes JM. TEM studies of Nb2O5

catalyst in ball-milled MgH2 for hydrogen storage. J AlloyComp 2008;453:341e6. Copyright (C) 2015 AmericanChemical Society (ACS). All Rights Reserved.

[31] Francke M, Hermann H, Wenzel R, Seifert G, Wetzig K.Modification of carbon nanostructures by high energy ball-milling under argon and hydrogen atmosphere. Carbon2005;43(6):1204e12.

[32] Ong TS, Yang H. Effect of atmosphere on the mechanicalmilling of natural graphite. Carbon 2000;38(15):2077e85.

[33] Rietsch JC, Gadiou R, Vix-Guterl C, Dentzer J. The influence ofthe composition of atmosphere on the mechanisms ofdegradation of graphite in planetary ball millers. J AlloyComp 2010;491(1e2):L15e9.

[34] Xiao X, Chen L, Wang X, Wang Q, Chen C. The hydrogenstorage properties and microstructure of Ti-doped sodiumaluminum hydride prepared by ball-milling. Int J HydrogenEnergy 2007;32(13):2475e9.

[35] Alsabawi K, Webb T, Gray EM, Webb C. Kineticenhancement of the sorption properties of MgH2 with theadditive titanium isopropoxide. Int J Hydrogen Energy2017;42(8):5227e34.










































































































































































































144

CHAPTER SEVEN

The effect of transition metal oxide additives on the sorption kinetics of MgH2 for

hydrogen storage

(As a Journal paper submitted to International Journal of Hydrogen Energy)

7. Introduction

The discovery that the transition organo-metallic oxide, TTIP, was just as effective as

the bench mark Nb2O2 (chapter 5) [1] on the sorption kinetics of MgH2 raised a number

of questions. Was the kinetic enhancement due or partly due to the use of the transition

metal titanium? Did the oxygen content of TTIP also play a role in improving the

kinetics? Was there a contribution from the carbon and/or hydrogen content of the

organic? Or was it simply that TTIP is a liquid at room temperature and was therefore

better dispersed throughout the MgH2 during milling? Or various combinations of these

reasons.

These questions relate to the role of the additive in the kinetic enhancement of the Mg-

H interaction, and, although it was shown that TTIP was another effective (transition

metal based high valence oxide) additive, it did not help explain the role this, or other

additives, play in the Mg-H system.

Understanding the role of additives in enhancing the kinetics of MgH2 is still a

challenge. Most published kinetic experiments involve several or many parameters

simultaneously, including ball milling parameters (gas environment, pressure and

temperature in the vial, milling time, vial and ball material, ball to powder ratio, sample

volume ratio [2], type of mill, rotation speed, milling/rest time ratios), the nature of the

additive and the MgH2 (liquid/solid, powder size, crystalline/amorphous, contaminants,

especially oxides) and the experimental procedures (sample cell materials, sample size,

sample temperature, especially during hydrogenation/dehydrogenation due to

exo/endothermic reactions, quality of vacuum, temperature ramp rate, back pressure

during desorption) – some of which are not measured or not communicated.

As part of a larger plan to relate the properties of additives to their measured

enhancement in order to evaluate which properties are effective and which properties do

145

not help, another investigation into similar but slightly different additives (liquid/solid,

titanium/vanadium, oxide/non-oxide) was undertaken to establish which additives work,

and under what conditions.

This chapter consists of an experimental paper which has been submitted to the

International Journal of Hydrogen Energy. Six different new additives, not previously

employed in the sorption kinetics of MgH2, were used as well as TTIP and Nb2O2 for

comparison purposes. Altogether, the eight additives included three powder samples

and five liquids, and the transition metals Nb, Ti and V. All but one additive were

oxides, with the other being an organo-metallic titanium-based chloride. Apart from the

Nb2O2 all the samples were organo-metallics, containing carbon, hydrogen and oxygen

(except the chloride additive which did not contain oxygen in the chemical

composition).

The same experimental procedure was used as in the previous TTIP investigation.

Different amounts of additives (0.5- 10 mol%) were ball milled with MgH2 for different

times, where the milling parameters and the apparatus used for the kinetics

measurements were similar to that used previously (the full details are in the submitted

paper). Overall both Ti and V based additives enhanced the sorption kinetics of MgH2,

where the V-based oxides had faster absorption and higher uptake when compared to

Ti-based oxides. A small amount of 1 mol% Ti-chloride based additive milled for short

time (1 h) had better desorption than all additives investigated in this work but with very

poor absorption.

Summarising, this investigation also raises further questions. While it did not

specifically address the physical or chemical role of the additive in the

adsorption/desorption process, it has established the general benefit of a liquid additive

over comparable solid additives. However, contradicting this, the chloride additive, a

solid, had the best desorption kinetics. This additive also suggests that an oxide additive

is not necessary for fast desorption. Interestingly, the extremely poor absorption results

for the chloride additive, to the extent that all of the composite chloride additive

samples absorbed more slowly than the ball-milled MgH2 without additive, support the

idea that different additives can enhance absorption and desorption differently, and that

combinations of additives might be an effective avenue to pursue [3].

146

7.1. References

1. Alsabawi K, Webb TA, Gray EM and Webb CJ. Kinetic enhancement of the



34;10.1016/j.ijhydene.2016.12.072.







106;10.1016/j.jpcs.2014.06.014.

7.2. Paper

147





The effect of transition metal oxide additives on the sorption kinetics of MgH2 for hydrogen

storage. International Journal of Hydrogen Energy. (submitted)





(Signed) _______________________________(Date)_ 12/02/2019_____________

Khadija Alsabawi

(Countersigned) ________________________(Date)_ 12/02/2019_____________


(Countersigned) ________________________(Date) 12/02/2019______________


148

The effect of transition metal oxide additives on the sorption kinetics of MgH2 for

hydrogen storage



Brisbane Australia

Abstract

The effect of a number of additives on the kinetics of absorption and desorption of

magnesium hydride were studied using TPD, TDS and single aliquot absorption in a

Sieverts instrument. The additives were mainly transition metal oxides (3 x Ti-based

organo-metallic oxides, 3 x V-based organo-metallic oxides and Nb2O5 as the

benchmark for comparison). The purpose was to investigate each additive to determine

the relative effect of the different transition metals, the effect of the organo-metallic

components, and the difference between liquid and powder additives. In addition, a Ti-

based organo-metallic chloride (C16H22Cl2Ti) was also tested to compare an oxide and

non-oxide additive. A systemic survey was performed for ball-milled MgH2 mixed with

various amounts of additives for different times. Most of the additives used were liquids

and it is possible to conclude that liquids are effective at improving the sorption kinetics

of MgH2, possible due to better dispersion during milling. The Ti-based chloride

(BisTiCl2) had the best desorption kinetics but the slowest absorption, raising questions

about the role of additives and differences between absorption and desorption kinetics.

* Corresponding author: Tel.: +61 7 37353625 E-mail address: [email protected]

149

Introduction

Magnesium Hydride (MgH2) remains an attractive material for hydrogen storage, due

its high volumetric capacity (~0.1g/cm3) and gravimetric (7.6 wt%) capacity and low

cost [1-3]. Unfortunately, the high thermodynamic stability (∆H= -75 kJ mol-1) [4],

requiring high temperatures for hydrogen release, hinders practical use of magnesium

for hydrogen storage and poor kinetics makes the rate of the hydrogen absorption and

desorption slow under moderate conditions [5-7].

Research has shown that decreasing the crystallite size of MgH2 through ball milling or

other forms of severe deformation [8] helps to improve the sorption kinetics. Ball

milling introduces additional grain boundaries which reduce the diffusion lengths for

hydrogen [9, 10] and creates fresh surfaces free from oxide layers that slow

hydrogenation.

The addition of small amounts (up to 10 mol%) of various materials added to MgH2,

sometimes described as catalysts, during milling, enhances the sorption kinetics [11-15].

Transition metal oxides have been of a particular interest, although other materials,

including the transition metal halides and elemental transition metals, have also

significantly improved the kinetics of MgH2 absorption and desorption [16-18].

Elemental transition metals such as V and Nb do not form intermetallic compounds

with magnesium [19], which makes them heterogeneous catalysts [20]. The influence of

a wide range of metal oxides on the sorption kinetics of nanocrystalline Mg-based

systems were investigated, by Oelerich et al. [21] where they showed that only

transition metal oxides additives lead to a significant enhancement of hydrogen sorption

kinetics. One of the earliest, and still one of the most effective additives, is Nb2O5[22],

which is regarded as the benchmark for Mg-H system kinetics. Other materials

producing similar results include TiO2 [23, 24], V2O5 [21, 25, 26] and titanium tetra-

isopropoxide TTIP [27].

Transition metals, thought to enable dissociation and re-association of hydrogen on the

magnesium surface, may also facilitate the reduction of MgH2 particle size during

milling [7, 28]. Several authors have also suggested that the electronic structure of

transition metal ions determines their catalytic effect on MgH2, where metal oxides with

higher valence state have greater catalytic effect [21, 22]. Different niobium oxides

(NbO, NbO2 and Nb2O5) additives were investigated by Barkhordarian, and it was

found that the catalytic efficiency increased with the valence state of the niobium [29].

150

However, the exact catalytic mechanism of transition metal oxides on the sorption

properties of MgH2 is still undetermined.

Metal oxide surfaces tend to passivate the metal, preventing reactive gases from

reaching the pure metal; hence the slow diffusion of hydrogen through a MgO layer

slows the hydrogenation of Mg [6, 30]. However, Aguey-Zinsou et al. [28, 31] showed

that MgO acts as a ‘process control agent’ during milling, by preventing cold welding

and reducing particle agglomeration, and thus helps to decrease the particle size of

MgH2, which increases H diffusion and improves the rate of H-sorption in magnesium.

Identifying the nature of the oxide phase after milling, and then after cycling, is still a

challenge, in the case of Nb2O5 some studies have found no structural transformation of

the Nb2O5 phase and that it only acts as a catalyst [22, 32], whereas others have shown

that Nb2O5 acts as an additive where they report a reduction in the Nb oxide species

after cycling and a new ternary oxide phase MgNbxOy observed by XRD [33-36].

One problem with understanding the role of additives is that many factors are

(sometimes necessarily) incorporated into each experiment. Most studies involve ball-

milling, but the parameters vary and are not always well-reported. Sometimes negative

results help to indicate what is effective and what isn’t. TiN is also an effective additive,

suggesting oxides are not essential [37]. There are studies that show some additives

enhance desorption but not absorption (and vice versa), suggesting that they support a

rate mechanism involved in one process, but not the other. For this reason, it has been

suggested [7] that strategic combinations of additives may be effective.

Jung et al .[38] in 2006 synthesized composites of MgH2 with a range of metal oxides

(V2O5, Cr2O3, Fe3O4, and Al2O3) by ball milling where they followed the steps of

Oelerich et al. [21] but milled for shorter time, to investigate the hydriding properties.

They confirmed the work done previously [21], and found that at lower temperatures

(250-200 ºC) the MgH2/V2O5 sample showed even faster kinetics and higher absorption

capacities up to 3.2 and 2.25 wt% compared to unmodified MgH2. Korablov et al. [39],

found that V2O5 reduces readily on contact with hydrogen and during milling with

MgH2. V2O5 composite samples were cycled twice, and XRD patterns collected,

showing V2H and V peaks, which suggest that V2H and V act as active catalysts, since

they did not form intermetallics or alloys.

151

Another oxide that has been widely studied is TiO2. Wang et al.[40] showed that ball

milling MgH2 with TiO2 improved the sorption kinetics, where they claimed that the

improvement was due to absence of oxide reduction and the existence of n-TiO2, which

was also found later [23], while others reported reduction of TiO2 during milling [24,

41-43]. Jung et al. [23] showed that a uniform distribution occurs during milling

provided the rutile TiO2 concentration is below 5 mol%. Croston et al. [42] prepared

TiO2 through calcination at different temperatures from 70 to 800 ºC, where they found

that the calcination temperature has an effect on the dehydrogenation temperature of

MgH2 - with 70 ºC calcination temperature having the best result. They also found that

anatase was more effective than the rutile at lowering the dehydrogenation temperature

of MgH2. However, Vujasin et al. [44] showed that rutile TiO2 additives decreased the

desorption activation energy while anatase TiO2 had a negligible effect. Alsabawi et al.

[45] used TTIP as an additive for MgH2 and found that it has similar effect on the

sorption kinetics as the bench mark Nb2O5. The fact that TTIP is a liquid may have

facilitated a more homogeneous distribution and therefore higher surface contact with

smaller well-distributed particles, thus improving the kinetics of MgH2.

Although the mechanisms by which additives enhance the sorption kinetics of the Mg-H

system are still unknown, and despite some contradictory results, some aspects of

additives have been confirmed. The success of transition metal compounds is well-

established, although this is not a requirement. The suggestion of higher order oxides

has been supported by the initial results of Nb2O5 and the recent success of TTIP

(Ti(OCH(CH3)2)4). With these results in mind, and including the prospect of better

dispersion for a liquid additive, the following compounds have been chosen for this

study:

• Titanium Ethoxide, Ti(OC2H5)4, a Ti based organo-metallic liquid for comparison

with TTIP as another liquid with Ti and 4 oxygens,

• Titanium(IV) methoxide Ti(OCH3)4 as a powder Ti based additives to compare

with the liquid additives,

• Vanadium(V) oxytriisopropoxide OV(OCH(CH3)2)3,

• Vanadium(V) oxytripropoxide OV(OC3H7)3,

• V-oxytriethoxide OV(OC2H5)3

(All the vanadium based additives in this work are liquids at room temperature, with

different hydrocarbon chain lengths, for comparison with each other as well as the Ti

based organo-metallic liquid.)

152

• Bis(isopropylcyclopentadienyl) titanium dichloride C16H22Cl2Ti, is a powder, Ti-

based chloride additive used to compare the effect of a Cl-based to the oxide based

additives investigated.

• TTIP,

• and Nb2O5 were repeated as benchmarks and for comparison.

Table 1: A list of material and chemical used for this studies, indicating suppliers, state, purity and

chemical formula

Materials/ chemicals Formula Purity

(%)

Amount

(mol%)

Milling

time

(h)

State Supplier

Magnesium hydride MgH2 98 20 Powder ABCR

Niobium pentoxide Nb2O5 (20 nm) _ 1 2 powder NOVacentrix

Titanium isopropoxide Ti(OCH(CH3)2)4 97 1 2 Liquid Sigma-

Aldrich

Titanium(IV) methoxide Ti(OCH3)4 95 1 2 powder Sigma-

Aldrich

Titanium(IV) ethoxide Ti(OC2H5)4 97 0.5-

10

2 Liquid Sigma-

Aldrich

Vanadium(V)

oxytriisopropoxide

OV(OCH(CH3)2)3 _ 1 2 Liquid Sigma-

Aldrich

Vanadium(V) oxytripropoxide OV(OC3H7)3 98 1 2 Liquid Sigma-

Aldrich

Vanadium(V) oxytriethoxide OV(OC2H5)3 95 1 2 Liquid Sigma-

Aldrich

Bis(isopropylcyclopentadienyl

) titanium dichloride

C16H22Cl2Ti 95 1-2 1-10 powder Sigma-

Aldrich

Experimental

As-sourced MgH2 was pre-milled in a PQ-N04 planetary mill (Across International) for

20 h using a 50 ml zirconia vial and 10×10 mm zirconia balls with a ball-to-powder

ratio (bpr) of 7:1 and a vial filling factor [46] of 17.4%. Pre-milled MgH2 samples

153

(preMgH2) were prepared in 5 g batches, as per our previous studies [45, 47, 48]. The list

of additives studied, as well as chemical formula and source, is presented in Table 1.

Additives were subsequently milled with the preMgH2 (1 g sample, 6 ZrO2 balls, 600

rpm, 21:1 bpr, vial filling factor 7.7%).

For most additives, 1 mol% was added to the MgH2 and ball-milled for 2 h, as this has

been shown to be sufficient for Nb2O5 and TTIP [45, 49].

Bis(isopropylcyclopentadienyl) titanium dichloride samples were prepared with a range

of ball milling times (1- 10 h) and additive quantity in order to investigate the effect of

these parameters on the sorption kinetics. Different quantities of the Titanium (IV)

ethoxide additive were also used to determine the optimal amount. All materials were

handled in a dry Ar atmosphere MBraun glovebox with O2 and H2O levels below 1

ppm.

A modified custom Sieverts apparatus was used to perform the hydrogen cycling. The

as-prepared hydride samples started under dynamic vacuum, 3×10-6 mbar, before

initiating the TPD studies, where the samples were heated from room temperature to

400 °C at a ramp rate of 2 °C/min [45, 50]. To ensure complete desorption, the sample

temperature was kept at 400 °C for a further 2 h. Then the sample temperature was

reduced to 230 °C, and a single hydrogen pressure aliquot was introduced to reach an

initial pressure of 50 bar on the sample. The pressure of the sample was constantly

monitored by a Quartzdyne DL series pressure sensor with an overall accuracy of

0.015% FSR (340) bar.

Results and discussion:(1 mol%) Ti-based oxide additives milled for 2 h

Different Ti-based oxide additives were milled with preMgH2 in order to compare their

effect on the sorption kinetics of MgH2. Fig 1 shows the desorption kinetics of the Ti-

based oxides, where it can be seen that all additives have a positive effect on the

kinetics when compared to the preMgH2 without additive. The best performing additive

is TTIP confirming our previous studies [45]. This composite released hydrogen in the

temperature range (120-310 ºC) with a maximum peak at 250 ºC. Ti-ethoxide was

slightly less effective than TTIP, starting desorption at a later temperature (166 ºC) and

peaking at 270 ºC. The Ti-methoxide composite sample required a higher temperature

before hydrogen was released (200 ºC), and for all the additives the desorption was

complete by 310 ºC (ie before the ball-milled MgH2 without additive reached its peak in

the hydrogen release). Both TTIP and Ti-ethoxide are liquid at room temperature while

154

Ti-methoxide is in powder form, lending some support to the suggestion that liquids are

better dispersed during milling, helping to improve the sorption kinetics of MgH2.

Fig 1:Desorption kinetics of Pre

MgH2 + 1 mol% Ti based-oxides milled for 2h

under Ar.

The absorption kinetics of the Ti-based oxides are shown in Fig 2. It can be seen that all

composite samples reached their maximum capacity in the first 10 min. Ti-methoixde

and Ti-ethoxide additives have the highest uptake ~6.5 wt%, however the latter was

faster. The preMgH2 without additives has the highest hydrogen uptake but the slowest

kinetic rate. The TTIP composite sample has the lowest uptake which could be due to

its higher molecular weight compared to the other additives. These results show that the

additives make a significant enhancement to the absorption kinetics of MgH2, as well as

the desorption kinetics. Due to the fast kinetics and high capacity, the Ti-ethoxide

composite sample was further investigated to determine the optimal amount of additive.

155

Fig 2: Absorption kinetics of Pre

MgH2 + 1 mol% Ti based-oxide milled for 2 h

under Ar.

Different amount of Ti-ethoxide milled for 2 h

The effect of different amounts (0.5- 10 mol%) of Ti-ethoxide additive on the

desorption kinetics is shown in Fig 3. It can be seen that desorption kinetics differ

significantly depending on the amount added and an optimal amount was found to be 1

mol%. Decreasing the amount of Ti-ethoxide to 0.5 mol% yielded poorer results; with

broadening of the peak shape indicating the release of hydrogen by the composite

sample over a greater range of temperature. For the higher quantities, 2 and 10 mol%,

there appears to be two overlapping processes where the second peak of 10 mol% is

matches the pre-milled MgH2 peak, suggesting a portion of the composite sample is not

affected by the additive, even though there is more of it.

156

Fig 3: Desorption kinetics of Pre

MgH2 + X mol% Ti-ethoxide milled for 2h under

Ar

Fig 4 shows the effect of different amounts of Ti-ethoxide on the absorption kinetics

together with pre-milled MgH2 without additives for comparison. The 0.5 mol% Ti-

ethoxide sample has the highest hydrogen capacity and fastest kinetic rate, whereas the

10 mol% has the lowest capacity and slowest kinetics – significantly worse than the

ball-milled pure MgH2. There is little difference between the 0.5 mol% and 1 mol%

composite samples, other than the final capacity, which is simply due to the amount of

non-absorbing additive used. Overall, these comparative results match the

corresponding desorption kinetics.

157


MgH2 + X mol% Ti-ethoxide milled for 2h under

Ar

1 mol% V-based oxide additives milled for 2 h

In order to further investigate the effect of liquid oxides on the kinetic enhancement, V-

based liquid oxide additives were milled under the same conditions to draw a

comparison between the Ti- and V-based additives. Fig 5 shows the desorption kinetics

of 1 mol% V-based organo-metallic oxide composite samples milled for 2 h under Ar.

All samples showed similar kinetics, covering approximately the same temperature

range. The broadening of the peak shape indicates that hydrogen is released over larger

temperature range, in the case of V-oxytriisopropoxide there appear to be overlapping

processes, evidenced by a small peak/plateau at about 250 ºC. While the V-oxide liquid

additives improved the desorption kinetics when compared to the preMgH2 without

additives, they did not achieve the same desorption enhancement as TTIP or Nb2O5

which is in agreement with the literature [51].

158


MgH2 + 1 mol% V based-oxide milled for 2h under

Ar. (Scaled by a factor 3)

The effect of the V-based oxide additives on the absorption kinetics of MgH2 is shown

in Fig 6. All additives show fast kinetics, both V-oxytriethoxide and V-oxytripropoxide

reached their maximum capacity in the first minute, whereas V-oxytriisoprpoxide

reached its maximum after 8min but had the highest capacity of 6.68 wt%. While the

capacity is largely determined by the weight of the 1 mol% of the additive, since the

composite samples were close to fully absorbed after 20 minutes, the slower result for

V-oxytriisopropoxide confirms that additives affect the absorption and desorption

process differently.

159

Fig 6:Absorption kinetics of Pre

MgH2 + 1 mol% V-oxide milled for 2 h under Ar.

BisTiCl2 (Bis(isopropylcyclopentadienyl) titanium dichloride) additive

To compare the effect of a Cl-based additive to the oxide based additives investigated,

BisTiCl2 was also studied. This additive contains chlorine, where various halides have

been shown to kinetically enhance MgH2, but no oxygen. Fig 7, shows the effect of 1

mol% BisTiCl2 milled for different times (1 – 10 h) under Ar, and 2 mol% milled for 5

h, in order to investigate the effect of different amounts of additive and milling time.

The best desorption kinetics were observed for the 1 mol% of BisTiCl2 sample milled

for 1 h, with the hydrogen released in the temperature range (200 – 280 ºC), whereas the

2 mol% milled for 5 h had poor kinetics and a very wide peak over a large temperature

range. In terms of when the desorption finished, it did not show a significant

improvement over the un-modified preMgH2. Overall, the longer milling times appear to

have significantly reduced the kinetic enhancement, pointing to a possible

decomposition of the additive during milling.

160


MgH2 + X mol% BisTiCl milled for different times

under Ar

Fig 8 shows the absorption kinetics of the BisTiCl composite samples milled for

different times. All the composite samples have significantly poorer absorption kinetics

than the Pre MgH2 without additives, indicating the additive has inhibiting absorption

rather than facilitating it. The sample with 1 mol% milled for 10 h is the best of the

composite samples; the shorter milling times produce poorer results. The composite

sample with 1 mol% milled for 1 h, which had the best desorption kinetics, had the

slowest absorption kinetics and the least hydrogen uptake.

161


MgH2 + X mol% BisTiCl milled for different times

under Ar

Comparison of all additives

In order to make comparisons with the literature and to put the improvement of the

desorption kinetics of organo-metallic liquid oxides milled with MgH2 in perspective,

the desorption kinetics of 1 mol% of all additives used in this study milled for 2 h,

except for BisTiCl which was milled for 1 h, are plotted in Fig 9. The best composite

sample for desorption was the BisTiCl2, a powder additive for which the desorption was

finished by 270 ºC. This was followed by the TTIP and Nb2O5 composite samples.

162

Fig 9: Desorption kinetics of 1 mol% all additives, milled for 2 h (except BisTiCl 1

h) with MgH2

Conclusion

The effect of various amounts of different liquid oxide (Ti-based and V-based)

additives milled with MgH2, as well as two Ti-based powder additives, one an oxide

and the other a chloride, was systematically surveyed by measuring the desorption and

absorption kinetics. TPD studies were performed under vacuum followed by a single

absorption pressure aliquot at 50 bar H2 at 230 ºC. The benchmark Nb2O5 additive was

used to compare the effect of these additives on MgH2, as well as the recently studied

TTIP. It was found that when comparing the effect of the Ti and V based oxide

additives on the kinetics of MgH2, that Ti-based oxide additives improved the desorption

kinetics whereas the V-based oxide have a positive effect on the absorption kinetics.

The 1 mol% BisTiCl2 composite sample, milled for 1 h, which was used as a non-oxide

additive to compare with the oxide samples, had the best desorption kinetics when

compared to all the additives surveyed in this work, but, interestingly, it had the slowest

absorption and lowest hydrogen uptake, lower than the undoped MgH2, demonstrating

the different effect of additives on the two different sorption processes.

Part of this investigation was to examine the difference between liquid and powder

additives, and, generally the liquid additives perform well, as expected if the

163

distribution of additive through the magnesium/hydride is an important consideration.

However, while a number of additives compare favourably, or even equally, with

Nb2O5, a powder, they are not significantly better, questioning whether the liquid state

is of benefit, or whether Nb2O5 has other characteristics that outweigh poorer

distribution due to being in powder form.

The sole chloride sample, with excellent enhancement of desorption and significant

inhibition of absorption, provides further evidence that additives can influence the

different absorption and desorption processes in different ways, suggesting the possible

use of two or more additives that complement each other and maximize the net effect on

the sorption kinetics.

Acknowledgements

K.A acknowledges receipt of an Australian Government Research Training Program

Scholarship.

164

References:

1. Züttel A. Materials for hydrogen storage. Materials today. 2003;6(9):24-33.

2. Grochala W and Edwards PP. Thermal decomposition of the non-interstitial

hydrides for the storage and production of hydrogen. Chemical Reviews.

2004;104(3):1283-316;10.1021/cr030691s




4. Pukazhselvan D, Perez J, Nasani N, Bdikin I, Kovalevsky AV and Fagg DP.

Formation of Mgx Nby Ox+y through the Mechanochemical Reaction of MgH2

and Nb2O5, and Its Effect on the Hydrogen-Storage Behavior of MgH2.

Chemphyschem. 2016;17(1):178-83;10.1002/cphc.201500620.



2004;364(1):242-6; 10.1016/s0925-8388(03)00530-9.



25;10.1016/s0925-8388(99)00073-0.



106;10.1016/j.jpcs.2014.06.014.




9. Huot J, Liang G and Schulz R. Mechanically alloyed metal hydride systems.

Applied Physics A. 2001;72(2):187-95; 10.1007/s003390100772

10. Zhu M, Lu Y, Ouyang L and Wang H. Thermodynamic Tuning of Mg-Based

Hydrogen Storage Alloys: A Review. Materials. 2013;6(10):4654-74;

10.3390/ma6104654



2007;32(9):1121-40;10.1016/j.ijhydene.2006.11.022.

165

12. Reule H, Hirscher M, Weißhardt A and Kronmüller H. Hydrogen desorption

properties of mechanically alloyed MgH2 composite materials. Journal of Alloys

and Compounds. 2000;305(1-2):246-52;10.1016/s0925-8388(00)00710-6.

13. Jain I, Lal C and Jain A. Hydrogen storage in Mg: a most promising material.

International Journal of Hydrogen Energy. 2010;35(10):5133-44;

10.1016/j.ijhydene.2009.08.088

14. Crivello J-C, Dam B, Denys R, Dornheim M, Grant D, Huot J, et al. Review of

magnesium hydride-based materials: development and optimisation. Applied

Physics A. 2016;122(2):1-20; 10.1007/s00339-016-9602-0.

15. Crivello J-C, Denys R, Dornheim M, Felderhoff M, Grant D, Huot J, et al. Mg-

based compounds for hydrogen and energy storage. Applied Physics A.

2016;122(2):1-17; 10.1007/s00339-016-9601-1.




17. Bobet J. Study of Mg-M (M=Co, Ni and Fe) mixture elaborated by reactive

mechanical alloying — hydrogen sorption properties. International Journal of

Hydrogen Energy. 2000;25(10):987-96;10.1016/s0360-3199(00)00002-1.

18. Varin RA, Li S, Chiu C, Guo L, Morozova O, Khomenko T, et al.

Nanocrystalline and non-crystalline hydrides synthesized by controlled reactive

mechanical alloying/milling of Mg and Mg–X (X = Fe, Co, Mn, B) systems.

Journal of Alloys and Compounds. 2005;404-406:494-

8;10.1016/j.jallcom.2004.12.176.

19. Kattner U and Massalski T. Binary alloy phase diagrams. ASM International,

Materials Park, OH. 1990;147.


Cycling of Niobium and Vanadium Catalyzed Nanostructured Magnesium.

Journal of the American Chemical Society. 2005;127(41):14348-

54;10.1021/ja051508a.



Compounds. 2001;315(1-2):237-42;10.1016/s0925-8388(00)01284-6.



7;10.1016/s1359-6462(03)00259-8.

166

23. Jung KS, Kim DH, Lee EY and Lee KS. Hydrogen sorption of magnesium

hydride doped with nano-sized TiO2. Catalysis Today. 2007;120(3-4):270-

5;10.1016/j.cattod.2006.09.028.


ray absorption spectroscopic study on valence state and local atomic structure of

transition metal oxides doped in MgH2. The Journal of Physical Chemistry C.

2009;113(30):13450-5; 10.1021/jp901859f.

25. Du A, Smith SC, Yao X, Sun C, Li L and Lu G. The role of V2O5 on the

dehydrogenation and hydrogenation in magnesium hydride: an ab initio study.

Applied Physics Letters. 2008;92(16); 10.1063/1.2916828.


Fast hydrogen sorption from MgH 2 –VO 2 (B) composite materials. Journal of





34;10.1016/j.ijhydene.2016.12.072.




29. Barkhordarian G, Klassen T and Bormann R. Catalytic Mechanism of

Transition-Metal Compounds on Mg Hydrogen Sorption Reaction. J Phys Chem

B. 2006;110(Copyright (C) 2015 American Chemical Society (ACS). All Rights

Reserved.):11020-4;10.1021/jp0541563.

30. Manchester F and Khatamian D, editors. Mechanisms for activation of

intermetallic hydrogen absorbers. Materials Science Forum; 1988: Trans Tech

Publ.






41;10.1016/j.jallcom.2005.02.043.



167


10;10.1016/j.actamat.2005.08.024.

34. Porcu M, Petford-Long AK and Sykes JM. TEM studies of Nb2O5 catalyst in

ball-milled MgH2 for hydrogen storage. J Alloys Compd. 2008;453(Copyright

(C) 2015 American Chemical Society (ACS). All Rights Reserved.):341-

6;10.1016/j.jallcom.2006.11.147.

35. Aurora A, Mancini MR, Gattia DM, Montone A, Pilloni L, Todini E, et al.

Microstructural and kinetic investigation of hydrogen sorption reaction of

MgH2/Nb2O5 nanopowders. Mater Manuf Processes. 2009;24(Copyright (C)

2015 American Chemical Society (ACS). All Rights Reserved.):1058-

63;10.1080/10426910903022361.





15;10.1016/j.ijhydene.2016.04.029.

37. Wang Y, Li L, An C, Wang Y, Chen C, Jiao L, et al. Facile synthesis of TiN

decorated graphene and its enhanced catalytic effects on dehydrogenation

performance of magnesium hydride. Nanoscale. 2014;6(12):6684-91;

10.1039/C4NR00474D.

38. Jung KS, Lee EY and Lee KS. Catalytic effects of metal oxide on hydrogen

absorption of magnesium metal hydride. Journal of Alloys and Compounds.

2006;421(1-2):179-84;10.1016/j.jallcom.2005.09.085.




2015;30(S1):S9-S15; 10.1017/s0885715615000056.

40. Wang P, Wang AM, Zhang HF, Ding BZ and Hu ZQ. Hydrogenation

characteristics of Mg–TiO2 (rutile) composite. Journal of Alloys and

Compounds. 2000;313(1-2):218-23;10.1016/s0925-8388(00)01188-9.





168




8;10.1016/j.jallcom.2009.10.199.

43. Pukazhselvan D, Nasani N, Yang T, Ramasamy D, Shaula A and Fagg DP.

Chemically transformed additive phases in Mg2TiO4 and MgTiO3 loaded

hydrogen storage system MgH2. Applied Surface Science. 2018; 99-104;

10.1016/j.apsusc.2018.04.052.




11;10.1016/j.ijhydene.2016.01.095.




34;10.1016/j.ijhydene.2016.12.072.





47. Alsabawi K, Webb TA, Gray EM and Webb CJ. The effect of C60 additive on



48. Alsabawi K, Gray EM and Webb CJ. The effect of ball-milling gas environment

on the sorption kinetics of MgH2 with/without additives for hydrogen storage.


80;10.1016/j.ijhydene.2018.12.026.

49. Isobe S, Umeda A, Wakasugi T, Ma T, Yamagami R, Hino S, et al. Microscopic

Study on Hydrogenation Mechanism of MgH2 Catalyzed by Nb2O5. Materials

Transactions. 2014; 1175-1178 ;10.2320/matertrans.MG201417.

50. Alsabawi K, Webb TA, Gray EM and Webb CJ. The effect of C 60 additive on



169

51. Bellemare J and Huot J. Hydrogen storage properties of cold rolled magnesium

hydrides with oxides catalysts. Journal of Alloys and Compounds.

2012;512(1):33-8;10.1016/j.jallcom.2011.08.085.

170

CHAPTER EIGHT

8. Summary and Conclusion

Hydrogen is considered to be a readily available energy vector that can be produced

from renewable energy sources and used to generate clean, affordable energy [1, 2].

One of the main barriers in establishing a strong hydrogen economy is a safe,

economical and practical storage method. To eliminate this barrier, new designs for

hydrogen storage materials must consider weight, volume, safety and cost as well as

good reversibility for release and uptake of hydrogen in practical applications. Several

methods have been utilised or suggested to store hydrogen, including compressed and

liquid hydrogen, liquids such as ammonia and methylcyclohexane, and solid state

storage materials that can readily take up and release large quantities of hydrogen. Solid

state materials, including simple metal and intermetallic hydrides, complex hydrides,

metal organic materials, such as MOFs, and adsorptive materials such as carbons,

zeolites and polymers, are all capable of storing hydrogen with different volumetric and

gravimetric densities, but none of them fulfil all the requirements for facilitating

hydrogen storage for applications such as hydrogen fuel cell vehicles or smoothing

supply/demand requirements for intermittent renewable energy generators.

This project focussed on the synthesis and characterisation of composite materials to

investigate the kinetic properties of the Mg-H system, by milling MgH2 with different

additives. Magnesium is a promising hydrogen storage material containing 7.6 wt% of

hydrogen as magnesium dihydride, MgH2. Practical applications of MgH2 for hydrogen

storage are limited by poor absorption/desorption kinetics and the high dissociation

temperature of MgH2, due to the thermodynamic stability of the hydride (approx. 75

kJ/mol.H2). Additives in small amounts have been shown to have a considerable effect

on the kinetics of absorption and desorption of hydrogen, although the effect on the

thermodynamics is minimal or negligible. The additives chosen for this work have not

previously been investigated for their effect on hydrogen sorption of MgH2. However,

the choice of additive was based on results of previous studies.

Many studies have shown the beneficial effect of transition metal oxides as additives for

the enhancement of hydrogen sorption kinetics e.g., Nb2O5 and V2O5. The dramatic

kinetic improvement by addition of Nb2O5 was found in 2003 [3] and this material is

171

now considered a benchmark for comparing additives. Following an exploration of

different Nb [3-6] and V oxides [7-9] it was suggested that the higher valence oxides

have a greater effect on the sorption kinetics [10, 11]. Transition metal halides have also

been successfully used as additives for improving the kinetics e.g., VCl3 [12]and TiF3

[13] as well as many other materials [14].

Finally, an important aspect of the synthesis of composite samples is ensuring a good

interaction between the hydrogen sorption material and the additive. The use of ball-

milling is common in studies of the effect of additives on MgH2, and this can decrease

the size of the MgH2 and additive particles, creating more surface area for interaction,

and also disperses the additive throughout the MgH2. A liquid may more readily

disperse throughout the MgH2 powder, and for this reason a number of liquid (at RT)

additives were chosen for this work.

Since Nb2O5 and V2O5 have already been reported in the literature, organo-metallic

materials based on oxides of transition metals offer another set of additives worthy of

investigation. While the role of oxygen in MgH2 is still unclear, some of the organo-

metallic oxides used contain more oxygen per transition metal atom than Nb2O5. In

addition, some of these are liquid at RT, providing another set of comparisons - between

that of powder and liquid additives. The additives chosen for this work were transition

metal organo-metallic oxides in liquid and powder forms as well as a Ti-based organo-

metallic chloride. The transition metals were Ti and V. Nb2O5 was used as the

benchmark for comparisons.

The composite sample materials were prepared through ball milling under different

conditions to determine the optimal parameters. Ball milling time varied from 0.5h to 20

h depending on the additive used, to explore the effect of milling time on the composite.

In addition, composite samples were ball milled under different gases (H2 and Ar), in an

attempt to understand the effect of the milling environment on the sorption kinetics.

The positive effect of ball milling with additives on enhancing the temperature onset

and rates of hydrogen sorption for the MgH2 system has been widely studied, and the

studies reported here confirm the previous articles. This work resulted in a new set of

information, extending the knowledge regarding transition metal oxide additives for

MgH2, including the use of novel additives and may help to understand some of the

mechanisms by which the additives influence the sorption kinetics of MgH2.

Comparisons of Ti and V oxides, Nb oxide with Ti and V organo-metallic oxides, liquid

172

v solid additives and ball-mill gas environments all contribute to the knowledge in this

area.

Sievert apparatus 300 bar:

In chapter 3, a detailed description was given of the design and implementation of a

Sieverts type apparatus for determining hydrogen uptake and release at pressures up to

340 bar and temperatures from 77-773 K. An existing instrument was re-developed

during this project to enhance the temperature stability, improve the ratio of reference

and cell volumes and add TPD and TDS capabilities, in order to perform all the required

kinetic and capacity measurements in a single environment.

As part of the re-development, characterisation and verification of the apparatus was

performed. Multiple volume calibrations (full volume calibration, empty cell isotherm

and divided-volume calibration) were made, to fully characterise the instrument

volumes. The performance of the instrument for isotherm measurements and

calculations was validated by running standard materials such as LaNi5 and a low

density material (FiltraSorb-400). Details of the instrument design and validation were

written as a journal paper and submitted to IJHE.

Milling MgH2 with additives under different environment

In this work MgH2 was milled with additives under Ar (1 bar) and H2 (10bar)

atmospheres (Chapter 6). The as-sourced MgH2 was milled first for 20 h without

additives and then milled for another 2 h with the addition of 1-2 mol% of the additives

TTIP, Nb2O5 and C60, while varying the gas environment for both ball-milling stages.

TPD experiments were carried out to determine the desorption kinetics, which was then

followed by a single pressure aliquot of hydrogen for the absorption kinetics

measurement. The milled composites revealed that the milling environment had little to

no effect on the desorption kinetics, although, the Ar milled TTIP composite desorbed

most of the hydrogen at a slightly lower temperature than that milled under H2.

Importantly, significant differences in the absorption kinetics were observed for Nb2O5

sample milled under Ar then H2. Both Nb2O5 and C60 composites samples milled under

Ar had superior absorption than those milled under H2. Conversely, the TTIP composite

sample milled under Ar showed poorer absorption kinetics than that milled under H2.

Based on these findings, the milling environment has been shown to have a significant

effect on the sorption kinetics, but there is no easy formula to determine which gas

173

produces the best results as it was dependent on the additives used. This work has been

published in IJHE [15].

Milling MgH2 with TTIP

A systematic survey of the effect of ball milling MgH2 with TTIP (organo-metallic

liquid) additive on the absorption and desorption kinetics was conducted. TTIP has been

used previously as an additive to improve the sorption kinetics of LiBH3-MgH2 system,

but had not been used before to improve the kinetics of MgH2 alone. Chapter 5 shows a

comparison between the effect of TTIP and the benchmark additive Nb2O5 on the

sorption kinetics of MgH2. Various amounts (0.5 – 2 mol%) of both additives were

milled with MgH2 for different times (0.5- 10 h), as well as hand mixing 1 mol% TTIP

with the pre-milled MgH2. A TPD experiment followed by a single pressure aliquot of

hydrogen were performed to measure the desorption and absorption kinetics,

respectively. TTIP was found to be equally effective as Nb2O5 as an additive and 0.5

mol% of TTIP was sufficient to enhance the kinetics of MgH2. In addition, the hand

mixed TTIP composite sample produced the best desorption results, but had poorer

absorption compared to the milled samples. The underlying mechanism by which the

additive TTIP enhances the sorption kinetics is still unclear. One of the possible

contributing reasons proffered is that the liquid state of TTIP above 290 K facilitates

good dispersion through the MgH2. This work has been published in the International

Journal of Hydrogen Energy [16]

Kinetics of MgH2 with different oxide/non-oxide Ti/V based liquid and powder

additives

To gain further information which might improve the understanding of the role of

transition metal oxides in enhancing the sorption kinetics of MgH2, a number of

selected transition metal organo-metallic oxide (Ti and V) additives as well as a

titanium-based organo-metallic chloride (C16H22Cl2Ti) were investigated and compared

to Nb2O5. In order to follow up the favourable results of using the liquid TTIP additive

(chapter 5), as well as to compare the relative effect of liquid and solid additives, the

majority of these additives investigated were also liquid at room temperature.

From the kinetics measurements, a number of conclusions can be drawn from this study.

For Ti-based additives it was found that TTIP and Ti-ethoxide which are liquid at room

temperature had greater improvement of the desorption kinetics compared to Ti-

174

methoxide which is in powder form. The V-based oxides, all in liquid form, were found

to improve the kinetics compared to the preMgH2 without additives and all showed

similar enhancement of the desorption kinetics, but their absorption kinetics differed.

Finally, the Ti-chloride additive (1 mol% of BisTiCl2 in powder form milled for 1 h)

had the best desorption kinetics compared to Ti-based and V-based oxides, but,

interestingly, showed significantly poorer absorption kinetics than even the Pre MgH2

without additives. Overall the Ti-based oxides achieved better desorption enhancement

than the V-based oxides, whereas, V-based oxides had faster absorption and higher

uptake.

8.1. Future work

In this work different composite materials were synthesised to investigate their effect on

the sorption kinetics of MgH2. Different transition metal-based additives were chosen

(organo-metallic liquid/solid oxides, non-oxides and bench mark Nb2O5 for

comparison) and it was clear that each composite synthesised shared some common and

some different effects on the sorption kinetics of MgH2. In total, all the additives used in

this work (except for the Ti-chloride), made a significant improvement to the sorption

kinetics when compared to MgH2 without additives. However, to determine the

underlying mechanism by which these additives improve the kinetics needs more

research. Even with the benchmark Nb2O5, which has been extensively studied for more

than a decade, the debate on the exact mechanism is still open. Similarly, some of the

experimental work can be questioned and there are contradictions from one study to

another. Some studies have found that Nb2O5 does not undergo any structural

transformation during milling or cycling and only acts as a catalyst, and others have

found that Nb2O5 reduces during milling and/or further reduces during cycling with a

new ternary oxide phase MgNbxOy observed with XRD.

Generally, the existence of a thin surface oxide layer (MgO) covering the magnesium

metal particles, slows the kinetics of MgH2, due to poor diffusion rates of hydrogen

through MgH2, and yet some studies showed that milling MgO with MgH2 showed an

improvement in the hydrogen sorption of MgH2 by acting as dispersive agent during

milling reducing the particle size. Considering the difficulty of completely eliminating

oxygen contamination, and the excellent kinetic enhancement by various oxides, the

role of oxygen in the Mg-H system may be critical to the understanding of how

additives work.

175

The kinetics of absorption and desorption have different pathways with different rate

limiting steps– so different additives may perform differently for absorption and

desorption. An interesting result presented in this work, when Ti-chloride additive, a

solid, was compared with other transition metal oxides, it showed the best desorption

kinetics, but had extremely poor absorption results which was even slower than the ball-

milled MgH2 without additive. This outcome supports those studies that report that

some additives enhanced absorption and desorption differently. Future work might

combine different additives to optimally enhance the sorption kinetics of MgH2.

The search for a hydrogen material that meets the target capacities and practical

operating conditions, including moderate temperatures, such as could be integrated with

a fuel cell for on board vehicle applications is still ongoing. The materials investigated

in this study do not meet the optimal targets, but the studies provide additional

information which may assist in further understanding of their properties and their

effect on the sorption kinetics of MgH2.

176

8.2. Reference

1. Schlapbach L and Zuttel A. Hydrogen-storage materials for mobile applications.

Nature. 2001;414(6861):353-8;10.1038/35104634.





7;10.1016/s1359-6462(03)00259-8.

4. Huot J, Pelletier JF, Lurio LB, Sutton M and Schulz R. Investigation of

dehydrogenation mechanism of MgH2–Nb nanocomposites. Journal of Alloys

and Compounds. 2003;348(1-2):319-24;10.1016/s0925-8388(02)00839-3.

5. Pelletier JF, Huot J, Sutton M, Schulz R, Sandy AR, Lurio LB, et al. Hydrogen

desorption mechanism in MgH2−Nb nanocomposites. Physical Review B.

2001;63(5):052103;10.1103/PhysRevB.63.052103.

6. Bhat VV, Rougier A, Aymard L, Darok X, Nazri G and Tarascon JM. Catalytic

activity of oxides and halides on hydrogen storage of MgH2. Journal of Power




Compounds. 2001;315(1-2):237-42;10.1016/s0925-8388(00)01284-6.





Fast hydrogen sorption from MgH2 –VO2 (B) composite materials. Journal of



Stroppa DG, et al. Evolution of reduced Ti containing phase(s) in MgH2 /TiO2

system and its effect on the hydrogen storage behavior of MgH2. Journal of


177



43;10.3390/catal2030330.




2015;30(S1):S9-S15;10.1017/s0885715615000056.

13. Ma LP, Kang XD, Dai HB, Liang Y, Fang ZZ, Wang PJ, et al. Superior catalytic

effect of TiF3 over TiCl3 in improving the hydrogen sorption kinetics of MgH2:

Catalytic role of fluorine anion. Acta Materialia. 2009;57(7):2250-

8;10.1016/j.actamat.2009.01.025.



106;10.1016/j.jpcs.2014.06.014.

15. Alsabawi K, Gray EM and Webb CJ. A Sieverts apparatus with TPD/TDS for

accurate, high-pressure hydrogen uptake and kinetics measurements.

International Journal of Hydrogen Energy. 2019;Submitted.




34;10.1016/j.ijhydene.2016.12.072.

The Effect of Metal Oxides Additives on the Absorption ...

Documents

Transcript of The Effect of Metal Oxides Additives on the Absorption ...