Hybrid Magnesium Based Materials for Hydrogen Energy Storage

by Eki Jaya Sasmita Setijadi

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Chemical Engineering

The University of New South Wales

Sydney, Australia

2014 PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Setijadi

First name: Eki Other name/s: Jaya Sasmita

Abbreviation for degree as given in the University calendar: PhD

School: Chemical Engineering Faculty: Engineering

Title: Hybrid Magnesium Based Materials for Hydrogen Energy Storage

Abstract 350 words maximum: (PLEASE TYPE)

Nanostructuring metal hydride has been identified as a potential approach to overcome kinetics and thermodynamic limitations due to the large surface area and high surface energy of nanomaterials. However, in practice the synthesis of such nanosized materials with controlled properties is a real challenge. In particular, the high reactivity of magnesium - a promising material for hydrogen storage - challenges its synthesis at the nanoscale. Hence, this thesis aims to explore different strategies based on wet synthesis methods to synthesize and stabilize magnesium hydride (MgHz) nanoparticles.

Thermal decomposition of organomagnesium is a promising method to obtain magnesium nanoparticles in simple step and with high yield. Yet, the resulting decomposition products would depend on the precursors, conditions, and medium during decomposition. Di-n-butylmagnesium is the best precursor investigated in the study. The mediums also determined the physical properties of MgHz from di-n-butylmagnesium; hydrogenolysis in dry solid conditions led to materials capable to store 7.1 wt% hydrogen capacity with fast desorption kinetics at 300 •c. Similar kinetics also being observed in the material obtained from hydrogenolysis of di-n-butylmagnesium in cyclohexane but with only 5.5 wt% capacity due to more hydrocarbon residue from solvent. Other promising method is through catalytical hydrogenation of MgAnthracene.3THF complex which could produce MgH2 nanoparticles in high yield and with good economical value.

Despite having kinetic improvements, the thermodynamic limitations are still causing high temperature requirement for hydrogen desorption from these materials. Indeed some theoretical studies showed that significant destabilization can only occur when the nanoparticles sizes are less than 5 nm. The methods were further extended by introducing other compounds such as surfactants and polymers to obtain much smaller size nanoparticles. Herein, the hybrid magnesium polystyrene nanocomposite was successfully synthesised and proven to give protection against oxidations. However, the size did not become smaller with polystyrene but we found the thermodynamic could be altered by the functional groups on the polystyrene. To achieve smaller particle size, polystyrene wtth different nanostructures such as star, dendrimers, and hyperbranched were synthesised and used as templates for limiting the particle growth.

Declaration relating to dlspoeltion of project thnlsldlssertatlon

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

...... ~ .~ . I. .~. .!..'?:. 9.~ f ...... / -::~ :1 . ~·~Witness· ··~ · · · Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS COPYRK3HTSTATBMENT

'I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in al tonns of media, now or here after known, subject to the provisions of the Copyright Ad 1968. I retain al proprietary r1ghts, such as patent rights. I also retain the right to use in Mure works (such as artides or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; v.tlere permission has not been granted I have applied/WiU apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ··· ·· ·· ···~··· ·· · ····· · · ········ ··· ·· · ··· · ··· · · ··· ········· 2oc(" Date ...... -z.-~...... I~...... I ......

AUTHENllCITY STATEMENT

'I oeftlfy that the Ubrary deposit digital COf1f Is a direct equivalent of the final officialy approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.'

Signed ...... ~ ......

Date ...... ?:. "'::!.. / .. ~ .. !.!:-. .'?.~.f...... ORIGINALITY STATEMENT

'I hereby dedare that this submission Is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explldtly acknowledged in the thesis. I also dedare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or In style, presentation and linguistic expression Is acknowledged.'

Signed ...... ~ ...... 2.?> I~ I ·?.-o 15' Date

Abstract

ABSTRACT

– a promising material for hydrogen storage - challenges its synthesis at the nanoscale.

Hence, this thesis aims to explore different strategies based on wet synthesis methods to synthesize and stabilize magnesium hydride (MgH2) nanoparticles.

Thermal decomposition of organomagnesium is a promising method to obtain magnesium nanoparticles in simple steps and with high yield. Yet, the resulting decomposition products would depend on the precursors, conditions, and medium during decomposition. Di-n- butylmagnesium is the best precursor investigated in the study. The mediums also determined the physical properties of MgH2 from di-n-butylmagnesium; hydrogenolysis in dry solid conditions which led to materials capable of storing 7.1 wt% hydrogen capacity with fast desorption kinetics at 300 °C. Similar kinetics were also observed in the material obtained from hydrogenolysis of di-n-butylmagnesium in cyclohexane but with only 5.5 wt% capacity due to more hydrocarbon residue from solvent. Another promising method is through catalytical hydrogenation of MgAnthracene.3THF complex which could produce

MgH2 nanoparticles in high yield and with good economical value.

Despite having kinetic improvements, the thermodynamic limitations still cause a high temperature requirement for hydrogen desorption from these materials. Indeed some theoretical studies showed that significant destabilization can only occur when the nanoparticles sizes are less than 5 nm. The methods were further extended by introducing

Abstract

other compounds such as surfactants and polymers to obtain much smaller sized nanoparticles. Herein, the hybrid magnesium polystyrene nanocomposite was successfully synthesised and proven to give protection against oxidation. However, the size did not become smaller with polystyrene but we found the thermodynamic could be altered by functional groups on the polystyrene. To achieve smaller particle size, polystyrene with different nanostructures such as star, dendrimers, and hyperbranched were synthesised and used as templates for limiting the particle growth.

Acknowledgements

ACKNOWLEDGEMENTS

First and foremost, praises and thanks to the God, the Almighty, for His showers of blessings throughout my research work to complete the research successfully.

Looking back, five years ago when I just finished my undergraduate school, I never thought that I would end up doing a PhD or stayed in the same university for almost another five years. But here I am, writing one last section left on my thesis after more than three and half years’ worth of works and dedication of research. But throughout this journey, I have been given opportunity to grow and develop myself scientifically and personally. It would not have been obtained without the support of many people, who I like to address specifically in this acknowledgements. My apologies to those not specifically named.

I would like to give the utmost gratitude to A/Prof. Kondo-Francois Aguey-Zinsou for his supervision. His passion, guidance, discipline, and work ethics have been exemplary and indispensable to my growth as a scientist and a person on this span of years. I’m really grateful for his tremendous supports as I realised that it is quite rare to see such huge dedication to make sure students are prepared for whatever the next step in their journeys may be. Also, I like to express my gratitude to A/Prof. Cyrille Boyer for being the second supervisor of this thesis. Without him and his recommendation I would not be able to undertake this thesis at all. Of course, his inputs and ideas throughout the thesis were valuable beyond expression.

I also owe thanks to the rest members of MERlin, it has been a great pleasure to be part of a growing group throughout this thesis. It started only three of us with my only senior in the group, Meganne Christian. I like to thank her for the help and friendliness especially during

vii

Acknowledgements

my early time of this thesis. Our trip to the Kyoto conference was memorable at least for me.

Then, Wei Liu joined not long after and I thanked him for his inputs and discussions about our research together. Lei Wang, I wanted to specially acknowledge him for his entertaining feats throughout the time I spent at work also our good after hour chats will surely be missed. I want to thank Daniel Rivero to become such a good friend in the group and introduced the trivia night to me. Also to Chao Xi and Yvonne especially for the last six months when they replaced my job to clean up the lab.

Besides the group members, I have met a lot of people with different backgrounds in the

Chemical Science building and they have given me personal or professional support throughout the years. Starting from the people in the offices (I moved offices three times in the span of this theis) who have been great to me, there was no drama or whatsoever. The technical and admin officers in the School of Chemical Engineering school helped me enormously too. I like to give special thanks the people who have assisted me and contributed some of the results presented in this thesis. They are Katie Levick for her guidance on using TEM instrument ever since I was an undergraduate, James Hook and

Aditya Rawal for the works on NMR analyses, Bill Gong for the XPS measurements, and Yu

Wang for the training on XRD instrument.

My life outside the thesis has been great and I really thankful that I have this other life to experience too. I’m especially grateful that I was a member of Bel Cantare family which I gained so many things such as stress relief. I had such good memory with the group including those special friends within the group which I truly cherish and spent a lot of time together until now. Also to few number of friends that I knew ever since I came to Sydney,

I’m glad to keep our friendship together and stayed in touch even just occasionally. I like to mention special thanks to Dr. Ridwan Setiawan who gave many advises and offering his ears patiently to listen my matters both personally and scientifically. Also to my housemate,

viii

Acknowledgements

roommate, and bedmate, Mr. Andy Setyawan who has been really supportive and gave me wise perspectives to the problems I have to overcome in life.

In this occasion, I want to especially thank my family who have shaped me into who I am now. My older brother, Eko Setijadi and his wife (and soon a nephew too!) I want to express my gratitude for his care as an older brother despite living far apart (especially for special incentives to ensure my wellbeing as a poor PhD student). My beloved sister, Odilia Rosalin who lives with me throughout this thesis so we became really close personally. Also who would have thought that our 7 years apart could possibly allow us to go and work in the same lab together! going to the same uni together and on the same field. This also bring the special thanks to my relatives those in Indonesia and in Sydney who made me feel belong to such big and warm heart family wherever I go. My parents deserve special mention here, because without their constant supports and unconditional love, the completion of this thesis would not be possible. I always remembered their wisdoms and advises that encouraged me to go forward each time. I love you both Mom and Dad.

The last but not least is dedicated to the most important person in my life right now, Jess.

Through her love, patience, support and unwavering belief in me, I have been able to complete this long journey. She has patiently endured many, many long hours alone while I worked on this thesis. We have laughed and cried, travelled and in the end, she has given me so many happy and beautiful memories throughout this journey. There are no words that can express my gratitude and appreciation for all you have done and been for me. Thank you with all my heart and soul.

“To accomplish great things we must not only act, but also dream; not only plan, but also

believe.” – Anatole France

List of Publications

LIST OF PUBLICATIONS

Journal Paper

1. Setijadi, E.J. , Boyer, C., Aguey-Zinsou, K.F. (2012), “Remarkable hydrogen storage

properties for nanocrystalline MgH2 synthesised by the hydrogenolysis of Grignard reagents” Physical Chemistry Chemical Physics 14 (32), 11386-11397

2. Setijadi, E.J. , Boyer, C., Aguey-Zinsou, K.F. (2013), “MgH2 with different morphologies synthesized by thermal hydrogenolysis method for enhanced hydrogen sorption” International Journal of Hydrogen Energy 38 (14), 5746-5757

3. Setijadi, E.J. , Boyer, C., Aguey-Zinsou, K.F. (2014), “Switching the thermodynamics of

MgH2 nanoparticles through polystyrene stabilisation and oxidation” RSC Advances 4 (75), 39934-39940

4. Liu, W. , Setijadi, E.J. , Aguey-Zinsou, K.F. (2014), “Tuning the Thermodynamic Properties

of MgH2 at the Nanoscale via a Catalyst or Destabilizing Element Coating Strategy” J. Phys. Chem. C, 2014, 118 (48), pp 27781–27792

5. Setijadi, E.J., Boyer, C., Aguey-Zinsou, K.F. (2015), “Substituent electronic effects on

nanocrystalline MgH2 for enhanced hydrogen sorption”. In preparation

6. Setijadi, E.J., Aguey-Zinsou, K.F. (2015), “ Layered MoS2 nanosheets for wormlike MgH2 nanostructures formation”. In preparation

List of Abbreviations

LIST OF ABBREVIATIONS

BET Brunauer-Emmet-Teller (surface area measurement) DOE US Department of Energy DSC Differential Scanning Calorimetry EDS Energy Dispersive X-Ray Spectroscopy GPC Gel Permeation Chromatography HPS Hyperbranched Polystyrene ICP-OES Inductively Coupled Plasma-Optical Emission Spectroscopy

MgA.(THF)3 Magnesium Anthracene Complex MS Mass Spectrophotometry MW Molecular Weight NMR Nuclear Magnetic Resonance PCI Pressure Composition Isotherm PDI Polydispersity Index PEM Proton Exchange Membrane PMMA Poly Methyl Methacrylate PST Polystyrene

PSTN Polystyrene N3 RAFT Reversible addition−fragmentation chain-transfer S(PS) Star Polystyrene TBAB Tetrabutyl Ammonium Chloride TEM Transmission Electron Microscopy TGA Thermogravimetric Analysis THF Tetrahydrofuran TPD Temperature Programmed Desorption VBA Vinyl Benzyl Ammonium XPS X-Ray Photoelectron Spectroscopy XRD X-Ray Diffraction

Table of Contents

TABLE OF CONTENTS

ABSTRACT ...... V

ACKNOWLEDGEMENTS ...... VII

LIST OF PUBLICATIONS ...... X

LIST OF ABBREVIATIONS ...... XI

TABLE OF CONTENTS ...... XII

LIST OF TABLES ...... XIV

LIST OF FIGURES ...... XVI

1 INTRODUCTION ...... 1

2 LITERATURE REVIEW ...... 4

2.1 BACKGROUND ...... 5

2.2 SOLID BASED MATERIALS FOR HYDROGEN STORAGE APPLICATION ...... 10

2.3 MGH2 DEVELOPMENT AND IMPROVEMENT FOR HYDROGEN STORAGE ...... 17

2.4 HYBRID MG-POLYMER SYNTHESIS AND STRATEGIES ...... 25

2.5 CONCLUSION ...... 32

2.6 REFERENCES ...... 33

3 EXPERIMENTAL AND CHARACTERISATION TECHNIQUES ...... 42

3.1 INTRODUCTION ...... 43

3.2 SYNTHESIS INSTRUMENTS AND PROCEDURES ...... 43

3.3 PHYSICAL AND STRUCTURAL CHARACTERISATION TECHNIQUES ...... 47

3.4 HYDROGEN STORAGE PROPERTIES CHARACTERISATION TECHNIQUES ...... 55

3.5 CONCLUSION ...... 63

3.6 REFERENCES ...... 64

xii

Table of Contents

4 NOVEL SYNTHESIS METHOD OF MGH2 NANOPARTICLES BY WET CHEMISTRY APPROACHES ...... 65

4.1 INTRODUCTION ...... 66

4.2 THERMAL DECOMPOSITION OF DIFFERENT GRIGNARD REAGENTS ...... 67

4.3 HYDROGENOLYSIS OF DI-N-BUTYLMAGNESIUM IN DIFFERENT ENVIRONMENTS ...... 86

4.4 INVESTIGATION OF MG PRODUCED FROM MGA.(THF)3 ...... 104

4.5 CONCLUSION ...... 120

4.6 REFERENCES ...... 122

5 HYBRID POLYSTYRENE-MGH2 NANOPARTICLES ...... 126

5.1 INTRODUCTION ...... 127

5.2 SWITCHING THE THERMODYNAMICS OF MGH2 NANOPARTICLES THROUGH POLYSTYRENE STABILISATION AND OXIDATION ...... 128

5.3 FUNCTIONAL GROUPS EFFECTS ON H2 DESORPTION BEHAVIOURS OF MGH2 NANOPARTICLES 140

5.4 POLYMER NANOSTRUCTURES AS SUPPORTS FOR MGH2 NANOPARTICLES ...... 159

5.5 CONCLUSION ...... 177

5.6 REFERENCES ...... 179

6 SUMMARY AND RECOMMENDATIONS ...... 184

xiii

List of Tables

LIST OF TABLES

Table 2.1.1 - Performances of different H2 storage systems and DOE targets26 ...... 8

Table 2.2.1 - Properties of selected metal hydrides 56 60 ...... 15

Table 2.2.2 - Selected complex hydrides with their gravimetric hydrogen storage capacity and desorption temperature 56 64 ...... 17

Table 4.2.1- Temperature used for the hydrogenolysis of the Grignard reagents and particle/crystallite sizes of the nanoparticles obtained...... 70

Table 4.2.2 - Possible decomposition paths for the selected Grignard reagents and associated mass loss ...... 76

Table 4.2.3 - Elemental surface composition (atomic percentage, %) as determined by XPS for the materials obtained after the decomposition of the selected Grignard under Ar or H2 atmosphere. MgO and Mg(OH)x may result from a partial oxidation of the materials during transfer in air to the instrument. C-C at 284.9 eV corresponds to hydrocarbon contamination. a amorphous form bcrystalline...... 81

Table 4.3.1 - Summary of the synthetic conditions used, yields ...... 88

Table 4.3.2 - MgH2 particle size/morphology as determined by TEM and crystalline size as determined by XRD analysis using the Scherrer equation. Associated activation energy (Ea) and enthalpy (H) for the hydrogen desorption reaction...... 90

Table 4.3.3 - Elemental surface composition (atomic percentage, %) as determined by XPS for the materials obtained after the decomposition of di-n-butylmagnesium. XPS analysis was carried out after the 3rd cycle...... 93

Table 4.3.4 – Magnesium and aluminium content as determined by elemental analysis. .100

Table 4.4.1 - MgH2 crystalline size as determined by XRD analysis using the Scherrer equation. Associated activation energy (Ea), enthalpy (∆H), and entropy (∆S) for the hydrogen desorption reaction...... 119

Table 5.2.1 - Summary of the physical properties of the synthesized material. Particle morphology/size was determined by TEM and crystallite size from XRD analysis ...... 131

xiv

List of Tables

Table 5.2.2 - Summary of the enthalpy (∆H), entropy (∆S), and activation energy (Ea) of synthesised materials ...... 133

Table 5.3.1. - Hammett constants of para and meta substituted obtained from experiments and calculated Induction/Field (F) and Resonance (R) parameters of functional groups on Polystyrene (PST) used in this study. COOH value would not correlate with these other three groups, but due to its less bulky structures it should give higher value in terms of the induction. Hence the order of the F and R with ascending order towards electron withdrawing ability is NMe2

Table 5.3.2 - Summary of the crystallite size and particle size of synthesised materials ..149

Table 5.3.3 - Elemental surface composition (atomic percentage, %) as determined by XPS for the materials obtained from hydrogenolysis of di-n-butylmagnesium in cyclohexane with or without polystyrene with different functional groups. XPS analysis for cycled materials was carried out after the 3rd cycle...... 149

Table 5.3.4 - Summary of the enthalpy (∆H), entropy (∆S), activation energy (Ea), crystallite size and particle size of synthesized materials ...... 155

Table 5.4.1 - GPC and BET results summary ...... 166

List of Figures

LIST OF FIGURES

Figure 2.2.1 Bond strength of different physisorption and chemisorption based materials and desirable range required for room temperature hydrogen storage...... 10

Figure 2.2.2 Lennard-Jones potential energy curve corresponding to energy barriers required during absorption/desorption in a metal hydride, (i) energy for physisorption, (ii) energy for dissociation, (iii) energy for chemisorption, (iv) energy for hydrogen penetration in subsurface, (v) energy for hydrogen diffusion, and (vi) energy for nucleation and growth of the hydride phase ...... 11

Figure 2.2.3 Correlation between surface area (Langmuir method, N2) and saturation hydrogen uptake at 200 °C (data were obtained from various synthetic MOFs) 45...... 12

Figure 2.2.4 (Left) Single wall carbon nanotube from a rolled single graphene sheet and (Right) Multiple wall carbon nanotube made from rolled multiple single graphene sheets51 ...... 13

Figure 2.2.5 Different modifications on MOFs to gain favourable hydrogen storage capacities (a) inclusion of adsorbent particles such as fullerene C60 inside the pore (b) open metal sites as the metal ions will favour axial coordination of adsorbing atoms 53 55 ...... 13

Figure 2.2.6 Hydrogenation mechanism of MgH2 showing dissociation of hydrogen molecules to atoms and diffusion into α-phase and converted to β-phase through lattice deformation and expansion...... 14

Figure 2.3.1 ∆H values of MgH2 at different particle sizes and different surface plane. MgH2

(110) has surface energy 1.3 J m-2 resulting in a drastic decrease in enthalpy while MgH2 (001) has surface energy of 0.48 J m-2 resulting in increased enthalpy formation. 61 ...... 18

Figure 2.3.2 Density of nucleation in bulk metal hydride (top) shows that it has multiple nucleation sites that grow and form a closed loop to prevent hydrogen from diffusing faster and the same goes during dehydrogenation as it needs to diffuse through a thick layer of β- phase. While in nano metal hydrides (bottom), diffusion through α-phase is still possible even in a larger fraction of β-phase for hydrogenation reaction, while quicker hydrogen diffusion to the outside is expected. 82 ...... 19

xvi

List of Figures

Figure 2.3.3 Mg nanoparticles synthesised by (left) electrochemical methods (5 nm)98, (middle) electroless reduction (5 nm)100, (right) digestive ripening (3 nm)99 ...... 24

Figure 2.3.4 di-n-butylmagnesium impregnation and subsequent formation of Mg in (left) carbon aerogel 107 and (right) porous carbon 108 ...... 25

Figure 2.4.1 Influence of anthracene concentration on the rate of formation of MgA.(THF)3 at 25 °C (left) and 60 °C (right) in THF, for the initial anthracene concentrations of 0.02 (-x- ), 0.005 (-+-) and 0.1 mol/l (-ʘ-) 117 ...... 27

Figure 2.4.2 Examples of block copolymer micelles with different self-assembly structures to obtain Fe nanoparticles. Different structures were obtained by adjusting the solvent polarity ratios133 ...... 31

Figure 2.4.3 Schematic illustration of the internal network of hyper-cross-linked polystyrene (HPS). The speckled phenyl rings reside in a different plane relative to the unspeckled ones in the cross-linked material, and the circle identifies a postulated cavity in which Co nanoparticles could grow.136 ...... 31

Figure 3.1.1 The overall procedure of the experiments ...... 43

Figure 3.2.1 Schematic diagram of the pressure reactor vessel (left) and the bench top setup during high temperature and pressure reaction (right) ...... 45

Figure 3.2.2 (Left) Schematic diagram of the high pressure manual sorption apparatus for hydrogen storage studies and the corresponding apparatus setup (Right) ...... 46

Figure 3.3.1 Schematic diagram of an X-ray diffractometer measuring 2휃 ...... 47

Figure 3.3.2 (A) XRD sample holder with kapton foil used in this study (B) Typical XRD pattern of crystalline MgH2 and sample holder with kapton foil...... 48

Figure 3.3.3 Schematic diagram of TEM instrument ...... 50

Figure 3.3.4 (A) EDS spectrum of MgCl2 and MgH2 particles , (B) EDS mapping of Mg element on MgH2 heterogeneous particles ...... 51

Figure 3.3.5 Schematic diagram of the XPS process of electron from the 1s shell ...... 52

Figure 3.4.1 (left) TPD result of a commercialised MgH2 with theoretical [H2] of 0.1 mol and the area under the curve was used to plot the calibration line (right) ...... 56

Figure 3.4.2 Component diagram of conventional apparatus and magnetic suspension balance 9 ...... 59

xvii

List of Figures

Figure 3.4.3 (A) Pressure-Composition-Isotherm (PCI) diagram shows plateau pressures of the phase transition from α-phase to β-phase measured at different temperatures, (B) corresponding Van’t Hoff plot: logarithm of the equilibrium pressures vs inverse temperature to determine enthalpy and entropy of hydrogenation...... 61

Figure 3.4.4 Schematic drawing of the pressure composition isotherm system: an automatic gas reaction controller manufactured by the Advanced Materials Corporation...... 63

Figure 4.2.1 – TGA/DSC curves of the different Grignard compounds investigated...... 71

Figure 4.2.2 – Evolution of volatile organic matter and hydrogen (followed by MS) resulting from the thermal decomposition of the selected Grignard reagents during their TGA analysis...... 72

Figure 4.2.3 – XRD patterns of the materials obtained after the thermal decomposition of the Grignard reagents: (A) under an Ar atmosphere, and (B) under hydrogen pressure, i.e. after hydrogenolysis. The temperatures used for these syntheses are tabulated in Table 4-1. A partial oxidation of the materials may have occurred during measurement...... 74

Figure 4.2.4 – SEM images of the materials obtained after hydrogenolysis of the Grignard reagents...... 78

Figure 4.2.5 – TEM images of the materials obtained after hydrogenolysis of the Grignard reagents and each corresponding to the EDS analysis ...... 79

Figure 4.2.6 – Evolution of volatile organic matter and hydrogen (followed by MS) from the materials obtained after hydrogenolysis of the selected Grignard reagents. Measurements were carried out at a heating rate of 10 C.min-1 and under a 25 ml.min-1 Ar flow...... 82

MS during the thermal decomposition at 10 C.min-1 under a 25 ml.min-1 Ar flow of H2-

DibutylMg after the first cycle, and (D) TEM image of H2-DibutylMg after 3 cycles. No gases except hydrogen were detected from H2-DibutylMg after the first cycles...... 84

Figure 4.3.1 – SEM images of the material obtained from the hydrogenolysis of di-n- butylmagnesium: (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether...... 89

Figure 4.3.2 – TEM images of the material obtained from the hydrogenolysis of di-n- butylmagnesium: (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether...... 90

xviii

List of Figures

Figure 4.3.3 – XRD patterns of the material obtained from the hydrogenolysis of di-n- butylmagnesium: under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether...... 91

Figure 4.3.4 XRD showing the formation of aluminium from triethylaluminium under synthetic conditions similar to those used for the hydrogenolysis of di-n-butylmagnesium in cyclohexane...... 91

Figure 4.3.5 – 13C NMR spectra the material obtained from the hydrogenolysis of di-n- butylmagnesium: under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether...... 92

Figure 4.3.6 – MS of gases evolving from the thermal decomposition of the materials obtained (Left) and corresponded TG-DSC spectras (Right) after the hydrogenolysis of di-n- butylmagnesium: (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether. The fragments m/z 27, 41 and 56 correspond to the decomposition of the butyl group. The additional m/z 84 and 74 correspond to the release of cyclohexane and diethyl ether, respectively...... 94

Figure 4.3.7 – XRD of the materials obtained from the hydrogenolysis of di-n- butylmagnesium under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether after two hydrogen absorption/desorption cycles (top) after hydrogen absorption and (bottom) after hydrogen desorption. The MgO peak observed by XRD after desorption may correspond to a partial oxidation of the material during measurement. .. 96

Figure 4.3.8 - : TGA/DSC after the 3rd hydrogen absorption of the materials resulting from the hydrogenolysis of di-n-butylmagnesium :(A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether...... 97

Figure 4.3.9 - MS of the gases evolving from the thermal decomposition of the materials after the 3rd hydrogen absorption. Materials were synthesized from the hydrogenolysis of di-n- butylmagnesium: : (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether...... 98

Figure 4.3.10 – Hydrogen desorption kinetics of the materials obtained from the hydrogenolysis of di-n-butylmagnesium under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether. The 3rd hydrogen desorption cycle at 300 °C measured in High pressure TGA is shown...... 99

Figure 4.3.11 - 13C NMR spectrum of the materials after hydrogen cycling and XPS wide-scan of the material obtained from the hydrogenolysis of di-n-butylmagnesium under hydrogen

xix

List of Figures

pressure after synthesis and hydrogen cycling. A similar reduction in carbon content was observed for all the other materials ...... 101

Figure 4.3.12 – TEM of the materials after the 3rd hydrogen absorption. Materials were synthesized from the hydrogenolysis of di-n-butylmagnesium: : (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether. .102

Figure 4.4.1 XRD spectra of (A) MgH2-Mg(excess)A, (B) MgH2-MgA, ...... 108

Figure 4.4.2 TEM analysis of (A) MgH2-Mg(excess)A and (B) MgH2-MgA as synthesised .109

Figure 4.4.3 (A) materials precipitated from MgA in cyclohexane (B) materials from decomposition of MgA under vacuum ...... 110

Figure 4.4.4 (A) materials precipitated from MgA in cyclohexane hydrogenated at 300 °C 30 bar H2, (B) materials precipitated from MgA in diethyl ether hydrogenated at 300 °C 30 bar

H2, (C) materials from decomposition of MgA under vacuum and hydrogenated at 300 °C

30 bar H2 ...... 111

Figure 4.4.5 TPD analysis of MgH2-Mg(excess)A (left) TGA-DSC analysis and the corresponding MS spectra (right) ...... 112

Figure 4.4.6 TPD analysis of MgH2-MgA (left) TGA-DSC analysis and the corresponding MS spectra (right) ...... 113

Figure 4.4.7 XRD spectra of (A) cycled MgH2-Mg(excess)A, (B) cycled MgH2-MgA ...... 114

Figure 4.4.8 TEM analyses of (A) MgH2-Mg(excess)A cycled and (B) MgH2-MgA cycled ...115

Figure 4.4.9 Hydrogen desorption kinetics of MgH2-Mg(excess)A and MgH2-MgA. The 3rd hydrogen desorption cycle at 300 °C measured in Sievert Instrument is shown...... 115

Figure 4.4.10 TPD analysis of cycled MgH2-Mg(excess)A with TGA-DSC analysis (left) and the corresponding MS spectra (right) ...... 116

Figure 4.4.11 TPD analysis of cycled – MgH2-MgA with TGA-DSC analysis (left) and the corresponding MS spectra (right) ...... 117

Figure 4.4.12 Van’t Hoff plots of the materials obtained from catalytic hydrogenation of MgA and decomposition of di-n-butylmagnesium under hydrogen pressure in dry conditions and in cyclohexane. Equilibrium pressures were obtained from the PCI measurements...... 117

Figure 4.4.13 Plot of enthalpy and entropy changes of the (A) MgH2-Mg(excess)A, (B) MgH2-

MgA, (C) MgH2-H2MgBu2, (D) MgH2-Cyclohexane MgBu2 reflecting the compensation effects ...... 118

List of Figures

Figure 5.2.1– TEM images of: (a) MgH2/C and (b) MgH2/PSTN as-synthesized and after hydrogen absorption/desorption cycling at 300 °C (c) MgH2/C including EDX analysis insert and (d) MgH2/PSTN...... 130

Figure 5.2.2– XRD patterns of MgH2/C and MgH2/PSTN as-synthesized and after hydrogen desorption. The diffraction patterns after hydrogen absorption are identical to that of the as-synthesised materials...... 131

Figure 5.2.3 – Hydrogen desorption profiles as measured by TGA/MS of (a) MgH2/C and (b)

MgH2/PSTN as-synthesised, after hydrogen cycling, and after oxidation for 24 h in air. ..132

Figure 5.2.4 – (A) TGA profile of PSTN shows the complete decomposition of the polymer between ~330-450 oC and (B) the corresponding MS spectrum shows different m/z have peaked during the decomposition ...... 132

Figure 5.2.5 – Hydrogen desorption kinetics measured at 300 °C for (a) MgH2/C and (b)

MgH2/PSTN, as synthesised and after oxidation for 24 h in air...... 134

Figure 5.2.6 - P-C-I curves at different isotherms of the MgH2/C (top) and MgH2/PSTN (bottom) before and after exposure to air ...... 135

Figure 5.2.7 – Van’t Hoff plots of MgH2/C and MgH2/PSTN as synthesized and oxidised. Equilibrium pressures were obtained from the PCI measurements...... 136

Figure 5.2.8 – TEM images of: (a) MgH2/C and (b) MgH2/PSTN as-synthesized and oxidised for 24 h in air and associated images after hydrogen cycling at 300 °C (c) MgH2/C and (d)

MgH2/PSTN...... 136

Figure 5.2.9 – XPS peak profiles (a) Mg2p and (b) O1s of MgH2/C and MgH2/PSTN as- synthesized and oxidised...... 137

Figure 5.2.10 – XRD patterns of MgH2/C and MgH2/PSTN as-synthesized and oxidised after hydrogen cycling, PCI measurements and absorption...... 138

Figure 5.3.1 – TEM images and particle sizes distribution of: (A) MgH2@PS-NMe3+ , (B)

MgH2@PS-NMe2, (C) MgH2@PS-OH, (D) MgH2@PS-COOH ...... 147

Figure 5.3.2 XRD patterns of MgH2 as-synthesized from hydrogenolysis of MgBu2 in cyclohexane with different functional groups of polystyrene...... 148

Figure 5.3.3 – TGA/DSC of MgH2@PS-NMe3+, MgH2@PS-NMe2, MgH2@PS-OH, and

MgH2@PS-COOH as synthesized ...... 151

Figure 5.3.4 TGA shows the different functional group polystyrenes decomposition ...... 152

xxi

List of Figures

Figure 5.3.5 – TEM of the materials after the 3rd hydrogen absorption...... 152

Figure 5.3.6 - TGA/ DSC after the 3rd hydrogen absorption of the materials ...... 153

Figure 5.3.7 – MS of the gases evolved from the materials including PS free MgH246 ...... 154

Figure 5.3.8 – Hydrogen desorption kinetics of the materials. The 3rd hydrogen desorption cycle is shown...... 155

Figure 5.3.9 P-C-I absorption measurements of MgH2 synthesized with different functional groups polystyrene at various temperatures ...... 156

Figure 5.3.10 Van’t Hoff plots of MgH2 from MgBu2 in cyclohexane and with different functional group polystyrenes ...... 157

Figure 5.3.11 – Thermodynamics plot, enthalpy vs entropy values of the materials ...... 157

Figure 5.4.1 (left) NMR polystyrene of RAFT polystyrene, (right) GPC results of linear polystyrene, SPS, and HPS ...... 165

Figure 5.4.2 BET analyses results of HPS and S(PS) showing (A) the volume absorption profiles indicate the surface areas and (B) the pore distribution of the materials ...... 166

Figure 5.4.3 Nickel nanoparticles generated in HPS (top) and S(PS) (bottom) with corresponding EDS spectra ...... 167

Figure 5.4.4 TEM and EDX analyses of the resulting materials of hydrogenolysis of di-n- butylmagnesium in cyclohexane with (A) S(PS) and (B) HPS ...... 168

Figure 5.4.5 TPD analyses of the (A) MgBu2/S(PS) and (B)MgBu2/HPS consist of TGA-black, DSC-blue, MS-red ...... 169

Figure 5.4.6 TEM images of MgH2 produced from catalytic hydrogenation of MgA in THF with (A) HPS, (B) HPS-Ni, (C) S(PS), (D) S(PS-PVBA) ...... 170

Figure 5.4.7 TPD analyses of materials synthesised from catalytical hydrogenation of isolated MgA.(THF)3 with addition of (A) S(PS) and (B) HPS ...... 171

Figure 5.4.8 1H NMR of (left) poly(Sty-co-VBC) and (right) poly(Sty-co-VBA) ...... 173

Figure 5.4.9 (A) TEM and EDS analyses of the material synthesised from catalytic hydrogenation of MgA.(THF)3 with addition of S(PS-PVBA) and the corresponding TPD analysis of the freshly synthesised material (B) ...... 174

Figure 5.4.10 (A) TEM and EDS analyses of the material synthesised from catalytic hydrogenation of MgA.(THF)3 with addition of HPS-Ni and the corresponding TPD analysis of the freshly synthesised material (B) ...... 175

xxii

List of Figures

xxiii

List of Figures

“Philosophy is written in this grand book, the universe, which stands continually open to our gaze. But the book cannot be understood unless one first learns to comprehend the language and read the letters in which it is composed. It is written in the language of mathematics, and its characters are triangles, circles, and other geometric figures without which it is humanly impossible to understand a single word of it.”

Galileo Galilei, 1623

xxiv

Introduction

1 INTRODUCTION

1.1 Background

Hydrogen economy is one of the most promising energy solutions for the continuity of modern living society. However, the hydrogen storage technology still remains an obstacle for the full implementation of the hydrogen economy. A number of strategies and materials have been designed to accommodate the storage of hydrogen in a safe and efficient way. In particular, hydrogen storage in solid materials has been the pinpoint of many researches because it has much potential in terms of economics, safety, and efficiency. Magnesium hydride is highly considered to be the best candidate for the solid hydrogen storage material because of its high availability, non-toxicity, good gravimetric and volumetric hydrogen capacity, and high reversibility for hydrogen storage. However the hydrogenation/dehydrogenation processes of MgH2 have problems associated with slow kinetics and high temperature requirement. Extensive researches have been undergone to solve these problems by making Mg alloys with other metals, catalysts, and size reductions up to micron sizes. Recently, a few theoretical and experimental based studies of MgH2 indicate that ultimate hydrogen storage properties would be achieved if the MgH2 size is

Introduction

down to sub-nanometer scale. Indeed, this provides new challenges to synthesise such small

Mg/MgH2 nanoparticles and the stabilisation process to avoid agglomeration of these nanoparticles after several hydrogen cycles. Yet, advancement in nanoparticle science and inorganic-organic synthesis offers a lot of potential to achieve such goals.

1.2 Objectives

Therefore, the primary aim of this research is to explore the novel synthesis methods of

MgH2 nanoparticles and study their structural behaviours. This would give valuable information that is useful for further modifications to apply stabilisation and improvement towards the overall properties. Specifically, the key objectives that will be addressed are as follows:

1. To develop novel synthetic methods for MgH2 nanoparticles preferably

based on the wet synthesis method feasible for scale up productions

2. To identify the suitable stabilisation and improvement methods especially

by utilising polymers as the additives

3. To understand and gain more information on how the MgH2 is synthesised

under influences of external factors such as synthetic conditions or additives

1.3 Thesis Structure

The structure of the thesis is as follows.

 Chapter 2 presents a review of relevant literature on general technical information

around hydrogen energy, development of solid hydrogen storage research over the

past decades, current state of MgH2 research for hydrogen storage including the

Introduction

size effect discussions, and several potential strategies from past studies that could

contribute to the synthetic process design of hybrid-MgH2 nanostructures

 Chapter 3 presents experimental methods to synthesise and characterise the

materials studied in this thesis

 Chapter 4 is divided into three sub-chapters. The first part investigates different

Grignard reagents as potential precursors to produce MgH2 via the hydrogenolysis

method. The second part investigates the effect of different mediums upon the MgH2

properties produced from the hydrogenolysis of di-n-butylmagnesium, the last part

investigates the Mg/MgH2 synthesised from different methods based on the

MgA.THF3

 Chapter 5 investigates the uses of different polystyrenes to improve hydrogen

storage properties of MgH2 produced in Chapter 4. This chapter is also divided into

three sub-chapters. The first part focuses on the influence of polystyrene to

withstand oxidation and the overall effects on hydrogen storage properties. The

second part investigates the effect of different functional groups attached to

polystyrene on the overall hydrogen storage properties. The third part utilises well

defined nanostructured polystyrene to control the morphologies of MgH2.

 The thesis is concluded in Chapter 6 with a summary of key results and

contributions. In addition, a discussion of some possible extensions to this work is

presented.

Literature Review

2 LITERATURE REVIEW

2.1 BACKGROUND ……………………………………………………………………………………………….5 2.1.1 Energy storage and hydrogen economy …………………………………………………..6 2.1.2 Hydrogen economy and storage challenges …………………………………………………..7 2.2 SOLID BASED MATERIALS FOR HYDROGEN STORAGE APPLICATION ……………………….10 2.2.1 Porous materials ……………………………………………………………………………………..12 2.2.2 Metal and complex hydrides ………………………………………………………………………....14

2.3 MGH2 DEVELOPMENT AND IMPROVEMENT FOR HYDROGEN STORAGE ………………………..17 2.3.1 Size effects improvement and nanosizing strategies ……………………………………..18 2.3.2 Nanosizing challenges and problems ……………………………………………………………..20

2.3.3 Current development of MgH2 nanoparticles …………………………………………………22 2.4 HYBRID MG-POLYMER SYNTHESIS AND STRATEGIES ………………………………………………25

2.4.1 Mg/MgH2 synthesis strategies …………………………………………………………….25 2.4.2 Novel stabilisation strategies …………………………………………………………….29 2.5 CONCLUSION …………………………………………………………………………………………….32 2.6 REFERENCES …………………………………………………………………………………………….33

Literature Review

2.1 Background

Today humanity consumes a lot of resources and produces a large quantity of waste products to satisfy its needs. However, nature has a finite amount of natural resources to meet these demands. Global trends show that since the 1970s humanity has an annual consumption of natural resources exceeding what the Earth can regenerate each year.1 At the current rate of consumption, humanity uses the equivalent production of 1.5 planets per year. Consequently, if current growths in the population and consumption trends are to continue, by 2030, two planets will be required to sustain the current way of life in developed society and in emerging economies. Therefore, some major changes in the way our society lives will need to be made to prevent this trend and shift it closer to an equilibrium between nature’s resources and ecological footprints. Obviously, the growth in population and consumption will be difficult to avoid especially if humanity wants to maintain its current comfortable and luxury living lifestyle. In fact, there is a correlation between High Development Index (HDI) as a measurement of prosperity in life and energy/electricity consumptions to afford those luxuries.2 3 The blame for this issue can be pointed towards the over-reliance on the use of fossil fuels as non-renewable resources. Oil, coal, and gas are the fossil fuel resources that become primary energy sources for a wide range of applications in this modern era.4 However, the world has experienced “peak oil”, indicated by the soaring prices of oil beginning in the 1970s when it was increased by a factor of ten, from $3.50 to $35 per barrel.5 It is also expected that gas and coal will follow the same trend with a depletion predicted to occur in 37 and 107 years respectively at the current consumption rate.6 Moreover, another problem caused by these energy resources is that they produce greenhouse gases which contribute to climate change and air pollution.

These will cause many unfavourable environmental and health problems if they are not mitigated.7 8 Therefore, there is no other choice for humanity but to have alternative energy resources replace the use of fossil fuels.

Literature Review

Indeed, many alternative energy resources are available which can be divided into two types, non-renewable and renewable.9 Nuclear energy is one of the options as a new non- renewable source that potentially supplies a large amount of energy. However, the finite amount of uranium resources and safety concerns relating to nuclear power plants are a hindrance to increasing the use of nuclear energy.10 Nevertheless, the alternative option lies with renewable resources. Solar, wind, geothermal, and hydropower are a few examples of resources that can be considered to be environmentally friendly and non-finite. However, the use of renewable resources is limited by their intermittency. For example, the output from a wind turbine depends on wind speed which varies across different places and time scales. Similarly, PV can only be relied upon when the sun shines in the area.11 Hence, the solution for this lies in a new energy storage device capable of storing the renewable energy when it is not being generated.

2.1.1 Energy storage and hydrogen economy

Energy storage can be considered to have played an important role in balancing the supply and demand of energy especially over the past century when electricity was introduced to society.12 13 14 Improvements to battery technology still continues to develop and just over the past decade has made remarkable progress through the discovery of the lithium ion(Li- ion) battery technology.15 The technology used Li ions to migrate across the electrolytes in between positive and negative electrodes. The advantages of the Li-ion battery lies in its high energy density (0.36-2.23 MJ/kg and 0.90-2.23 MJ/L) and also its long lifecycle and rate capability as compared to the common lead acid batteries.12 These properties attracted many uses in electronic applications and also make it regarded as one of the battery choices for powering electric vehicles.16 However, even with the latest state of the art battery technology, electric cars still cannot match the same travel distance as the conventional gasoline vehicles.17 In addition, for the mobile application of energy storage, batteries have

Literature Review

several limitations including the cost, short life cycle and limited energy density.18 This would favour hydrogen as a viable alternative to store energy with high density and in a clean fashion.

Hydrogen is considered to be the primary fuel for fuel cell technology which also allows a vehicle to run on electricity.19 20 Indeed, hydrogen is another form of energy storage system which can be obtained by electrolysing water from renewable energy sources. This concept is called the ‘hydrogen economy’ which was proposed back in the 1970s as a potential solution to substitute for fossil fuel use.21 Hydrogen, being one of the most abundant elements on earth and having a high energy density, is deemed to be a suitable energy carrier. In addition, hydrogen is also a cleaner form of fuel since the by-product is water when hydrogen is utilised to produce energy either in a fuel cell or by combustion. Current hydrogen production mostly still relies on hydrocarbons as raw materials thus having a high carbon footprint, but progress in water electrolysis and fuel cell technologies shows that a hydrogen economy is fully feasible.22

2.1.2 Hydrogen economy and storage challenges

In the past, hydrogen economy progress has been stalled mainly by technical issues such as a system to store hydrogen and economic issues such as the high cost of hydrogen compared to the fossil fuels.21 In this case, a good hydrogen storage system has several requirements which need to be satisfied such as gravimetric and volume capacity, cost and the refuelling ability.23 24 25 26 Wide ranges of options and material classes have been considered and each has their own advantages and disadvantages in being potential hydrogen storage materials.

To guide these materials developments, the US Department of Energy (DOE) set up targets as the guidelines which are summarised in Table 2.1.1 together with the current achievements of each class of materials.27 The targets include gravimetric and volumetric

Literature Review

density, reversibility, recharge time, minimum delivery pressure, minimum flow rate, life cycles and fuel impurities.26

Table 2.1.1 - Performances of different H2 storage systems and DOE targets26

Gravimetric Volumetric Costs Storage Systems (kWh/kg sys) (kWh/L sys) ($/kWh) 700 bar compressed (Type IV) 1.7 0.9 19 350 bar compressed (Type IV) 1.8 0.6 16 Cryo-compressed (276 bar) 1.9 1.4 12

Metal Hydrides (NaAlH4) 0.4 0.4 N.A. Porous materials (AX-21 carbon, 200 bar) 1.3 0.8 N.A. Chemical Hydrogen Storage (AB-liquid) 1.3 1.1 N.A. 2017 Targets 1.8 1.3 12 Ultimate Targets 2.5 2.3 8

Indeed, storing hydrogen is quite a complex matter. Hydrogen has a very low density (0.09 g/L) at normal atmospheric pressure resulting in really low volumetric density which is important for vehicles (11 litres to store 1 kg hydrogen that is needed to run a car for 100 km). In order to decrease this volume requirement, it is usually compressed in a high pressure tank which typically operates at 150 to 300 bar.23 However, much higher pressure

(700 bar) is necessary to reach the ideal volumetric/ gravimetric density for hydrogen fuel cell vehicles and this also requires more robust pressure vessels to sustain such high pressure.28 Currently there are some hydrogen storage researches that focus on improving the robustness of the tank to sustain even higher hydrogen pressure.29 High pressure gaseous hydrogen (HPGH2) storage system has the advantage of its simplicity and fast filling-releasing rate.30 Mainly, Type III (metallic wrapped liner) and IV (non-metallic wrapped liner) vessels are used for HPGH2. Typically, lightweight materials such as aluminium for Type III and carbon fibers for Type IV are used. However, further improvement is still required especially on several aspects such as the durability and performances of components in contact with high pressure hydrogen, and safety issues, such as leakage risks associated with the use of high pressure hydrogen.31 32

Literature Review

Another way to increase the overall volumetric storage density is by hydrogen liquefaction.

Liquid hydrogen has been commercially available since it was used as spacecraft fuel in the past.33 However, in order to maintain its density of 71 g/L at -253 °C, it needs to be kept in a vacuum insulated vessel. Furthermore, liquid hydrogen would lose some hydrogen

(between 30-40%) due to the boil-off at the lower heating value (LHV) compared to HPGH2

(5-20%).31 34 This led to the development of cryo-compressed systems (LLNL) which have thermal endurance improvements that can prevent these evaporative losses to some extent.

This system also reduced the requirement for high pressure and more ambient temperature compressed hydrogen. However, further work is still needed to solve some issues that are mainly similar with those of HPGH2 vessels.30 35

Another alternative to storing hydrogen as a fuel is by recombining it with other elements or compounds. One option is to make a synthetic hydrocarbons fuel with H2 and decompose them to release hydrogen. For example, formic acid has been considered for hydrogen storage applications since it can decompose directly to gaseous CO2 and H2.36 Similarly, other hydrocarbons such as methanol and octane could also be considered as hydrogen storage with high storage density. However, CO formation during the H2 release would contaminate the PEM cell that could lead to a decrease in the fuel cell’s efficiency.37

Moreover, the reforming of these hydrocarbons is extremely difficult. Methanol has the most potential over all hydrocarbons due to a lower temperature requirement for its reforming (250-300°C and lower in the presence of catalyst).38

Recently, a new type of synthetic fuel based on boron-nitrogen-hydride has attracted more attention since these materials would not produce greenhouse gases as by-products.39 40

Ammonia borane or AB contains 19.6 wt% H2 and is thermally stable and safe enough at ambient temperatures to transport and store.39 40 However, it faces some challenges especially in terms of the compound regeneration after releasing hydrogen and production of borate by-products that can disrupt PEM fuel cell systems.

Literature Review

Indeed, the reversibility of materials to store hydrogen is another important factor to consider and in particular, that more energy shall not be expended to produce or synthesise the material. Hence, the last option is to simply charge H2 as fuel to a material that can reversibly release and also absorb hydrogens. This can be done mostly in solid materials either by forming hydrides (metal and complex) or through physical adsorption on porous materials.

2.2 Solid based materials for hydrogen storage application

The principle of hydrogen storage material in solid form is to bind hydrogen with a material.

This can be achieved mainly in two ways, through Van der Waals interaction/ surface adsorption (physisorption) within a porous material or through chemisorption with hydrogen atoms forming chemical bonds with other materials (Figure 2.2.1). The chemisorption materials can be divided further depending on the bond strength, certain conditions such as temperature and hydrogen pressure are needed to break and form the bonds.

Bond strength (kJmol-1) 0 - 5 10 - 60 50 - 200 100 - 300 > 300

Desirable

Physisorption Metal hydrides Complex hydrides Chemical hydrides

Figure 2.2.1 Bond strength of different physisorption and chemisorption based materials and desirable range required for room temperature hydrogen storage.

Ideally, as shown in Figure 2.2.1, the material should have hydrogen bond strength of 10-60 kJ.mol-1 to be stable at ambient temperature and able to undergo H2 release at temperature

Literature Review

less than 100°C. However, neither physisorption nor chemisorption materials have this requirement.

In physisorption materials, it is not necessary for the hydrogen molecule to be dissociated into atoms as it can directly adsorb onto the surface hence it does not require high activation energy (Figure 2.2.2). As shown in the Lennard-Jones potential energy curves, it is also weakly bonded on the surface which helps the reversibility of the materials. However, due to such weak physical bonds, a low temperature around -200 °C is required to afford higher hydrogen molecules adsorption.41 On the contrary, chemisorption materials such as metal hydrides are required to overcome several steps including the dissociation energy barrier, energy for chemisorption, hydrogen penetration in subsurface, hydrogen diffusion energy and for the nucleation and growth process (Figure 2.2.2). These steps require some energy that accumulates and result in a higher temperature for H2 release of the chemisorption materials.

(ii)

(i) (iii) (iv) (v) (vi

)

Literature Review

2.2.1 Porous materials

For physisorption to occur with high capacity storage, high surface materials are required.

Many different types of porous materials such as zeolites, carbon based materials like graphitic nanofibres (GNF) or carbon nanotubes (CNTs), and Metal-organic frameworks

(MOFs) have been considered for hydrogen storage applications because of their high surface area.42 43 44 As shown in Figure 2.2.3, there are correlations with H2 uptake by physisorption and surface areas of the materials but only at -200 °C. At room temperature, the bonds are too weak to keep the H2 confined within the structure.45

Figure 2.2.3 Correlation between surface area (Langmuir method, N2) and saturation hydrogen uptake at 200 °C (data were obtained from various synthetic MOFs) 45

Lately, synthetic carbon porous materials such as carbon nanotubes (CNT) and graphene sheets have been extensively researched.46 47 48 CNTs which can be defined as a rolled graphene sheet, have been considered as a hydrogen storage material since their discovery in 1993.46 Besides a single walled CNT, multiple numbers of graphene sheets can be rolled together to make multi number walled CNTs (Figure 2.2.4) hence improving the adsorption sites and their porosity. Some have reported that it can be adsorbed onto the interior and exterior of the structure which leads to 3.3 wt% H2 at -200 °C.46 An alternative to increase

H2 capacity is to decorate porous carbon with transition metals to help dissociate hydrogen

Literature Review

molecules to atoms. Hydrogen atoms would have more volumetric density compared to hydrogen molecules thus leading to an increase in adsorption.49 50

Figure 2.2.4 (Left) Single wall carbon nanotube from a rolled single graphene sheet and (Right) Multiple wall carbon nanotube made from rolled multiple single graphene sheets51

Other types of porous materials that have attracted a lot of attention in recent years are the

MOFs. An MOF is a crystalline porous solid architecture consisting of organic ligands connected to metal ions. The advantages of MOFs are their versatility in obtaining high surface areas synthetically and their ability to be modified to increase the hydrogen uptake.

Figure 2.2.5 shows several different modifications that can be performed on MOFs to increase the H2 uptakes.52 53 54 55 One of the ideas is to combine the MOF with another MOF with other porous material inside the pore of the MOF to increase H2 uptake (Figure 2.2.5-

(a)). 53

(a) (b)

Literature Review

Another idea is opening up the metal component to become metal ions for allowing more

H2 binding as shown in Figure 2.2.5-(b). Overall, although porous carbons and MOFs can afford reasonable H2 uptake (Figure 2.2.3), the application still remains hindered as low temperature is required for H2 uptake and release.

2.2.2 Metal and complex hydrides

Metal and complex hydrides are the two distinct types of hydrides which should be reviewed separately. For metal hydrides, the nature of hydrogen sorption involved two step processes which include dissociation and recombination of hydrogen molecules and atoms, and diffusion of hydrogen atoms within the metal lattices.56 As shown in Figure 2.2.2, high surface energy is needed for Mg to dissociate hydrogen molecules into atoms (~ 432 kJ mol-

1 H2 ).57 These hydrogen atoms will then diffuse in and fill the space within the metal lattice in the phase called -phase. The diffusion process continues until the space between the metal expands and -phase of metal hydrides starts to nucleate and grow. Once it reaches hydrogen and metal equilibrium concentration it will be fully converted to a metal hydride

(Figure 2.2.6).

α - phase β - phase

Figure 2.2.6 Hydrogenation mechanism of MgH2 showing dissociation of hydrogen molecules to atoms and diffusion into α-phase and converted to β-phase through lattice deformation and expansion.

Literature Review

Most of the heavy metals such as Pd and LaNi5 are able to reversibly release and absorb hydrogen at moderate conditions. However, because of their weight, the gravimetric hydrogen capacity is really low i.e. 0.6 wt% for PdH0.6.58 On the other hand, light metals such as Li, Ca, Na, Mg, and Al will have high hydrogen capacity that satisfies the DOE targets but they store H2 under conditions of temperature and pressure that are not practical. This is mainly due to the unique bonds between these metals with hydrogen, for example Li-H is ionic and requires a temperature of more than 700 °C for releasing hydrogen.59

Table 2.2.1 - Properties of selected metal hydrides 56 60

Metal hydrides Storage Capacity (wt%) Tdes (°C) ∆H (kJ.mol-1)

LiH 12.7 720 -116.3

CaH2 4.79 600 -181.5

NaH 4.20 425 -56.5

MgH2 7.66 330 -75.3

AlH3 10.1 150 -46.0

In the case of AlH3, a high temperature is not required to release its hydrogen capacity that accounts for 10.1 wt%. However, the main challenge for it is the hydrogenation of aluminium which requires over 105 bars H2 pressure.60 Among these materials, storing H2 with Mg seems more feasible since Mg/MgH2 reaction occurs under relatively mild conditions of temperature and pressure (i.e. 400 oC, 10 bar and 7.6 wt%).61

Meanwhile, complex hydrides usually consist of light metals cation that form ionic bonds with anion groups such as [AlH4] – or alanate, [BH4]- or borohydride, and [NH2] – or amide.

Unlike metal hydrides, in complex hydrides the hydrogen atoms reside at the corner and are covalently bound to a central atom (e.g. B or Al).59 Table 2.2.2 shows selected complex hydrides with their hydrogen storage capacities and temperature for H2 release. Most of the materials have high storage capacities due to the high number of hydrogen atoms on the

Literature Review

crystal structures.62 However, the applications are inhibited by kinetics, thermodynamics, and reversibility issues. In particular, these complex hydrides decompose in several steps hence making a reversible hydrogenation difficult since it needs to recombine the constituent elements. For example, NaAlH4 undergoes its first decomposition reaction at

185 °C and its second at 250 °C having total hydrogen desorption of 5.6 wt%.63

3NaAlH4 → Na3AlH6 + Al + H2 (2-1)

Na3AlH6 → 3NaH + Al + 1.5H2 (2-2)

Another important group of complex hydrides are the borohydrides which have attracted a lot of attention owing to their high hydrogen capacity (up to 18.4 wt% for LiBH4). However the release of hydrogen must occur at high temperatures as the LiBH4 decomposes according to the following pathway.

ퟑ LiBH4 → LiH +B + H2 (2-3) ퟐ

ퟏ LiH →Li + H2 (2-4) ퟐ

LiBH4 releases three of the four hydrogen atoms from the compounds when it melts at 280

°C and decomposes into LiH and boron.64 The reaction can be reversed at 600 °C with 155 bar while kinetically limited.65 It is possible that this reaction does not go to completion and instead forms other intermediate products such as B2H6. This compound is toxic and can disrupt the PEM cell.66 Recently, a technique has been developed to confine the decomposition of by-products in the core shell structures with other metals as a shell which resulted in a reversibility reaction of the compounds.67 The material is a NaBH4 which is reversible with 5 wt% H2 capacity at 350 °C. However, the high temperature requirement for hydrogen release and huge decreases on the overall capacity still remain problems.

Literature Review

Table 2.2.2 - Selected complex hydrides with their gravimetric hydrogen storage capacity and desorption temperature 56 64

Complex Hydrides Storage Capacity (wt%) Tdes (°C)

LiBH4 13.5 380

NaBH4 10.8 400

Mg(BH4)2 13.7 280

LiAlH4 7.9 160, 180

NaAlH4 5.6 210, 250

Mg(AlH4)2 9.3 140-200

Li2NH 6.5 280

2.3 MgH2 development and improvement for hydrogen storage

Over the past decades, MgH2 has dominated metal hydrides research for hydrogen storage applications since it has many promising properties such as light, abundance, high hydrogen storage capacity (7.6 wt%) and good reversibility. However there are many challenges for the on-board applications especially in vehicles, because there is a need to improve the high temperature requirement and the slow hydrogen absorption and desorption kinetics.

Over the past decades, kinetic improvements were mainly done through the addition of different catalysts such as transition metals and increasing surface areas or creating defects through mechanical millings. Many transition metals have been investigated for their catalytical properties to dissociate the hydrogen molecules into atoms. For example, Nb or

Nb2O5 catalyst milled together with MgH2 achieved hydrogen desorption below 5 minutes

68 69 at 300 °C. Similar kinetics improvement have also been observed when V2O5 was used as a catalyst to desorb MgH2 with 7 wt% capacity.

Other improvements usually include destabilisation of the strong Mg-H bonds (76 kJ.mol-

1.H2-1) by alloying another element such as Ni that forms weak Ni-H bonds.70 71 72 A well- known example is Mg2Ni alloys which formed Mg2NiH2 after hydrogenation and desorbed

Literature Review

-1 -1 73 hydrogen at 245 °C as a result of enthalpy decreases to 65 kJ.mol .H2 . Other metals such as Si, Al are also known for their ability to destabilise Mg-H bonds.74 However, introducing these alloys to the MgH2 means the overall capacity is greatly reduced i.e. Mg2NiH4 has 3.6 wt% H2. As a result, extensive research was conducted to find the best composition of these other elements in the system.

2.3.1 Size effects improvement and nanosizing strategies

As mentioned earlier, some kinetic improvements observed on MgH2 is that their size has been reduced by mechanical milling. It is known that decreasing the surface areas of particles will improve the kinetics by the increased number of dissociation sites and decrease of the diffusion barrier. It is also assumed that by milling particles, the defects and grain boundary created within the particle would lead to more H2 diffusion and nucleation of the hydride phase.69 However, the desorption temperature is still too high (300 °C) due to the enthalpy of the MgH2 remaining too strong.

H Mg (001)

(110)

Figure 2.3.1 ∆H values of MgH2 at different particle sizes and different surface plane.

MgH2 (110) has surface energy 1.3 J m-2 resulting in a drastic decrease in enthalpy while MgH2 (001) has surface energy of 0.48 J m-2 resulting in increased enthalpy formation. 61

Literature Review

It has been theoretically predicted that the size effects can cause significant destabilisation only when the particle size decreases to the nanoscale, < 2nm.75 This is the result of excess surface energy on Mg nanoparticles which become more significant when they reached subnano size. This excess surface energy can be optimised by altering the morphologies of the nanoparticles 76 or even the crystal surface plane.77 As shown in Figure 2.3.1 (001) the surface plane will increase the overall enthalpy formation as the particle sizes decreases.

On the contrary, (110) the surface plane will decrease the enthalpy formation to the value of more than those nanoparticles without a specific plane.61 These differences are due to the certain surface plane which could have a higher coordination number of hydrogen than the others.78 79

Figure 2.3.2 Density of nucleation in bulk metal hydride (top) shows that it has multiple nucleation sites that grow and form a closed loop to prevent hydrogen from diffusing faster and the same goes during dehydrogenation as it needs to diffuse through a thick layer of β-phase. While in nano metal hydrides (bottom), diffusion through α-phase is still possible even in a larger fraction of β-phase for hydrogenation reaction, while quicker hydrogen diffusion to the outside is expected. 82

Another possible advantage of MgH2 at subnanosize can be related to the differences in nucleation and growth mechanisms of nanosize MgH2 and bulk.80 81 Figure 2.3.2 shows the proposed mechanism of nucleation and growth of Mg to become MgH2 in bulk and nanoparticles. The nucleation of MgH2 would always start near the surface where the highest hydrogen concentrations are located. At bulk size, it would grow to become small grain size creating a closed hydride layer formation and may block the hydrogen to further diffuse inside the metal phase.82 On the contrary, at a much smaller size Mg (< 5 nm), the possibility to form a layer of MgH2 is minimised and hydrogen diffuses in at a much shorter

Literature Review

distance.82 This of course does not always occur since it depends on the growth mode of the

MgH2.

2.3.2 Nanosizing challenges and problems

Indeed, MgH2 nanoparticles offer great promises to improving the hydrogen storage properties but there are still many issues requiring further investigation. One of them is related to the thermodynamics properties of the MgH2 nanoparticles. Although theoretical studies have shown that destabilisation can occur by the size effects, experimental results did not show the predicted value of improvements. For example although ~15 °C reduction is predicted when the enthalpy (∆H) reduced by 2.84 kJ, experimentally it only caused 6 °C reduction in desorption temperature.83 This is due to the enthalpy-entropy compensation effect which refers to the specific linear relationship found between the ∆H and the change in entropy (∆S) of reaction, in this case the hydrogenation reaction. The relationship between ∆H and ∆S can be expressed in Gibbs free energy (∆G) which is defined as:

∆푮 = ∆푯 − 푻∆푺 (2-5)

This means that as ∆H becomes less negative (weaker bonding), ∆S tends to increase as the system becomes more disordered.84 However, there is no clear mechanism to explain this phenomenon although it has been discussed in many literatures including those on MgH2.85

86 87 Hence, this issue is subjected to our investigation in this study.

Another challenge is related to the synthesis and stabilisation of MgH2 with the size of less than 5 nm. The initial process of the synthesis technique is usually the most important aspect in order to obtain a really small nanoparticle unit. In general, synthesis of nanoparticles can be achieved in two approaches which are bottom-up and top-down approaches. Top-down approaches e.g. mechanical milling and vapour deposition usually involve high energy intensive processes to obtain size reduced particles. However, the

Literature Review

improvement from the ball milled materials do not last in many hydrogen cycles since nanograins usually rapidly grow after the 1st H2 cycle.88 In addition, scaling up the top-down processes is far more difficult than the bottom-up approaches in general.

The more feasible approach is by the bottom-up synthesis through nucleation and growth steps of particles in solution.89 90 Nucleation can occur only when the growth species have reached supersaturated condition or there is an excess of dissolved solute in the solution.

The degree of supersaturation (푆) can be adjusted and quantified based on the following equation

풂 풂 푪 푺 = 푨 푩 = (2-6) 푲푺풑 푪풆풒

Where 푎퐴 and 푎퐵 represent the activities of solutes A and B, Ksp is the solubility product constant, C and Ceq are the solute concentrations at saturation and equilibrium respectively.

Therefore, based on the equation (2-6), supersaturation can be altered in a number of ways including altering the concentration of solute, cooling of saturated solution and application of pressure.91 After nucleation, the nuclei continue to grow until it reaches a critical radius,

Rcr where it reaches a maximum change in free energy of the system. Indeed, this maximum change corresponds to the activation energy required to form the cluster as it is obtained from the free Gibbs energy change, ∆퐺 which can be expressed by the following equation.

ퟒ ∆푮 = − 흅푹ퟐ 풌 푻풍풏(푺) + ퟒ흅풓ퟐ풚 (2-7) 풗 풄풓 풃

Where 푣 is the molecular volume of the growth species, 푘푏 is the Boltzmann constant, 푦 is the specific surface energy. From equation (2-7), it can be concluded that the higher degree of supersaturation will decrease the critical radius as well as the particle size.

After nucleation, the nuclei would start the growth process which is usually done in two steps, first is the mass transfer of the monomer/ growth species to the particle surface

(diffusion limited) and second is the reaction that occurs on the surface to start the growth

(reaction limited). 92 93 As this growth process continues, the supersaturation will decrease

Literature Review

slowly as well as the overall nucleation and growth. In the end, the final size distribution of the nanoparticles is highly influenced by the growth process. Normally, the growth process occurs slowly leading the nanoparticles to become more monodispersed after nucleation is finished.94 Further growth can occur via Ostwald ripening whereby smaller particles would combine with other particles to become thermodynamically stable.95

Another challenge is to prevent the formation of oxidation layers on the surface layers of the Mg nanoparticles. Every light metal is prone to oxidation and is pyrophoric when it is exposed to air. In bulk or other micro-size Mg, the oxide layers formed will not be as detrimental when compared to nanosize Mg. This is because MgO typically forms an oxide passivation layer in the range of 1-2 nm and it also could reduce the hydrogenation rate.96

To initiate hydrogen absorption, the oxide layer on the magnesium surface needs to be cracked. It is usually done by annealing the Mg under H2 pressure and cycling the material.

Therefore, synthesis methods of Mg nanoparticles need to be done under inert atmosphere and with a clean environment to minimise the impurities. This would become an additional challenge for the technical applications of these nanomaterials in the fuel cell and it should be considered as one of the important aspects in designing a synthetic process of Mg nanoparticles.

2.3.3 Current development of MgH2 nanoparticles

As previously mentioned, subnanoparticles (< 5 nm) of MgH2 can only be obtained by tuning the nucleation and growth processes in the bottom up approaches. Based on those factors, control over the growth of nanoparticles is possible by manipulating the balance of energy and entropy of the system. In the case of gas vapour based synthetic methods, growth termination can be done by immediate quenching/cooling of the system or adjusting other conditions.97 However, in the liquid system when the nanoparticles start to precipitate it will need the presence of stabiliser to impede the further growth process/ Ostwald ripening.

Literature Review

Indeed, the stabilisation process holds the most important role in order to obtain small Mg nanoparticles. So far there have been several attempts on Mg nanoparticles synthesis that have used different types of stabilising agents to make small size nanoparticles.98 99 An electrochemical synthesis of Mg was performed and was able to produce 5 nm size nanoparticles. In this case, tetrabutylammonium bromide (TBAB) was used as a stabiliser to control the size of nanoparticles. This is a type of surfactant that has a weak cappling ligand to avoid strong bindings with Mg surfaces hence minimising the restrictions towards hydrogen adsorption. Surprisingly, the materials showed partial hydrogen desorptions at really low temperatures and at 85 °C where it was desorbing 1.38 wt% of H2. The rest of H2 could not be released until the temperature was above 300 °C since larger MgH2 (>80 nm) were also observed within the materials.98 These different sizes may be due to the agglomeration of the particles as a result of a limited surfactants capping to restrict the particle growth.

Another successful attempt was performed by Kalindidi et al who was able to synthesise highly monodisperse colloidal Mg with radii of 3 nm through a method called digestive ripening.99 In general, the method simply utilises hexadecylamine (HDA) as a stabiliser to confine the growth of precipitates Mg from Mg-THF colloid. This Mg-THF colloid was prepared by the solvated metal atom dispersion (SMAD) process that co-condensates metal atoms and solvents on a reactor maintained at a cryrogenic temperature. The solvent was then evaporated to initiate the Mg nanoparticle precipitations and growth. The results showed initial desorption of the material to occur at 115 °C although the major hydrogen peaks also occurred at above 300 °C.99

Another stabilisation method besides using surfactants is the nanoconfinement technique which utilises a nanoporous support to facilitate the confinement of small particles inside the pore and which can sustain the heat treatments or cycling due to the high thermal

Literature Review

stability of the matrix.101 102 103 Not all porous materials can be used as the support for hydrogen storage applications.

Figure 2.3.3 Mg nanoparticles synthesised by (left) electrochemical methods (5 nm)98, (middle) electroless reduction (5 nm)100, (right) digestive ripening (3 nm)99

The same requirements for other hydrogen storage materials are also applied. First it must be light so it would not compromise the gravimetric capacity, and it should be abundant and low-cost. The materials should also be highly porous and have a high surface area to allow high loadings of the hydrides. However, it needs to be inert so it will not form any reaction with magnesium. Some examples include nanoporous carbon, mesoporous oxides, zeolites, and MOF.104 The deposition of the metal hydrides inside the porous matrix can be done in several ways. A method that is widely used is by solution impregnation. It is done by concentrating a soluble precursor in solvent and using the capillary forces to draw the Mg precursor within the matrix. For example, the infiltration of porous carbon material was done with di-n-butylmagnesium dissolved in heptane or hexane which is reacted with hydrogen at an elevated temperature to form nanoconfined MgH2. 103 105 However, the maximum loading of these nanoparticles inside the matrixes is still limited (up to 20wt%

Mg loading). Most of the time, multiple impregnations are required but nucleation and growth process of the reactants may occur after the solvent removal hence blocking the pore for further reactants deposits.106 Another approach is to impregnate the matrix by melt infiltration technique resulting in 2-5 nm MgH2 nanoparticles but also with only 33 wt% Mg loading.102

Literature Review

Figure 2.3.4 di-n-butylmagnesium impregnation and subsequent formation of Mg in (left) carbon aerogel 107 and (right) porous carbon 108

A new promising strategy was attempted by Jeon et al to confine the Mg nanoparticles in a polymer matrix.100 This study was performed by reduction of MgCp2 with lithium naphthalene while mixed with poly(methyl methacrylate) or PMMA in THF. It resulted in the formation of 5 nm Mg nanocomposite and remarkably it also shows good stability against oxidations because of the properties of PMMA. Moreover, these nanoparticles were able to desorb 5 wt% hydrogen at a temperature of 200 °C in several cycles.100 The advantage of using polymers such as PMMA is that it has high selectivity towards hydrogen while impeding oxygen or water from entering the shell.100 109 Indeed, this method could become the perfect strategy for immobilizing such small MgH2 nanoparticles whilst improving the overall properties to tackle all challenges that have been discussed until now.

2.4 Hybrid Mg-Polymer synthesis and strategies

2.4.1 Mg/MgH2 synthesis strategies

Several strategies that have been discussed so far showed that besides the stabilisation method, synthesis methods and precursors also play important roles in determining the morphology and size of Mg nanoparticles. Some of the most common methods employed

Literature Review

are based on the reduction and thermal decomposition of organomagnesium compounds as the precursors.

It is known that metal cation can be reduced into its monoatomic form through reduction reactions which eventually will form metal nanoparticles through the nucleation and growth process when it becomes supersaturated.110 The reduction is only possible if there is enough electron transfer by oxidation process to counteract the reduction process.

However, Mg itself is a strong reducing agent and only lighter metals such as K and Li have

E° potential low enough to reduce Mg.110 111 The use of electron carriers such as naphthalene which combined with metals to be used as reduction agents effectively reduces organomagnesium compounds.112 113 One example, magnesocene (MgCp2) was used as a precursor and dissolved in glyme to be reduced by different combinations of K and different electron carriers. It was found that these electron carriers could determine the sizes nanoparticles produced, i.e. potassium naphthalide (38 nm particles), potassium phenanthrene (32 nm particles), potassium biphenyl (25 nm particles). Eventually, size effects were observed from the hydrogenation/ dehydrogenation kinetics of these MgH2. It was found that the 25 nm particles absorbed hydrogen 7 times faster than the 38 nm particles proving the kinetic improvements caused by the size reductions.114

Reduction can also be performed through electrochemistry, where the reduction is carried out by the electric current until the voltage exceeds the standard reduction potential of the

Mg. In a normal electrochemistry setup, usually the reduced metal cations will be electro- deposited in the electrode (cathode). Haas et al performed this electrochemical reduction combined with ultrasonication to disperse Mg nanoparticles.115 This study used several

Grignard reagents as precursors that can be dissolved in THF or diglyme and it led to a 4 nm

Mg nanoparticles.

Another method is by utilising complex compounds of MgA.(THF)3 (A = C14H10 or anthracene). MgA.(THF)3 compounds can be prepared from magnesium and anthracene in

Literature Review

THF as described by Bogdanovic et al.116 The rate of formation depends on the magnesium surface area and on the anthracene concentration. Moreover, the whole reaction exists in a reversible temperature-dependent equilibrium, for example the formation of MgA.(THF)3 is favoured at lower temperatures.117 This equilibrium plays an important role in the method developed by Bogdanovic to prepare large scale MgH2 from hydrogenating this MgA.(THF)3 at 60 °C.116 Figure 2.4.1 shows changes in MgA.(THF)3 formation equilibrium when the reaction occurred at 25 or 60 °C.

Figure 2.4.1 Influence of anthracene concentration on the rate of formation of

MgA.(THF)3 at 25 °C (left) and 60 °C (right) in THF, for the initial anthracene concentrations of 0.02 (-x-), 0.005 (-+-) and 0.1 mol/l (-ʘ-) 117

At 25 °C, all anthracene in the solution mixture would be converted to MgA.(THF)3.

However, during heating to 60 °C some of the MgA.(THF)3 abruptly decomposed back to Mg and anthracene because the MgA.(THF)3 formation equilibrium had shifted. This portion of anthracene then formed a new MgA.(THF)3 batch until its concentration became 0.01 mol/L which then stopped the overall conversion of Mg to MgA.(THF)3. Therefore, a total concentration of anthracene in solution needs to be higher than 0.01 mol/L in order to keep this reaction sequence repeating continuously.117

With this repeating process (Scheme 2-1), the amount of anthracene and catalyst (CrCl3) used for hydrogenation can be minimised. In the end, there were some considerable

Literature Review

amounts of by-products produced by this method such as MgCl, CrCl3 , and anthracenes.

However, the amount of these by-products could be negligible if more Mg is initially used and more MgH2 produced in the end.117 For example, a study of 1 kg of active material would produce MgH2 that is able to have reversible hydrogen capacity from 6.4 to 6.6 wt%.118

Apparently, the hydrogen profiles of these materials showed hydrogen take up and release occurred at 230-350 °C also with good kinetics performances. Although the size was not specified in the study it reported that the materials had high specific surface areas (100-130 m2g-1).118

Scheme 2-1 Catalytic hydrogenation of Mg from MgAnthracene.3THF complex adopted from Bogdanovic 119

A more simple method to make Mg nanoparticles can be achieved by thermal decomposition of organomagnesium compounds. This is a simple process that only involves heating on a precursor to undergo decomposition and form Mg or MgH2.120 As shown before, a decomposition of di-n-butylmagnesium under H2 pressure leads to formation of MgH2. Yet there are other organomagnesium compounds capable of doing the same. The past studies have shown that hydrogenolysis of the Grignard reagents in ether solvent would also produce MgH2 and Mg halides (MgX) as the by-products.121 122 123 The process of the decomposition mechanism is still not clear as there are possibilities for them to form HMgX

Literature Review

compounds as intermediates during the decomposition.123 Moreover, the effect of organic groups within the Grignard reagents may affect the precipitation process and properties of the MgH2 produced. Obviously, more work is needed including the investigation of hydrogen storage properties of these materials which has not been performed yet.

2.4.2 Novel stabilisation strategies

Nanoparticles can be stabilised either by steric or electrostatic repulsion mechanisms. In steric repulsion, nanoparticles can be placed in an inert environment such as inorganic matrix and polymers or bound with organic or inorganics ligand thus it will not interact with other particles. Electrostatic repulsion is typically done by applying electric charges with the same type of charge on each nanoparticle so they will repel each other to avoid agglomeration. Parameters involved in this stabilisation will be the key to determining the size and morphology of the nanoparticles produced.124 For example if a nanoparticle is bound with larger molecules or a greater number of molecules then it may result in small size particles due to increases in steric hindrance effect.124

This produces the idea to use polymers instead of normal surfactants as polymers generally have a longer carbon chain and higher molecular weight. Like surfactants, polymers can have different functional groups that are chemically bound on the structures. Active functional groups may be incorporated into polymer chains by direct polymerisation of monomers containing the desired functional groups or by chemical modification of a preformed polymer.125 However, a limited range of monomers that are both synthetically available with a controlled polymerisation technique make the latter become the more viable option. Click chemistry is one of the postfunctionalization techniques that has found many useful applications.126 127 128 It is known for its versatility to introduce the functional group on the selected part of the polymers such as the polymer backbone or in the chain end.

Literature Review

Indeed, thiol-ene chemistry has emerged to become the most important click chemistry technique. It involves an addition reaction of the thiol group with an alkene group through free radical reaction or in the presence of Michael catalyst (Scheme 2-2).129

Scheme 2-2 Thiol-ene click reaction where a single thiol reacts with a single ene through either free radical reaction or Michael catalyst addition reaction.

The importance of this click chemistry is due to the versatility of thiols that can be easily obtained in a wide range of polymerisation techniques. This includes from the cleavage of thiocarbonylthio end groups as the chain transfer agent (CTA) that is used for polymer synthesis via reversible addition-fragmentation chain transfer polymerization (RAFT). As shown in Scheme 2-3 where the RAFT-end group forms thiol after undergoing aminolysis process.

Scheme 2-3 Overview of the RAFT polymerization and subsequent aminolysis of RAFT end-groups to form thiol group

Over the years, RAFT polymerization has been increasingly used to create well defined polymer architectures on the nanotechnology applications.130 These nanostructures can be used as a template for the nanoparticle synthesis to control the growth of nanoparticles and particle size distribution.110 131 132 One way to do it is to form metal nanoparticles in a core of block copolymers micelle. Block copolymer micelle typically is synthesised from amphiphilick block copolymers that are dissolved in selective solvents (one block needs to be soluble and the other insoluble).134 135 The core size and shape will determine the nanoparticle morphology and it can be controlled by adjusting the nature of the blocks such as molecular weight and compatibility towards the solvent (Figure 2.4.2).

Literature Review

Figure 2.4.2 Examples of block copolymer micelles with different self-assembly structures to obtain Fe nanoparticles. Different structures were obtained by adjusting the solvent polarity ratios133

Another way is to use crosslinkers to create hyperbranched structures of polymer which create a polymer matrix with cavities (pore). Then, nanoparticles can form in this cavity with a really small size and narrow size distribution (Figure 2.4.3). For example, hypercrosslinked polystyrene has a large inner surface area (usually 1000 m2g-1) and pore size distribution of 2 nm. It has been used to form cobalt and platinum nanoparticles.136 137

Literature Review

2.5 Conclusion

A study of the current literature has revealed that Mg is a very promising hydrogen storage material due to its good reversibility for storing a high amount of H2. However, the challenges remain in the high temperature conditions for H2 release as well as slow kinetics.

Theoretical works and limited practical investigations have predicted that size reduction could be a means of improving the properties of the MgH2 for practical applications.

Preliminary studies have shown improvements in kinetic properties at the nanoscale but limited improvements were seen in the thermodynamic properties. Mg with a size less than

5 nm would have considerable improvement on destabilisation of Mg-H strength.

However, it is synthetically challenging to obtain such small Mg as free standing nanoparticles are highly unstable. Good stabilising agents are required and other methods developed in inorganic-organic synthesis may provide this opportunity.

This project will therefore aim to develop and optimise new methods for synthesising hybrid Mg nanoparticles. Their hydrogen storage properties will be evaluated in order to provide a pathway to the practical storage of hydrogen, particularly for on-board application.

Literature Review

2.6 References

1. Living Planet Report 2014; WWF International: 2014. 2. Li, S.; Gong, C., Revealed Comparative Economies of Scope and Development Index. 1995. 3. Jogoo, V. K., The Political Economy of Transitioning to a Green Economy in Mauritius. Transitioning to a Green Economy: Political Economy of Approaches in Small States 2014, 128. 4. Archer, D., Fate of fossil fuel CO2 in geologic time. Journal of Geophysical Research: Oceans (1978–2012) 2005, 110 (C9). 5. Hall, C. A.; Powers, R.; Schoenberg, W., Peak oil, EROI, investments and the economy in an uncertain future. In Biofuels, solar and wind as renewable energy systems, Springer: 2008; pp 109-132. 6. Shafiee, S.; Topal, E., When will fossil fuel reserves be diminished? Energy Policy 2009, 37 (1), 181-189. 7. Marland, G.; Boden, T. A.; Andres, R. J.; Brenkert, A.; Johnston, C., Global, regional, and national fossil fuel CO2 emissions. Trends: A compendium of data on global change 2003, 34-43. 8. Robertson, G. P.; Paul, E. A.; Harwood, R. R., Greenhouse gases in intensive agriculture: contributions of individual gases to the radiative forcing of the atmosphere. Science 2000, 289 (5486), 1922-1925. 9. Manne, A. S.; Richels, R. G., An alternative approach to establishing trade-offs among greenhouse gases. Nature 2001, 410 (6829), 675-677. 10. Mudd, G. M.; Diesendorf, M., Sustainability of uranium mining and milling: toward quantifying resources and eco-efficiency. Environmental Science & Technology 2008, 42 (7), 2624-2630. 11. Painuly, J. P., Barriers to renewable energy penetration; a framework for analysis. Renewable energy 2001, 24 (1), 73-89. 12. Dunn, B.; Kamath, H.; Tarascon, J.-M., Electrical energy storage for the grid: a battery of choices. Science 2011, 334 (6058), 928-935. 13. Conway, B. E., Transition from “supercapacitor” to “battery” behavior in electrochemical energy storage. Journal of the Electrochemical Society 1991, 138 (6), 1539-1548. 14. Cao, J.; Schofield, N.; Emadi, A. In Battery balancing methods: A comprehensive review, Vehicle Power and Propulsion Conference, 2008. VPPC'08. IEEE, IEEE: 2008; pp 1-6. 15. Malini, R.; Uma, U.; Sheela, T.; Ganesan, M.; Renganathan, N., Conversion reactions: a new pathway to realise energy in lithium-ion battery—review. Ionics 2009, 15 (3), 301-307. 16. Fetcenko, M.; Ovshinsky, S.; Reichman, B.; Young, K.; Fierro, C.; Koch, J.; Zallen, A.; Mays, W.; Ouchi, T., Recent advances in NiMH battery technology. Journal of Power Sources 2007, 165 (2), 544-551. 17. Hwang, J.-J., Sustainability study of hydrogen pathways for fuel cell vehicle applications. Renewable and Sustainable Energy Reviews 2013, 19, 220-229.

Literature Review

18. Van den Bossche, P.; Vergels, F.; Van Mierlo, J.; Matheys, J.; Van Autenboer, W., SUBAT: An assessment of sustainable battery technology. Journal of power sources 2006, 162 (2), 913- 919. 19. Turner, J. A., A realizable renewable energy future. Science 1999, 285 (5428), 687-689. 20. Barton, J. P.; Infield, D. G., Energy storage and its use with intermittent renewable energy. Energy Conversion, IEEE Transactions on 2004, 19 (2), 441-448. 21. Crabtree, G. W.; Dresselhaus, M. S.; Buchanan, M. V., The hydrogen economy. Physics Today 2004, 57 (12), 39-44. 22. Schug, C., Operational characteristics of high-pressure, high-efficiency water-hydrogen- electrolysis. International journal of hydrogen energy 1998, 23 (12), 1113-1120. 23. Züttel, A., Materials for hydrogen storage. Materials Today 2003, 6 (9), 24-33. 24. Weidenthaler, C.; Felderhoff, M., Solid-state hydrogen storage for mobile applications: Quo Vadis? Energy & Environmental Science 2011, 4 (7), 2495. 25. Durbin, D. J.; Malardier-Jugroot, C., Review of hydrogen storage techniques for on board vehicle applications. International Journal of Hydrogen Energy 2013, 38 (34), 14595-14617. 26. Stetson, N. Hydrogen Storage U.S. Department of Energy: 2013. 27. Law, K.; Rosenfeld, J.; Han, V.; Chan, M.; Chiang, H.; Leonard, J. US Department of Energy Hydrogen Storage Cost Analysis; TIAX LLC: 2013. 28. Paggiaro, R.; Bénard, P.; Polifke, W., Cryo-adsorptive hydrogen storage on activated carbon. I: Thermodynamic analysis of adsorption vessels and comparison with liquid and compressed gas hydrogen storage. international journal of hydrogen energy 2010, 35 (2), 638-647. 29. Dalebrook, A. F.; Gan, W.; Grasemann, M.; Moret, S.; Laurenczy, G., Hydrogen storage: beyond conventional methods. Chem. Commun. 2013, 49 (78), 8735-8751. 30. Ahluwalia, R.; Hua, T.; Peng, J.-K.; Lasher, S.; McKenney, K.; Sinha, J.; Gardiner, M., Technical assessment of cryo-compressed hydrogen storage tank systems for automotive applications. International journal of hydrogen energy 2010, 35 (9), 4171-4184. 31. Zheng, J.; Liu, X.; Xu, P.; Liu, P.; Zhao, Y.; Yang, J., Development of high pressure gaseous hydrogen storage technologies. International Journal of Hydrogen Energy 2012, 37 (1), 1048-1057. 32. Hong, J.-H.; Han, M.-G.; Chang, S.-H., Safety evaluation of 70MPa-capacity type III hydrogen pressure vessel considering material degradation of composites due to temperature rise. Composite Structures 2014, 113, 127-133. 33. Scott, J. H., The development of fuel cell technology for electric power generation: From NASA's manned space program to the" hydrogen economy". Proceedings of the IEEE 2006, 94 (10), 1815-1825. 34. Zhou, L.; Zhou, Y.; Sun, Y., Enhanced storage of hydrogen at the temperature of liquid nitrogen. International Journal of Hydrogen Energy 2004, 29 (3), 319-322. 35. Aceves, S. M.; Espinosa-Loza, F.; Ledesma-Orozco, E.; Ross, T. O.; Weisberg, A. H.; Brunner, T. C.; Kircher, O., High-density automotive hydrogen storage with cryogenic capable pressure vessels. International Journal of Hydrogen Energy 2010, 35 (3), 1219-1226. 36. Joó, F., Breakthroughs in hydrogen storage—Formic acid as a sustainable storage material for hydrogen. ChemSusChem 2008, 1 (10), 805-808.

Literature Review

37. Eberle, U.; Felderhoff, M.; Schueth, F., Chemical and physical solutions for hydrogen storage. Angewandte Chemie International Edition 2009, 48 (36), 6608-6630. 38. Palo, D. R.; Dagle, R. A.; Holladay, J. D., Methanol steam reforming for hydrogen production. Chemical reviews 2007, 107 (10), 3992-4021. 39. Marder, T. B., Will we soon be fueling our automobiles with ammonia–borane? Angewandte Chemie International Edition 2007, 46 (43), 8116-8118. 40. Stephens, F. H.; Pons, V.; Baker, R. T., Ammonia–borane: the hydrogen source par excellence? Dalton Transactions 2007, (25), 2613-2626. 41. Nijkamp, M.; Raaymakers, J.; Van Dillen, A.; De Jong, K., Hydrogen storage using physisorption–materials demands. Applied Physics A 2001, 72 (5), 619-623. 42. Dillon, A.; Heben, M., Hydrogen storage using carbon adsorbents: past, present and future. Applied Physics A 2001, 72 (2), 133-142. 43. Suh, M. P.; Park, H. J.; Prasad, T. K.; Lim, D. W., Hydrogen storage in metal-organic frameworks. Chemical reviews 2012, 112 (2), 782-835. 44. Germain, J.; Frechet, J. M.; Svec, F., Nanoporous polymers for hydrogen storage. Small 2009, 5 (10), 1098-111. 45. Collins, D. J.; Zhou, H.-C., Hydrogen storage in metal–organic frameworks. Journal of materials chemistry 2007, 17 (30), 3154-3160. 46. Cheng, H.-M.; Yang, Q.-H.; Liu, C., Hydrogen storage in carbon nanotubes. Carbon 2001, 39 (10), 1447-1454. 47. Barghi, S. H.; Tsotsis, T. T.; Sahimi, M., Chemisorption, physisorption and hysteresis during hydrogen storage in carbon nanotubes. International Journal of Hydrogen Energy 2014, 39 (3), 1390-1397. 48. Shiraishi, M.; Takenobu, T.; Yamada, A.; Ata, M.; Kataura, H., Hydrogen storage in single- walled carbon nanotube bundles and peapods. Chemical physics letters 2002, 358 (3), 213- 218. 49. Demirocak, D. E.; Ram, M. K.; Srinivasan, S. S.; Kumar, A.; Goswami, D. Y.; Stefanakos, E. K., Spillover enhancement for hydrogen storage by Pt doped hypercrosslinked polystyrene. International Journal of Hydrogen Energy 2012, 37 (17), 12402-12410. 50. Zacharia, R.; Rather, S.-u.; Hwang, S. W.; Nahm, K. S., Spillover of physisorbed hydrogen from sputter-deposited arrays of platinum nanoparticles to multi-walled carbon nanotubes. Chemical physics letters 2007, 434 (4), 286-291. 51. Reilly, R. M., Carbon nanotubes: potential benefits and risks of nanotechnology in nuclear medicine. Journal of Nuclear Medicine 2007, 48 (7), 1039-1042. 52. Kesanli, B.; Cui, Y.; Smith, M. R.; Bittner, E. W.; Bockrath, B. C.; Lin, W., Highly interpenetrated metal–organic frameworks for hydrogen storage. Angewandte Chemie International Edition 2005, 44 (1), 72-75. 53. Chae, H. K.; Siberio-Pérez, D. Y.; Kim, J.; Go, Y.; Eddaoudi, M.; Matzger, A. J.; O'Keeffe, M.; Yaghi, O. M., A route to high surface area, porosity and inclusion of large molecules in crystals. Nature 2004, 427 (6974), 523-527. 54. Dinca, M.; Dailly, A.; Liu, Y.; Brown, C. M.; Neumann, D. A.; Long, J. R., Hydrogen storage in a microporous metal-organic framework with exposed Mn2+ coordination sites. Journal of the American Chemical Society 2006, 128 (51), 16876-16883.

Literature Review

55. Kong, X.; Scott, E.; Ding, W.; Mason, J. A.; Long, J. R.; Reimer, J. A., CO2 dynamics in a metal– organic framework with open metal sites. Journal of the American Chemical Society 2012, 134 (35), 14341-14344. 56. Sakintuna, B.; Lamari-Darkrim, F.; Hirscher, M., Metal hydride materials for solid hydrogen storage: a review. International Journal of Hydrogen Energy 2007, 32 (9), 1121-1140. 57. B. Arboleda Jr, N.; Kasai, H.; Nobuhara, K.; A. Diño, W.; Nakanishi, H., Dissociation and sticking of H2 on Mg (0001), Ti (0001) and La (0001) surfaces. Journal of the Physical Society of Japan 2004, 73 (3), 745-748. 58. Yamauchi, M.; Ikeda, R.; Kitagawa, H.; Takata, M., Nanosize effects on hydrogen storage in palladium. The Journal of Physical Chemistry C 2008, 112 (9), 3294-3299. 59. Schüth, F.; Bogdanović, B.; Felderhoff, M., Light metal hydrides and complex hydrides for hydrogen storage. Chemical communications 2004, (20), 2249-2258. 60. Zidan, R.; Garcia-Diaz, B. L.; Fewox, C. S.; Stowe, A. C.; Gray, J. R.; Harter, A. G., Aluminium hydride: a reversible material for hydrogen storage. Chem. Commun. 2009, (25), 3717- 3719. 61. Aguey-Zinsou, K.-F.; Ares-Fernández, J.-R., Hydrogen in magnesium: new perspectives toward functional stores. Energy & Environmental Science 2010, 3 (5), 526. 62. Orimo, S.-i.; Nakamori, Y.; Eliseo, J. R.; Züttel, A.; Jensen, C. M., Complex hydrides for hydrogen storage. Chemical reviews 2007, 107 (10), 4111-4132. 63. Bogdanović, B.; Felderhoff, M.; Kaskel, S.; Pommerin, A.; Schlichte, K.; SCHüTH, F., Improved Hydrogen Storage Properties of Ti‐Doped Sodium Alanate Using Titanium Nanoparticles as Doping Agents. Advanced materials 2003, 15 (12), 1012-1015. 64. Li, H.-W.; Yan, Y.; Orimo, S.-i.; Züttel, A.; Jensen, C. M., Recent progress in metal borohydrides for hydrogen storage. Energies 2011, 4 (1), 185-214. 65. Mauron, P.; Buchter, F.; Friedrichs, O.; Remhof, A.; Bielmann, M.; Zwicky, C. N.; Züttel, A., Stability and reversibility of LiBH4. The Journal of Physical Chemistry B 2008, 112 (3), 906- 910. 66. Rude, L. H.; Nielsen, T. K.; Ravnsbaek, D. B.; Bösenberg, U.; Ley, M. B.; Richter, B.; Arnbjerg, L. M.; Dornheim, M.; Filinchuk, Y.; Besenbacher, F., Tailoring properties of borohydrides for hydrogen storage: a review. physica status solidi (a) 2011, 208 (8), 1754-1773. 67. Christian, M. L.; Aguey-Zinsou, K.-F. o., Core–Shell Strategy Leading to High Reversible Hydrogen Storage Capacity for NaBH4. ACS nano 2012, 6 (9), 7739-7751. 68. Friedrichs, O.; Sánchez-López, J. C.; López-Cartes, C.; Klassen, T.; Bormann, R.; Fernandez, A., Nb2O5 “pathway effect” on hydrogen sorption in Mg. The Journal of Physical Chemistry B 2006, 110 (15), 7845-7850. 69. Aguey-Zinsou, K.-F.; Ares Fernandez, J.; Klassen, T.; Bormann, R., Effect of Nb 2 O 5 on MgH 2 properties during mechanical milling. International Journal of Hydrogen Energy 2007, 32 (13), 2400-2407. 70. Nogita, K.; Ockert, S.; Pierce, J.; Greaves, M.; Gourlay, C.; Dahle, A., Engineering the Mg– Mg 2 Ni eutectic transformation to produce improved hydrogen storage alloys. International Journal of Hydrogen Energy 2009, 34 (18), 7686-7691. 71. Moroz, S.; Tan, X. F.; Pierce, J.; Greaves, M.; Duguid, A.; Dumur, K.; Ng, J., Systems based on hypo-eutectic Mg–Mg 2 Ni alloys for medium to large scale hydrogen storage and delivery. Journal of Alloys and Compounds 2013, 580, S329-S332.

Literature Review

72. Kappes, M.; Iannuzzi, M.; Carranza, R. M., Hydrogen Embrittlement of Magnesium and Magnesium Alloys: A Review. Journal of the Electrochemical Society 2013, 160 (4), C168- C178. 73. Orimo, S.; Fujii, H., Materials science of Mg-Ni-based new hydrides. Applied Physics A 2001, 72 (2), 167-186. 74. Vajo, J. J.; Mertens, F.; Ahn, C. C.; Bowman, R. C.; Fultz, B., Altering hydrogen storage properties by hydride destabilization through alloy formation: LiH and MgH2 destabilized with Si. The Journal of Physical Chemistry B 2004, 108 (37), 13977-13983. 75. Wu, Z.; Allendorf, M. D.; Grossman, J. C., Quantum Monte Carlo simulation of nanoscale MgH2 cluster thermodynamics. Journal of the American Chemical Society 2009, 131 (39), 13918-13919. 76. Nanda, K.; Maisels, A.; Kruis, F.; Fissan, H.; Stappert, S., Higher surface energy of free nanoparticles. Physical review letters 2003, 91 (10), 106102. 77. Dai, J. H.; Song, Y.; Yang, R., Intrinsic mechanisms on enhancement of hydrogen desorption from MgH2 by (001) surface doping. International Journal of Hydrogen Energy 2011, 36 (20), 12939-12949. 78. Pozzo, M.; Alfe, D., Hydrogen dissociation and diffusion on transition metal (= Ti, Zr, V, Fe, Ru, Co, Rh, Ni, Pd, Cu, Ag)-doped Mg (0001) surfaces. International journal of hydrogen energy 2009, 34 (4), 1922-1930. 79. Wu, G.; Zhang, J.; Wu, Y.; Li, Q.; Chou, K.; Bao, X., First-principle calculations of the adsorption, dissociation and diffusion of hydrogen on the Mg (0001) surface. Acta Physico- Chimica Sinica 2008, 24 (1), 55-60. 80. Pundt, A., Hydrogen in Nano‐sized Metals. Advanced Engineering Materials 2004, 6 (1‐2), 11-21. 81. Shirinyan, A.; Wautelet, M., Phase separation in nanoparticles. Nanotechnology 2004, 15 (12), 1720. 82. Berube, V.; Radtke, G.; Dresselhaus, M.; Chen, G., Size effects on the hydrogen storage properties of nanostructured metal hydrides: a review. International Journal of Energy Research 2007, 31 (6‐7), 637-663. 83. Paskevicius, M.; Sheppard, D. A.; Buckley, C. E., Thermodynamic changes in mechanochemically synthesized magnesium hydride nanoparticles. Journal of the American Chemical Society 2010, 132 (14), 5077-5083. 84. Dunitz, J. D., Win some, lose some: enthalpy-entropy compensation in weak intermolecular interactions. Chemistry & biology 1995, 2 (11), 709-712. 85. Krug, R.; Hunter, W.; Grieger, R., Enthalpy-entropy compensation. 1. Some fundamental statistical problems associated with the analysis of Van't Hoff and Arrhenius data. The Journal of Physical Chemistry 1976, 80 (21), 2335-2341. 86. Mulder, F. M.; Singh, S.; Bolhuis, S.; Eijt, S. W., Extended solubility limits and nanograin refinement in Ti/Zr fluoride-catalyzed MgH2. The Journal of Physical Chemistry C 2012, 116 (2), 2001-2012. 87. Tang, W. S.; Chotard, J.-N.; Raybaud, P.; Janot, R., Enthalpy–Entropy Compensation Effect in Hydrogen Storage Materials: Striking Example of Alkali Silanides MSiH3 (M= K, Rb, Cs). The Journal of Physical Chemistry C 2014, 118 (7), 3409-3419.

Literature Review

88. Paik, B.; Walton, A.; Mann, V.; Book, D.; Jones, I.; Harris, I., Microstructure of ball milled MgH 2 powders upon hydrogen cycling: An electron microscopy study. International Journal of Hydrogen Energy 2010, 35 (17), 9012-9020. 89. Finney, E. E.; Finke, R. G., Nanocluster nucleation and growth kinetic and mechanistic studies: A review emphasizing transition-metal nanoclusters. Journal of colloid and interface science 2008, 317 (2), 351-374. 90. Mer, V. K. L., Nucleation in Phase Transitions. Industrial & Engineering Chemistry 1952, 44 (6), 1270-1277. 91. Murray, C. B.; Kagan, C.; Bawendi, M., Synthesis and characterization of monodisperse nanocrystals and close-packed nanocrystal assemblies. Annual Review of Materials Science 2000, 30 (1), 545-610. 92. Von Weimarn, P., The Precipitation Laws. Chemical reviews 1925, 2 (2), 217-242. 93. Barlow, D. A.; Baird, J. K.; Su, C.-H., Theory of the von Weimarn rules governing the average size of crystals precipitated from a supersaturated solution. Journal of crystal growth 2004, 264 (1), 417-423. 94. Reiss, H., The growth of uniform colloidal dispersions. The Journal of Chemical Physics 1951, 19 (4), 482-487. 95. Vetter, T.; Iggland, M.; Ochsenbein, D. R.; Hänseler, F. S.; Mazzotti, M., Modeling Nucleation, Growth, and Ostwald Ripening in Crystallization Processes: A Comparison between Population Balance and Kinetic Rate Equation. Crystal Growth & Design 2013, 13 (11), 4890-4905. 96. Friedrichs, O.; Sánchez-López, J. C.; López-Cartes, C.; Dornheim, M.; Klassen, T.; Bormann, R.; Fernández, A., Chemical and microstructural study of the oxygen passivation behaviour of nanocrystalline Mg and MgH2. Applied Surface Science 2006, 252 (6), 2334-2345. 97. Swihart, M. T., Vapor-phase synthesis of nanoparticles. Current Opinion in Colloid & Interface Science 2003, 8 (1), 127-133. 98. Aguey-Zinsou, K.-F.; Ares-Fernández, J.-R., Synthesis of colloidal magnesium: a near room temperature store for hydrogen. Chemistry of Materials 2007, 20 (2), 376-378. 99. Kalidindi, S. B.; Jagirdar, B. R., Highly monodisperse colloidal magnesium nanoparticles by room temperature digestive ripening. Inorganic chemistry 2009, 48 (10), 4524-4529. 100. Jeon, K. J.; Moon, H. R.; Ruminski, A. M.; Jiang, B.; Kisielowski, C.; Bardhan, R.; Urban, J. J., Air-stable magnesium nanocomposites provide rapid and high-capacity hydrogen storage without using heavy-metal catalysts. Nature materials 2011, 10 (4), 286-90. 101. Gross, A. F.; Vajo, J. J.; Van Atta, S. L.; Olson, G. L., Enhanced hydrogen storage kinetics of LiBH4 in nanoporous carbon scaffolds. The Journal of Physical Chemistry C 2008, 112 (14), 5651-5657. 102. Jongh, P. E. d.; Wagemans, R. W.; Eggenhuisen, T. M.; Dauvillier, B. S.; Radstake, P. B.; Meeldijk, J. D.; Geus, J. W.; Jong, K. P. d., The preparation of carbon-supported magnesium nanoparticles using melt infiltration. Chemistry of Materials 2007, 19 (24), 6052-6057. 103. Zhang, S.; Gross, A. F.; Van Atta, S. L.; Lopez, M.; Liu, P.; Ahn, C. C.; Vajo, J. J.; Jensen, C. M., The synthesis and hydrogen storage properties of a MgH2 incorporated carbon aerogel scaffold. Nanotechnology 2009, 20 (20), 204027. 104. Jia, Y.; Sun, C.; Cheng, L.; Abdul Wahab, M.; Cui, J.; Zou, J.; Zhu, M.; Yao, X., Destabilization of Mg-H bonding through nano-interfacial confinement by unsaturated carbon for

Literature Review

hydrogen desorption from MgH2. Physical chemistry chemical physics : PCCP 2013, 15 (16), 5814-20. 105. Konarova, M.; Tanksale, A.; Norberto Beltramini, J.; Qing Lu, G., Effects of nanoconfinement on the hydrogen desorption properties of MgH 2. Nano Energy 2013, 2 (1), 98-104. 106. de Jongh, P. E.; Adelhelm, P., Nanosizing and nanoconfinement: new strategies towards meeting hydrogen storage goals. ChemSusChem 2010, 3 (12), 1332-1348. 107. Zhao-Karger, Z.; Hu, J.; Roth, A.; Wang, D.; Kübel, C.; Lohstroh, W.; Fichtner, M., Altered thermodynamic and kinetic properties of MgH2 infiltrated in microporous scaffold. Chemical communications 2010, 46 (44), 8353-8355. 108. Bogerd, R.; Adelhelm, P.; Meeldijk, J. H.; de Jong, K. P.; de Jongh, P. E., The structural characterization and H2 sorption properties of carbon-supported Mg1− xNix nanocrystallites. Nanotechnology 2009, 20 (20), 204019. 109. Aguey‐Zinsou, K. F.; Boyer, C., Synthesis and Stabilisation of MgH2 Nanoparticles by Self‐ Assembly. ChemPlusChem 2012, 77 (6), 423-426. 110. Cushing, B. L.; Kolesnichenko, V. L.; O'Connor, C. J., Recent advances in the liquid-phase syntheses of inorganic nanoparticles. Chemical reviews 2004, 104 (9), 3893-3946. 111. Rieke, R. D.; Bales, S. E., Activated metals. IV. Preparation and reactions of highly reactive magnesium metal. Journal of the American Chemical Society 1974, 96 (6), 1775-1781. 112. Rieke, R. D.; Li, P. T.-J.; Burns, T. P.; Uhm, S. T., Preparation of highly reactive metal powders. New procedure for the preparation of highly reactive zinc and magnesium metal powders. The Journal of Organic Chemistry 1981, 46 (21), 4323-4324. 113. Song, M.-R.; Chen, M.; Zhang, Z.-J., Preparation and characterization of Mg nanoparticles. Materials Characterization 2008, 59 (5), 514-518. 114. Norberg, N. S.; Arthur, T. S.; Fredrick, S. J.; Prieto, A. L., Size-dependent hydrogen storage properties of Mg nanocrystals prepared from solution. Journal of the American Chemical Society 2011, 133 (28), 10679-81. 115. Haas, I.; Gedanken, A., Synthesis of metallic magnesium nanoparticles by sonoelectrochemistry. Chem. Commun. 2008, (15), 1795-1797. 116. Bogdanovic, B., Magnesium anthracene systems and their application in synthesis and catalysis. Accounts of Chemical Research 1988, 21 (7), 261-267. 117. Bogdanović, B.; Liao, S. T.; Mynott, R.; Schlichte, K.; Westeppe, U., Rate of formation and characterization of magnesium anthracene. Chemische Berichte 1984, 117 (4), 1378-1392. 118. -systems for hydrogen storage. International journal of hydrogen energy 1987, 12 (12), 863-873. 119. Bogdanović, B., Catalytic synthesis of organolithium and organomagnesium compounds and of lithium and magnesium hydrides—Applications in organic synthesis and hydrogen storage. Angewandte Chemie International Edition in English 1985, 24 (4), 262-273. 120. Faust, J.; Whitney, E.; Batha, H.; Heying, T.; Fogle, G., Catalytic preparation of magnesium hydride. Journal of Applied Chemistry 1960, 10 (4), 187-188. 121. Becker, W. E.; Ashby, E. C., Hydrogenolysis of the Grignard Reagent. The Journal of Organic Chemistry 1964, 29 (4), 954-955.

Literature Review

122. Wiberg, E.; Bauer, R., Der Magnesiumwasserstoff MgH2. Chemische Berichte 1952, 85 (6), 593-605. 123. Ashby, E. C.; Kovar, R.; Kawakami, K., Existence of HMgX compounds. Inorganic chemistry 1970, 9 (2), 317-324. 124. Artuso, F.; D'Archivio, A. A.; Lora, S.; Jerabek, K.; Kralik, M.; Corain, B., Nanomorphology of polymer frameworks and their role as templates for generating size-controlled metal nanoclusters. Chemistry 2003, 9 (21), 5292-6. 125. Akelah, A.; Sherrington, D. C., Application of functionalized polymers in organic synthesis. Chemical reviews 1981, 81 (6), 557-587. 126. McQuade, D. T.; Pullen, A. E.; Swager, T. M., Conjugated polymer-based chemical sensors. Chemical reviews 2000, 100 (7), 2537-2574. 127. Ramanathan, T.; Abdala, A.; Stankovich, S.; Dikin, D.; Herrera-Alonso, M.; Piner, R.; Adamson, D.; Schniepp, H.; Chen, X.; Ruoff, R., Functionalized graphene sheets for polymer nanocomposites. Nature nanotechnology 2008, 3 (6), 327-331. 128. Fournier, D.; Hoogenboom, R.; Schubert, U. S., Clicking polymers: a straightforward approach to novel macromolecular architectures. Chemical Society Reviews 2007, 36 (8), 1369-1380. 129. Lowe, A. B., Thiol-ene “click” reactions and recent applications in polymer and materials synthesis. Polymer Chemistry 2010, 1 (1), 17-36. 130. Roth, P. J.; Boyer, C.; Lowe, A. B.; Davis, T. P., RAFT polymerization and thiol chemistry: a complementary pairing for implementing modern macromolecular design. Macromolecular rapid communications 2011, 32 (15), 1123-43. 131. Chen, Z.-X.; Xu, G.-Y.; Yang, G.-C.; Wang, W., Preparation of non-cross-linked polystyrene- supported quaternary ammonium salts and use as phase transfer catalysts under microwave. Reactive and Functional Polymers 2004, 61 (1), 139-146. 132. Lanza, G.; Fragalà, I. L.; Marks, T. J., Ligand substituent, anion, and solvation effects on ion pair structure, thermodynamic stability, and structural mobility in “constrained geometry” olefin polymerization catalysts: An ab initio quantum chemical investigation. Journal of the American Chemical Society 2000, 122 (51), 12764-12777. 133. Hu, J.; Qian, Y.; Wang, X.; Liu, T.; Liu, S., Drug-loaded and superparamagnetic iron oxide nanoparticle surface-embedded amphiphilic block copolymer micelles for integrated chemotherapeutic drug delivery and MR imaging. Langmuir : the ACS journal of surfaces and colloids 2011, 28 (4), 2073-2082. 134. Zhang, L.; Niu, H.; Chen, Y.; Liu, H.; Gao, M., Preparation of platinum nanoparticles using star-block copolymer with a carboxylic core. Journal of colloid and interface science 2006, 298 (1), 177-82. 135. Gao, H., Development of star polymers as unimolecular containers for nanomaterials. Macromolecular rapid communications 2012, 33 (9), 722-34. 136. Sidorov, S.; Bronstein, L.; Davankov, V.; Tsyurupa, M.; Solodovnikov, S.; Valetsky, P.; Wilder, E.; Spontak, R., Cobalt nanoparticle formation in the pores of hyper-cross-linked polystyrene: control of nanoparticle growth and morphology. Chemistry of materials 1999, 11 (11), 3210-3215.

Literature Review

137. Doluda, V. Y.; Sulman, E.; Matveeva, V.; Sulman, M.; Lakina, N.; Sidorov, A.; Valetsky, P.; Bronstein, L., Kinetics of phenol oxidation over hypercrosslinked polystyrene impregnated with Pt nanoparticles. Chemical Engineering Journal 2007, 134 (1), 256-261.

3 EXPERIMENTAL AND CHARACTERISATION TECHNIQUES

3.1 INTRODUCTION ……………………………………………………………………………………………….43 3.2 SYNTHESIS INSTRUMENTS AND PROCEDURES……………………………………………………….. …..43 3.2.1 Pressure Reactor Vessel ……………………………………………………………………………44 3.2.2 Pressure Reactor Rig ……………………………………………………………………………45 3.3 FUNDAMENTAL CHARACTERISATION TECHNIQUES …………………………………………………47 3.3.1 X-ray diffraction (XRD) ……………………………………………………………………………47 3.3.2 Electron Microscopy ……………………………………………………………………………49 3.3.3 XPS Analysis ………………………………………………………………………………………..51 3.3.4 NMR Analysis ………………………………………………………………………………………..53 3.3.5 Elemental Analysis ……………………………………………………………………………53 3.3.6 BET analysis ………………………………………………………………………………………..54 3.4 HYDROGEN STORAGE PROPERTIES CHARACTERISATION TECHNIQUES ………………………….55 3.4.1 TPD analysis ………………………………………………………………………………………..55 3.4.2 Gravimetric technique ……………………………………………………………………………58 3.4.3 Sievert / Volumetric method ……………………………………………………………….60 3.5 CONCLUSION ……………………………………………………………………………………………….63 3.6 REFERENCES ……………………………………………………………………………………………….64

Experimental and Characterisation Techniques

3.1 Introduction

The aim of this chapter is to describe the range of analytical techniques and methods used to characterise the materials synthesised. In this work, the characterisation techniques are divided into two sections. The first section is related to the investigation of the material’s fundamental properties while the second section concerns the analysis of their hydrogen storage properties. Typically, the as-prepared materials are physically characterised before and after dehydrogenation and hydrogenation cycles. The overall experimental procedures are shown in Figure 3.1.1.

Synthesis

Structural characterisation

XRD TEM XPS Hydrogen storage properties characterisation

TPD PCT Kinetics

Figure 3.1.1 The overall procedure of the experiments

3.2 Synthesis instruments and procedures

As mentioned in the literature review section, oxidations and impurities can be detrimental towards the properties of nanostructured MgH2. This is because MgH2 nanoparticles are extremely sensitive to air and moisture which cause the particles to be easily oxidised.

Experimental and Characterisation Techniques

Therefore, all experiments in this study were performed using clean and dry glassware and high purity commercial reagents. Solvents were ensure with H2O < 5 ppm and O2 < 1 ppm while prepared by drying solvent units and molecular sieves. Furthermore, all handling and storage of the reagents and samples was carried out in an inert atmosphere. Most of the experiments were conducted in a glovebox running under high purity Ar (less than 1 ppm

O2 and H2O) or in other sealed environments. A high pressure reactor was used for the liquid based synthetic methods because high temperature and pressure was required. On the other hand a high pressure rig instrument was used for the dry synthetic methods and hydrogen treatment of all synthesised materials before further characterisation.

3.2.1 Pressure Reactor Vessel

The majority of the materials investigated in this thesis were obtained from liquid phase reactions at high temperatures and pressures. This set of experiments was carried out in a stainless steel pressure reactor vessel (Series 4560, Parr Instrument Company, capacity 600 ml). Figure 3.2.1 shows the component diagram of the reactor and its typical experimental setup. It has a steering shaft positioned near the bottom of the reactor vessel that ensures proper mixing during the liquid phase reaction. The stirring speed used for most experiments was set and held constant at 150 rpm. Heating was done by a furnace equipped with a temperature controller (Series 4840 Modular Controllers, Parr Instrument

Company). During the synthesis, the reactor vessel was inserted into the furnace and the reaction temperature was monitored by the temperature controller equipped with a J-type thermocouple. Other important components of the reactor include the pressure gauge, the gas inlet and the release valves for pressure adjustment.

Typically, the starting materials were transferred into the reactor inside the glovebox before it was assembled and installed outside on the bench. After the connection of each inlets/outlets valve, the vessel was purged with H2 gas at least 5 times to replace the Ar. A

Experimental and Characterisation Techniques

leak test was performed prior to the start of heating as it was crucial not to let the air and oxygen enter the reactor during the experiments. An initial pressure of 10 bar was set to prevent solvent from boiling off. The pressure was then increased to the desired experimental value until it had reached the final temperature of the synthesis. At the end of the synthesis, the reactor was cooled to the room temperature by stopping the heating and removing the vessel from the furnace. The hydrogen gas was released when the vessel was at room temperature, and the reactor was transferred back into the glovebox to be disassembled. Occasionally, the experiments were carried out without solvents such that solid samples were collected by directly scraping them out through the wall of vessel. When the experiments were carried out in solvents, the solid residues/samples were collected by centrifugation and washed with appropriate solvents. Centrifugation was usually done in the range of 10,000-15,000 rpm for 10 to 15 min. The samples were allowed to dry under the Schlenk line for 24 h before further characterisations.

Gas outlet valve thermocouple Pressure Gas inlet valve gauge

stirrer

furnace

Figure 3.2.1 Schematic diagram of the pressure reactor vessel (left) and the bench top setup during high temperature and pressure reaction (right)

3.2.2 Pressure Reactor Rig

Hydrogenation/dehydrogenation cycles of the synthesised materials were done in a pressure rig chamber. This instrument was used to anneal and clean the materials from

Experimental and Characterisation Techniques

unwanted by-products of the reactions. It was done by manually controlling the hydrogenation/dehydrogenation process where the rig chamber was heated to a certain temperature either under vacuum or H2 pressure. A schematic drawing of the apparatus is shown in Figure 3.2.2. It has a thermocouple that is connected to the temperature controller and attached to the outside of the sample chamber to measure the temperature of the sample. The temperature controller regulates the furnace to heat the sample chamber to the set temperature. The pressure in the gas reservoir was monitored using a pressure gauge and was adjusted with manual valves. Typically, a sample was first dehydrogenated under a dynamic vacuum to ensure the removal of by-products for 3-5 h. Then, the rest of the dehydrogenation steps were performed in a closed static vacuum. The length of the dehydrogenation time was determined from the kinetics measurements of the materials with either a Sievert or Gravimetric instrument. The sample usually underwent 3 hydrogenation/dehydrogenation cycles before it was characterised further as a cycled material.

Figure 3.2.2 (Left) Schematic diagram of the high pressure manual sorption apparatus for hydrogen storage studies and the corresponding apparatus setup (Right)

Experimental and Characterisation Techniques

3.3 Physical and structural characterisation techniques

3.3.1 X-ray diffraction (XRD)

XRD is a powerful qualitative analysis tool to confirm the phase of Mg/MgH2 formed after the synthesis and hydrogen cycling. This analysis is commonly used to determine material phases with a crystalline or semi-crystalline structure. It is achieved by measuring the scattered intensity of X-ray beams from a sample as function of incident and scattered angle and wavelength. This phenomena is governed by Bragg’s law which is represented by the following equation.

ퟐ풅 풔풊풏 휽 = 풏흀 (3-1) where 휃 is the incident beam angle, 푑 is the distance between two successive crystallographic planes, 휆 is the wavelength of the X-ray, and n is an integer number. Based on this law, the interference would occur if the path difference of the beams diffracted from two adjacent planes creating patterns which are recorded by an X-ray diffractometer.1 As shown in Figure 3.3.1 , the diffraction pattern generated as a function of 2휃 which provides information on the crystal structure properties of the materials. The diffraction pattern is typically compared to data from the database of known materials for specific identification of a given phase.

Figure 3.3.1 Schematic diagram of an X-ray diffractometer measuring ퟐ휽

Experimental and Characterisation Techniques

Besides phase identifications, other crystal information such as a sample’s crystalline size can be obtained from the recorded XRD patterns data. However, it only applies for crystallites in < 100 nm in size where the broadening of the X-ray diffraction pattern corresponds to the actual size of the particles.2 It can be calculated with Scherrer’s equation which is expressed as:

푲흀 푫 = (3-2) 휷풄풐풔휽 where D is the crystallite size diameter, 훽 is the full width of half maxima (FWHM), and 휃 is the incident beam angle, and K is the dimensionless shape factor. The value of K depends on the shape of the particles analysed, but throughout this study, the value was set to be fixed at 0.89 which represents the K factor of spherical shape.2 The XRD instrument used in this study was a Philips X’pert Multipurpose XRD system which operated at 40 mA and 45 kV with a monochromated Cu Kα radiation. It should be noted that the value of 휆 corresponding to Cu is equal to 1.541 Å.

(A) (B)

Figure 3.3.2 (A) XRD sample holder with kapton foil used in this study (B) Typical

XRD pattern of crystalline MgH2 and sample holder with kapton foil.

The samples used throughout this study were in the form of powder and they were loaded onto a stainless steel sample holder as shown in Figure 3.3.2-A. Kapton foil was used to protect the sample from exposure to the air. Figure 3.3.2-B shows the typical XRD patterns of fully crystalline MgH2 with the protective film. MgH2 has main peaks at 27.947 °2휃 (110),

35.744 °2휃 (100), 39.856 °2휃 (200) and 54.617 °2휃 (211). Crystallography data of these

Experimental and Characterisation Techniques

peaks such as FWHM (obtained from the X’pert High Score Plus software) and 휃 were used to calculate the average crystallite sizes for MgH2.

3.3.2 Electron Microscopy

Electron microscopy is an important tool to characterise the physical structures of nanomaterials. There were two types of electron microscopy used in this study, Scanning

Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM). SEM is useful to observe the sample’s morphology and topography but it can be quite challenging if the nanoparticle’s size is less than 100 nm. TEM allows for imaging at higher resolutions and the sample’s morphology can also be clearly observed. However, the topography of the nanoparticle’s surface is impossible to determine.

Throughout this study, TEM was used more extensively as compared to SEM since the synthesised particles were really small. TEM is also capable of performing supplementary characterisations to obtain chemical compositions or crystallographic information by

Energy Dispersive X-Ray Spectroscopy (EDS) and Selected Area Electron Diffraction (SAED) analyses.

Unlike normal microscopy, TEM uses a high energy electron beam rather than optical light.

As shown in Figure 3.3.3, a beam of electrons is transmitted and passes through an ultra- thin specimen on which materials are deposited. This results in the formation of an image which is then magnified and focused onto an imaging device captured by a CCD camera. TEM instrument used in this study was a Phillips CM200 microscope which operated with the electrons energy at 200 keV. The optimum resolution is 0.24 nm. The mode of operation for all TEM analyses in this study was in bright field imaging mode. In this mode, the thicker regions of the sample or regions with a higher atomic number appeared dark while regions with no sample in the beam path appeared bright.

Experimental and Characterisation Techniques

Figure 3.3.3 Schematic diagram of TEM instrument

EDS analysis can be performed by TEM to determine the element composition of the sample’s selected regions. The analysis can involve just a simple scan that shows the element compositions on the regions or it can be extended to the mapping of the elements on the image. The examples of these EDS analyses can be seen in Figure 3.3.4. The EDS analyses were performed regularly during the TEM image experiments to ensure the images corresponded to the right materials. Dust or other small contaminant particles could be observed during an analysis since the sample preparation was not done in a free contaminations environment. The sample preparations were done inside the glovebox by suspending a small amount of materials in THF solution. Tetrabutylammonium Bromide

(TBAB) was usually added to the suspension to act as a dispersant and it was then sonicated for a few seconds. Subsequently, a drop of the suspension was subjected to a carbon coated copper TEM grid. Most of the time, the sample preparation was done 3-4 h prior to the analysis in order to allow the solvent to dry.

Few SEM analyses were performed during this study. It was carried out by using a Hitachi

S-900 with an accelerating voltage of 2kV. It has resolution ranging from 0.8 nm at 30 kV and 3-4 nm resolution at 1 kV. A very small amount of fine powder samples was sprinkled

Experimental and Characterisation Techniques

onto a carbon tape which was stuck on the gold sample rod. It was then placed inside the instrument where it was exposed to a current of electron beams. As a result, electron signals were emitted from the specimen for direct imaging mostly as secondary electron (SE) signals. SE signals are highly dependent on the surface features of the specimen and only emitted from the top thin layer of the specimen due to the low energies of SE.3

(A) (B)

Figure 3.3.4 (A) EDS spectrum of MgCl2 and MgH2 particles , (B) EDS mapping of Mg element on MgH2 heterogeneous particles

3.3.3 XPS Analysis

Surface analyses of nanoparticles usually contain useful and important information which can be related to many of the surface properties and their capability to split H2. In this case, the chemistry on nanomaterial surfaces often is correlated with the observed hydrogen storage properties. Therefore, one of the most useful techniques to investigate solid surface materials is the X-ray Photoelectron Spectroscopy (XPS) of which the elemental composition, empirical formula, and chemical or electronic state of the elements can be obtained.4

Experimental and Characterisation Techniques

Photoelectron

EFermi level

2p Binding Energy Photon 2s

Figure 3.3.5 Schematic diagram of the XPS process of electron from the 1s shell

This technique works by irradiating sample materials with monochromatic photons under a high vacuum which in turns leads to the generation of electrons by the irradiated materials with certain kinetic energy. Since each element has a unique group of binding energies, this can be used to quantify the different elements present in the sample and know their chemical activities. There are two types of electrons emission resulting from this irradiation process. These are of the normal photoemission process (XPS) or of the Auger electron emission process (AES). In the XPS process, electrons are emitted from the sample after they are ejected from their atomic energy level by an X-ray photon.5 The overall process can be seen in the Figure 3.3.5 showing the emission of an electron from the 1s shell of an atom which is then recorded by the spectrometer.

In this study only the XPS process was applied and the instrument used was EscaLab 220-

IXL from VG Scientific, which operated with a monochromated Al-Ka radiation at 1486.60 eV and a power source of 120 W. A spot size of 0.5 mm in diameter with a pass energy of

100 eV was used for wide scans and a pass energy of 20 eV was used for narrow scans of particular elemental peaks. Only a small amount of sample was needed for the experiment since it had spatial resolution. Surfaces with up to 5 nm of the material were analysed and the exposure to air was always minimised during the sample transfer into the instrument.

Experimental and Characterisation Techniques

3.3.4 NMR Analysis

Nuclear Magnetic Resonance (NMR) analysis was performed to identify the chemical structures of organic molecules and other compounds. Different atoms within a molecule resonate at different frequencies at a given field strength. This is influenced by the element and the surrounding species. NMR provides valuable information of the overall chemical structures.

Most of the organic compounds synthesised during this study were dissolved in deuterated solvents, mostly d-chloroform (CDCl3). NMR spectra of these compounds were obtained from a Bruker 300 MHz spectrometer.

Nevertheless, inorganic samples such as MgH2 based materials are often insoluble in any solvents hence their NMR analysis must be done in the solid state. This was performed with another instrument; Bruker Avance III NMR spectrometer. Typically, the samples (approx.

50–80 mg) were packed into 4 mm o.d. zirconia rotors, capped with Kel-F rotor caps. Then, the samples were spun at the magic angle and at 12 kHz under the control of a Bruker MAS

II unit (to 2 Hz) to eliminate potential interference from spinning side-bands. Spectra were acquired at the 13C frequency of 75.47 MHz at the magic angle spinning (MAS) with cross polarisation (CP). CP was initiated by a 2.5 ms 1H pulse, followed by a 2 ms contact time, ramped for 1H, and a 5–20 s recycle time; or with a 13C excitation pulse of 2.0 ms (451 pulse) with HP decoupling on 1H, and 6–30 s recycled time. However, the detection limit of the solid NMR in a favourable condition is about 1% and anything beyond that is impossible to detect.

3.3.5 Elemental Analysis

Quantification of a material’s element composition is important to give valuable information about the purity of the material, which can be useful for calculating its hydrogen capacity.

Experimental and Characterisation Techniques

This is possible by performing the quantitative measurement of Mg content and comparing it with H2 storage capacity measured by TPD analysis or a Sievert instrument. The measurement is done by Inductively Coupled Plasma Optical Emission Spectrometry (ICP-

OES). The instrument used throughout this study was a Perkin Elmer OPTIMA7300 ICP-OES

Instrument capable of measuring Mg concentrations in a solution. Typical sample preparation involved the hydrogenation/dehydrogenation cycling of the materials at least twice to remove most of its impurities. Then, the desorbed state material was digested in

0.5 mL of concentrated nitric acid and diluted 20 times with high purity water. It was then analysed by the ICP-OES to obtain the Mg concentrations in the solution, which was then compared to the solid concentration of the solution.

3.3.6 BET analysis

Part of this study involved the synthesis and use of well structured polystyrene with star or hyperbranched architectures. One of the characterisation techniques applied for these materials was the Brunauer–Emmett–Teller (BET) that obtained their porosity and surface areas. During this study, a Micromeritics TriStar 3000 Analyzer from Micrometrics

Instrument Corporation was used and sample materials were degassed for several hours at

70 °C prior to the analysis. The analysis can be performed by adsorption isotherm of N2 on the sample. Typically, a specific amount of N2 was introduced to the manifold where there are samples inside the sample tube. The pressure, temperature and quantity of the N2 are recorded before the sample valve is closed. Then the adsorption is allowed to proceed until it reaches equilibrium., The amount of adsorbed gas was then calculated from the quantity of N2 introduced minus the residual N2 left inside the manifold.6

Experimental and Characterisation Techniques

3.4 Hydrogen storage properties characterisation techniques

Hydrogen storage properties that are investigated in this thesis include the hydrogen capacity of the material, the kinetics of hydrogen desorption and absorption, and the thermodynamics of the reactions. There are several methods that can be used to investigate those properties. Throughout the course of this study, Temperature-Programmed

Desorption (TPD) analyses were performed extensively to obtain the hydrogen desorption information of all materials including the synthesised and cycled materials. In-depth characterisations of kinetics and thermodynamic properties were performed by both gravimetric and volumetric techniques. Furthermore, theoretical calculations of those properties will also be discussed in detail.

3.4.1 TPD analysis

In this study, a TPD analysis was performed by a tandem system consisting of Mettler

Toledo Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC) system which were coupled to an Omnistar Mass Spectrometry (MS). These instruments were installed inside a glove box filled with high purity Ar (low O2 ppm).

In general, TGA allows for a simple measurement that monitors the weight loss of the material while ramping the temperature up. Besides the weight loss, DSC also monitors the heat flow that corresponds to the inert reference material during the temperature ramps.

In this way, all thermal events of the materials can be observed on the DSC analysis. These events include the hydrogen desorption event (endothermic process) and other structural change events such as melting. The most important part of TPD was the spectroscopic desorption measurement which was performed by the MS instrument connected to the

TGA-DSC instrument chamber. During heating, the hydrogen and other volatiles from the materials were released into the evacuation line which was connected to the MS. They were

Experimental and Characterisation Techniques

then analysed to determine the relative composition of the desorbed gas based on their specific molecular weight. Using the MS, masses between m/z = 2 to 99 were selectively monitored and measured. For example, m/z= 2 corresponded to the hydrogen released by the materials.

A typical sample measurement involved a loading of a small amount of samples (4 mg) into an alumina or aluminium crucible to undergo heating on the furnace. Then, a 10 minute settling time was used to stabilise the initial temperature before the temperature was ramped up to 500 °C. The heating rate was set at 10 °C.min-1 under a flow of 25 ml.min-1 of

Ar.

Several important results were obtained from this analysis. One in particular was the hydrogen capacity of the overall materials which was calculated from two sources; the weight loss captured by the TGA analysis and the calculation based on the hydrogen evolution monitored by the MS. However, this is only applicable when there are relatively no impurities on the material as observed in the MS. Therefore, it was usually done only with the materials after hydrogen cycling, while the calculation based method was done by integration of H2 peak from MS after a 1 point calibration of the MS with standard MgH2.

Standard MgH2 used is commercially available MgH2 (99%, Aldrich) with the calibration results shown in Figure 3.4.1.

Figure 3.4.1 (left) TPD result of a commercialised MgH2 with theoretical [H2] of 0.1 mol and the area under the curve was used to plot the calibration line (right)

Experimental and Characterisation Techniques

Total Hydrogen storage capacities are typically presented in units of weight percentage

(wt%), defined as:

풎풂풔풔 풘풕% = 푯ퟐ 풙 ퟏퟎퟎ% (3-3) 풎풂풔풔 +풎풂풔풔 풔풂풎풑풍풆 푯ퟐ

Other useful information that can be obtained is the activation energy of the material to undergo decomposition with temperature change, which can be expressed as

풅휶 풅휶 = 휷 = 풌(푻). 풇(휶) (3-4) 풅풕 풅푻

Where 훼 is the fraction of conversion, 훽 = 푑푇/푑푡 is the heating rate. In this study, we adopted the Kissinger method to obtain the kinetic and subsequently the activation energy through Arrhenius equation 7 8 , which is expressed as follows

푬 (− 풂) 풌(푻) = 푨풆 푹푻 (3-5) where 푅 is the gas constant, T is temperature in Kelvin, and A is constant representing the frequency factor for a molecule to have activation energy 퐸푎 participate in the reaction. It is assumed that the degradation reaction follows the n-th order reaction expressed as:

풇(휶) = (ퟏ − 휶)풏 (3-6)

By combining and integrating equation (8) and (9) , the Kissinger method can be rewritten as:

풏−ퟏ 휷 풏(ퟏ−휶풎) 풁푹 푬풂 퐥퐧 ( ퟐ ) = 풍풏 [ ] − (3-7) 푻풅풎 푬풂 푹푻풅풎

where 푇푑푚 is the absolute temperature at the maximum rate of thermal decomposition and

훼푚 is the weight loss at the maximum decomposition rate. In the Kissinger method, the value of the reaction order is constant and not dependent on the heating rate. Therefore, the

훽 value of activation energy can be calculated from the plot of ln ( 2 ) versus 1/푇푑푚. Different 푇푑푚 heating rate (훽) measurements were performed on the cycled materials to minimise the

Experimental and Characterisation Techniques

influences of impurities towards the hydrogen desorption. The different heating rates performed were 5, 10, and 15 °C.min-1. The TGA and DSC curves shifted to a higher temperature as the heating rate increased.

3.4.2 Gravimetric technique

Measurement of hydrogen absorption/desorption by a gravimetric method is similar to the

TGA analysis in the DTA. However, in this case the technique has more sophisticated aspects to measure the weight changes more accurately. Throughout the course of this study, a

Rubotherm gravimetric system equipped with a magnetic suspension balance was used.

The sample was placed in a sample chamber which was isolated from the external atmosphere and the weight change was measured in situ. As shown in Figure 3.4.4, the sample weight was measured via the interaction between a permanent magnet connected to the sample inside and an electromagnet on the outside. With this configuration, the weight of the sample can be accurately measured at different conditions including the temperature and pressure with the buoyancy correction. Buoyancy correction is required to correct the deviation of results caused by the flow rates and type of gas used during the measurements. This correction was done by performing buoyancy calibration which was performed on each measurement for different materials.

Each hydrogen absorption/desorption experiment typically ran for 5-7 days, which included several hydrogen absorption/desorption cycles of the sample materials. During those times, besides the absorption/desorption measurements there were blank and buoyancy measurements. The blank measurement was performed on the empty sample holder which ran under the same conditions as that of the sample. This was to obtain the accurate sample volume required for the sample’s density calculations. For the sample preparation, the materials were loaded into the sealed sample holder with a support inside the glovebox before it was placed on the instrument. The sample was then settled down on

Experimental and Characterisation Techniques

a support as the zero point position. This corresponded to the weight of the suspension magnet coupled with empty balance. This zero point position was used for the internal calibration to correct the weight difference automatically and it was repeated throughout the measurements.

Figure 3.4.2 Component diagram of conventional apparatus and magnetic suspension balance 9

The gravimetric balance was used to perform sample weight measurements with a resolution of up to 1 μg to ensure the accurate measurement of the hydrogen absorbed and desorbed. The amount of materials that was analysed ranged from 20 to 50 mg which should absorb/ desorb 1.5 to 4 mg H2. Once the sample holder was installed and the support was removed, the sample was subjected to a high vacuum (10-8 Bar) for 1 h at 300 °C prior to the first desorption measurement. Then it was followed with the sets of other measurements subjected to different temperature and pressure conditions. The balance readings of these measurements were composed of the following elements:

∆풎 = 풎풔풄 + 풎풔 + 풎푨 − (푽풔풄 + 푽풔 + 푽푨) ∙ 흆(풑, 푻, 풚 (3-8)

where mSC and VSC are the mass and volume of the empty sample container, mS and VS are the mass and volume of the sample, and mA and VA are the mass and volume of hydrogen absorbed.

Experimental and Characterisation Techniques

mSC and VSC were determined from a linear regression of the balance reading in a blank measurement:

∆풎 = 풎풔풄 − 푽풔풄 ∙ 흆 (3-9)

The total volume of the sample with sample holder (푉푠푐+푠) was determined by the buoyancy calibration which was performed by measuring the sample weight at different helium pressures.9 These different points were plotted to form a linear regression which corresponded to:

∆풎 = 풎풔풄+풔 − (푽풔풄+풔) ∙ 흆 (3-10)

Then, these values obtained were applied to the initial balance reading that had been corrected for the buoyancy effect to obtain 푚퐴 :

풎푨 = ∆풎 + (푽풔풄+풔) ∙ 흆 − 풎풔풄 (3-11)

Finally, the specific uptake of the material (푚퐴퐷푆) was determined by a comparison to the sample mass (푚푠) :

풎푨−풎풔 풎푨푫푺 = (3-12) 풎풔

In most cases, results reported in this study were obtained after samples were cycled more than twice to ensure that impurities were removed.

3.4.3 Sievert / Volumetric method

In this study, Sievert apparatus was used to obtain both the kinetics and thermodynamic properties of the materials. The basic concept of the volumetric method is to determine the amount of hydrogen absorbed by the sample from a change in pressure during the measurement. As shown in Figure 3.4.4, the apparatus consists of a gas reservoir connected to a specimen reactor of which its pressure, volume and temperature are monitored. In the

Experimental and Characterisation Techniques

end, the concentration of H2 was determined by the evolution in pressure where pressure changes were measured inside the instrument. This required the calibration of the volume of the sample holders to obtain an accurate measurement of the gas volume in the system.

Moreover, the temperature applied throughout the measurement should be known and constant so that the pressure changes are directly related to the concentration of H2 in the system.

One intrinsic advantage of the volumetric method is it allows a small amount of H2 to be introduced to the sample. This enables the measurement of Pressure-Temperature-

Isotherm (PCI) measurements which is useful to determine the thermodynamic properties of the material as shown in Figure 3.4.3-A.

The values of both ∆H and ∆S can be obtained from a Van’t Hoff plot (Figure 3.4.3-B). It is a logarithmic scale plot of the equilibrium pressure against the inverse of the temperature as defined by:

Experimental and Characterisation Techniques

∆푯 ∆푺 퐥퐧 푷 = − (3-13) 풆풒 푹푻 푹

The value of the equilibrium pressure corresponds to the whole phase transition from α- phase to β-phase at a certain temperature. The equilibrium is portrayed as a plateau pressure where the length determines the reversible hydrogen capacity of the whole material. This implies that by applying pressure above this plateau temperature, the material will undergo hydrogenation. On the other hand, below the plateau pressure, the β- phase becomes too unstable such that the material will undergo dehydrogenation. These plateau pressures change accordingly depending on the temperature that the material is exposed to during the phase change.

Another important aspect of constructing the Van’t Hoff plots from PCT curves is that they are dependent on whether the plateau pressures observed are from absorption or desorption measurements. These values will differ due to hysteresis even if the equilibrium is achieved, which generally causes a lower pressure plateau during desorption measurements.10 In this study, most of the PCT measurements were done for the absorption measurements due to its faster kinetics as compared to those of desorption.

The Sievert Instrument used in this study was a Gas Reaction Controller (Advanced

Materials Corporation). The temperature and pressure were controlled and monitored by

GrcQD software. There were two channels used to measure the sample. Typically the first channel required ~100 mg of material while the second channel which was more sensitive, required only ~20 mg of material. After the sample was loaded into the sample holder, quartz wools were placed to prevent the material from migrating into the instrument. It was then sealed and connected to the instrument apparatus. Absorption was done under 30 bar while desorption under 0.5 bar. The largest errors in volumetric measurements were caused by the determination of the volume of the sample and the measurement of the temperature of a material undergoing a highly exothermic reaction. The sample volume was calculated from its density obtained from the gravimetric method or occasionally from bulk

Experimental and Characterisation Techniques

-3 11 density of MgH2 (1.45 g.cm ). Moreover, errors in the temperatures were minimised by having a better configuration in the instrument (Figure 3.4.4) in which the temperatures were monitored and adjusted in both the gas reservoir and on the sample chamber.

Figure 3.4.4 Schematic drawing of the pressure composition isotherm system: an automatic gas reaction controller manufactured by the Advanced Materials Corporation.

3.5 Conclusion

It is the intent of this chapter to present the overview of the fundamental and measurement considerations of the experiments in this study. In particular, it will clarify the recommended practices and limitations in performing high-quality experiments to measure the hydrogen storage properties of the investigated materials. Most of the results on the following chapters will be presented as discussed in this chapter and additional methodologies will be explained if necessary.

Experimental and Characterisation Techniques

3.6 References

1. Suryanarayana, C.; Norton, M. G., X-ray diffraction: a practical approach. Springer: 1998. 2. Patterson, A., The Scherrer formula for X-ray particle size determination. Physical review 1939, 56 (10), 978. 3. Brundle, C. R.; Evans, C. A.; Wilson, S., Encyclopedia of materials characterization: surfaces, interfaces, thin films. Gulf Professional Publishing: 1992. 4. Watts, J. F.; Wolstenholme, J., An introduction to surface analysis by XPS and AES. An Introduction to Surface Analysis by XPS and AES, by John F. Watts, John Wolstenholme, pp. 224. ISBN 0-470-84713-1. Wiley-VCH, May 2003. 2003, 1. 5. Splinter, S.; McIntyre, N.; Lennard, W.; Griffiths, K.; Palumbo, G., An AES and XPS study of the initial oxidation of polycrystalline magnesium with water vapour at room temperature. Surface science 1993, 292 (1), 130-144. 6. Lowell, S., Characterization of porous solids and powders: surface area, pore size and density. Springer: 2004; Vol. 16. 7. Boswell, P., On the calculation of activation energies using a modified Kissinger method. Journal of Thermal Analysis and Calorimetry 1980, 18 (2), 353-358. 8. Criado, J.; Ortega, A., Non-isothermal transformation kinetics: remarks on the Kissinger method. Journal of non-crystalline solids 1986, 87 (3), 302-311. 9. Rubotherm Magnetic Suspension Balances. http://www.rubotherm.com/magnetic- suspension-balances.html. 10. Gross, K.; Carrington, K.; Barcelo, S.; Karkamkar, A.; Purewal, J.; Ma, S.; Zhou, H.; Dantzer, P.; Ott, K.; Burrell, T., Recommended Best Practices for the Characterization of Storage Properties of Hydrogen Storage Materials. National Renewable Energy Laboratory, Available Online: http://www1. eere. energy. gov/hydrogenandfuelcells/pdfs/bestpractices_h2_storag e_materials. pdf 2008. 11. Nielsen, T. K.; Manickam, K.; Hirscher, M.; Besenbacher, F.; Jensen, T. R., Confinement of MgH2 nanoclusters within nanoporous aerogel scaffold materials. ACS nano 2009, 3 (11), 3521-3528.

4 NOVEL SYNTHESIS METHOD OF MGH2 NANOPARTICLES BY WET CHEMISTRY APPROACHES

4.1 INTRODUCTION ...... 66 4.2 THERMAL DECOMPOSITION OF DIFFERENT GRIGNARD REAGENTS ...... 67 4.2.1 Experimental details ...... 69 4.2.2 Results and Discussion ...... 70 4.2.2.1 Decomposition behaviour of the selected Grignard compounds ...... 70 Hydrogenolysis of the selected Grignard reagents ...... 77 4.2.2.2 Hydrogen desorption properties ...... 83 4.2.3 Conclusion ...... 85 4.3 HYDROGENOLYSIS OF DI-N-BUTYLMAGNESIUM IN DIFFERENT ENVIRONMENTS ...... 86 4.3.1 Experimental details ...... 87 4.3.2 Results and discussion ...... 88 4.3.3 Conclusion ...... 103 4.4 INVESTIGATION OF MG PRODUCED FROM MGA.(THF)3 ...... 104 4.4.1 Experimental details ...... 105 4.4.2 Results and discussion ...... 107 4.4.3 Conclusion ...... 119 4.5 CONCLUSION ...... 120 4.6 REFERENCES ...... 122

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

4.1 Introduction

This chapter describes several strategies to obtain MgH2 nanoparticles with novel synthesis methods and versatility to be subjected for further modifications that shall lead to improvements in hydrogen storage properties. Such strategy required full understanding and control over the synthesis method since parameters such as reagents and their concentration, temperature, medium of the reaction and even impurities would eventually influence the overall properties of the nanomaterials produced.1 2 3

In this study, we firstly focused on the investigation of different types of organomagnesium compounds to be used as precursors for the magnesium hydride synthesis. Grignard reagents and di-n-butylmagnesium were selected for the preliminary investigation because of their availability in the chemical industry.4 In addition, several reports have shown the potential to produce MgH2 nanoparticles from these compounds through thermal decomposition under H2 pressure.5 6 7 However, there are limited investigations especially on the physical properties of MgH2 produced as well as their correlation with the overall hydrogen storage properties of the materials. This will be discussed thoroughly in the first section of this chapter.

The subsequent section of this chapter focusses on the effect of different mediums where the organomagnesium compounds, in this case di-n-butylmagnesium underwent the hydrogenolysis reaction to form MgH2. Two series of experiments were performed in dry conditions either under inert gas (Ar) or H2 gases while the other two series of experiments were performed in different solvents which had opposite polarity with the other, they are cyclohexane and diethyl ether. These different mediums in the end would affect the physicality and morphology of the MgH2 formed and subsequently their effects towards hydrogen storage properties were investigated.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

The last section of this chapter then focusses on a different precursor to yield Mg/MgH2 in nanosize scale. Magnesium anthracene complex or C14H10Mg.(THF)3, (MgA.(THF)3 was selected as a potential precursor as is capable of forming highly reactive Mg with high yield from several different routes.8 Several Mg based materials produced from these routes were investigated and the relationship between morphologies and hydrogen storage properties of these materials were discussed.

4.2 Thermal decomposition of different Grignard Reagents

The recent trend in the searching for a novel synthesis method of Mg nanoparticles by wet synthesis method has made Grignard reagents the attractive option especially because of their availability. Grignard reagents can be considered as the most widely used organometallic reagents in many industrial processes.4 The main role in undergoing a

Grignard reaction is to prepare many chemical reagents by forming carbon-carbon bonds when an alkyl or aryl magnesium halide is added to aldehydes, ketones, esters and amides.

The Grignard reagents are also known as the source for obtaining a monoatomic form of Mg through chemical or electrochemical reduction. This has been proven by several studies on the Mg/MgH2 nanoparticles for hydrogen storage material by these two methods.9 10

However, the overall yield of reduction methods is usually low and a significant amount of impurities is formed from the use of reducing agents.

There is another prospect for obtaining Mg from the Grignard reagents through thermally decomposing them under H2 pressure (hydrogenolysis process). This process has been proposed in the past by Becker et al where the Grignard reagents would undergo the following process 11 12:

2RMgX + 2H2  2RH + MgX2 + MgH2 (4-1)

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Reaction (4-1) was found to proceed at relatively mild temperatures (75-230 C) but often harsh conditions of hydrogen pressures in excess of 100 bar were used.11 12 Since this first report, the exact mechanism and the properties of MgH2 that were obtained still remains unclear because they would depend on the types of Grignard reagent that was being used.

It is known that such a reagent belongs to a covalent bound organometallic compound group where interaction of metals with carbon and hydrogen are important. As a result, intermediate compounds will be formed during hydrogenolysis due to the bonding of hydrogen molecules with metal bonded carbon as shown in Scheme 4-1.13

Scheme 4-1 Possible intermediate compound formed during the hydrogenolysis of Grignard reagents

This scheme may lead to a unique growth process and surface property of MgH2 that is obtained from each different Grignard reagent. To investigate further this hypothesis, we have selected four Grignard compounds bearing different ligands in addition to di-n- butylmagnesium for undergoing hydrogenolysis under the same conditions (Scheme 4-2).

Hydrogenolysis of di-n-butylmagnesium has been used to obtain stabilised MgH2 nanoparticles by nanoconfinement method.14 15 Similarly, little is known about the properties of Mg that is produced by this precursor alone without any presence of a stabiliser.

Therefore in this chapter, the thermal decomposition of these five compounds is reported as well as possible decomposition paths to form intermediate compounds. Further analyses were conducted to determine the physical/chemical properties of the magnesium nanoparticles generated through the hydrogenolysis of these reagents.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Scheme 4-2 Chemical structure of the different Grignard reagents selected: (a) di-n- butylmagnesium, (b) tert-butylmagnesium chloride, (c), allylmagnesium bromide (d) m-tolylmagnesium chloride and (e) methylmagnesium bromide.

4.2.1 Experimental details

Materials Di-n-butylmagnesium (1.0 M in heptane), tert-butyl magnesium chloride (2.0 M in diethyl ether), m-tolylmagnesium chloride (1.0 M, in tetrahydrofuran), allylmagnesium bromide

(1.0 M in diethyl ether), methylmagnesium bromide (3.0 M in diethyl ether) were obtained from Sigma-Aldrich. All the Grignard reagents were dried under vacuum for solvent removal which results in the starting reagents being in a powder form.

Synthesis of MgH2 nanoparticles by hydrogenolysis of the Grignard reagents Hydrogenolysis was carried out with the Grignard reagents in their dry form to minimise possible solvent effects. We found that the yield of MgH2 was significantly reduced when the reaction was carried out in organic solvents. Typically, 1 g of dry Grignard reagent was decomposed under 50 bar H2 pressure to facilitate the formation of MgH2. After 24 h, the material obtained was collected and washed several times with tetrahydrofuran to remove hydrocarbons and Mg halides by-products. Then it was dried under vacuum for another 24 h for solvent removal before being further characterised. The temperatures used for the hydrogenolysis of the Grignard reagents are summarised in Table 4.2.1. As a control, the

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

thermal decomposition of the Grignard reagents was also carried out under an inert atmosphere (Ar, 1 bar) mimicking the conditions of the TGA/DSC analysis.

Table 4.2.1- Temperature used for the hydrogenolysis of the Grignard reagents and particle/crystallite sizes of the nanoparticles obtained.

H - H - H - H - H - 2 2 2 2 2 DibutylMg TertbutylMgCl AllylMgBr TolylMgCl MethylMgBr Temperature 200 150 160 250 250 (C) Particle size 25-170 50-300 2-70 7-50 20-50 (nm) Crystalline size 32 - - - - (± 5 nm)

4.2.2 Results and Discussion

Decomposition behaviour of the selected Grignard compounds

The temperature required to decompose the selected Grignard reagents was investigated by the TPD analyses. As shown in Figure 4.2.1, all Grignard reagents started to decompose at a relatively low temperature (40-130 C) following two or three steps leading to full decomposition at 500 C. From TGA, di-n-butylmagnesium could be seen to undergo decomposition from 120 C in a sharp step leading to a mass loss of 81.7 % at 280 C. This step was accompanied with an endothermic peak at 260 C on DSC analysis (Figure 4.2.1-

A). The exact mechanism of di-n-butylmagnesium decomposition had been proposed by

Wiberg et al where the reaction involves an intramolecular mechanism to form a concerted four-centre intermediate, also known as β-elimination (4-2).5, 16

  H - Mg +R2

R1CH2CH2MgR2   2RCHCH2 + MgH2 (4-2)   RC +H C -H2

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

This reaction would result in the formation of MgH2 and olefin as by-products. This hypothesis is confirmed by the experimental observation of 1-butene masses in MS analysis

(Figure 4.2.2-A) during the first decomposition step. Indeed, another small decomposition step (2.8% mass loss) appeared at 350 C but significant increase of hydrogen evolution was observed. This would correspond to the MgH2 decomposition although it is noteworthy that hydrogen was not the only gas that was being released during the 2nd decomposition step.

1-butene masses were detected during this decomposition which indicate the formation of surface contamination of the MgH2 consisting of intermediate compounds that were being discussed earlier (Scheme 4-3).

(A) 20 (B) 10 100 Di-n-butylmagnesium 100 Tert-butylmagnesium chloride

80 10 80 5

60 335 60 0

Mass loss (%)

0 Heat flow (mW) 40 350

Heat flow (mW)

Mass loss (%) 40 Endothermic

260 -10

20 Endothermic

50 (C) (D) 20 150 15-5 Allylmagnesium bromide 100 m-tolylmagnesium chloride 100 60 12

80 9 90 40

80 20 60 3

Mass loss (%)

Heat flow (mW)

Endothermic

Mass loss (%) 70 0 Heat flow (mW) 40 0

Endothermic

175 -3

110 60 -20 20 100 200 300 400 500 100 200 300 400 500 Temperature (C) Temperature (C)

Methylmagnesium bromide (E) 100 80

60 90 40

Mass loss (%)

285 20

Heat flow (mW)

Endothermic 70 0

270

100 -20 60 100 200 300 400 500 Temperature (C)

Figure 4.2.1 – TGA/DSC curves of the different Grignard compounds investigated.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

H (m/z 2) H2 (m/z 2) 2 Di-n-butylmagnesium Tert-butylmagnesium chloride (A) + (B) Cl (m/z 71) C2H3 (m/z 27) 2 + C H + (m/z 27) C3H5 (m/z 41) 2 3 + C H (m/z 41) C4H8 (m/z 56) 3 5

C4H8 (m/z 56)

Intensity (a.u.) Intensity

Allylmagnesium bromide H2 (m/z 2) m-tolylmagnesium chloride H2 (m/z 2) (C) (D) Cl (m/z 71) + 2 C2H3 (m/z 27) C H (m/z 78) + 6 6 C3H5 (m/z 41) C H CH (m/z 92) Br (m/z 79) 6 5 3

Intensity (a.u.)

Intensity (a.u.) Intensity

100 200 300 400 500 100 200 300 400 500  Temperature ( C ) Temperature ( C )

H2 (m/z 2) Methylmagnesium bromide + (E) CH3 (m/z 15) Br (m/z 79)

Intensity (a.u.) Intensity

100 200 300 400 500 o Temperature ( C)

Figure 4.2.2 – Evolution of volatile organic matter and hydrogen (followed by MS) resulting from the thermal decomposition of the selected Grignard reagents during their TGA analysis.

Decomposition of di-n-butylmagnesium experiment at 200 C under an Ar atmosphere

(DibutylMg) confirmed the formation of -MgH2 through the β-elimination process without the presence of H2 pressure (Figure 4.2.3-A). This observation was in agreement with the very first report on the thermal decomposition of diethylmagnesium proving the possibility of forming crystalline MgH2 under very mild conditions.17

The origins of surface contaminations on the MgH2 in DibutylMg were further investigated by XPS analysis. It showed a significant amount of carbon that may correspond to surface contamination (Table 4.2.3). Lefrancois and Gault reported that a long hydrocarbon chain

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

including pentene and hexane could be formed even during the pyrolysis of ethylmagnesium halides.16 Such hydrocarbons might be formed from radicals as the decomposition by-products.

On the other hand, the hydrogenolysis/decomposition process of Grignard reagents to obtain MgH2 is not as simple as with dialkylmagnesium. Previous reports mentioned the need of very high H2 pressures (up to 350 bar) for the formation of MgH2 from the proposed mechanism assuming the cleavage of the Grignard by hydrogen (Scheme 4-1).11 However, the thermal decomposition of tert-butylmagnesium chloride behaved similarly with di-n- butylmagnesium where it proceeded in one main step starting from 80 C and ending at 180

C (Figure 4.2.1-B). But there were three endothermic peaks which appeared at 50, 150, and

335 C. The first endothermic event may correspond to the melting of the compound since no weight loss was observed. The second endothermic peak would refer to the decomposition of tert-butylmagnesium chloride. This is confirmed by the detection of isobutylene masses by MS at that range of temperatures (Figure 4.2.2-B). Isobutylene is the by-product of the decomposition of tert-butylmagnesium chloride which undergoes a similar mechanism with di-n-butylmagnesium. However, the excessive mass loss observed for the first decomposition step (61.8 % instead of the theoretical hydrocarbon content of

48.8%, (Table 4.2.3) would also indicate the simultaneous release of chloride as observed by MS (Figure 4.2.2-B). Moreover, the third endothermic peak should represent the decomposition of MgH2 which was proven by the increases in hydrogen desorption at temperatures above 300 C. However, besides MgH2 there were MgCl2 which appeared as the main decomposition products. This was confirmed by the presence of MgCl2 phase in

XRD analysis of tert-butylmagnesium chloride decomposed under Ar at 150 C

(TertbutylMgCl) (Figure 4.2.3-A). The XPS analysis confirmed the formation of the MgCl2 phase since Mg-Cl binding energy appeared during analysis. (Table 4.2.3). Furthermore, since no Mg2p peak appeared by XPS, it can be assumed that the thermal decomposition of

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

tert-butylmagnesium chloride most likely resulted in MgH2 particles being buried within the MgCl2 matrice (the penetration depth of the photoelectron was 5 nm). Similarly in the di-n-butylmagnesium case, the XPS also indicated some carbon contamination (Table 4.2.3), despite the low temperature required for the decomposition of tert-butylmagnesium chloride. Hence, it seems that surface contamination cannot be avoided. This may be due to the inherent reactivity of the magnesium surface.

 -MgH (A)  DibutylMg  2 

   *  *     * MgO

MgCl   2   TertbutylMgCl      MgBr      2

 MgBr .6H O AllylMgBr  2 2               MgClOH     Kapton

    TolylMgCl

Intensity (a.u.)

    MethylMgBr       

  * * *

20 40 60 80 2 (degree) -MgH  2 (B)   H -DibutylMg  2

 * MgO       MgCl 2       H -TertbutylMgCl  2 MgBr   2        MgBr .6H O  2 2   H -AllylMgBr  2 MgClOH                Kapton

  H -TolylMgCl   2

Intensity (a.u.)

  H -MethylMgBr    2

 

  *  * 

20 40 60 80 2 (degree)

Figure 4.2.3 – XRD patterns of the materials obtained after the thermal decomposition of the Grignard reagents: (A) under an Ar atmosphere, and (B) under hydrogen pressure, i.e. after hydrogenolysis. The temperatures used for these syntheses are tabulated in Table 4.2.1. A partial oxidation of the materials may have occurred during measurement.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Furthermore, the decomposition of allylmagnesium bromide involved two main steps accompanied by two endothermic peaks at 110 and 175 C, respectively (Figure 4.2.1-C).

During these two events, propylene and its associated fragments were detected from the

MS while no bromine was detected (Figure 4.2.2-C). XRD analysis of the material obtained after the decomposition of allylmagnesium bromide under Ar at 160 C (AllylMgBr) revealed the formation of MgBr2 also as the main phase (Figure 4.2.3-A).

The formation of -MgH2 however was not clear in this XRD analysis and there was no typical endothermic peak of MgH2 being observed in DSC. Indeed there was an increase of hydrogen evolution at temperatures above 300 C observed by MS Figure 4.2.2-A. It is noteworthy that the hydrogen evolutions have multiple peaks above 250 C. This release of hydrogen may still be due to the formation of an MgH2 phase because metallic magnesium was detected in AllylMgBr by XPS (Table 4.2.3). Furthermore, the carbon content as determined by XPS analysis for AllylMgBr was found to be the highest of all the Grignard precursors investigated. These results may indicate that high temperatures (> 200 C) and/or a hydrogen atmosphere are required to generate MgH2 from allylmagnesium bromide.

Similarly, the thermal decomposition of m-tolylmagnesium chloride and methlymagnesium bromide proceeded without the formation of any crystalline MgH2 phase (Figure 4.2.3-A).

The only phases detected by XRD after thermal decomposition at 250 C under Ar

(TolylMgCl) were those of the respective magnesium halide. The decomposition of m- tolylmagnesium chloride occurred in three steps from 120 C until as high as 330 C (Figure

4.2.1-D). This started with the release of chlorine followed by the decomposition of the tolyl group as the whole structure without fragmentation into benzene and methyl moieties

(Figure 4.2.2-D). Again, the formation of an MgH2 phase may occur at elevated temperatures because hydrogen evolution is stronger and the only one that appears at temperatures 300

C (Figure 4.2.2-D).

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Table 4.2.2 - Possible decomposition paths for the selected Grignard reagents and associated mass loss

Theoretical Temperature Mass content (%) Decomposition (step 1 and 2) range (C) loss % in compounds

DibutylMg (C4H9)2Mg → (2-x)(C4H8)g + x(C4H8)surfMgH2 120-280 81.7 (hydrocarbon:

82.4) (C4H8)surfMgH(s) → (C4H8)g + MgH2(s) → Mg(s) + H2 280-400 2.8

TertbutylMgCl 2(CH3)3CMgCl → 2(CH3)2CCH2 + x/2Cl2 + 80-200 61.8 x/2MgH2 + (1-x/2)MgCl2 (hydrocarbon: 48.8) MgH2(s) → Mg(s) + H2 200-400 1.8 (Chloride: 30.4)

3 C3H5MgBr → C3H5 + C3H4 + x(C3H5)MgH2(s) + (1- AllylMgBr 70-175 20.1 x)MgBr2 (hydrocarbon:

28.2) 2 (C3H5)MgH2(s) → C3H5 + C3H4 + MgH2(s) → Mg(s) 175-500 15.9 (Bromide: 55.0) + H2

TolylMgCl 2CH3C6H4MgCl → x(CH3C6H4Mg) + x/2Cl2 + (1- 130-300 26.1 x)MgCl2 (hydrocarbon:

60.4) 2(CH3C6H4Mg) → CH3C6H4 + 2(CyHz)MgH(s) → 200-400 41.8 (Chloride: 23.5) 2(CyHz)Mg(s) + H2

MethylMgBr 2CH3MgBr → CH3- + xMgBr2 + Mg + x/2Br2 40-210 24.1

(hydrocarbon:

13.1) Mg + CH3- ↔ (CH3)surf Mg 210-500 7.6

(Bromide: 67)

On the other hand, methlymagnesium bromide showed the lowest onset decomposition temperature (40 C), however the release of the methyl group was found to occur at temperatures above 250 C with no indication of a significant increase in hydrogen evolution (Figure 4.2.2-E). Indeed, two main decomposition steps were observed and multiple DSC peaks such as two endothermic and one exothermic events occurred at 100,

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

270 and 285 C, respectively (Figure 4.2.1-E). The first endothermic peak can be related to the decomposition of the Grignard since simultaneous release of the methyl and bromide group were observed. However the origin of the endothermic peak at 270 C and exothermic event remains uncertain. It is possible that the exothermic peak observed corresponds to the reaction of metallic magnesium with the organic volatile matter released from the

Grignard. This is because a peak corresponding to Mg2p was obtained by XPS analysis of methlymagnesium bromide decomposed under Ar at 250 C (MethylMgBr) (Table 4.2.3).

The analysis of thermal decomposition of the selected Grignard reagents as well as the possible decomposition paths are summarised in Table 4.2.2. It appears that the electronic density of the hydrocarbon group strongly influences the thermal decomposition behaviour of these Grignard compounds. Hence, due to their respective inductive effects, the methyl group with the lowest electronic density led to the lowest initial decomposition temperature (i.e. 40 C) whereas the tolyl with the highest electronic density resulted in the highest initial decomposition temperature (i.e. 130 C). However, only Grignard precursors with a butyl group led directly to the formation of MgH2 at relatively low temperatures (<

200 C) even without any hydrogen pressure. Other compounds may need a hydrogen atmosphere for the hydrogenolysis reaction to effectively produce MgH2.16

Hydrogenolysis of the selected Grignard reagents

The hydrogenolysis of the selected Grignard reagents was carried out under a hydrogen pressure of 50 bar and at temperatures corresponding to an advanced stage of thermal decomposition of the Grignard precursors (Table 4.2.1). The early report of Becker and

Ashby indicated full hydrogenolysis of most Grignard compounds at 125 C.12 Herein, based on the TGA analysis, temperatures higher by 25-125 C were chosen to ensure full hydrogenolysis of the selected Grignard precursors and a high yield of MgH2.

After reaction of the precursors with hydrogen for 24 h, the materials obtained were collected and characterised. Analysis by SEM revealed different morphologies for the

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

hydrogenolysed materials (Figure 4.2.4). The material obtained from the hydrogenolysis of di-n-butylmagnesium (H2-DibutylMg) displayed small particles (100 nm) assembled into larger aggregates (Figure 4.2.4-A). For tert-butylmangesium chloride (H2-TertbutylMgCl) similar agglomerates containing smaller spherical particles ( 70 nm) were observed

(Figure 4.2.4-B). However the material resulting from the hydrogenolysis of allylmagnesium bromide (H2-AllylMgBr) only showed large agglomerates made of flakes (Figure

4.2.4-C and D). For m-tolylmagnesium chloride (H2-TolylMgCl) and methylmagnesium bromide (H2-MethylMgBr), the structures obtained were a mixture of small particles (20-

40 nm) and bigger flakes assembled into large agglomerates (Figure 4.2.4-E and F).

Figure 4.2.4 – SEM images of the materials obtained after hydrogenolysis of the Grignard reagents.

Moreover, further analysis by TEM revealed H2-DibutylMg has quite distinct features of nanoparticles but with irregular sizes. The size varied between 30 nm in radii and occasionally it assembled into larger nanostructures as big as 170 nm (Figure 4.2.5A). It is noteworthy that there is uneven spread of darker areas within the nanoparticles. This could

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

be from the partial formation of an MgO layer due to the carbon contaminants on the nanoparticle surface that avoid them being oxidised. The in-situ degradation of the hydrocarbons released during the hydrogenolysis process could possibly lead to the formation of carbon entities building up into “walls” at the surface of the magnesium nanoparticles. Hence, similar carbon/magnesium interactions leading to nanoconfined

MgH2 particles may occur during the hydrogenolysis of di-n-butylmagnesium in a closed vessel.

Figure 4.2.5 – TEM images of the materials obtained after hydrogenolysis of the Grignard reagents and each corresponding to the EDS analysis

The formation of such nanoarchitecture was only observed for H2-DibutylMg. For H2-

TertbutylMgCl mainly large particles (120 nm) with rectangular shapes were observed

(Figure 4.2.5-B). For H2-AllylMgbr very small (2 nm) and larger particles (50 nm) were obtained (Figure 4.2.5-C). Finally, H2-TolylMgCl displayed large spherical nanoparticles and

H2-MethylMgBr nanoparticles of a cubic or rectangular shape (Figure 4.2.5-E and F). The various particle sizes obtained are summarised in Table 4.2.1. During these TEM analyses,

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

EDS were performed and revealed Mg in all hydrogenolysed materials and the corresponding halides (e.g. Cl for H2-TertbutylMgCl) as the main constituents of the nanoparticles imaged, except for H2-DibutylMg for which only Mg was observed.

Further analysis by XRD also revealed that only di-n-butylmagnesium and tert- butylmagnesium chloride led to the formation of -MgH2 (Figure 4.2.3-B). The other

Grignard precursors only showed the corresponding magnesium halide as the main crystalline phase and this is more or less similar with the materials decomposed under Ar.

Hence, only H2-DibutylMg would consist of MgH2 nanoparticles while H2-TertbutylMgCl would consist of MgCl2 and MgH2 nanoparticles. These results prove that the hydrogenolysis of Grignard reagents does not necessarily lead to the facile formation of MgH2.

Because the MgH2 phase on the XRD spectra is distinctive, its crystalline size of H2-

DibutylMg can be calculated by the Scherrer equation. It was found that H2-DibutylMg had average crystalline size of 32 ± 5 nm (Table 4.2.1). This value corresponds to the size of the nanoparticles confined within the larger nanostructures observed by TEM (Figure 4.2.5-A).

Hence H2-DibutylMg would display a nanocrystalline structure similar to that of nanocrystalline MgH2 produced by ball-milling.18 19 However, this method is proven to be more energy efficient and feasible for large scale production than the ball-milling. Also, H2-

DibutylMg consists of individual particles that organised themselves by restricting the number of individuals that are in direct contact. This may be due to the hydrocarbon layer on the surface which causes inter-particle distances and limits the growth sizes of the MgH2.

Indeed, XPS analyses proved the existence of this carbon contamination even after hydrogenolysis for all Grignard precursors (Table 4.2.3). H2-TertbutylMgCl showed the lowest carbon contamination, but the material contained a significant amount of MgCl2 as suggested by the amount of Mg-Cl bonds. Similarly the XPS analysis of H2-TolylMgCl, H2-

AllyMgBr and H2-MethylMgbr confirmed that the materials contained MgCl2 and MgBr2, respectively. Furthermore, three different chemical states of Mg were observed by XPS:

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

metallic magnesium (Mg) at 50.70 eV, amorphous magnesium oxide (MgO) or hydroxide

(Mg(OH)x) at 51.22eV and crystalline MgO or Mg(OH)x at 52.67 eV. MgO and Mg(OH)x are likely due to a partial oxidation of the material from exposure to air when it was transferred into the instrument. The thickness of this oxide layer also may have influenced the detection of metallic magnesium in particular for H2-TertbutylMgCl (Table 4.2.3).

Mg 2p 20 21 Br 3d Cl 2p3 C 1s 21 22 O 1s 21

Mg MgO/Mg MgO/ Mg-Br Mg-Cl C-C MgO MgOH

(OH)xa Mg(O (50.7 (69.3 (199.1 (284.9 (529. (532.9

H)xb eV) (51.2 eV) eV) eV) 7 eV) eV)

DibutylMg Ar 21.93 -eV) 18.43(52.7 - - 18.15 7.09 24.98

H2 12.85 - 26.88eV) - - 11.62 7.16 29.58

Tertbutyl Ar - 10.28 22.16 - 22.76 14.20 2.92 0.53

MgCl H2 - 13.63 19.8 - 32.15 6.03 18.85 22.76

TolylMgCl Ar - 11.80 15.93 - 26.64 19.12 - 18.44

H2 - 17.82 7.52 - 24.65 27.16 - 15.04

AllylMgBr Ar 13.69 13.73 - 16.84 - 28.68 7.33 9.88

H2 15.14 10.31 - 15.46 - 30.61 1.92 16.22

MethylMg Ar 19.55 11.82 - 17.19 - 11.83 - 30.55

Br H2 19.69 17.54 - - - 11.21 - 29.44

In an attempt to quantify the amount of this carbon contamination, the hydrogenolysed materials were analysed by TGA/MS. As shown in Figure 4.2.6, most of the hydrocarbon contamination is released at the same time as hydrogen is released. For H2-DibutylMg and

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

H2-TertbutylMgCl, this occurred at relatively high temperatures (> 300 C). This confirmed the previous hypothesis that the hydrocarbon layer is at the surface of the MgH2 nanoparticles and would need to be broken to allow the hydrogen release to proceed.

However, for H2-AllylMgBr, H2-TolylMgCl, and H2-MethylMgBr, they showed similar observations despite MgH2 formations not being confirmed by XRD. Indeed, MgH2 may still be present on both materials but it just was not detected on XRD for several reasons. These could be that the MgH2 phase was amorphous or the ratio of Mg halides within the materials is much larger compared to the MgH2. Another possibility is that no MgH2 was formed and the hydrogen came from the fragmentation of the various hydrocarbons evolving from the surface of MgBr2 or MgCl2 itself. (Figure 4.2.6-C, D, and E).

H2 (m/z 2) H -DibutylMg H (m/z 2 Cl (m/z 71) 2 2 2 + (A) (B) + C2H3 (m/z 27) C2H3 (m/z 27) + + C3H5 (m/z 41) H -TertbutylMgCl C3H5 (m/z 41) 2

C4H8 (m/z 56) C4H8 (m/z 56)

Intensity (a.u.) Intensity (a.u.)

H2 (m/z 2) H (m/z 2) H -AllylMgBr 2 H -TolylMgCl (C) + 2 (D) 2 C2H3 (m/z 27) Cl (m/z 71) 2 + C H (m/z 41) C H (m/z 78) 3 5 6 6 79 Br (m/z 79) C H CH (m/z 92) 6 5 3 C H (m/z 42) 3 6

Intensity (a.u.) Intensity (a.u.)

100 200 300 400 500 100 200 300 400 500 o o Temperature ( C) Temperature ( C) H (m/z 2) 2 H -MethylMgBr (E) + 2 CH3 (m/z 15) Br (m/z 79) C3H6 (m/z 42)

Intensity (a.u.)

100 200 300 400 500 Temperature (oC)

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Hydrogen desorption properties

The two materials H2-DibutylMg and H2-TertbutylMgCl that clearly showed the formation of -MgH2 were further characterised for their hydrogen storage properties. Absorption/ desorption measurements were done on both materials for several times by the Gravimetric method. XRD analysis confirmed the reversibility of the reaction since Mg and -MgH2 were the only phases observed after hydrogen desorption and absorption, respectively (Figure

4.2.7-B). Furthermore, this hydrogen cycling also confirmed only having hydrogen that was released from the material (Figure 4.2.7-C). No other gases were detected and the hydrogen released peak is consistent with previous TGA/MS analysis of the material obtained after hydrogenolysis (Figure 4.2.6-A). Therefore, the hydrogen desorption/absorption cycle at

300 C was sufficient to remove most of the organic contamination and it still has remarkably high storage capacity of 6.8 mass % (Figure 4.2.7-A).

H2-TertbutylMgCl displayed extremely slow hydrogen desorption kinetics. Such slow kinetics may be due to the large amount of MgCl2 embedding the MgH2 nanoparticles as sluggish desorption kinetics were also observed for MgH2 nanoparticles embedded in LiCl23 contrary to H2-DibutylMg where it displayed very fast desorption kinetics. Less than 15 min were required to fully release hydrogen from the material at 300 C and even at 250 C it was still possible to fully desorb the material in 120 min. In comparison nanocrystalline magnesium produced by ball milling usually requires more than 30 min at 300 C and more than 40 h at 250 C to fully release hydrogen.18 Therefore, the hydrogen desorption kinetics of nanocrystalline magnesium produced by the hydrogenolysis of di-n-butylmagnesium are far superior to that of ball milled magnesium. It is also noteworthy that the desorption kinetics of H2-DibutylMg are much faster than that of nanoconfined MgH2 which required

10 h to fully release hydrogen.24 25 This extremely slow hydrogen desorption kinetic may be due to the long diffusion ranges inherent to porous networks.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

(A) 1 (B)  Mg  -MgH MgO 0  2 *  Kapton -1 H -tertbutylMgCl at 300 oC 2 -2 H -DibutylMg at 250 oC 2    Desorbed -3 o     H -DibutylMg at 300 C *  2 *

-4  

-5 Intensity (a.u.)   Absorbed     -6       

Hydrogen desorbed (mass %) -7

0 50 100 150 200 20 40 60 80 time (min) 2(degree)

C4H8 (m/z 56)

Intensity (a.u)

100 200 300 400 500 o Temperature ( C)

Figure 4.2.7 – (A) Hydrogen desorption kinetic of H2-TertbutylMgCl and H2-DibutylMg measured under isothermal conditions and at a pressure of 0.1 bar, (B) XRD of H2- DibutylMg after hydrogen desorption and hydrogen absorption, (C) Evolution of hydrogen followed by MS during the thermal decomposition at 10 C.min-1 under a 25 ml.min-1 Ar flow of H2-DibutylMg after the first cycle, and (D) TEM image of H2-

DibutylMg after 3 cycles. No gases except hydrogen were detected from H2-DibutylMg after the first cycles.

Further analysis by TEM showed that H2-DibutylMg has undergone significant morphological evolutions after cycling. In particular, the disappearance of the voids observed within the nanoparticles assemblies (Figure 4.2.5-B and Figure 4.2.7-D) would suggest some sintering of the nanoconfined particles during hydrogen cycling or shrinkage of the walls. Scherrer analysis of the XRD pattern of H2-DibutylMg after the third absorption and desorption cycles gave an average crystallite size of 28 and 30 ± 5 nm, respectively.

Hence, no significant change in crystallite size occurred during the first three hydrogen cycles. This behaviour is significantly different to that of nanocrystalline magnesium

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

produced by ball milling which undergoes a drastic growth of crystallite size upon the first and subsequent hydrogen absorption/desorption cycles. Growth of crystallite size from 10 nm after ball milling to 200 nm after cycling have been reported.26 27 Limiting such crystallite growth has been the cornerstone of many investigations because smaller crystals should lead to higher hydrogen diffusion rates and possibly some improvement in the thermodynamics of the H2/Mg reaction.28 29 30 To date, the main approach to limit such crystallite growth has only been done by adding metal oxides such as Nb2O5 during mechanical milling.26 Producing isolated nanoparticles would also lead to faster desorption kinetics but kinetics usually rapidly deteriorate as the nanoparticles agglomerate and sinter into large particles upon a few cycles.31 32 33 H2-DibutylMg is however very stable and deterioration of the hydrogen kinetics did not occur after as many as 20 cycles. The stability of crystallite and particle size could therefore be at the origin of the superior kinetics of H2-

DibutylMg as compared to milled magnesium. It is noteworthy that the desorption kinetics of H2-DibutylMg observed were not supported by any catalyst and this further highlights the possibility of designing effective hydrogen storage materials from magnesium by synthesizing the right nanostructure.

4.2.3 Conclusion

In this study, a thorough investigation over the thermal decomposition and hydrogenolysis of Grignard reagents and di-n-butylmagnesium was performed. Different Grignard reagents and a di-n-butylmagnesium were selected as the potential precursors to produce MgH2 nanoparticles with better hydrogen storage properties. It was found that decomposition and hydrogenolysed products of all Grignard reagents led to the formation of Mg halides as impurities. Attempts to remove these Mg halides were done such as by dissolution through

THF but the overall removal was minimal due to the partial solubility of these compounds in organic solvents. In the end, the presence of these halides would decrease the overall

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

hydrogen storage capacity and slowed the hydrogen desorption kinetics possibly due to the embodiment of the Mg inside the Mg halides shell. On the other hand, it was found that the thermal decomposition of di-n-butylmagnesium in inert atmosphere led to the formation of

-MgH2 from the β-elimination process. More remarkably, addition of H2 pressure during the decomposition process would lead to the formation of small isolated MgH2 nanoparticles which performed superior hydrogen desorption that is comparable to those ball milled MgH2 with catalyst. The di-n-butylmagnesium is thus depicted to be the best precursor and worth further investigation on this study compared to other Grignard reagents.

4.3 Hydrogenolysis of Di-n-butylmagnesium in different environments

As previously shown, hydrogenolysis products of di-n-butylmagnesium gave the most promising physical and hydrogen storage properties when compared to those from the

Grignard reagents. Even without the presence of any hydrogen pressure it would still form

MgH2 through β-elimination mechanism.5 This occurs from a relatively low temperature of

110 C with excessive degradation above 150 C. Hence a temperature of 200 C is sufficient for the rapid formation of MgH2 through β-elimination (4-3).

(C4H9)2Mg → 2C4H8(g) + MgH2(s) (4-3)

Since MgH2 is highly reactive toward organic molecules 34, a process involving the thermal degradation of organomagnesium for the synthesis of magnesium would be sensitive to the synthetic medium. This should lead to different magnesium structures and hydrogen performances. Moreover, the use of polar solvent as a synthetic medium is crucial since most polymers are soluble in such solvent therefore it can be utilized in the hybrid magnesium-polymers synthesis. Herein, the nucleation and growth of MgH2 via the

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

hydrogenolysis of di-n-butylmagnesium is investigated on different synthesis media, i.e. an argon atmosphere, hydrogen pressure, non-polar solvent (cyclohexane) or polar solvent

(diethyl ether). Remarkably, various morphologies of MgH2 were obtained and they had significant differences in terms of hydrogen sorption properties.

4.3.1 Experimental details

Materials

Di-n-butylmagnesium (1.0 M in heptanes with 1 mass % of triethylaluminium) anhydrous diethyl ether and cyclohexane (> 99%), were obtained from Sigma-Aldrich.

MgH2 synthesis by dry method

Di-n-butylmagnesium was dried under vacuum for solvent removal. The resulting material was then decomposed at 200 °C for 24 h in a pressure reactor vessel under an argon atmosphere (1 bar) or a hydrogen pressure of 30 bar. After the reaction, a grey material was obtained in both cases. The amount of starting material used and yields are summarised in

Table 4.3.1.

MgH2 synthesis by wet method

Di-n-butylmagnesium solution in heptane was added to 100 mL of diethyl ether or cyclohexane in a pressure reactor vessel. The solution was then stirred at 150 rpm and heated to 200 °C under a hydrogen pressure of 30 bar. After 24 h, a white precipitate was obtained from the decomposition of di-n-butylmagnesium in diethyl ether, whereas a grey precipitate formed in cyclohexane. The precipitates were washed several times with diethyl ether and cyclohexane, respectively. The obtained materials were further dried at room temperature under dynamic vacuum on a Schlenk line. The amount of starting material used and yields also are summarised in Table 4.3.1.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Table 4.3.1 - Summary of the synthetic conditions used, yields

Synthetic medium Diethyl Argon Hydrogen Cyclohexane Ether

Di-n-butylmagnesium used as dry dry liquid liquid Amount of 0.663 0.503 2.852 2.852 di-n-butylmagnesium used (g) Yield after synthesis (g) 0.080 0.070 0.092 0.144 Yield after hydrogen cycling (g) 0.055 0.046 0.081 0.131

4.3.2 Results and discussion

Different synthetic environments were used to determine the effect of the synthetic medium on the properties of the MgH2 generated from the decomposition of di-n-butylmagnesium.

These included under an inert atmosphere of Ar and H2 pressure to enhance the formation of MgH2. To improve diffusion rates, the same synthesis was also carried out in two solvents, i.e. cyclohexane and diethyl ether with a hydrogen pressure of 30 bar. The reactivity of the solvent with magnesium and its polarity may also affect the nucleation and growth of crystals along preferential directions depending on the interaction at the solvent/crystal interface as observed in solvothermal syntheses.35 Both cyclohexane and diethyl ether were selected because they are completely opposite in polarity. A study has shown that cyclohexane is inert toward magnesium surfaces while ether compounds could react with magnesium to form intermediate and stable Grignard compounds.34

Figure 4.3.1 shows the images obtained by SEM of the material after synthesis.

Decomposition of di-n-butylmagnesium under an atmosphere of argon led to the formation of long rods with a length of ~150-200 nm (Figure 4.3.1-A). However, under hydrogen pressure particles 25-170 nm with a spherical shape were obtained (Figure 4.3.1-B). In

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

cyclohexane, the particles remained spherical but were found to be of a smaller size 15-50 nm (Figure 4.3.1-C). On the other hand, the use of diethyl ether as a synthetic solvent resulted in the formation of large aggregates of flakes (Figure 4.3.1-D).

Further analysis by TEM confirmed the formation of these distinct morphologies (Figure

4.3.2). The observed sizes and morphologies of the particles are summarised in Table 4.3.2.

From these observations, it is apparent that the synthetic medium strongly influences the growth behaviour of MgH2 as di-n-butylmagnesium undergoes hydrogenolysis. The exact mechanisms governing the formation of these different morphologies are not clear but it is well known that external factors including chemical environment, solvent polarity, temperature and pressure would strongly affect nucleation and growth behaviour.36 37 38

Hence, the presence of hydrogen pressure may reduce the growth of nuclei according to the morphology of particles synthesised under hydrogen pressure (Figure 4.3.2).

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Analysis by XRD confirmed that the synthesised materials corresponded to crystalline -

MgH2 (Figure 4.3.3). In addition, the Scherrer equation found that those decomposed in cyclohexane have the smallest average crystalline size with 17 ± 3 nm while those decomposed in diethyl ether are the biggest with 40 ± 5 nm. The rest of the results are summarised in Table 4.3.2.

Crystallite size (nm)

Synthetic Particle Ea medium Morphology/size (nm) After After (kJ.mol-1 H2) synthesis cycling

Argon Rods, 10-50/150-200 24 ± 4 23 ± 3 157 ± 8

Hydrogen Round like, 25-170 32 ± 5 30 ± 3 166 ± 8

Cyclohexane Round like, 15-50 17 ± 3 23 ± 3 133 ± 10

Diethyl Ether Flakes 40 ± 5 37 ± 4 170 ± 10

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.3.3 – XRD patterns of the material obtained from the hydrogenolysis of di-n- butylmagnesium: under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether.

Elemental analyses were also performed on each material decomposed in all synthetic mediums and the results confirmed the presence of Al in all of them. Al concentrations on materials synthesised in H2, Ar, cyclohexane, and diethyl ether were found to be 3, 5, 12, and

3 % respectively. Indeed, material that was synthesised in cyclohexane has the most Al content assuming they are in their crystalline form as it appeared in XRD spectra.

 o 90% triethylaluminium o MgH  2   o o Al o  o o o oo o  o o o Kapton  50% triethylaluminium o o  o    o o o oo o o o o

Intensity o o o 1% triethylaluminium

 o   o o oo o   o o o

30 40 50 60 70 80 90

2(degree)

Figure 4.3.4 XRD showing the formation of aluminium from triethylaluminium under synthetic conditions similar to those used for the hydrogenolysis of di-n- butylmagnesium in cyclohexane.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Indeed, the presence of this aluminium would contribute to the decrease of the material overall hydrogen storage capacity. This will be discussed later in the section of this chapter.

Meanwhile, other additional impurities are also present mainly due to the hydrocarbon contaminants which formed from the decomposition of the organic component of di-n- butylmagnesium and also from solvents when it is used as synthetic medium. In order to investigate further the chemical structures of these carbon contaminants, solid NMR analyses of these materials were performed. The results (Figure 4.3.5) showed the presence of butylene oligomers when the decomposition was carried out under an argon atmosphere, hydrogen pressure or cyclohexane. However, the material that was obtained under diethyl ether showed peaks at chemical shifts that corresponded to the diethyl ether. As a highly polar solvent, diethyl ether most likely was adsorbed at the surface of the MgH2 particles during the growth process.

Figure 4.3.5 – 13C NMR spectra the material obtained from the hydrogenolysis of di-n- butylmagnesium: under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether.

Further analysis by XPS also confirmed the presence of unique C-O bonding from the diethyl ether while other similar carbon contamination bondings were observed in the rest of all

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

materials. (Table 4.3.3). This carbon contamination would be minimised by an annealing process before hydrogen cycling. This will be discussed after exploring the hydrogen storage properties of these materials.

Mg2p C1s O1s Synthetic Mg C-C C-O MgO MgOH medium MgO/Mg(OH)x (51 (284.9 (286 (529.7 (52 eV) (532 eV) eV) eV) eV) eV) As- 21.93 18.43 18.15 2.45 7.09 24.98 Argon synthesised Cycled 37.71  8.03 1.26 12.27 31.92 As- 12.85 26.88 11.62 0.99 7.16 29.58 Hydrogen synthesised Cycled 41.36  5.79 0.94 10.62 32.11 As- 24.3 20.29 6.17 0.71 7.07 31.82 Cyclohexane synthesised Cycled 45.91  3.02 0.35 15.86 28.24 As- 31.06  14.47 12.82  38.52 Diethyl synthesised Ether Cycled 42.59  3.78 1.1 21.87 24.9

Figure 4.3.6 (Left) showed the gases evolution of the synthesised materials from the TPD analyses. Similar mass of gases were detected on all materials which depict the fragmentation of 1-butene as the by products of the decomposed di-n-butylmagnesium. In addition, there were also small amounts of gases from cyclohexane or diethyl ether from the materials decomposed in those solvents. The corresponding TGA/DSC analyses depicted several decomposition steps in all the materials and with only one main endothermic peak most likely related to the decomposition of MgH2 with hydrogen release.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.3.6 –MS of gases evolving from the thermal decomposition of the materials obtained (Left) and corresponded TG-DSC spectras (Right) after the hydrogenolysis of di-n-butylmagnesium: (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether. The fragments m/z 27, 41 and 56 correspond to the decomposition of the butyl group. The additional m/z 84 and 74 correspond to the release of cyclohexane and diethyl ether, respectively.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

It is noteworthy that the mass loss recorded for all the materials was in excess of the theoretical hydrogen storage capacity of MgH2, i.e. 7.6 mass %. The lowest mass loss was found to be of 9 mass % for MgH2 synthesized in cyclohexane. For MgH2 synthesised under argon or hydrogen pressure this mass loss was 13 %. As expected, for MgH2 prepared in diethyl ether, the mass loss was the highest with 37 % of the material decomposed at 400

C due to more carbon contaminant degradation. At this temperature magnesium would not evaporate and the significant mass losses observed can only be attributed to the degradation of the organic matter at the surface of the MgH2 particles as observed by MS

(Figure 4.3.6). This organic contamination would be strongly bound to the surface of the magnesium particles since its degradation occurred simultaneously with the release of hydrogen.

Hydrogen desorption properties The materials were then subjected to hydrogen absorption/desorption cycles for their hydrogen storage properties evaluation. The ability of these materials to cycle hydrogen was confirmed by XRD which showed the transition from -MgH2 to Mg upon heating at 300

C under vacuum (Figure 4.3.). However, materials produced in cyclohexane, diethyl ether and argon as synthetic mediums did not have a full transition. These incomplete releases of hydrogen could be due to various hydrogen bonding energies that would co-exist within the materials. With small size particles, as observed in the material with the cyclohexane medium, the low coordination number of the surface atoms would result in the formation of strong metal-hydrogen bonds at the nanoparticle surface. This led to hydrogen sorption upon widespread temperature range. The same phenomenon was observed with other nanosize hydrogen storage materials such as Pd 39 and self-assembled MgH2 nanoparticles.7

In other materials like those that synthesised in diethyl ether, the bigger particle sizes and carbon contamination contributions towards hydrogen sorption are the main causes of such poor cycling performance.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

These hypotheses are further confirmed by TPD analyses of these materials after cycling.

MgH2 produced in cyclohexane showed hydrogen release in two steps at 335 and 400 C from the material after cycling (Figure 4.3.-C and Figure 4.3.-C). MgH2 produced in argon and diethyl ether also shows high temperatures requirement for their hydrogen sorption, >

379 C as observed in TPD analyses (Figure 4.3. and Figure 4.3.). Accordingly, the size effects have been observed where particle sizes and hydrogen desorption peaks are correlated with each other. In this case the hydrogen desorption properties were improved as the particle sizes decrease.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.3.8 : TGA/DSC after the 3rd hydrogen absorption of the materials resulting from the hydrogenolysis of di-n-butylmagnesium :(A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether.

It is noteworthy that the kinetics of the hydrogen cycling show a similar trend to the particle sizes and hydrogen desorption peaks observed in TPD analyses, i.e. the kinetics decreased in the order: cyclohexane hydrogen > argon diethyl ether as the temperature required for hydrogen desorption increased (Figure 4.3.7). As shown in Figure 4.3.10, MgH2 that were synthesised under hydrogen pressure and cyclohexane showed the fastest desorption kinetics. It took less than 10 min to fully release hydrogen from these materials. Hydrogen desorption from MgH2 produced under an argon atmosphere or within diethyl ether had slower desorption kinetics as it took 40 and 30 min to fully release hydrogen, respectively.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.3.9 - MS of the gases evolving from the thermal decomposition of the materials after the 3rd hydrogen absorption. Materials were synthesized from the hydrogenolysis of di-n-butylmagnesium: : (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether.

Indeed, this kinetic trend also was reflected in their activation energy values (Ea) (Table

4.3.2). Ea for MgH2 obtained from the synthesis in cyclohexane was found to be 123 kJ.mol-

1, a value close to that reported for nanocrystalline magnesium (ball milled).40 However, for all the other materials, Ea was found within a range of 157-180 kJ.mol-1, i.e. values corresponding to that of polycrystalline magnesium.40 Another possibility is that the crystalline aluminium contained in material produced in cyclohexane could be beneficial towards the hydrogen desorption properties. There are a few studies which reported a destabilisation and kinetics improvements caused by the addition of Al into MgH2.41 42

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.3.7 – Hydrogen desorption kinetics of the materials obtained from the hydrogenolysis of di-n-butylmagnesium under an argon atmosphere, under hydrogen pressure, in cyclohexane and in diethyl ether. The 3rd hydrogen desorption cycle at 300 °C measured in High pressure TGA is shown.

However, as mentioned earlier, the presence of this Al could be detrimental towards the storage capacity. Nevertheless, the storage capacity of the materials prepared would depend upon the synthetic medium. The highest hydrogen storage capacity of 7.1 mass % was achieved for MgH2 synthesised under H2 pressure and the lowest for MgH2 prepared in diethyl ether (3.5 mass %). MgH2 generated under an argon atmosphere and cyclohexane showed intermediate hydrogen capacity of 6.0 and 5.2 mass %, respectively. These differences in hydrogen storage capacity may be due to the effectiveness of the hydrogenolysis process of di-n-butylmagnesium, the variations in the amount of generated

MgH2, and the amount of magnesium still ‘active’ for hydrogen storage within the materials, i.e. not further reacted during synthesis. Also, a synthetic done in solvents tends to produce more carbon contaminant that is added to the overall weight of the ‘inactive’ material on the samples.

By elemental analysis, the content of magnesium was found to vary depending on the synthesis method (Table 4.3.4) and all hydrogen storage capacities were below that of the materials’ theoretical capacity. For example, the elemental analysis of MgH2 prepared in argon and hydrogen pressure revealed 94 and 97 % magnesium content, respectively.

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

However, the measured hydrogen capacity was lower than theoretical values, i.e. 6 instead of 7.1 mass % for argon and 7.1 instead of 7.4 mass % for the synthesis under H2 pressure.

On the other hand, in cyclohexane the amount of magnesium as determined by elemental analysis was 85 % and only 69 % in diethyl ether (Table 4.3.1). These would correspond to theoretical storage capacities of 6.4 and 5.2 mass %, respectively. However, the measured hydrogen storage capacities are in fact lower (5.2 and 3.5 mass %, respectively). These are due to the additional impurities such as aluminium especially for material synthesised in cyclohexane and hydrocarbon for material synthesised in diethyl ether.

Table 4.3.4 – Magnesium and aluminium content as determined by elemental analysis. Synthetic medium Diethyl Argon Hydrogen Cyclohexane Ether Mg content by elemental 94 97 85 69 analysis after cycling (%)

Al content by elemental analysis 5 3 12 3 after cycling (%)

Freshly generated magnesium is known to readily react with many organic entities, and this has been the basis of numerous advances in organic chemistry 11 43 44 Thus as magnesium is generated during the synthetic process, it may further react to form bonds with hydrocarbons hence becoming un-reactive toward hydrogen. It has been reported if the carbon chain linked to the magnesium atom in the organomagnesium is saturated and long enough then it can be decomposed even at low temperatures.17 Analysis by NMR after cycling (Figure 4.3.8) revealed a decomposition of the organic matter within the materials and this was confirmed by XPS in agreement with previous findings.45 Thus upon heating at sufficient high temperatures most of the organic contamination would be removed.

However, the fact that overall hydrogen capacities of these materials are less than theoretical values, this means those carbons are still there and may be spread and

100

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

undetectable by NMR and XPS. Exceptionally high concentrations of organic contamination was observed in the material synthesised in diethyl ether. The effects of these contaminations are reflected on the morphologies and structural evolution of the materials after cycling.

Figure 4.3.8 - 13C NMR spectrum of the materials after hydrogen cycling and XPS wide- scan of the material obtained from the hydrogenolysis of di-n-butylmagnesium under hydrogen pressure after synthesis and hydrogen cycling. A similar reduction in carbon content was observed for all the other materials

TEM analysis of the materials after cycling (Figure 4.3.9) showed no significant evolution of the morphology except with the MgH2 synthesized under hydrogen pressure. The evolution of the morphology observed could be due to the shrinkage of the wall during the cycling thus it looks like it has been sintered. However, the rest of the materials retained similar morphologies and have relatively the same size as synthesized, eg. Argon – rods, cyclohexane – spheres, diethyl ether- flakes. These stabilities could be due to the persistent organic matters in those materials that stayed within the materials even after cycling, especially with those materials synthesised in diethyl ether. MgH2 synthesised under hydrogen pressure have the least amount of carbon contaminations and the most hydrogen storage capacity hence the deteriorate morphologies were expected. Despite these observations, as previously shown no materials showed any sign of sintered materials or deterioration in desorption properties. There was no significant change in crystallite sizes

101

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

after the 3rd hydrogen cycles (Table 4.3.2). Hence, it can be concluded that the particles remained mostly unchanged and relatively insensitive to sintering upon 3 hydrogen cycles.

Figure 4.3.9 – TEM of the materials after the 3rd hydrogen absorption. Materials were synthesized from the hydrogenolysis of di-n-butylmagnesium: : (A) under an argon atmosphere, (B) under hydrogen pressure, (C) in cyclohexane and (D) in diethyl ether.

It is noteworthy that this behaviour significantly differs from that of milled magnesium which undergoes significant crystalline growth from 10 to 200 nm after the first hydrogen absorption/desorption cycle at 300 C.26 27 46 The presence of this carbon contamination gave stabilization effects to the synthesised MgH2 nanoparticles limiting their agglomeration and sintering. However, the amount of carbon contaminants would also affect the hydrogen storage performance as it influenced the particles morphology. It has long been recognised that any presence of the impurities would modify the crystal growth process. Many has even tried to tailor the impurity as the additives to provide more control over the morphology.53 In this case, effect of solvent is essential as the tailor additives in the form of carbon contaminants which lead to the formation of relatively monodispersed and stable particles. However, the results show that the role of solvent-surface interactions are proven to be quite complex in determining the morphology of the particles. At least, based

102

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

on just the two samples produced in cyclohexane and THF, it shows that the surface interaction between them and Mg may even depended on their chemical structures. This mean that it is not solely based on the solvent polarity but in fact that the functional group of the solvent has some role, for example THF is reacted more to the Mg as it has the oxygen group. Another important point to consider is also the compatibility of the precursor and

MgH2 in these solvent selections. Theoretically, a better compatibility of the precursor or the Mg/MgH2 on the sample will enhance the supersaturation level for the nucleation process to occur. The chance for burst nucleation to occur is increased hence the smaller nanoparticle sizes would form. Indeed, these hypotheses support the results from various characterisations which lead to a conclusion that the material resulted from hydrogenolysis of di-n-butylmagnesium in cyclohexane is the smallest nanoparticle and it is superior in terms of kinetics and hydrogen desorption performances.

4.3.3 Conclusion

Further investigations on the hydrogenolysis products of di-n-butylmagnesium in different synthetic mediums have been performed. Indeed, remarkable changes in morphologies were observed within the resulting materials and their influences on the hydrogen storage properties were discussed. Several main findings were obtained in this section of the chapter. First, hydrogenolysis (addition of H2 pressure) during the decomposition of a precursor would help to enhance the nucleation and growth process of MgH2 and lead to favourable structures (smaller size and spherical morphology) compared to when it was done under inert atmosphere (Ar). Secondly, Mg is really reactive towards organic substances and caused them to have more carbon contaminant deposited in the materials that were formed in solvents compared to dry conditions. The types of solvents also determined the amount of carbon contaminants that were formed and it was shown that a more polar solvent such as diethyl ether has more carbon contaminations compared to

103

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

when a non-polar such as cyclohexane was used. These also caused distinctive differences in morphologies due to the formation of these carbon contaminations on the surfaces that influence the nucleation and growth process. In the end, the materials that were produced in cyclohexane are the most stable and smallest nanoparticles with homogeneous morphology, and also lead to the best hydrogen desorption behaviours.

4.4 Investigation of Mg produced from MgA.(THF)3

As mentioned in Chapter 2.4.2, MgA.(THF)3 is a unique precursor that is able to generate monoatomic Mg or MgH2 in several ways, including the steps that involve a mild temperature condition. Such steps can be crucial in order to avoid a chemical reaction between magnesium with organic substances at high temperature, which was observed in

Chapter 4.3. A mild condition is also important too since there is a possibility of using a stabiliser which would decompose at high temperature conditions. Indeed, the MgA.(THF)3 is known for its capability to reversibly decompose to the original precursors (anthracene,

Mg, THF) and form back to MgA.(THF)3 at between 20-60 C (Scheme 4-4).

Scheme 4-4 Reversible reaction of MgAnthracene.3THF complex

In fact, because of this equilibrium reversible reaction, MgA.(THF)3 can become a source of atomic magnesium while at the same time react as a di-nucleophile in solution.44 These properties made it particularly attractive for the synthesis of other organomagnesium compounds such as Grignard reagents.47 Furthermore, studies by Bogdanovic et al have investigated the possibility of producing MgH2 from MgA.(THF)3 including the potential for hydrogen storage. The result showed promising hydrogen storage properties with kinetics

104

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

48 comparable to those ball milled MgH2. Nevertheless, limited studies on the physical properties of these MgH2 encourage us to investigate further this MgH2 and eventually identify the possibility of generating smaller nanoparticles including the use of stabiliser.

4.4.1 Experimental details

Materials

Mg powder (20-30 mesh, 99.7%), anthracene (99%), chromium chloride, anhydrous tetrahydrofuran, cyclohexane and diethyl ether (> 99%) were obtained from Sigma-Aldrich.

Ethyl bromide (97%) was obtained from Ajax APS.

Catalytic hydrogenation of excess Mg with MgA.(THF)3 equilibrium

Mg powders (0.25 g, 1.04 x 10-2 mol) were stirred with ethyl bromide (0.1 ml, 1.34 x 10-3 mol) in THF (20 ml, 2.47 x 10-1 mol) for 1 h at room temperature. Anthracene (0.64 g, 3.61 x 10-3 mol) were added to the mixture and stirred for 2 h at 60 °C. The solution mixture changed its colour from colourless to green slowly and eventually an orange precipitate started to appear which indicated the formation of MgA.(THF)3. The precipitates solution was further reacted at room temperature overnight to ensure the equilibrium of the reaction had been reached. CrCl3 (7 mg, 4.41 x 10-5 mol) was added to the mixture with another addition of 20 mL THF. Then the mixture was transferred to a pressure reactor vessel to be hydrogenolysed with 30 bar H2 pressure and temperature of 60 °C for 24 h. After reaction, the solution was centrifuged and then washed with THF and cyclohexane twice to remove the by-products and excess reagents. Grey powders were obtained after solvent removal under vacuum (Yield 0.246 g, 91 wt%).

Mg synthesis from decomposition and hydrogenation of MgA.(THF)3

There are several different methods to decompose MgA.(THF)3 and release monovalent Mg.

The first approach is through catalytic hydrogenation similar to the previous method but this time a known amount of MgA.(THF)3 was used instead. The MgA.(THF)3 was

105

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

synthesised and isolated before going through the hydrogenation process with catalyst. The synthesis process involved the reaction of Mg and anthracene with the same molar ratio.

Mg powders (1 g, 4.16 x 10-2 mol) were stirred with ethyl bromide (0.2 ml, 2.68 x 10-3 mol) in 140 ml THF at room temperature for 1 h. Anthracene (7.49 g, 4.20 x 10-3 mol) was added to the mixture and stirred for 2 h at 60 °C. The mixture was stirred further for 48 h to ensure that all Mg was fully converted to MgA.(THF)3. The orange suspension was filtered and washed with THF several times to remove any unreacted anthracene. The isolated orange solid was then characterised by 1H NMR. 1H-NMR (300 MHz, CDCl3) d (ppm from TMS); 1.8-

2.0 (6H, m, -O CH2CH2-), 3.51 (1H, s, -CH-Mg) 3.6 (6H, m, -O CH2CH2-), 5.9-6.0 (8H, m, CH-

Ph). Then, this isolated MgA.(THF)3 (2 g, 4.78 x 10-3 mol) was suspended in 100 mL THF and mixed with 0.007 g of CrCl3 (7 mg, 4.41 x 10-5 mol). The mixture was transferred to the pressure reactor vessel and hydrogenated with 30 bar H2 at 60 °C. The reaction was kept for

15 h and the products were collected as the black greenish precipitates after several washes with THF (Yield 0.1223 g, 98%)

The second approach is to decompose MgA.(THF)3 in different several organic solvents such as cyclohexane, toluene, and diethyl ether. Typically, MgA.(THF)3 (1 g, 2.39 x 10-3 mol) were suspended in 50 mL of the selected solvent and stirred at 60 °C for 24 h. Depending on the amount and type of solvent, the orange suspension eventually became a grey suspension.

The suspension was centrifuged and grey precipitates were collected after being washed several times with THF. The dried solids were hydrogenated with 30 bar H2 at 300 °C for 24 h before being further characterised.

The third approach is to thermally decompose MgA.(THF)3 under vacuum to remove the anthracene and THF which would leave only Mg at the end. Typically, MgA.(THF)3 (1 g, 2.39 x 10-3 mol) was heated at 200 °C while subjected to dynamic vacuum in the pressure rig.

THF was released to the vacuum pump and anthracene was recovered as sublimate. After

106

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

10 h decomposition reaction, the heating was increased to 300 °C to perform hydrogenolysis with 30 bar H2.

4.4.2 Results and discussion

The catalytic hydrogenation cycles of MgA.(THF)3 had been proven to be an economical and efficient method to generate large amounts of MgH2 since the use of small quantities of anthracene and THF.48 As has been discussed in Chapter 2.4.2, the anthracene and THF can be recycled only if the equilibrium concentration of anthracene and Mg in the reaction mixture is enough to form MgA.(THF)3. Indeed, the concentration of anthracene in the mixture should always be higher than 0.01 mol/L to allow the continuous formation of

MgA.(THF)3. In this study, the concentration of anthracene used was 0.09 mol/L to convert

0.72 mol/L of bulk Mg powder to MgH2 through MgA.(THF)3 as intermediate. The conditions were chosen to ensure that all Mg was converted to MgH2 via MgA.(THF)3 in just 24 h of reaction.

Indeed, a pure β-MgH2 phase (Figure 4.4.1-A) was observed from the XRD analysis of the material that was produced from this approach. From here on, the material produced through this method will be called MgH2-Mg(excess)A. For comparison, the catalytic hydrogenation with the same conditions also performed using the isolated form of

MgA.(THF)3 which was synthesised by the same molar ratio of Mg and anthracene. The result of this hydrogenated material will be called MgH2-MgA.

107

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.4.1 XRD spectra of (A) MgH2-Mg(excess)A, (B) MgH2-MgA,

During the hydrogenation, all MgA.(THF)3 was expected to fully convert to MgH2 and nonetheless the XRD analysis confirmed that the MgH2-MgA consisted of the pure crystalline phase of β-MgH2 (Figure 4.4.1-B). Further analyses of these spectra by Scherrer equation revealed the crystalline sizes for β-MgH2 phases in both materials are between 16-18 nm

(Table 4.4.1).

The TEM images showed the morphology of these materials to be cubical/ rods particles in

20-30 nm sizes for MgH2-Mg(excess)A (Figure 4.4.2-A), and smaller nanoparticles with 5-

15 nm in size for MgH2-MgA but without any specific morphologies (Figure 4.4.2-B). The formation of these small size nanoparticles were quite unexpected since they were formed without any presence of stabilising agents during the synthesis process in order to stop the growth of these nanoparticles after precipitation. But as shown in Chapter 4.3, a similar phenomenon occurred with the materials obtained from the hydrogenolysis of di-n- butylmagnesium in solvents. We mentioned that the surface of magnesium could react with solvents and cause the formation of carbon contaminations on the surface that stopped the growth of these nanoparticles. Overall, there were many factors that could influence the amount of this contamination hence the morphology of the grown nanoparticles. These include the solvent type (THF), overall concentrations of anthracene, H2 pressure and temperature (60 °C).

108

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.4.2 TEM analysis of (A) MgH2-Mg(excess)A and (B) MgH2-MgA as synthesised

Other attempts to obtain precipitates Mg from the MgA.(THF)3 is through an anti- precipitation solvents method where the MgA.(THF)3 was suspended in other solvents such as cyclohexane, toluene and diethyl ether. The equilibrium between Mg, THF, and anthracene is strongly dependent on the solvent composition and the temperature.8 In this study, the isolated MgA.(THF)3 was decomposed at 60 °C in 50 ml of selected solvent for 24 h. Among all the solvents that were tried, cyclohexane is depicted as the best organic solvent to fully convert the MgA.(THF)3 to Mg. It was based on experimental observations where the suspension undergoes changes in colour from its original colour, orange (MgA.(THF)3) to grey (Mg) after a certain period of time. In diethyl ether, the suspension only changed to a dark brown colour which indicates the presence of MgA.(THF)3 even after more than 24 h reaction. While in toluene, the suspension did not undergo any colour change after 24 h reaction which indicates the decomposition of MgA.(THF)3 did not occur at all. Moreover, such low decomposition rates are not preferable since the long exposure to the solvent would increase the formation of hydrocarbon contamination significantly. In addition, the use of a non-polar solvent such as cyclohexane works better to minimise the reaction with

Mg surfaces.

Nevertheless, the precipitates collected from the anti-precipitation methods of the

MgA.(THF)3 in cyclohexane were further analysed. TEM analysis of this material (Figure

4.4.3-A) depicted a large dendritic structure formed during the precipitation in cyclohexane.

109

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

The dendritic structures growth is common during the solidification of metal as a result of ripening due to surface energy.49 However, the slow precipitation process led to the formation of large particle sizes. The precipitation rates can be increased in several ways by adjusting external factors such as diluting the concentration of the MgA.(THF)3 or increasing temperature. However eventually, some sort of stabiliser or template is required to stop the growth in order to obtain small size Mg nanoparticles from this method.

(A) (B)

Figure 4.4.3 (A) materials precipitated from MgA in cyclohexane (B) materials from decomposition of MgA under vacuum

Another potential method for obtaining Mg from the MgA.(THF)3 is by subliming the anthracene and evaporating off the THF component. It was done by heating the MgA.(THF)3 at 200 °C under vacuum for 24 h which is enough to perform the sublimation. Theoretically, this process would give a higher decomposition rate than the anti-precipitation method.50

Indeed, TEM analysis of this material revealed the formation of smaller particles with the size between 100-200 nm with unique polygon shapes (Figure 4.4.3-B). The broad size distribution of the particles is mainly due to the lack of control over the growth process during the decomposition. Narrow size distribution and a smaller size could be achieved if the isolated MgA.(THF)3 are subjected to high surface areas for deposition and rapid cooling process which is usually applied in CVD method. 51 Similarly, this method is highly inefficient or scalable compared to the wet synthesis method.

110

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Hydrogen sorption properties

The materials produced from different decomposition methods of MgA.(THF)3 were subjected to hydrogenation under the same temperature and H2 pressure. As expected of such large structure Mg, it behaved like bulk Mg which would not easily absorb hydrogen at

300 °C and with only 30 bar H2 pressure. The XRD analyses (Figure 4.4.4) shows Mg phases dominate the XRD patterns of the hydrogenated materials obtained from anti-precipitation methods while the material from sublimation of MgA.(THF)3 was able to be partially hydrogenated to become MgH2 after 24 h hydrogenation.

(A)

(B)

(C)

Figure 4.4.4 (A) materials precipitated from MgA in cyclohexane hydrogenated at 300

°C 30 bar H2, (B) materials precipitated from MgA in diethyl ether hydrogenated at

300 °C 30 bar H2, (C) materials from decomposition of MgA under vacuum and hydrogenated at 300 °C 30 bar H2

These results show the versatility of the MgA.(THF)3 as precursor to form Mg where presence of the stabiliser or template to stop the growth process would help to impede the growth process of the particle.

Furthermore, a deep investigation was done on MgH2-Mg(excess)A and MgH2-MgA since they were the only materials that consist of nanostructured MgH2 materials. The TPD analyses of MgH2-Mg(excess)A is shown in Figure 4.4.5.

111

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.4.5 TPD analysis of MgH2-Mg(excess)A (left) TGA-DSC analysis and the corresponding MS spectra (right)

A small mass loss was observed (<5 wt%) at low temperature region (100 °C) which was due to the residual solvent within the material sample. It started to undergo significant mass loss (nearly 35 wt%) from 280 °C until 400 °C. This massive mass loss was mainly due to the hydrocarbon releases from the materials. It is noteworthy that two decomposition events were observed during this mass loss where the first event (25 wt% mass loss) occurred at 280-300 °C while accompanied by the exothermic peaks from the DSC signal.

This is not the usual DSC signal for MgH2 decomposition which usually shows the endothermic peaks. Indeed, this event should correspond to the decomposition of anthracene. This is confirmed by the observed evolution of the anthracene peaks on the correlated MS spectra (Figure 4.4.5). Meanwhile, the MgH2 decomposition event occurred at higher temperature and it was proven by the presence of a broad endothermic peak at

372 °C from the DSC signal. The mass loss during this decomposition step is around 8-10 wt% and occurred at slower rate (320-400 °C) displayed by the broad endothermic peak.

Also, this was followed by the observation of hydrogen evolution at those peaks by MS alongside the gases associated with fragmentations of THF during this temperature. As observed previously, solvent contaminations usually released at the same time during the hydrogen desorption of MgH2 if they are located on the Mg surface. This explained the higher mass loss that was observed compared to the theoretical capacity of Mg (7.6 wt%).

112

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

On the other hand, the TPD analysis of MgH2-MgA (Figure 4.4.6) shows that it also underwent significant mass loss (25 wt%) from 250-320 °C in the TPD analysis. However, the decomposition of the anthracene is not as significant as the MgH2-Mg(excess)A considering there was no exothermic peak observed. The DSC signal also only depicts a narrow endothermic peak at 351 °C which should correspond to the decomposition of MgH2 event. However, the MS spectra revealed the evolution of anthracene and THF fragmentations before and during this decomposition of MgH2. This observation indicates the possibility of anthracene binding within the surface of MgH2 nanoparticles similar to solvent binding where it was released alongside the hydrogen.

Figure 4.4.6 TPD analysis of MgH2-MgA (left) TGA-DSC analysis and the corresponding MS spectra (right)

This hypothesis can be correlated with physical structures of these materials observed in

TEM (Figure 4.4.3) where the MgH2-MgA particle sizes are smaller compared to the MgH2-

Mg(excess)A. It is possible that having a smaller concentration of anthracene during the synthesis (MgH2-Mg(excess)A) might lead to less restricted growth of MgH2 and thus led to bigger particles while on the contrary, the more concentrated anthracene (MgH2-MgA) resulted in a higher chance for anthracene to bind on the MgH2 that led to smaller particles.

Obviously, the smaller particles (MgH2-MgA) has the hydrogen release at lower temperature than the bigger particles (MgH2-Mg(excess)A ).

113

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Both materials were then subjected to hydrogen absorption/desorption cycles to evaluate their hydrogen storage properties further. The reversibility of the materials was confirmed by the XRD analyses shown in Figure 4.4.7 where both were able to fully form β- MgH2 after going through several hydrogen cyclings at 300 °C and with 30 bar H2 pressure. However, the β- MgH2 peaks became significantly narrower compared to the freshly synthesised materials. This indicated the growth of the crystal size after cycling. The crystalline sizes determined by Scherrer equations showed increases to 48 nm and 30 nm for MgH2-

Mg(excess)A and MgH2-MgA respectively (Table 4.4.1).

Figure 4.4.7 XRD spectra of (A) cycled MgH2-Mg(excess)A, (B) cycled MgH2-MgA

Increase in the particle sizes was also observed in TEM of both materials after they were cycled (Figure 4.4.8). The agglomeration occurred more obviously on the cycled MgH2-

Mg(excess)A where large grain particle sizes (>200 nm) was revealed. However, cycled

MgH2-MgA consisted of small size nanoparticles but with more compactness within the particles. Cycled MgH2-MgA may be able to retain the small particle sizes due to the higher amount of carbon contaminations (anthracene) that are bound on the surface to prevent the sintering process occurring. This was the same observation from those MgH2 that were obtained from di-n-butylmagnesium decomposition in Chapter 4.3.

114

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Figure 4.4.8 TEM analyses of (A) MgH2-Mg(excess)A cycled and (B) MgH2-MgA cycled

Desorption kinetic of the materials was analysed and Figure 4.4.9 depicts two hydrogen cycles. MgH2-Mg(excess)A released hydrogen at a slower rate compared to MgH2-MgA.

MgH2-Mg(excess)A took over 100 minutes before fully releasing its hydrogen capacity (~5 wt%) while MgH2-MgA fully released ~4wt% H2 at only ~45 min. Again, the size effects played a role in in the hydrogen release kinetic of these materials. However, although having superior kinetics the capacity of MgH2-MgA suffered from the high amount of carbon contaminations.

Figure 4.4.9 Hydrogen desorption kinetics of MgH2-Mg(excess)A and MgH2-MgA. The 3rd hydrogen desorption cycle at 300 °C measured in Sievert Instrument is shown.

TPD analyses of the cycled materials are shown in Figure 4.4.10. Cycled MgH2-Mg(excess)A shows 5 wt% mass loss starting at 350 °C until above 400 °C. The endothermic signal peaked at 371 °C and a smaller peak was observed at temperatures higher than 400 °C. These multiple peaks were observed more clearly on the hydrogen evolution in corresponding MS

115

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

spectra. This worsening of the hydrogen desorption properties was due to the agglomeration of the MgH2-Mg(excess)A after being cycled as shown earlier. The fact that it has multiple peaks should indicate there were stable forms of MgH2 even after several cycles.

Figure 4.4.10 TPD analysis of cycled MgH2-Mg(excess)A with TGA-DSC analysis (left) and the corresponding MS spectra (right)

On the other hand, the cycled MgH2-MgA (Figure 4.4.11) underwent a smaller amount of mass loss (3 wt%) from 320-370 °C while accompanied by a narrow endothermic peak at

351 °C. This would be similar to those the freshly synthesised material had. However, the mass loss from lower temperatures until 300 °C (1.5 wt%) is not clear since the MS did not observe any fragments of ions being released during this mass loss. It is possibly due to the melting of the hydrocarbon impurities within the material which decreases the material’s density. Indeed, the overall H2 capacity of the MgH2-MgA is low due to the high carbon contaminations content, especially from the anthracene that bound to the magnesium surface. These hydrocarbons led to the possibility of the magnesium carbide formation which decreases the active part of Mg for hydrogen release and uptake.

Further investigation of these materials would focus on the relationship between the structural properties and the hydrogen storage properties. These properties of both MgH2-

Mg(excess)A and MgH2-MgA were compared with materials obtained from hydrogenolysis of di-n-butylmagnesium at dry conditions and in cyclohexane (MgH2-H2MgBu2 and MgH2-

116

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

CyclohexaneMgBu2). The comparisons are summarised in Table 4.4.1 including the associated crystallite sizes, activation energies, enthalpy and entropy values.

Figure 4.4.11 TPD analysis of cycled – MgH2-MgA with TGA-DSC analysis (left) and the corresponding MS spectra (right)

Figure 4.4.12 shows the Van’t Hoff plots that were used to generate the ∆H and ∆S values of these materials.

Surprisingly, materials made from MgA.(THF)3 have larger ∆H and ∆S values despite the smaller particle sizes that they have. MgH2-MgA has similar ∆H to bulk MgH2, ~76 kJ.mol-

1.H2 and ∆S of 133.5 J.mol-1 K-1.H2.19 MgH2-Mg(excess)A has an even higher value compared

117

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

-1 to the bulk MgH2 (∆H is 83 kJ.mol ) which is unexpected since its hydrogen sorption properties are definitely better than those of bulk MgH2. But we found that its ∆S also underwent big increases to 151 J.mol-1 K-1.H2 where it may counteract the increase of ∆H thus resulting in minimum change in hydrogen desorption behaviour. On the contrary,

MgH2-H2MgBu2 and MgH2-CyclohexaneMgBu2 have low ∆H and ∆S compared to the bulk

MgH2. In general, these thermodynamic values also corroborated well with the crystalline size of the materials after hydrogen cycling. The decreasing ∆H and ∆S values also followed with the decrease in crystalline size, e.g. cycled MgH2-Mg(excess)A has the biggest crystalline size (48 nm) and cycled MgH2-CyclohexaneMgBu2 has the smallest size (23 nm).

However, such big decreases in ∆H did not reflect the temperature of hydrogen release of these materials. Again, the compensation of both ∆H and ∆S mitigated the big improvements on the hydrogen desorption properties. As has been discussed in Chapter 2.3.2, enthalpy- entropy compensation effects have been the subject of many studies including on the size effects on metal hydrides.52 In here, the effect of both enthalpy-entropy compensation is confirmed by the linear relationship of ∆H and ∆S of the materials investigated as shown in

Figure 4.4.13.

(A) (B) (C)

(D)

Figure 4.4.13 Plot of enthalpy and entropy changes of the (A) MgH2-Mg(excess)A, (B)

MgH2-MgA, (C) MgH2-H2MgBu2, (D) MgH2-Cyclohexane MgBu2 reflecting the compensation effects

118

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

Meanwhile, the decomposition activation energy, Ea values of the materials revealed that

MgH2 produced from MgA.(THF)3 also have relatively larger Ea than MgH2- Cyclohexane

MgBu2 (also shown in Table 4.4.1). Indeed, this result reflected the kinetics of the materials as shown earlier where both MgH2-Mg(excess)A and MgH2-MgA desorbed hydrogen at a slower rate compared to the more superior kinetic of MgH2- Cyclohexane MgBu2 (Figure

4.3.7 and Figure 4.4.9). However MgH2- H2MgBu2 has the highest calculated Ea with 166 kJ.mol-1 although its kinetic was comparable with MgH2- Cyclohexane MgBu2. The broad distribution in particle sizes of MgH2- H2MgBu2 may be the reason for this. In addition, although MgH2-H2MgBu2 has the highest Ea among all materials, it has the lowest amount of contaminant (proven by its high H2 capacity ~7 wt%) which may not impede the hydrogen diffusion so much compared to other materials during its release. This would explain the slower desorption kinetics of the MgH2-Mg(excess)A and MgH2-MgA.

E ∆H ∆S a Materials (kJ.mol- (kJ.mol-1. (J.mol-1 After After 1 H ) H ) K-1. H ) synthesis cycling 2 2 2

MgH2-Mg(excess)A 18 ± 2 48 ± 1 144 ± 8 -83 ± 2 151 ± 4

MgH -MgA 2 16 ± 2 30 ± 3 152 ± 7 -74 ± 3 139 ± 6

MgH - H MgBu 2 2 2 32 ± 5 30 ± 3 166 ± 8 -64 ± 3 120 ± 5 MgH - Cyclohexane 133 ± 2 17 ± 3 23 ± 3 -53 ± 5 102 ± 8 MgBu2 10

4.4.3 Conclusion

In this sub chapter, we have considered utilising MgA.(THF)3 as an alternative precursor to generate active Mg. In fact, several methods were proven to be of high potential for large

119

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

scale production of active Mg/MgH2 from this particular precursor. The first method is the catalytic hydrogenation approach using a transitional metal catalyst (CrCl3) to facilitate the conversion of MgA.(THF)3 to MgH2. Two types of experiments were performed either by using an excess amount of Mg or the isolated form of MgA.(THF)3 intermediate. Other approaches are by thermally decomposing the isolated form of MgA.(THF)3 where it can be done either by anti precipitation in different organic solvents or evaporated off both THF and anthracene under vacuum at temperature of 200 °C. The results revealed only catalytic hydrogenation methods capable of producing small size particles without the presence of stabilisers. However, the thermal decomposition methods would result in Mg with a lower concentration of contaminant which offers other alternatives to produce small nanoparticles when combined with good stabilizing strategies. Without the presence of any any stabilisers, only hydrocarbon contaminations will act as the stabiliser and determine the size of the particles formed. This is shown by the material made from the catalytic hydrogenation using the isolated form of MgA.(THF)3 however the persistent carbon contaminants are proven to be detrimental towards the H2 capacity. These findings challenged the solvothermal synthesis method of MgH2 to minimise the formation of the persistent organic contaminants to be removed. This would be one of the priorities to be tackled in the next chapter that discusses the modification attempts of these materials.

4.5 Conclusion

This chapter explored different organomagnesium precursors and methods to obtain MgH2 nanoparticles. The main findings show that the conditions of the synthesis methods and precursors would highly influence the physical properties as well as the hydrogen storage properties of the MgH2 produced. In general, the formation of impurities in MgH2 generated through wet synthesis method is nearly impossible to avoid because of the high reactivity

120

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

of Mg. Combinations of the polarity of solvent, temperature, and H2 pressure influenced the amount of impurities and morphology of the Mg formed. These hydrocarbon impurities were persistent and difficult to remove upon hydrogen cycling and eventually deteriorated the overall H2 storage capacity of the material. Nonetheless, the investigations show that a non-polar solvent such as cyclohexane could minimise the formation of these impurities. It was found that hydrogenolysis of di-n-butylmagnesium in this solvent led to the formation of highly uniform MgH2 nanoparticles (~20 nm in size). Remarkably, the nanoparticles were stable upon many hydrogen cycles due to the formation of a thin carbon layer on the surface.

A similar stabilisation process was observed in the materials made from MgA.(THF)3 complex as a precursor. In this way, smaller MgH2 nanoparticles (10-15 nm) with lower temperature (60 °C) compare to the hydrogenolysis of di-n-butylmagnesium (180 °C).

However, the amount of contamination is greater since the polar solvent (THF) and other organic substances (anthracene) are present during the synthesis.

121

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

4.6 References

1. Cushing, B.L., V.L. Kolesnichenko, and C.J. O'Connor, Recent advances in the liquid-phase syntheses of inorganic nanoparticles. Chemical reviews, 2004. 104(9): p. 3893-3946. 2. Semagina, N. and L. Kiwi‐Minsker, Recent Advances in the Liquid‐Phase Synthesis of Metal Nanostructures with Controlled Shape and Size for Catalysis. Catalysis Reviews, 2009. 51(2): p. 147-217. 3. Yin, Y. and A.P. Alivisatos, Colloidal nanocrystal synthesis and the organic-inorganic interface. Nature, 2005. 437(7059): p. 664-70. 4. Banno, T., Y. Hayakawa, and M. Umeno, Some applications of the Grignard cross-coupling reaction in the industrial field. Journal of organometallic chemistry, 2002. 653(1): p. 288- 291. 5. Wiberg, E. and R. Bauer, Der Magnesiumwasserstoff MgH2. Chemische Berichte, 1952. 85(6): p. 593-605. 6. Michalczyk, M.J., Synthesis of magnesium hydride by the reaction of phenylsilane and dibutylmagnesium. Organometallics, 1992. 11(6): p. 2307-2309. 7. Aguey‐Zinsou, K.F. and C. Boyer, Synthesis and Stabilisation of MgH2 Nanoparticles by Self‐ Assembly. ChemPlusChem, 2012. 77(6): p. 423-426. 8. Bogdanović, B., et al., Rate of formation and characterization of magnesium anthracene. Chemische Berichte, 1984. 117(4): p. 1378-1392. 9. Aguey-Zinsou, K.-F. and J.-R. Ares-Fernández, Synthesis of colloidal magnesium: a near room temperature store for hydrogen. Chemistry of Materials, 2007. 20(2): p. 376-378. 10. Haas, I. and A. Gedanken, Synthesis of metallic magnesium nanoparticles by sonoelectrochemistry. Chem. Commun., 2008(15): p. 1795-1797. 11. Ashby, E.C., R. Kovar, and K. Kawakami, Existence of HMgX compounds. Inorganic Chemistry, 1970. 9(2): p. 317-324. 12. Becker, W.E. and E.C. Ashby, Hydrogenolysis of the Grignard Reagent. The Journal of Organic Chemistry, 1964. 29(4): p. 954-955. 13. Podall, H., H. Petree, and J. Zietz, Relative Ease of Hydrogenolysis of Some Organometallic Compounds1. The Journal of Organic Chemistry, 1959. 24(9): p. 1222-1226. 14. Jongh, P.E.d., et al., The preparation of carbon-supported magnesium nanoparticles using melt infiltration. Chemistry of Materials, 2007. 19(24): p. 6052-6057. 15. Nielsen, T.K., et al., Confinement of MgH2 nanoclusters within nanoporous aerogel scaffold materials. ACS nano, 2009. 3(11): p. 3521-3528. 16. LefranÇois, M. and Y. Gault, Décomposition thermique des halogénures d'éthylmagnésium. Journal of Organometallic Chemistry, 1969. 16(1): p. 7-19. 17. Jolibois, P., Formula of the organo-magnesium derivative and magnesium hydride. Compt. rend, 1912. 155: p. 353-355. 18. Huot, J., et al., Structural study and hydrogen sorption kinetics of ball-milled magnesium hydride. Journal of Alloys and Compounds, 1999. 293: p. 495-500.

122

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

19. Zaluska, A., L. Zaluski, and J. Ström–Olsen, Nanocrystalline magnesium for hydrogen storage. Journal of Alloys and Compounds, 1999. 288(1): p. 217-225. 20. Friedrichs, O., et al., Chemical and microstructural study of the oxygen passivation behaviour of nanocrystalline Mg and MgH2. Applied Surface Science, 2006. 252(6): p. 2334- 2345. 21. Fotea, C., J. Callaway, and M.R. Alexander, Characterisation of the surface chemistry of magnesium exposed to the ambient atmosphere. Surface and interface analysis, 2006. 38(10): p. 1363-1371. 22. Wagner, C.D., et al., Handbook of X-Ray Photoelectron Spectroscopy. 1997: Perkin & Elmer Cooperation. 23. Paskevicius, M., D.A. Sheppard, and C.E. Buckley, Thermodynamic changes in mechanochemically synthesized magnesium hydride nanoparticles. Journal of the American Chemical Society, 2010. 132(14): p. 5077-5083. 24. Zhang, X., et al., Synthesis of magnesium nanoparticles with superior hydrogen storage properties by acetylene plasma metal reaction. International Journal of Hydrogen Energy, 2011. 36(8): p. 4967-4975. 25. Aguey-Zinsou, K.-F., et al., Using MgO to improve the (de) hydriding properties of magnesium. Materials Research Bulletin, 2006. 41(6): p. 1118-1126. 26. Friedrichs, O., et al., MgH with NbO as additive, for hydrogen storage: Chemical, structural and kinetic behavior with heating. Acta Materialia, 2006. 54(1): p. 105-110. 27. Paik, B., et al., Microstructure of ball milled MgH 2 powders upon hydrogen cycling: An electron microscopy study. International Journal of Hydrogen Energy, 2010. 35(17): p. 9012-9020. 28. Berube, V., et al., Size effects on the hydrogen storage properties of nanostructured metal hydrides: a review. International Journal of Energy Research, 2007. 31(6‐7): p. 637-663. 29. de Jongh, P.E. and P. Adelhelm, Nanosizing and nanoconfinement: new strategies towards meeting hydrogen storage goals. ChemSusChem, 2010. 3(12): p. 1332-1348. 30. Fichtner, M., Properties of nanoscale metal hydrides. Nanotechnology, 2009. 20(20): p. 204009. 31. Li, W., et al., Magnesium nanowires: enhanced kinetics for hydrogen absorption and desorption. Journal of the American Chemical Society, 2007. 129(21): p. 6710-6711. 32. Shao, H., et al., Hydrogen storage properties of magnesium ultrafine particles prepared by hydrogen plasma-metal reaction. Materials Science and Engineering: B, 2004. 110(2): p. 221-226. 33. Jeon, K.J., et al., Air-stable magnesium nanocomposites provide rapid and high-capacity hydrogen storage without using heavy-metal catalysts. Nat Mater, 2011. 10(4): p. 286-90. 34. Lu, Z., et al., On the electrochemical behavior of magnesium electrodes in polar aprotic electrolyte solutions. Journal of Electroanalytical Chemistry, 1999. 466(2): p. 203-217. 35. Xu, L., et al., ZnO with different morphologies synthesized by solvothermal methods for enhanced photocatalytic activity. Chemistry of Materials, 2009. 21(13): p. 2875-2885. 36. Biswas, S., S. Kar, and S. Chaudhuri, Effect of the precursors and solvents on the size, shape and crystal structure of manganese sulfide crystals in solvothermal synthesis. Materials Science and Engineering: B, 2007. 142(2): p. 69-77.

123

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

37. Burda, C., et al., Chemistry and properties of nanocrystals of different shapes. Chemical reviews, 2005. 105(4): p. 1025-1102. 38. Demazeau, G., et al., Contribution of solvothermal processes to the synthesis of novel nitrides and the development of shaping processes. ChemInform, 2009. 40(13): p. i. 39. Yamauchi, M., et al., Nanosize effects on hydrogen storage in palladium. The Journal of Physical Chemistry C, 2008. 112(9): p. 3294-3299. 40. Aguey-Zinsou, K.-F. and J.-R. Ares-Fernández, Hydrogen in magnesium: new perspectives toward functional stores. Energy & Environmental Science, 2010. 3(5): p. 526. 41. Domenech-Ferrer, R., et al., Hydrogenation properties of pure magnesium and magnesium–aluminium thin films. Journal of power sources, 2007. 169(1): p. 117-122. 42. Song, Y., Z. Guo, and R. Yang, Influence of selected alloying elements on the stability of magnesium dihydride for hydrogen storage applications: A first-principles investigation. Physical Review B, 2004. 69(9): p. 094205. 43. Bartmann, E., et al., Active Magnesium from Catalytically Prepared Magnesium Hydride or from Magnesium Anthracene and its Uses in the Synthesis. Chemische Berichte, 1990. 123(7): p. 1517-1528. 44. Bogdanović, B., Catalytic synthesis of organolithium and organomagnesium compounds and of lithium and magnesium hydrides—Applications in organic synthesis and hydrogen storage. Angewandte Chemie International Edition in English, 1985. 24(4): p. 262-273. 45. Konarova, M., et al., Porous MgH 2/C composite with fast hydrogen storage kinetics. International Journal of Hydrogen Energy, 2012. 37(10): p. 8370-8378. 46. Aurora, A., et al., Microstructural and kinetic investigation of hydrogen sorption reaction of MgH2/Nb2O5 nanopowders. Materials and Manufacturing Processes, 2009. 24(10-11): p. 1058-1063. 47. Raston, C.L. and G. Salem, Magnesium anthracene: an alternative to magnesium in the high yield synthesis of Grignard reagents. Journal of the Chemical Society, Chemical Communications, 1984(24): p. 1702. 48. Bogdanovic, B., Magnesium anthracene systems and their application in synthesis and catalysis. Accounts of Chemical Research, 1988. 21(7): p. 261-267. 49. Flemings, M.C., Behavior of metal alloys in the semisolid state. Metallurgical Transactions B, 1991. 22(3): p. 269-293. 50. Gspann, J., On the phase of metal clusters. Zeitschrift für Physik D Atoms, Molecules and Clusters, 1986. 3(2): p. 143-145. 51. Zhu, C., et al., Shape-controlled growth of MgH2/Mg Nano/microstructures via hydriding chemical vapor deposition. Crystal Growth & Design, 2010. 10(12): p. 5123-5128. 52. Tang, W.S., et al., Enthalpy–Entropy Compensation Effect in Hydrogen Storage Materials: Striking Example of Alkali Silanides MSiH3 (M= K, Rb, Cs). The Journal of Physical Chemistry C, 2014. 118(7): p. 3409-3419. 53. Weissbuch, I.; Popovitz-Biro, R.; Lahav, M.; Leiserowitz, L., Understanding and control of nucleation, growth, habit, dissolution and structure of two-and three-dimensional crystals usingtailor-made'auxiliaries. Acta Crystallographica Section B: Structural Science 1995, 51 (2), 115-148.

124

Novel Synthesis Method of MgH2 Nanoparticles by Wet Chemistry Approaches

125

5 HYBRID POLYSTYRENE-MGH2 NANOPARTICLES

5.1 INTRODUCTION ...... 127 5.2 SWITCHING THE THERMODYNAMICS OF MGH2 NANOPARTICLES THROUGH POLYSTYRENE STABILISATION AND OXIDATION ...... 128 5.2.1 Experimental Details ...... 129 5.2.2 Results and Discussion ...... 129 5.2.3 Conclusion ...... 139 5.3 FUNCTIONAL GROUPS EFFECTS ON H2 DESORPTION BEHAVIOURS OF MGH2 NANOPARTICLES 140 5.3.1 Experimental Details ...... 144 5.3.2 Results and Discussion ...... 147 5.3.3 Conclusion ...... 158 5.4 POLYMER NANOSTRUCTURES AS SUPPORTS FOR MGH2 NANOPARTICLES ...... 159 5.4.1 Experimental Details ...... 161 5.4.2 Results and Discussion ...... 164 5.4.3 Conclusion ...... 176 5.5 CONCLUSION ...... 177 5.6 REFERENCES ...... 179

Hybrid Polystyrene-MgH2 Nanoparticles

5.1 Introduction

In the previous chapter, we have explored several routes to obtain self-assembly MgH2 nanoparticles. Several methods could even produce stable nanoparticles with uniform morphology without the presence of stabilizers. However, these nanoparticles still require further improvements to meet DOE’s targets as hydrogen storage material especially in terms of the operating conditions. Decreasing the sizes of these nanoparticles further could lead to better sorption properties but this has proven to be quite challenging so far. One approach that we have explored is through the formation of the hybrid nanostructures that are able to tune the physical and chemical properties of Mg. The hybrid system could involve the use of organic molecules such as polymers, whose synthesis techniques have been widely used in other applications.1 2 3 4

As has been discussed in Chapter 2.3.3, one of the early attempts to combine polymer and

MgH2 nanoparticles was achieved by Jeon et al.5 In their work, stabilised nanoparticles with

5 nm diameters were synthesised by using polymers as a stabilising agent. This polymer

(PMMA) also provided protection against oxidations.5 Indeed, oxidations and pyrophoric nature of Mg nanoparticles are the major issues for implementing them in real applications.

However, PMMA is prone to degradation upon exposure to high temperature. Hence, we considered the use of another alternative polymer that has higher thermal stability than

PMMA in this study. We chose polystyrene because it is one of the most stable and the most commercialised polymers in the industry.6 Polystyrene only starts to decompose at 350 °C

7, and thus provides a better temperature window for cycling magnesium compared to

PMMA which decomposes from 200 °C. Furthermore, like PMMA, polystyrene has low oxygen permeability with 2.4 and 3.3 Barrer, respectively 8 whilst offering high permeability towards H2 to allow its sorption. Therefore, we expect similar protection properties observed with hybrid polystyrene-MgH2 nanoparticles as hydrogen storage materials. This chapter focused on three investigations of a polystyrene-MgH2 hybrid system: the effects of

127

Hybrid Polystyrene-MgH2 Nanoparticles

polystyrene against oxidations in the hybrid-MgH2 nanoparticles, the effects of different functional groups attached on the polystyrene, and the uses of nanostructures polystyrene as nano templates of MgH2 nanoparticles.

5.2 Switching the thermodynamics of MgH2 nanoparticles through polystyrene stabilisation and oxidation

Throughout this study, linear polystyrene with low molecular weight and narrow polydispersity was used to protect MgH2 nanoparticles synthesised from di-n- butylmagnesium in cyclohexane. Low molecular weight polystyrene (MW = 5000) is necessary to make sure the polymer is fully soluble in solvent as the medium of the synthetic process. A non-polar solvent such as cyclohexane may not be able to dissolve high molecular weight polystyrene. Another reason to have smaller chains in the polymer is to keep the gravimetric capacity high when the polymer binds onto the nanoparticles’ surface.

Nonetheless, the polystyrene binding is expected to form smaller and stabilised Mg nanoparticles even upon hydrogen cycling and provide further protection against oxidation or moisture. Herein, the effects of polystyrene addition during the synthetic process were investigated including the thermodynamic and kinetic properties upon exposure to air and oxidation. As mentioned in the previous chapter, MgH2 nanoparticles produced by different methods show the indications of the enthalpy and entropy compensation effects. Notably, the material obtained from decomposition of di-n-butylmagnesium in cyclohexane showed a weaker interaction between molecules and having enthalpy reduced to only 53 kJ.mol-1.H2.

However, the increase of the configurational freedom of the system led to an increase of entropy thus creating the enthalpy and entropy compensation effects. This makes a small change in the free energy equation that contributes to hydrogen release properties.9

Therefore, introducing polystyrene into the system may lead to some changes in

128

Hybrid Polystyrene-MgH2 Nanoparticles

thermodynamics in a hope that the entropy and enthalpy reinforce each other resulting in a greater change in the free energy to help in releasing hydrogen.

5.2.1 Experimental Details

Materials

Di-n-butylmagnesium (1.0 M in heptane and up to 1 wt% triethylaluminium as viscosity reducer) and anhydrous cyclohexane were obtained from Sigma-Aldrich. Polystyrene

(PSTN) with a low molecular weight (MW = 5000) to ensure solubility in cyclohexane was obtained and synthesised by A/Prof. Cyrille Boyer.

MgH2/PSTN synthesis

Di-n-butylmagnesium was dried under vacuum for solvent removal. The powder obtained

(2.00 g) was resuspended in cyclohexane (100 mL) with PSTN (0.400 g). The hydrogenolysis of di-n-butylmagnesium was then carried out at 180 °C for 24 h under a H2 pressure of 30 bar. The resulting grey precipitates (MgH2/PSTN) were collected by centrifugation, washed several times with fresh cyclohexane and dried under vacuum at room temperature (0.577 g, 96% yield). This method was repeated without PSTN to leave

MgH2 only as a reference material (0.157 g, 85% yield). The materials were oxidised by exposing the powder obtained to air for 24 h. All further characterisations were then performed under a controlled atmosphere.

5.2.2 Results and Discussion

The morphology of the new composite, MgH2/PSTN, was investigated by TEM and compared to the MgH2 produced from hydrogenolysis of di-n-butylmagnesium only

(MgH2/C). Herein, the MgH2/C produced in slightly different steps compared to the

129

Hybrid Polystyrene-MgH2 Nanoparticles

materials that were discussed in Chapter 4.3. The di-n-butylmagnesium was used in solid form after being fully dried under vacuum and it was completely dissolved in the cyclohexane. The steps still resulted in the formation of nanoparticles with a size ranging from 25 to 50 nm (Figure 5.2.1-A). Furthermore, the addition of PSTN during the synthetic process led to the formation of bigger particles (~100 nm) with a rectangular shape (Figure

5.2.1-B and Table 5.2.1). These nanoparticles correspond to magnesium as confirmed by

EDS analysis (Figure 5.2.1-C).

Figure 5.2.1– TEM images of: (a) MgH2/C and (b) MgH2/PSTN as-synthesized and after hydrogen absorption/desorption cycling at 300 °C (c) MgH2/C including EDX analysis insert and (d) MgH2/PSTN.

The XRD analyses showed the crystalline nature of both materials was β-MgH2 (Figure

5.2.2). XRD in MgH2/PSTN also detected an aluminium phase but with less intensity compared to the MgH2/C. Using the Scherrer formula the crystallite size was determined for both materials and found to be relativity similar ~ 15 nm despite larger particles being observed in TEM for MgH2/PSTN (Table 5.2.1).

130

Hybrid Polystyrene-MgH2 Nanoparticles

Table 5.2.1 - Summary of the physical properties of the synthesized material. Particle morphology/size was determined by TEM and crystallite size from XRD analysis

Crystallite size (nm ± 3) Particle morphology/size Materials (nm) After After cycled synthesis

MgH2/C Round like, 25-50 12 28

MgH2/C oxidised Round like, 25-50 12 30

MgH2/PSTN Rectangular, 50x100 18 32

MgH2/PSTN oxidised Rectangular, ~50x100 16 36

TPD analyses of the MgH2/C revealed the hydrogen release start from 300 °C with a main peak at 343 °C (Figure 5.2.3-A). In comparison, desorption from MgH2/PSTN started at a slightly higher temperature of 330 °C with a main peak at 356 °C. This shift in temperature for MgH2/PSTN may be due to the larger magnesium particles that formed and/or due to the effects of polystyrene coating. Indeed, MS analyses of these materials show fragments corresponded to the decomposition of polystyrene, which evolved starting from 380 °C proving that some polymer remained within the material (Figure 5.2.3-B). The remaining polymer may be bound to the magnesium nanoparticles since its decomposition

131

Hybrid Polystyrene-MgH2 Nanoparticles

temperature is higher than that of pristine polystyrene, i.e. from 320 °C instead of the 380

°C observed for MgH2/PSTN (Figure 5.2.3-B and Figure 5.2.4-A). Magnesium is known to react with a range of organic molecules to form organomagnesium compounds;10 hence the formation of magnesium-polystyrene binds may be possible.

Figure 5.2.3 – Hydrogen desorption profiles as measured by TGA/MS of (a) MgH2/C and (b) MgH2/PSTN as-synthesised, after hydrogen cycling, and after oxidation for 24 h in air.

(A) (B)

Sievert apparatus was used to determine the storage capacity of the materials and hydrogen sorption kinetics. At 300 °C, MgH2/C was found to release 5 wt% of hydrogen in 70 min

132

Hybrid Polystyrene-MgH2 Nanoparticles

(Figure 5.2.5), which is significantly faster than the kinetics of ball-milled magnesium (> 200 min) measured under the same conditions.11 In comparison, MgH2/PSTN has the slower desorption kinetics as full hydrogen release was only achieved in 150 min. This slower kinetic rate is in agreement with the hydrogen desorption profiles measured by TPD measurement (Figure 5.2.3), and may be due to the larger particles size of MgH2/PSTN compared to MgH2/C (Table 5.2.1) and/or different activation energy (Ea). Larger particles would lead to slow kinetics due to longer diffusion distances, which explained the slower kinetics of MgH2/PSTN. However, the determination of Ea by the Kissinger method for both materials revealed higher activation energy for MgH2/PSTN as compared to MgH2/C (Table

5.2.2). Hence, the slower kinetics observed for MgH2/PSTN may predominantly be due to slower recombination rates of hydrogen at the surface of the magnesium particles covered with PSTN. Furthermore, XRD analysis confirmed the conversion of the tetragonal β-MgH2 into the hexagonal Mg phase upon hydrogen release proving that hydrogen was effectively stored upon hydrogen cycling (Figure 5.2.2). Analysis by TEM on the materials after hydrogen cycling also revealed no significant changes in the morphology of MgH2/C and

MgH2/PSTN (Figure 5.2.1-C and D).

Table 5.2.2 - Summary of the enthalpy (∆H), entropy (∆S), and activation energy (Ea) of synthesised materials

Materials ∆H (kJ.mol-1 H2) ∆S (J.mol-1. K-1 H2) Ea (kJ.mol-1 H2)

MgH2/C -51.5 ± 3.1 99.5 ± 4.5 133.0 ± 5.0

MgH2/C exposed -52.4 ± 3.6 101.8 ± 3.1 189.4 ± 4.9

MgH2/PSTN -68.1 ± 3.2 127.2 ± 4.1 171.3 ± 3.9

MgH2/PSTN exposed -52.3 ± 3.2 101.3 ± 4.5 196.2 ± 4.1

Ball-milled MgH2 12 13 -75.2 ± 1.8 139.0 ± 3.0 140-160

Surprisingly, the storage capacity of MgH2/PSTN was found to be higher than that of

MgH2/C, i.e. 5.8 instead of 5 wt %, respectively (Figure 5.2.5). Further investigations by

133

Hybrid Polystyrene-MgH2 Nanoparticles

TGA/MS showed that at 300 °C, only hydrogen was released from both materials with additional decomposition of the polymer for MgH2/PSTN still occurring from 380 °C after hydrogen release (Figure 5.2.3). Hence, the extra storage capacity of MgH2/PSTN was attributed to hydrogen release only. Such behaviour, observed for the first time with magnesium may be due to particle size effects. Indeed, similar effects have been observed with Pd for decreasing particle sizes. Hence, the narrowing of the equilibrium plateau pressure - or miscibility gap corresponding to the coexistence of the α and β phase - observed with decreasing Pd particle sizes, led to a reduction of the storage capacity, e.g. from 0.6-0.7 to 0.3 H/Pd for bulk and 3 nm Pd nanoparticles, respectively.14 This has been explained by a reduction of the nanoparticle volume that will transform into the hydride phase (β-phase) and the associated decrease of both enthalpy and entropy leading to weaker Pd-H bond strengths and also greater freedom of the hydrogen atoms within Pd nanoparticles as compared to bulk Pd.

Figure 5.2.5 – Hydrogen desorption kinetics measured at 300 °C for (a) MgH2/C and

(b) MgH2/PSTN, as synthesised and after oxidation for 24 h in air.

Further determination of the thermodynamic of MgH2/C and MgH2/PSTN by PCI measurements and associated Van’t Hoff plot (Figure 5.2.6 and Figure 5.2.7) revealed a significant decrease of both the enthalpy (∆H) and entropy (∆S) of MgH2/C as compared to

MgH2/PSTN and bulk MgH2 (Table 5.2.2). In comparison to bulk magnesium, this decrease

134

Hybrid Polystyrene-MgH2 Nanoparticles

-1 -1 -1 -1 is of ∆H = 23.7 kJ.mol H2 and ∆S = 39.5 kJ.mol . K H2 for MgH2/C, and ∆H = 7.1 kJ.mol H2 and ∆S = 8.8 kJ.mol-1. K-1 H2 for MgH2/PSTN. These differences were expected since the polystyrene binds would increase the enthalpy formation. A study has reported the linear relationship of a chain-length contribution in the thermodynamics of an inorganic framework.15 The enthalpy becomes more exothermic as longer chains stabilised the nanoparticles. Furthermore, the significant decrease in ∆H for MgH2/C is also compensated by a decrease in ∆S and led to an overall reduction of the desorption temperature Tdes =

∆H/∆S of 24 °C only at 1 bar hydrogen pressure. The MgH2/PSTN has relatively close value to this despite the increase in the ∆H values. Therefore, the enthalpy-entropy compensation effect may have occurred in MgH2/PSTN but not as significantly as in MgH2/C.

(b)

Figure 5.2.6 - P-C-I curves at different isotherms of the MgH2/C (top) and MgH2/PSTN (bottom) before and after exposure to air

135

Hybrid Polystyrene-MgH2 Nanoparticles

Figure 5.2.7 – Van’t Hoff plots of MgH2/C and MgH2/PSTN as synthesized and oxidised. Equilibrium pressures were obtained from the PCI measurements.

Effect of oxidation from air exposure

Figure 5.2.8 – TEM images of: (a) MgH2/C and (b) MgH2/PSTN as-synthesized and oxidised for 24 h in air and associated images after hydrogen cycling at 300 °C (c)

MgH2/C and (d) MgH2/PSTN.

Magnesium nanoparticles are expected to be highly flammable once exposed to air.

However, the materials were stable with no apparent degradation after 24 h. Indeed, TEM analysis did not show any significant evolution of particles morphology (Figure 5.2.8). XRD

136

Hybrid Polystyrene-MgH2 Nanoparticles

characterisation also did not show any formation of an MgO phase with diffraction patterns identical to the as-synthesised materials (Figure 5.2.2).

However, further characterisation by XPS analysis revealed that all materials were oxidised to some extent (Figure 5.2.9). Indeed, the O1s peak showed two different chemical states at

531.7 and 529.6 eV attributed to magnesium hydroxide and oxide, respectively.16 17

Nonetheless, the oxide layer was thin enough to allow the detection of metallic magnesium underneath as shown by the strong Mg2p peak at 49.9 eV alongside a peak at 51.9 eV attributed to magnesium hydroxide and/or oxide (Figure 5.2.9-A).16 17 It is noteworthy that even after 24 h of exposure to air the oxide layer did not seem to significantly grow since metallic magnesium was still detected for both oxidised materials. This is in agreement with previous reports claiming the rapid formation of a passivating oxide layer of 3-4 nm preventing further oxidation.18

Figure 5.2.9 – XPS peak profiles (a) Mg2p and (b) O1s of MgH2/C and MgH2/PSTN as- synthesized and oxidised.

From TGA/MS analysis (Figure 5.2.3), the hydrogen desorption profile of MgH2/C shifted to slightly higher temperatures with a main peak at ~350 °C instead of 343 °C (Figure 5.2.3-

A). However, a bigger shift of 24 °C was observed for MgH2/PSTN and this may be due to the oxide layer retarding hydrogen release. The determination of the storage capacity by

137

Hybrid Polystyrene-MgH2 Nanoparticles

PCT measurements also confirmed the oxidation of both materials. MgH2/C lost half its storage capacity, while MgH2/PSTN retains almost 70% of its original storage capacity

(Figure 5.2.5). Surprisingly, desorption kinetics were also found to be somewhat faster as compared to the non-oxidised materials with 95 % of hydrogen release from MgH2/C achieved in 50 min for the oxidised material instead of 70 min (Figure 5.2.6). Various reports have suggested partial oxidation of magnesium surfaces could effectively catalyse hydrogen sorption.19 20 However, the determination of the activation energy (Ea) by

Kissinger’s method showed a significant increase in Ea for the materials exposed to air

(Table 5.2.2). For example, Ea was found to increase from 133.0 ± 5.0 kJ.mol-1 H2 to 189.4 ±

4.9 kJ.mol-1 H2 for MgH2/C as synthesised and exposed to air, respectively. Assuming that the oxidation of magnesium particles will lead to the growth of a uniform surface oxide layer resulting in a smaller magnesium core, it is more likely that the faster kinetics are due to the smaller size of the active magnesium core within particles.

Figure 5.2.10 – XRD patterns of MgH2/C and MgH2/PSTN as-synthesized and oxidised after hydrogen cycling, PCI measurements and absorption.

The thermodynamic properties of the oxidised materials were also determined from the PCI measurements and the associated Van’t Hoff plots are reported in Figure 5.2.7. For MgH2/C no significant change in enthalpy and entropy was observed. However, both ∆H and ∆S for

MgH2/PSTN oxidised were found to drastically decrease to 52.3 ± 3.2 kJ.mol-1 H2 and 101.3

± 4.5 kJ.mol-1. K-1 H2, respectively. This is equivalent to the values observed for MgH2/C and the evolution of changes as function of particle sizes previously reported.11 Hence, based on this result only a significant decrease in particle size for MgH2/PSTN oxidised should have

138

Hybrid Polystyrene-MgH2 Nanoparticles

occurred during cycling. Indeed, TEM analysis revealed a change in morphology and much smaller magnesium particles sizes ~25-50 nm for MgH2/PSTN oxidised and cycled (Figure

5.2.8-D). Further characterisation by XRD of the oxidised materials after hydrogen cycling finally showed the formation of an MgO phase for MgH2/C only (Figure 5.2.10). For

MgH2/PSTN the oxide layer may remain amorphous or as a finely dispersed phase, since the oxidation of MgH2/PSTN resulted in a significant reconstruction of the magnesium particles upon hydrogen cycling. The driving forces for such a reconstruction remain unclear and are most likely driven by PSTN/ magnesium oxide interface at the surface of the magnesium nanoparticles and possible associated stress upon hydrogen absorption/desorption.

Decrepitation of metal hydrides is a well known phenomenon for microsized powders,21 22 but to the best of our knowledge this is the first time that it is observed for nanosized magnesium particles. For microsized powdered this leads to increased surface areas and thus enhanced hydrogen kinetics, at the nanosize it obviously provides an additional path to modify the thermodynamics properties of the magnesium/hydrogen reaction. As mentioned previously, the stabilisation effects of the polystyrene binds on the magnesium surface may have gone after the oxidation on its surface. This could also be the reason that the ∆H and ∆S values of oxidised MgH2/PSTN are close to MgH2/C values.

5.2.3 Conclusion

The thermal decomposition of di-n-butylmagnesium with polystyrene as a stabiliser led to the formation of relatively large magnesium nanoparticles (~ 100 nm) with a rectangular morphology. These nanoparticles had a higher storage capacity as compared to the nanoparticles (25-50 nm) generated from the hydrogenolysis of di-n-butylmagnesium alone but their thermodynamic properties were closer to that of bulk magnesium, i.e ∆H =

-68.1 ± 3.2 kJ.mol-1 H2 and ∆S = 127.2 ± 4.1 J.mol-1.K-1 H2. The difference in storage capacity correlated to particle size differences and the ability to store more H2 was due to the larger

139

Hybrid Polystyrene-MgH2 Nanoparticles

nanoparticles’ volumes. The enthalpy change is due to weaker hydrogen bond strengths in small nanoparticle sizes while a decrease in entropy results from the more weakly bonded and more mobile hydrogen atoms. Furthermore, polystyrene significantly reduced the detrimental effects of oxidation based from what was observed during the experiment, the polystyrene protected magnesium particles that were exposed to air for more than 24 h only lost 30 % of their initial storage capacity without any negative effects on kinetics. Under the same conditions, the unprotected magnesium particles lost 50% of their initial hydrogen storage capacity. More remarkably, the combined effect of oxidation and polystyrene coating was found to lead to a significant reconstruction of magnesium nanoparticles upon hydrogen cycling and a drastic shift of the thermodynamics with both enthalpy and entropy significantly decreasing to 52.3 ± 3.2 kJ.mol-1 H2 and 101.3 ± 4.5 J.mol-1.K-1 H2 respectively.

The changes should originate from the partial oxidation on the unbound surfaces of the nanoparticles or the degradation of polymer after exposure to high temperature conditions.

Nevertheless, this study shows that the thermodynamic properties of the magnesium/hydrogen reaction is tunable by controlling the oxidation on the surface by applying this polymer during the synthetic process.

5.3 Functional groups’ effects on H2 desorption behaviours of MgH2 nanoparticles

As seen previously, a polystyrene-MgH2 system has resulted in positive effects towards the hydrogen storage properties and overall stability of the materials. However, the interactions between the polystyrene and Mg/MgH2 surfaces are still not clear yet. In

Chapter 5.2, the polystyrene that was used (PSTN) has the functionality of N3 (azide) on each chain end. There are possibilities that different functional groups may react differently with MgH2 formed during the synthetic process and lead to different types of nanoparticles

140

Hybrid Polystyrene-MgH2 Nanoparticles

in the end. Hence, in this study we explored other possible functionalities to be attached to polystyrene and used them in the synthetic process of MgH2 in order to compare their structural and hydrogen desorption properties.

Besides the physical changes, there is a possibility that these functional groups could affect the H2 sorption through the electronic effects. It is widely known that ligands or functional groups of some compounds play an important role for their chemical reactivity especially in catalytic applications such as hydrocarbon oxidations 23, metathesis 24, polymerization 25

26 and redox chemistry.27 28 In particular, a linear relationship correlation between the redox potentials of Ni ions and electronically active ligand substituents has been reported. 29

Increases in the electron-donating power of the substituents on the ligand would lead to increases in electron density that resides on the metal atom. This means that, the more electron-donating substituents could enhance the catalytical activity whereas the electron- withdrawing substituent would have the opposite effect.26 A similar observation was found with the effect of metal ligands on the redox chemistry which was reported by a study of the pthalocyanine and different metal ligands.30 This study showed that the metal in the periphery could strongly influence the solution redox chemistry of these materials. It revealed the substituents with electron-donating groups led to a negative shift of the redox potentials and electron-withdrawing groups to a positive shift.30 Another example of the electronic effects are the induction effects caused by the delocalised electrons which might be able to destabilise the Mg-H bonds to facilitate the decomposition of the MgH2.31 The example of this improvement is commonly seen in the MgH2-transition metals alloys where a high number of d-electrons in transition metals act to destabilise both the H-H and Mg-H bond.32 33 34 The denser of the delocalised electrons caused strong bonding of the transition metals with hydrogen atoms hence weakening the atoms bound to Mg.32

Therefore, in this study we plan to determine possible electronic/inductive effects by substituent groups of organic compounds and herein we used polystyrenes with different

141

Hybrid Polystyrene-MgH2 Nanoparticles

functional groups attached on each chain end. A specific functionalised polystyrene was synthesised by a Reversible addition−fragmentation chain-transfer (RAFT) polymerisation which is able to afford polymer with a uniform molecular weight and narrow polydispersity.

These properties are important to optimise the investigations of each substituent and prevent the anomalies caused by steric interactions upon the MgH2 particles besides the electronic effects. Furthermore, this polymerisation technique can be extended for post modifications to introduce different functional groups; one of the most established techniques is the thiol-ene ‘click’ reaction.35 This technique is established in protective coatings or film industries and has been relevant to other applications such as biotechnology and electronics.36 37 38 Scheme 5-1 describes the overall functionalisation mechanism using the combination of RAFT polymerization of polystyrene (PST) and the thiol-ene chemistry.

Scheme 5-1 (1) RAFT polymerization of polystyrene and (2) post functionalisation (thiol-ene click reaction) with different acrylates compounds bearing different functional groups

142

Hybrid Polystyrene-MgH2 Nanoparticles

As mentioned earlier, the functional group effects could be sourced from the electronegativity differences between the substituents and atoms to which they are attached. However, it has been reported that this effect would be diminished along a chain of atoms.39 For larger chains, there is virtually no change in the charge, which indicates that transmission of electronegativity effects through a chain of atoms is almost completely damped out after the first or second atom. Hence, the changes in atom charge density at more remote atoms are mostly due to the substituent dipole to the reaction or measurement site. 40 Quantitative analyses of these effects are possible and the quantitative values usually depicted as Field Effect (F) values. These values depend on the overall chemical structures of the compounds, reaction conditions and with different reagents. Therefore, values for specific compounds differ and are not possible to be obtained from other sources unless exactly the same synthetic processes are applied. Instead, a qualitative comparison was made based on F values of specific functional groups that are available in literature. In this study, the F values used were obtained based on the calculations made from the Hammett and Taft constant of different functional groups. The Taft constant is the extended version of the Hammett constant which has been frequently used to predict equilibrium and rate constant for the organic reactions based on the reactivity of functional groups.39 The constants were generalised into two types of effects: induction and resonance effects.39 Taft proposed the following equation to quantify the field/ induction effects alone.

F = σI = 1.297 (±0.147)σm – 0.385 (±0.089) σp + 0.033 (±0.026) (5-1)

Where σm and σp are the Hammett constants obtained from the experiments considering the para and meta position of the substituent functional group. Combinations of these two constants were made to accommodate the steric effects on the reactivity. Furthermore, the method to calculate the resonance component effects was obtained by the following equation:

R = σR = σp - σI (5-2)

143

Hybrid Polystyrene-MgH2 Nanoparticles

The calculated values of these effects for the functional groups in this study were summarised in Table 5.3.1. However, only values for -NMe3+, -NMe2, and -OH were presented on the table since PS-COOH have different overall chemical structures that could not be comparable with the other three polystyrene (Scheme 5-1). Among these three, the order of the functional groups that would give the most electronic effects were -NMe3+ > -

NMe2 > -OH. It is noteworthy that with the same chemical structures applied the –COOH would place in between -NMe3+ and -NMe2.41 However, PS-COOH was modified with acrylic acid thus resulting in less dense structures compared to the other three PS-functional groups. In general, the more bulky structures or functional groups would lead to a lower F, thus PS-COOH was assumed to have the highest F and the most prominent electronic effects among the polystyrene that we used.39 Hence the order of the field effects strength would be -COOH>-NMe3+ > -NMe2 > -OH.

σm σp F R

+ 41 NMe3 0.88 0.82 0.86 -0.04

41 42 NMe2 -0.16 -0.83 0.15 -0.98

OH 41 0.12 -0.37 0.33 -0.70

5.3.1 Experimental Details

Polymer synthesis

144

Hybrid Polystyrene-MgH2 Nanoparticles

Materials

All reagents and solvents were supplied by Aldrich and used with no further purification unless otherwise stated. The initiator, 2,2-azobisisobutyronitrile (AIBN) was recrystallized twice from methanol prior to use. High purity nitrogen (Linde gases, 99.99%) was used for purging the reaction solutions before polymerization. Styrene (St, 99%), was deinhibited by percolating over a column of basic alumina (Ajax Chemical), hexylamine (99%), triethylamine (TEA, 99%), dimethylaminoethyl acrylate (DMEA, 99%), hydroxyethyl acrylate (HEA, (%), acrylic acid (99%), methyl bromide (99%). Toluene (Ajax Chemical), methanol (Ajax Chemical) and Dimethylformamide (DMF, Ajax Chemical) were used as received. RAFT agent 3-(benzylsulfanylthiocarbonylsulfanyl)-propionic acid (BSPA) was synthesized following the procedure in the literature.43

Methods

Synthesis of RAFT-Polystyrene (PS-RAFT)

Styrene (10.4 g, 0.100 mol, 1), AIBN (17 mg, 1.07 × 10-4 mol), BSPA (0.147 g, 5.40 × 10-4 mol,

2), and toluene (50 ml) were placed into a 100 ml round bottom flask, equipped with a magnetic stirrer bar. The reaction mixture was cooled on an ice bath, and degassed by purging with nitrogen for 20 minutes. The degassed solution was stirred at 70 °C for 16 hours. The reaction was sampled for GPC and 1H NMR analysis at this point. The remaining toluene was removed by rotary evaporation. The polymer (PS-RAFT) was precipitated several times in cold methanol yielding 8.0 g of polymer with Mn = 5 000 g/mol and PDI =

1.12 (by THF-GPC analysis). The polymer was then placed in an oven at 60 °C for 48 h to remove any trace of solvent. The polymer was characterized by 1H NMR and THF-GPC. 1H-

NMR (300 MHz, CDCl3) d (ppm from TMS); 1.5-1.6 (1H, s, CHCH2-Ph), 1.8-2.0 (2H, s, CHCH2-

Ph), 6.5-7.3 (5H, m, CHCH2-Ph).

145

Hybrid Polystyrene-MgH2 Nanoparticles

Aminolysis of the PST-RAFT and subsequent post-functionalization by thiol- ene reaction

An example of the method typically 2 g of PS-RAFT (4 × 10-4 mol) were dissolved in DMF

(20 ml). The solution was purged with nitrogen for 30 min. A solution (2 ml) of hexylamine

(2 × 10-3 mol/ml) and (4 × 10-3 mol/ml) in DMF was added under nitrogen. The solution became transparent after 4 h. Then a solution (5 ml) of acrylate compounds (DMEA or HEA or acrylic acid, 4 × 10-3 mol/ml) was added to the mixture also under nitrogen. After 12 h of stirring, the product was purified by precipitation in methanol and washed 3 times.

Further functionalization reaction was applied to dimethylamine functionalised polymer

(PS-NMe2) to obtain ammonium functionalized polymer. 2 g of PS-NMe2 (4.0 × 10-4 mol) were dissolved in THF (10 ml) and iodomethane (0.110 g, 8.0 × 10-4 mol) were added to the solution. The mixture was kept stirred for 24 h and it was purified again in cold methanol and washed 3 times. The functionalised PS was characterised with 1H NMR. 1H-NMR (300

MHz, CDCl3) d (ppm from TMS); 1.5-1.6 (1H, s, CHCH2-Ph), 1.8-2.0 (2H, s, CHCH2-Ph), 6.5-

7.3 (5H, m, CHCH2-Ph). Carboxylic Acid Polystyrene (PS-COOH), 2.6 (2H, t, CH2C(O)OH),

Hydroxyl Polystyrene (PS-OH), 3.5 (2H, t, CH2CH2OH), 4.2 (2H, t, CH2CH2OH), 3.7 (1H,

CH2CH2OH), Dimethylamine polystyrene (PS-NMe2), 2.8 (3H, CH2CH2N(CH3)2), 3.3 (2H,

CH2CH2N(CH3)2), 4.3 (2H, CH2CH2N(CH3)2), Trimethylammonium polystyrene (PS-NMe3+),

3.3 (3H, CH2CH2N+(CH3)3), 3.5 (2H, CH2CH2N(CH3)2), 4.3 (2H, CH2CH2N(CH3)2).

MgH2@PS synthesis

Materials

Di-n-butylmagnesium (1.0 M in heptane and up to 1 wt% triethylaluminium as viscosity reducer) and anhydrous cyclohexane (99%) were obtained from Sigma-Aldrich. High purity hydrogen gas (99.999%) was used throughout the experiments. All material handling, weighing, loading and washing was performed in a glove box filled with high purity argon

(O2 and H2O < 1 ppm) from LC Technology.

146

Hybrid Polystyrene-MgH2 Nanoparticles

Methods

Di-n-butylmagnesium solution in heptane (7.6 ml, 1x10-3 mol/ml) was added to cyclohexane (100 ml). Functionalized PS (PS-R, 20 mg, 4 x 10-6 mol/ml) was added to the mixture and it was transferred to a pressure reactor vessel. It was then hydrogenated at 180

°C for 24 h under a hydrogen pressure of 3 MPa. The resulting grey precipitates (MgH2@PS-

R) were collected by centrifugation, washed several times with fresh cyclohexane and dried under vacuum at room temperature (~0.192 g, 88% yield). All further characterisations were performed under a controlled atmosphere.

5.3.2 Results and Discussion

The previous hydrogenolysis attempt (Chapter 5.2) of di-n-butylmagnesium and polystyrene mixed in cyclohexane resulted in the formation of MgH2 with bigger particle sizes compared to the hydrogenolysed di-n-butylmagnesium alone. In this study, all the resulting materials underwent only slight increases in particle sizes from ~20 nm (Figure

4.3.2) to ~30-40 nm (Figure 5.3.1).

Figure 5.3.1 – TEM images and particle sizes distribution of: (A) MgH2@PS-NMe3+ , (B)

MgH2@PS-NMe2, (C) MgH2@PS-OH, (D) MgH2@PS-COOH

147

Hybrid Polystyrene-MgH2 Nanoparticles

Indeed, less concentration of polystyrene was used to produce these materials compared to those in Chapter 5.2 where the average size reached to ~100 nm. Nevertheless, all of the materials produced from different functional groups polystyrene have consistent size and morphologies and thus minimised the parameters influencing their hydrogen storage properties.

However, their crystallite sizes as calculated by the Scherrer equation (Table 5.3.2) are different for each material. These crystallite sizes were obtained from the crystalline β-

MgH2 phase peaks from their XRD analyses shown in Figure 5.3.2. Surprisingly, the crystallite sizes increased as the F values increased. This is possibly due to the electron withdrawing ability of the groups that elongated the Mg-H bonds leading to formation of larger crystallite sizes.44 45 The exception was the MgH2@PS-COOH, which had the largest calculated crystallite size. Indeed, in terms of the long chain or bulk density of these polystyrene, PS-COOH has less bulkier structures than the other polystyrene hence explaining the larger crystallite size formation compared to the other materials and non- conformance to the trend.

Figure 5.3.2 XRD patterns of MgH2 as-synthesized from hydrogenolysis of MgBu2 in cyclohexane with different functional groups of polystyrene.

148

Hybrid Polystyrene-MgH2 Nanoparticles

Table 5.3.2 - Summary of the crystallite size and particle size of synthesised materials

MgH @PS- MgH @PS- MgH @PS- MgH @PS- 2 2 2 2 NMe2 NMe3+ OH COOH Crystallite Size 24 ± 1 14 ± 1 20 ± 2 27 ± 3 (nm) Particle Size 35-64 30-41 31-48 26-61 (nm)

Mg2p C1s O1s

MgO/ C-C C-O MgO MgOH Mg Mg(OH) (286 (529.7 (532 (51 eV) x (284.9 (52 eV) eV) eV) eV) eV) After 24.3 20.29 6.17 0.71 7.07 31.82 Free PS synthesis MgH2 Cycled 45.91 - 3.02 0.35 15.86 28.24 After MgH2@ 41.89 - 11.07 1.18 14.72 24.91 PS- synthesis NMe2 Cycled 36.99 - 4.26 0.45 3.83 50.37 After 40.36 - 6.42 1.48 10.35 35.08 MgH2@ synthesis PS-OH Cycled 44.06 - 6.82 0.26 26.82 16.31 After MgH2@ 38.81 - 14.1 0.81 9.96 30.05 PS- synthesis COOH Cycled 40.45 - 8.7 0.51 15.44 27.09 After MgH2@ 42.27 - 6.97 0.81 12.65 28.69 PS- synthesis NMe3+ Cycled 49.61 - 3.42 1.45 25.81 15.45

XPS analyses were performed in order to further investigate the possible polymer interactions with the MgH2 surface. Table 5.3.3 is the summary of the XPS results. Compared to the free MgH2 (6.17%), there are more C-C observed based on the bond energy from the

MgH2 with polystyrene. However the differences are not huge especially for MgH2@PS-

149

Hybrid Polystyrene-MgH2 Nanoparticles

+ NMe3 (6.97%) and MgH2@PS-OH (6.42%). It is noteworthy that previous results in Chapter

5.2 showed the material with polystyrene possessed higher hydrogen capacity. Our hypothesis is that it was due to the formation of a smaller of amount hydrocarbon contaminant within the polymer-hybrid materials. It has proven to be difficult to distinguish the bond energy corresponding to either polymers or interstitial carbons. Different bond energies should be observed in Mg 2p when different functional groups are attached on the

Mg surface. However, such bond energies are not shown in the XPS results so the interactions between substituents might not have been detected due to their small amounts.

Hydrogen storage properties

TPD analyses were performed on the freshly synthesised materials. Figure 5.3.3 shows the

TGA/DSC analyses of these polystyrene/MgH2 nanocomposites. Losses of mass occurred at temperatures below 300 °C due to evaporation of some solvents and other hydrocarbons.

Moreover, typical hydrogen releases from the MgH2 occurred at temperatures higher than

300 °C and they were accompanied with the endothermic peaks from DSC signals. All of these materials revealed multiple or broad endothermic peaks from DSC signals. MgH2@PS-

COOH showed a distinct endothermic peak at around 331 °C and a possible extra peak with weak signal at higher temperature (380-400 °C). This small endothermic peak is accompanying the 1%wt loss observed in TGA. In general, the other materials have stronger endothermic peaks at higher temperatures, especially with MgH2@PS-NMe2 where the endothermic signal at lower temperature is much weaker and is accompanied with 1 wt% loss in TGA. This phenomenon could be associated with the degradation of the polystyrenes and their bindings on the MgH2 surface. As shown in Figure 5.3.4, the thermal decomposition of these polystyrenes could undergo partial decomposition from 200 °C until around 350 °C when all of them undergo rapid major mass loss indicating complete decomposition. However, this thermal stability should be improved when they covered or are bound to the MgH2 which should shift the rapid decomposition to more than 400 °C as

150

Hybrid Polystyrene-MgH2 Nanoparticles

proven in Chapter 5.2 and the work of Jeon et al with their Mg-PMMA nanocomposite.5 This means that the H2 desorption at the temperature range of 380-400 °C observed in the TPD analyses of the materials should be from the decomposition of MgH2 covered by polystyrene on their surface.

Figure 5.3.3 – TGA/DSC of MgH2@PS-NMe3+, MgH2@PS-NMe2, MgH2@PS-OH, and

MgH2@PS-COOH as synthesized

The amount of polystyrene covering the surfaces was determined by the mass losses occurring at this temperature region. Interestingly, we can also correlate this variant in mass losses with the electronic effects of the substituents that applied on the polystyrene used to synthesise these materials. For example, -NMe2 is the most electron acceptor substituent (lowest F) and the related resulting material underwent almost 4wt% loss in the higher temperature region in contrast to PS-COOH which we depicted to be the most electron withdrawing group which only underwent only 1wt% loss at the same temperature region.

151

Hybrid Polystyrene-MgH2 Nanoparticles

Figure 5.3.4 TGA shows the different functional group polystyrenes decomposition

The cyclability of the resulting materials were further investigated in a series of characterisation techniques. TEM analyses of the materials after hydrogen cycling (Figure

5.3.5) showed the same morphologies and particles sizes as the synthesised materials. This confirmed the previous finding where these nanoparticles are stable after being cycled. Our hypothesis was that this stability is due to the effects of the carbon materials resulting from the solvents reaction with Mg surfaces and decomposition products of the organic component of the precursor. However, capping of the nanoparticles with polystyrene may contribute to the stability as we have predicted.

Figure 5.3.5 – TEM of the materials after the 3rd hydrogen absorption.

152

Hybrid Polystyrene-MgH2 Nanoparticles

This was proven by the XPS analyses of the cycled materials (Table 5.3.3) where there was less carbon on these materials compared to the free PS MgH2. In particular, cycled

MgH2@PS-OH has even higher carbon content than the synthesised material indicating the stable polymer still covering the material’s surface. As we observed earlier, PS-OH was the only polystyrene that did not undergo partial decomposition until around 350 °C.

Meanwhile the MgH2@PS-NMe3+ showed the lowest amount of carbon after cycling (3.42%) since the PS-NMe3+ indeed is the least stable polystyrene as observed in Figure 5.3.4. The

TPD analyses of the cycled materials confirmed the thermal degradations of these polystyrene after several cycles (Figure 5.3.6).

Figure 5.3.6 - TGA/ DSC after the 3rd hydrogen absorption of the materials

Except for MgH2@PS-OH, other materials did not have mass losses at the higher temperature range (380-400 °C). Even in MgH2@PS-OH, the overall mass loss only accounted for 0.5 wt% which is significantly lower compared to 6 wt% mass loss observed on the freshly synthesised material. Interestingly, in the end the materials have relatively similar hydrogen desorption peaks (335-342 °C) and H2 capacity (4.7-5.1 wt%)

153

Hybrid Polystyrene-MgH2 Nanoparticles

corresponds to the same values observed on free PS material (Figure 4.3.10). Only

MgH2@PS-OH have a slightly increased desorption peak (353 °C) and a decrease in H2 capacity to only 2.7 wt%. Obviously this is due to the presence of polystyrene within the

MgH2.

The H2 evolution from the MS spectra of these cycled materials (Figure 5.3.7) proved what has been discussed so far. Only MgH2@PS-OH released H2 with substantial amounts at higher temperature. MgH2@PS-NMe2 and MgH2@PS-NMe3+ released few amount of hydrogen at higher temperature that could also be due to the slight presence of polystyrene on their surface.

Figure 5.3.7 – MS of the gases evolved from the materials including PS free MgH246

In terms of the kinetics for hydrogen release, Figure 5.3.8 shows the H2 desorption kinetics of all materials at 300 °C after 3 cycles. There is a slight variation across the materials but overall they have faster kinetics compared to the materials in Chapter 5.2 which have more polystyrene concentration and bigger particle sizes. However, this small variation does not correlate with the electronic effects trend we discussed earlier. At the end, MgH2@PS-COOH can be considered to have the fastest kinetics among all materials possibly because it has the least polymer covering its surface compared to others. MgH2@PS-COOH also has the

154

Hybrid Polystyrene-MgH2 Nanoparticles

lowest Ea 133-142 (kJ/mol). This value too is similar with the free PS MgH2 which differed from other materials which have more or less similar Ea values (Table 5.3.4).

Table 5.3.4 - Summary of the enthalpy (∆H), entropy (∆S), activation energy (Ea), crystallite size and particle size of synthesized materials

∆S ∆H (kJ/mol) Ea (kJ/mol) (J/mol.K)

MgH2@PS-NMe2 74 ± 5.9 135 ± 10 161-170

MgH2@PS-NMe3+ 63.8 ± 6.2 120 ± 11 158-165

MgH2@PS-OH 70.8 ± 4.7 132 ± 8 155-172

MgH2@PS-COOH 57.9 ± 4.9 106 ± 9 133-142

Figure 5.3.8 – Hydrogen desorption kinetics of the materials. The 3rd hydrogen desorption cycle is shown.

Furthermore, the thermodynamic aspect of these materials, ∆H and ∆S values (Table 5.3.3) were obtained from both Van Hoff’t plots and the PCI measurements at isothermal temperatures between 280 to 325 °C. Figure 5.3.9 showed the absorption PCI measurements to obtain the plateau pressures used in the Van Hoff’t plot. There were fluctuations in the plateau pressure during hydrogen absorption of these materials.

MgH2@PS-COOH and MgH2@PS-NMe2 had the lowest average equilibrium pressures at the same temperature isotherms. However, differences in the slope of the Van’t Hoff plots of both materials led to variation in the ∆H and ∆S values (Figure 5.3.10).

155

Hybrid Polystyrene-MgH2 Nanoparticles

We have reported earlier in Chapter 5.2 that the polystyrene-MgH2 system shifted the thermodynamic properties and became closer to that of bulk magnesium (68.1 ± 3.2 kJ.mol-

1 H2). Interestingly, this time the thermodynamic properties changes correlated to the electron affinity tendencies of the functional groups with these thermodynamic values. Both

∆H and ∆S decreases as the electron affinity becomes more positive which tends to be more electron withdrawing (Figure 5.3.11). MgH2@PS-NMe2 has the ∆H closer to that of bulk magnesium, i.e. 74 ± 5.9 kJ.mol-1 H2 while MgH2@PS-COOH is closer to the free PS MgH2, i.e.

57.9 ± 4.9 kJ.mol-1 H2.

Figure 5.3.9 P-C-I absorption measurements of MgH2 synthesized with different functional groups polystyrene at various temperatures

However, we could not conclude that the functional groups in the polystyrene are the cause of these changes. In fact, what we have observed so far is that polymers would inhibit the hydrogen diffusion and there is a possibility of the clamping effects elastic constraints occurred. This constraint imposed by the material capping the magnesium which would

156

Hybrid Polystyrene-MgH2 Nanoparticles

induce a repulsive H-H interaction instead of the usual attractive interaction leading to high desorption temperatures.47 48 49 This is why, MgH2@PS-COOH, which is considered to have the least amount of polymer, would behave similarly to those without any. Meanwhile, both

MgH2@PS-OH and MgH2@PS-NMe3+ are the only materials that have more polymer on the surfaces thus would have higher equilibrium pressure and higher temperature requirement for hydrogen release.

Figure 5.3.10 Van’t Hoff plots of MgH2 from MgBu2 in cyclohexane and with different functional group polystyrenes

Figure 5.3.11 – Thermodynamics plot, enthalpy vs entropy values of the materials

157

Hybrid Polystyrene-MgH2 Nanoparticles

5.3.3 Conclusion

Different functional groups of polystyrene were successfully synthesised by RAFT polymerisation and subsequently with thiol-ene click reactions for the post functionalisation. The substituents that were investigated in this study are -NMe3+, -OH, -

NMe2 which are in the order towards the tendency of electron donating respectively and another functional group, –COOH that could not be compared with other substituents since its functionalisation method led to less bulky chemical structures. However, it was assumed that –COOH polystyrene would have more prominent electronic effects towards electron withdrawing with such chemical structure. Nonetheless, each of these functional groups of polystyrene was mixed with di-n-butylmagnesium in cyclohexane and was hydrogenolysed.

It resulted in a slight increase of particle sizes from 20-40 nm and minimum differences in morphology were observed. TPD analyses of the materials showed the amount of polymer covering the surface of MgH2 particles increased when the polystyrene has more electron donating groups (-NMe2 has the most). However, this polymer on the surfaces led to hydrogen release at higher temperature due to inhibition of a hydrogen diffusion path on the surface by the polymers and possible clamping effects. In fact, MgH2@PS-COOH has the lowest amount of polymer covering the surface and it shows the fastest H2 desorption properties. However, a correlation between the electronic effect of the substituent and thermodynamics of the resulted materials has been observed. The enthalpy and entropy values are decreased when more electron withdrawing functional groups of polystyrene were applied and vice versa. Overall, these results show the tendencies of the MgH2 destabilisation process could be achieved simply by introducing functional groups or substituents. This is an important finding and opens a new direction to apply in order to improve the hydrogen storage properties of Mg/MgH2 based nanomaterials.

158

Hybrid Polystyrene-MgH2 Nanoparticles

5.4 Polymer nanostructures as supports for MgH2 nanoparticles

As shown earlier, introducing linear polystyrene during the hydrogenolysis of di-n- butylmagnesium did not result in a significant decrease in the particle size. In fact, it has led to the formation of larger size nanoparticles and the hydrogen sorption properties became worse. Therefore, a new strategy was sought to obtain smaller MgH2 nanoparticle that would lead to optimum improvements for hydrogen storage properties. One of the most available and well-studied strategies to synthesise inorganic nanoparticles is by a nanotemplate/ nanoreactor approach.50 51 This approach is similar to the nanoconfinement approach where the nanoparticles are stabilised within the porous carbon.14 52 53 In this study, instead of porous carbon materials, a well-defined polymer nanostructure such as dendrimers and hyperbranched polymers were used. These polymers can have numerous interior cavities and functional groups which can be efficient templates for nanoparticles.54

55 56 57 Moreover, in addition to the template role, a polymer such as hypercrosslinked polystyrene has the ability to store hydrogen through physisorption due to its high surface areas.58 59 As discussed in Chapter 2.2.1, improvements on the hydrogen adsorption through physisorption could be achieved by transition metal decorations on the structures of porous materials. This can be applied to the hypercrosslinked polymers where several studies have shown the metal nanoparticle spillovers indeed led to improvement in the hydrogen absorption.60 61 However, incorporating MgH2 on the structures of such hypercrosslinked polymers is still novel and the effect towards the hydrogen storage through physisorption also still under investigation. Several studies indicate the possibilities to cluster more H2 molecules through electrostatic charge caused by the alkali metal ions such as Li+ dispersed on the carbon nanostructures.62 63 For example, a study showed an increase of hydrogen uptake from 1.6 wt% to 6.1 wt% for the porous carbon decorated with 0.5 wt% Li and hydrogenated with 1 bar H2 pressure at -200°C.

159

Hybrid Polystyrene-MgH2 Nanoparticles

Furthermore, the enthalpies of this material underwent increases in adsorption enthalpies from 7.7 kJ.mol-1.H2 to 8.1 kJ. mol-1.H2.62 Therefore, the combination of MgH2 nanoparticles with high surface area polystyrene has the potential to improve their hydrogen storage ability through both physisorption and chemisorption.

Scheme 5-2 Strategies of the nano templates/reactors strategies to synthesize MgH2 nanoparticles

In this work, Star poly(St-DVB) or S(PS) and Hypercrosslinked Polystyrene (HPS) were synthesised and then utilised as templates to grow MgH2 nanoparticles from either hydrogenolysis of di-n-butylmagnesium in cyclohexane or through catalytical hydrogenation of MgA.(THF)3. Some modifications were performed on these nanostructures to promote the possibility of Mg growths on the templates. These include the inclusion of ammonium functional group in the polymer chain of the star polymer and nickel nanoparticles decorations within the HPS. The overall strategies are represented on

Scheme 5-2.

160

Hybrid Polystyrene-MgH2 Nanoparticles

5.4.1 Experimental Details

Polymer synthesis

Materials

All reagents and solvents were supplied by Aldrich and used with no further purification unless stated otherwise. The initiator, 2,2-azobisisobutyronitrile (AIBN) was recrystallized twice from methanol prior to use. High purity nitrogen (Linde gases, 99.99%) was used for purging the reaction solutions before polymerization. Styrene (St, 99%), vinyl benzyl chloride (VBC, 99%), divinylbenzene (DVB, 80%, containing p-tert-butylcatechol as inhibitor) were deinhibited by percolating over a column of basic alumina (Ajax Chemical).

Nickel(II) acetylacetonate (Ni(acac)2, 95%), triethylamine (TEA, 99%), toluene (99%), dimethyl formamide (DMF, 99%), anisole (99%), Diethylene glycol dimethyl ether (diglyme,

99.5%) and methanol (99%) were used as received. 4-cyanopentanoic acid dithiobenzoate

(CPADB, RAFT Agent) were synthesised by another researcher following other literature. 64

High purity N2 (Linde gases) was used for reaction solution purging.

Synthesis of S(PS)

Styrene (10.4 g, 0.100 mol), AIBN (45 mg, 2.83 × 10-4 mol), CPADB (0.237 g, 8.49 × 10-4 mol), and toluene (5 ml) were placed into a sealed tube, equipped with a magnetic stirrer bar. The reaction mixture was cooled on an ice bath, and degassed by purging with nitrogen for 20 minutes. The degassed solution was stirred at 75 °C for 16 hours. It was quenched and precipitated in methanol. It was washed several times with methanol to remove unreacted monomer yielding 9.0 g of solid precipitates with Mn = 6000 g/mol and PDI = 1.11 (by THF-

GPC analysis). Then it was dried in a vacuum oven at 60 °C for 48 h to remove any trace of solvent. The solid product (RAFT-PS) was characterised by 1H NMR and THF-GPC. 1H-NMR

(300 MHz, CDCl3) d (ppm from TMS); 1.5-1.6 (1H, s, CHCH2-Ph), 1.8-2.0 (2H, s, CHCH2-Ph),

6.5-7.3 (5H, m, CHCH2-Ph).

161

Hybrid Polystyrene-MgH2 Nanoparticles

RAFT-PS (2 g, 3.33 x 10-4 mol) was introduced into a vial equipped with a magnetic stirrer and dissolved in 5 mL DMF. DVB (0.38 ml, 2.67 × 10-3 mol) and AIBN (27 mg, 1.67 × 10-4 mol) were added and mixed. The vial was sealed and purged with nitrogen for 20 min at 0

°C. The reaction vial was then placed in an oil bath at 70 °C for 24 h. The reactant was purified by precipitation against methanol and the solvent was removed by evaporation under vacuum. It yielded 1.8 g of pink solid powders with Mn = 87000 g/mol and PDI of 1.39

(by THF-GPC analysis).

Synthesis of S(PS-PVBA)

RAFT-PS (1.25 g, 2.08 × 10-4 mol ) mixed with VBC (1.83 g, 1.20 × 10-4 mol) and AIBN (20 mg, 1.25× 10-5 ) in 5 ml anisole. The same procedure to synthesise RAFT-PS was repeated to obtain 2.50 g of Poly(St-VBC). It was characterized by 1H NMR. and GPC analyses. After purification and drying under vacuum, pink powders were obtained (RAFT-PS-PVBC) with

Mn = ~6000 and ~43-44% was PVBC.

RAFT-PS-PVBC (1.8 g, 2.00 × 10-4 mol), DVB (0.026 g, 2.00 × 10-4 mol), AIBN (25 mg, 1.58 ×

10-5) in 5 ml DMF were placed into a sealed tube, equipped with a magnetic stirrer bar. The reaction mixture was cooled on an ice bath, and degassed by purging with nitrogen for 20 minutes. The degassed solution was stirred at 70 °C for 16 hours. It was quenched and precipitated in methanol. It was washed several times with methanol to remove unreacted monomer. Pink solids were obtained (1.75 g) and it was characterized with 1H NMR GPC analysis (Mn = 87000). 1H-NMR (300 MHz, CDCl3) d (ppm from TMS); 1.5-1.6 (1H, s, CHCH2-

Ph), 1.8-2.0 (2H, s, CHCH2-Ph), 4.5-4.6 (-CH2-Cl)-6.5-7.3 (5H, m, CHCH2-Ph).

Star Poly(St-DVB-VBC) (1 g, 1.15 × 10-5 mol), and TEA (0.93 g, 5.74 × 10-3 ) were mixed in 5 ml DMF. It was stirred at 50 °C for 16 hours and purified to the same precipitation process and washings as before to obtain S(PS-PVBA) and characterised by 1H NMR. 1H-NMR (300

MHz, CDCl3) d (ppm from TMS); 1.0-1.3 (9H, t, -N(CH2CH3)3), 1.5-1.6 (1H, s, CHCH2-Ph), 1.8-

162

Hybrid Polystyrene-MgH2 Nanoparticles

2.0 (2H, s, CHCH2-Ph), 2.7-3.3 (6H, t, -N(CH2CH3)3 ), 3.5 (2H, s, Ph-CH2-N N(CH2CH3), 6.5-7.3

(5H, m, CHCH2-Ph).

Synthesis of HPS

Styrene (3.20 g, 3.0 × 10-2 mol), AIBN (16 mg, 9.60 × 10-5 mol), CPADB (0.054 g, 1.92 × 10-4 mol), DVB (0.125 g, 9.60 × 10-4 mol) and DMF (5 ml) were placed into a sealed tube, equipped with a magnetic stirrer bar. The reaction mixture was cooled on an ice bath, and degassed by purging with nitrogen for 20 minutes. The degassed solution was stirred at 80

°C for 36 hours. The solution started to become a pink gel over the time. After the reaction had finished, the gel was dissolved in DMF and precipitated in methanol. It was washed several times with methanol to remove unreacted monomer. It was characterized with

DMF-GPC and yielded 3.00 g of HPS with Mn = 30000 with PDI of 1.64.

Synthesis of HPS-Ni

HPS (1.50 g, 5.00 x 10-5 mol) was dissolved in 10 ml DMF and then Ni(acac)2 (0.040 g , 1.56 x 10-4 mol) was added to the solution. 0.2 ml NaBH4 solution (0.08 M in diglyme) dropwisely added into the solution while continuously mixed through stirring. It turned black slowly.

Then, the solution mixture was concentrated and quenched with methanol to afford precipitation of the polystyrene and washed further with methanol and acetone several times. This process yielded 1.0 g of white grey powder after being dried under vacuum and further characterised by TEM for investigating Ni nanoparticles deposition.

MgH2@PS synthesis

Materials

163

Hybrid Polystyrene-MgH2 Nanoparticles

weighing, loading and washing was performed in a glove box filled with high purity argon

(O2 and H2O < 1 ppm) from LC Technology.

Methods

Di-n-butylmagnesium solution in heptane (7.6 ml, 1x10-3 mol/ml) was added to cyclohexane (100 ml). HPS/SPS (0.100 mg, 1.42 – 3.33 x 10-6 mol) were added to the mixture and it was transferred to a pressure reactor vessel. It was then hydrogenated at 180 °C for

24 h under a hydrogen pressure of 30 bar. The resulting grey precipitates (MgH2@PS) were collected by centrifugation, washed several times with fresh cyclohexane and dried under vacuum at room temperature.

2 g of solid MgA (4.78 x10-3 mol) were suspended in 100 mL THF and mixed with of CrCl3

(0.007 g, 4.41x10-5) and HPS/SPS (0.100 g, 1.42 – 3.33 x 10-6 mol). The mixture was transferred to the pressure reactor vessel and hydrogenated with 3 MPa hydrogen pressure at 60 °C. The reaction was kept for 15 h and the products were collected as grey greenish precipitates after several washes with THF.

5.4.2 Results and Discussion

The uses of the RAFT polymerisation method to synthesise star and hyperbranched structures of polystyrene are well known and have been extensively studied.65 66 One of the approaches to form these structures is the ‘arm’ first approach where homopolystyrene was synthesised first before being post functionalised.67 The formation of these well structured polystyrene was characterised with NMR and GPC analyses (Figure 5.4.1). The NMR spectrum confirmed the pure form of RAFT polystyrene. This linear polystyrene formed star structures by further copolymerisation of high concentration DVB to form a highly crosslinked reaction thus creating a DVB core (Scheme 5-3).

164

Hybrid Polystyrene-MgH2 Nanoparticles

Scheme 5-3 Synthetic strategies of polystyrene based star and hyperbranched structures

Figure 5.4.1 (left) NMR polystyrene of RAFT polystyrene, (right) GPC results of linear polystyrene, SPS, and HPS

DVB has a strong affinity for styrene thus diffusing to the centre of the micelles where it reacts with living anions on the polystyrene chain ends to polymerise a poly(DVB).68

Eventually, a highly cross-linked region in the polystyrene core is formed by the inter- and intramolecular crosslinking of this DVB.69 The increase of molecular weight analysed by

GPC with relatively narrow PDI proved the formation of star or dendrimer structures. 68 67

On the contrary, the hyperbranched structure was synthesised directly through copolymerisation with low concentration DVB that caused a high degree of random

165

Hybrid Polystyrene-MgH2 Nanoparticles

crosslinking throughout the polymerisation (Scheme 5-3). The formation of hyperbranched structures can be seen through the GPC analysis where broad curves and high PDI values were observed.

Figure 5.4.2 BET analyses results of HPS and S(PS) showing (A) the volume absorption profiles indicate the surface areas and (B) the pore distribution of the materials

The GPC analyses results are summarised in Table 5.4.1 together with the BET analyses. The

BET analyses further confirmed the architectures of these synthesised polymers where

S(PS) has much lower surface areas compared to HPS. In addition, it also has average porosity smaller than HPS. Despite this, Figure 5.4.2 shows the pore distribution of both polymers where HPS have broad level porosity with high numbers of small pores and larger pores hence increasing the average porosity of the overall materials. However, it is still possible to generate the small size nanoparticles within this HPS through the smaller pores although the overall particle size distributions would not be as narrow as with the S(PS).

Table 5.4.1 - GPC and BET results summary

GPC analyses BET analyses

Mn (g mol-1) PDI Surface areas (m2/g) Average pore width (nm) S(PS) 87760 1.389 1.332 2.674 HPS 30042 1.640 59.213 4.732

Preliminary experiments were carried out to test the template roles of these different nanostructure polystyrene by a relatively simple inorganic nanoparticles synthesis. Nickel

166

Hybrid Polystyrene-MgH2 Nanoparticles

salts such as Ni(ac)2 can be reduced by a reducing agent (NaBH4) to generate nanoparticles depending on the types of stabiliser used.70 In this case, both S(PS) and HPS were used as the templates for controlling the Ni nanoparticles growths. TEM images on Figure 5.4.3 shows that both S(PS) and HPS roles as templates worked as expected. HPS gave small Ni nanoparticles with various sizes ranging from 3-10 nm in large spread areas while S(PS) gave consistent size clusters of 3-5 nm Ni nanoparticles. It is possible that these clusters were located in the core of DVB and this ‘cherry’ formation of nanoparticles may be resulted from small porosity formed within the cores caused by the high level crosslinking of DVB.

This is also proven by the size of porosity of S(PS) measured by BET analysis (Table 5.4.1) where the average porosity is much smaller than the size of the clusters observed by TEM.

Nevertheless, both polymers were able to perform as nano templates as expected and can be further used to synthesise the MgH2 nanoparticles.

Figure 5.4.3 Nickel nanoparticles generated in HPS (top) and S(PS) (bottom) with corresponding EDS spectra

The methods adopted in this study were those described in Chapter 4. The first method involved a simple mixing of either HPS and S(PS) into the solution of di-n-butylmagnesium in cyclohexane before being hydrogenolysed. However, it led to the formation of much

167

Hybrid Polystyrene-MgH2 Nanoparticles

bigger size particles similar to previous results when linear polystyrene were mixed during the hydrogenolysis (Figure 5.4.4).

Figure 5.4.4 TEM and EDX analyses of the resulting materials of hydrogenolysis of di- n-butylmagnesium in cyclohexane with (A) S(PS) and (B) HPS

The resulting materials with the S(PS) (MgBu2/S(PS)) has the particle size ranging from 50-

100 nm with rectangular morphology while with HPS (MgBu2/HPS), the material appeared to be agglomerated without any clear structures. It is possible that both polymers were reacting with di-n-butylmagnesium or MgH2 during the reaction at high temperature.

Hydrogen Storage properties

Figure 5.4.5 shows the TPD analyses of both materials. The MgBu2/S(PS) underwent around

24 wt% mass loss at between 300-450 °C. There were two decomposition steps of mass loss accompanied by endothermic peaks on DSC signal and hydrogen peaks on the MS. The first endothermic and hydrogen evolution peak occurred at ~375 °C while the second one occurred at ~425 °C. The latter would be caused by the decomposition of the S(PS) which accounted for ~5wt% loss. This indicates the polymer would still bind to the MgH2 resulting in hydrogen release at higher temperature. On the other hand, material with HPS underwent

~37% wt loss also at between 300-450 °C. This higher amount of polymers is expected with

168

Hybrid Polystyrene-MgH2 Nanoparticles

HPS since the polymer was insoluble in the solvent and would not be lost during the washing. Interestingly, multiple peaks for hydrogen release did not occur and instead a single hydrogen desorption peak at ~380 °C was observed. This shows that the MgH2 may not be inside the hyperbranched structures and the templates role did not fully work. This conformed to what is observed on the TEM analysis of the materials where it consists of larger particles. In fact, the particle sizes are bigger than the particles produced from the hydrogenolysis of di-n-butylmagnesium alone. This indicates the possibility of the precursor or the newly formed Mg reacting with the polymer causing some morphological changes during the growth.

(A) (B)

Figure 5.4.5 TPD analyses of the (A) MgBu2/S(PS) and (B)MgBu2/HPS consist of TGA- black, DSC-blue, MS-red

So far, the template method has faced a big challenge to have the precursor or Mg nucleate inside the porous matrix. The process has proven to be quite complex especially when it involves di-n-butylmagnesium as the precursor. The hydrogenolysis or thermal decomposition process it seems could not induce the nucleation and growth process for a large number of particles simultaneously. Moreover, the possibilities to have side reactions between the precursor or newly formed Mg and the polymers are greater since it involved such a high temperature (180 °C).

MgA.(THF)3 as precursor

In Chapter 4.4, we have explored the alternative method to use a lower temperature by using MgA.(THF)3 as precursor. It showed that the catalytical hydrogenation of the

169

Hybrid Polystyrene-MgH2 Nanoparticles

MgA.(THF)3 in the presence of a catalyst formed MgH2 nanoparticles with smaller sizes than those produced from hydrogenolysis of di-n-butylmagnesium. This means this synthetic process should have a sluggish decomposition rate compared to the decomposition of di-n- butylmagnesium. In addition, the overall process also involved a mild temperature (60 °C).

Nevertheless, Figure 5.4.6 shows the TEM analyses of the materials synthesised from this method with the addition of both S(PS) or HPS. Material produced with S(PS) (MgA/S(PS)) consists of particles clusters in the spherical area which indicated that they are in the core part of the star polymer. This is more or less similar to what has been observed earlier with the Ni and S(PS) system (Figure 5.4.3).

Figure 5.4.6 TEM images of MgH2 produced from catalytic hydrogenation of MgA in THF with (A) HPS, (B) HPS-Ni, (C) S(PS), (D) S(PS-PVBA)

However, there are apparent differences in the overall diameters of the clusters – 30-50 nm in diameters for the MgA compared to 15-30 nm in diameters for nickels. Of course, the average particle sizes inside the core are also bigger for the MgA/S(PS) (10-20 nm in diameter). This could be due to the deformation of the porous matrix due to the swelling of the core when dissolved in polar solvent such as THF. Particles with similar sizes were also observed in the material produced with HPS (MgA/HPS) and it shows the particles spread out and having interparticles distances which could indicate that these particles grew

170

Hybrid Polystyrene-MgH2 Nanoparticles

within the pores of HPS. However, as before, these particle sizes are bigger than the pores of the HPS we observed earlier. This means that the material underwent a similar growth mechanism as those synthesised with S(PS).

In terms of hydrogen desorption properties, Figure 5.4.7 shows the TPD analyses of both materials. Significant mass losses were still observed during the hydrogen desorption especially with the MgA/S(PS) where it underwent an overall 35 wt% loss. Around 20 wt% loss was from the polymer decomposition that occurred at high temperature range (>380

°C). The temperature peak for the hydrogen desorption improved to around 350 °C as expected for such particle sizes. However, MgA/HPS underwent much less mass loss (15 wt%) between 300-450 °C. A fraction of the loss was also from the hydrogen desorption as observed by MS and broad endothermic peak as observed from DSC on this temperatures range.

(A) (B)

Figure 5.4.7 TPD analyses of materials synthesised from catalytical hydrogenation of isolated MgA.(THF)3 with addition of (A) S(PS) and (B) HPS

These contrasting results should be caused by the particles’ locations and interactions with the polymers since both of the materials have very similar particle sizes. There was more polymer decomposition in MgA/S(PS) possibly due to the growth/shrink processes during hydrogen desorption causing the breakage of the cores and thus decomposing them.

Meanwhile, in MgA/HPS the particles are located on the porous matrix where there is more room for the particles to grow and thus it would not break the crosslinked structure that leads to decomposition of the polymers. Furthermore, the confinement of the particles

171

Hybrid Polystyrene-MgH2 Nanoparticles

inside the pores caused the hydrogen to be difficult to diffuse out and led to hydrogen release at higher temperature and slow kinetics. In the end, the hydrogen desorption failed to improve mainly because of the unexpected contraction of the pore sizes of the materials causing the formation of big particle sizes. But at least, the selection of MgA.(THF)3 was successful as a new precursor to form Mg particles with controlled morphologies on these nanotemplates.

Protonated (ammonium) core star polystyrene

Further modifications were applied to these nanostructures in order to improve the stability of the polymer core in S(PS) and eventually achieve smaller particle growth. We decided to incorporate a functional group to give electrostatic force that could contribute to stop the particle growth. This was observed in another study of the protonated core such as star PS-P2VP block copolymer. 71 In this case protonation was achieved by copolymerising the linear polystyrene with vinyl benzylchloride which then was modified to yield ammonium functionality by quaternarisation with TEA (Scheme 5-4).

Scheme 5-4 Copolymerisation and quaternarisation of the PS-PVBA for ammonium functionalisation of star polystyrene

Figure 5.4.8 shows the successful modifications as characterised with 1H NMR. The copolymerisation of vinyl benzylchloride was confirmed with the presence of polymer

172

Hybrid Polystyrene-MgH2 Nanoparticles

peaks at 4.5 ppm which correspond to the CH2-Cl of the vinylbenzyl chloride monomer.

Then in order to protonate the copolymer, a highly reactive trialkylamine was added to substitute for the chloride and give ammonium salt in the chain of the copolymer. Several types of trialkylamines were used but in the end, TEA performed the best reaction especially in terms of the yield. The quaternarisation is proven by the change observed in NMR spectra where 4.5 ppm peak was missing and instead new peaks were formed at 1.0-1.3 ppm, 2.7-

3.3 ppm, and 3.5 ppm confirmed the formation of -CH2-N+-(CH2CH3)3. Then through the exact same process as before, the quaternarised copolymer were further reacted with DVB with enough concentration to make a protonated core and star polymer structure (S(PS-

PVBA)).

Figure 5.4.8 1H NMR of (left) poly(Sty-co-VBC) and (right) poly(Sty-co-VBA)

This protonated star polymer then was used as the template for the MgH2 made from Mg

MgA.(THF)3 . Figure 5.4.9-A shows the TEM image of the MgH2 formed with this polymer. It still formed more or less similar clustering of MgH2 (as shown by the EDX analysis with Mg).

The size of the cluster and particles inside also looks similar compared to MgA/S(PS) sizes, the particles are not small enough to have lower desorption temperatures.

However, TPD analysis of the material revealed a narrower endothermic and hydrogen desorption peak compared to MgA/S(PS). Furthermore, MgA/S(PS-PVBA) underwent bigger mass loss , 60 wt% loss compared to 40% wt loss observed in MgA/S(PS). Almost 30

173

Hybrid Polystyrene-MgH2 Nanoparticles

wt% loss occurred at higher temperatures range which corresponded to the polymer decomposition. It appeared that the protonated chain would react more with the Mg surface and increase the concentration of polymer relative to Mg. Also, more uniform clusters of the

MgH2 nanoparticles were formed as observed in their TEM images. This is possibly due to the increases of this polymer concentrations giving more confinement effects towards the clusters. Indeed, this uniformity resulted in narrower endothermic and hydrogen desorption peak but did not lower the overall hydrogen desorption temperature requirements as the particle sizes within clusters did not change.

(A)

(B)

Nickel Decorated Hyperbranched Polystyrene

Previously, we utilised the HPS to synthesise supported Ni nanoparticles within the structures. Indeed, the transition metals such as Nickel are commonly accepted to boost the kinetics of hydrogenation and destabilising Mg-H bonds when Mg-Ni is formed. Therefore,

174

Hybrid Polystyrene-MgH2 Nanoparticles

we used HPS-Ni as the templates instead of unmodified HPS during the catalytic hydrogenation of MgA.(THF)3.

Figure 5.4.10-A revealed the TEM image of the resulting material (MgA/HPS-Ni). It revealed micelle clustering particles possibly formed due to the suspended Ni nanoparticles in the

HPS. To this date, this is the first time such MgH2 clusters are reported and the cause of this formation is still not clear. Possibly due to the attractive force which commonly seen between different metal atoms to form intermetallic bonds.72

(A)

(B)

The effect of this possible intermetallic formation of Ni-Mg led to just minimum improvement in the hydrogen desorption. TPD analysis of the material (Figure 5.4.10-B) showed the hydrogen desorption occurred at a lower temperature when compared to

MgA/HPS (~360 °C instead ~390 °C). Also, we could see clearly the polymer decomposition

(~2 wt%) at higher temperatures range. Huge mass loss that was observed at below 300 °C is possibly due to solvents and other contaminations in the HPS-Ni after synthesised from

175

Hybrid Polystyrene-MgH2 Nanoparticles

many process steps. Nevertheless, this result show the potential of combining more than two components (MgH2, organic compounds, and transition metals) to achieve more potent improvements in hydrogen sorption properties.

5.4.3 Conclusion

Different polystyrene based nanostructures have been successfully synthesised to accommodate the synthesis route of small MgH2 nanoparticles. Star and hypercrosslinked structured polystyrene were used as nanotemplates for MgH2 nanoparticles to grow. The hydrogenolysis of di-n-butylmagnesium was not the best approach for the polymers as the resultant materials proved to be mostly agglomerated materials. This is possibly due to the side reactions from the precursor and polymers at high temperature conditions. The catalytic hydrogenation of MgA.(THF)3 was a better method and the polymer proved to succeed in their templates role since the nanoparticles with specific clusters were formed.

However, the particle sizes within the clusters were not as small as predicted due to the rigidness of the polymer matrix hence increasing the pore sizes of the overall template cavities. Further attempts were made to functionalise the core of S(PS) with ammonium group in order to enhance the polymer bindings towards MgH2 particles. Indeed, the amount of polymers bound to the MgH2 cluster has increased but the particle sizes inside the clusters are more or less the same. Definitely different types of polymer with stronger mechanical properties are required to inhibit the growth of the MgH2. Another useful finding in this study was the Nickel nanoparticles’ role to attract MgH2 particles and give certain improvements especially by lowering the temperature requirement for hydrogen desorption properties through the destabilisation effect.

176

Hybrid Polystyrene-MgH2 Nanoparticles

5.5 Conclusion

Throughout this chapter, the potential of using polystyrene-MgH2 nanostructures has been discussed and investigated. First, the role of polystyrene to protect MgH2 particles from oxidation is confirmed. However the addition of polystyrene when the MgH2 was produced from hydrogenolysis of di-n-butylmagnesium led to an increase in their particle sizes and the higher the concentration of the polystyrene used the bigger the particle is. This is later identified as due to the high reactivity of Mg or the precursor with polymers at high temperatures. Further investigations were done on the effects of functional groups with the hydrogen storage properties of MgH2 nanoparticles. A series of material syntheses were performed in exactly the same conditions using polystyrenes with different functional groups as the only difference. The results show the correlation between the electronic effects caused by these functional groups towards the thermodynamic properties. The more electron withdrawing functional groups showed weaker enthalpies of the hydrogen desorption. This is the first time such effects have been observed on the nanosize metal hydrides. The last part of this chapter investigated the potential of well defined structures of star and hyperbranched polystyrenes as the nanotemplates for MgH2 nanoparticle synthesis. Initially, the hydrogenolysis of di-n-butylmagnesium method was chosen but high temperature conditions for this method became an obstacle for creating smaller MgH2 particles. A different method, a catalytic hydrogenation of MgA.(THF)3 was then chosen instead and to some extent succeeding in incorporating the particles within the polystyrene nanostructures. However, the overall size of the MgH2 were still too large and the pores of the polymer matrix grew during the MgH2 growth process. Other results proved that protonating star polystyrene could generate more particles inside the core but it did not prevent the growing size of the core. Meanwhile, similar effects of incorporating nickel on the structures could attract the micelles structuring of MgH2 clusters with Nickel nanoparticle as the core. Both of these modifications are novel and opened more

177

Hybrid Polystyrene-MgH2 Nanoparticles

possibilities for exploring the hybrid Mg/MgH2 based nanostructures with improved hydrogen storage properties.

178

Hybrid Polystyrene-MgH2 Nanoparticles

5.6 References

1. Zhang de, S.; Liu, X. Y.; Li, J. L.; Xu, H. Y.; Lin, H.; Chen, Y. Y., Design and fabrication of a new class of nano hybrid materials based on reactive polymeric molecular cages. Langmuir : the ACS journal of surfaces and colloids 2013, 29 (36), 11498-505. 2. Lim, D. W.; Yoon, J. W.; Ryu, K. Y.; Suh, M. P., Magnesium nanocrystals embedded in a metal-organic framework: hybrid hydrogen storage with synergistic effect on physi- and chemisorption. Angew Chem Int Ed Engl 2012, 51 (39), 9814-7. 3. Yuan, J.; Müller, A. H. E., One-dimensional organic–inorganic hybrid nanomaterials. Polymer 2010, 51 (18), 4015-4036. 4. Li, G. L.; Tai, C. A.; Neoh, K. G.; Kang, E. T.; Yang, X., Hybrid nanorattles of metal core and stimuli-responsive polymer shell for confined catalytic reactions. Polymer Chemistry 2011, 2 (6), 1368. 5. Jeon, K. J.; Moon, H. R.; Ruminski, A. M.; Jiang, B.; Kisielowski, C.; Bardhan, R.; Urban, J. J., Air-stable magnesium nanocomposites provide rapid and high-capacity hydrogen storage without using heavy-metal catalysts. Nature materials 2011, 10 (4), 286-90. 6. Chan, M. W.; Chew, K.; Dalby, C.; Gibson, V.; Little, I., Polystyrene supports for vanadium ethylene polymerisation catalysts. Chemical communications 1998, (16), 1673-1674. 7. Beyler, C. L.; Hirschler, M. M., Thermal decomposition of polymers. SFPE handbook of fire protection engineering 2002, 2. 8. Orme, C. J.; Stone, M. L.; Benson, M. T.; Peterson, E. S., Testing of polymer membranes for the selective permeability of hydrogen. Separation science and technology 2003, 38 (12- 13), 3225-3238. 9. Gallicchio, E.; Kubo, M. M.; Levy, R. M., Entropy-enthalpy compensation in solvation and ligand binding revisited. Journal of the American Chemical Society 1998, 120 (18), 4526- 4527. 10. Ashby, E., Grignard reagents. Compositions and mechanisms of reaction. Quarterly Reviews, Chemical Society 1967, 21 (2), 259-285. 11. Aguey-Zinsou, K.-F.; Liu, W., Size Effects and Hydrogen Storage Properties of Mg Nanoparticles Synthesised by Electroless Reduction Method. Journal of Materials Chemistry A 2014. 12. Aguey-Zinsou, K.-F.; Liu, W., Size Effects and Hydrogen Storage Properties of Mg Nanoparticles Synthesised by Electroless Reduction Method. Journal of Materials Chemistry A 2014, DOI: 10.1039/C4TA01108B 13. Varin, R.; Jang, M.; Czujko, T.; Wronski, Z., The effect of ball milling under hydrogen and argon on the desorption properties of MgH 2 covered with a layer of Mg (OH) 2. Journal of Alloys and Compounds 2010, 493 (1), L29-L32. 14. Zlotea, C.; Latroche, M., Role of nanoconfinement on hydrogen sorption properties of metal nanoparticles hybrids. Colloids and Surfaces A: Physicochemical and Engineering Aspects 2013, 439, 117-130. 15. Levchenko, A. A.; Yee, C. K.; Parikh, A. N.; Navrotsky, A., Energetics of self-assembly and chain confinement in silver alkanethiolates: Enthalpy-entropy interplay. Chemistry of materials 2005, 17 (22), 5428-5438.

179

Hybrid Polystyrene-MgH2 Nanoparticles

16. Splinter, S.; McIntyre, N.; Lennard, W.; Griffiths, K.; Palumbo, G., An AES and XPS study of the initial oxidation of polycrystalline magnesium with water vapour at room temperature. Surface science 1993, 292 (1), 130-144. 17. Wagner, C.; Muilenberg, G., Handbook of X-ray photoelectron spectroscopy. Perkin-Elmer: 1979. 18. Friedrichs, O.; Sánchez-López, J. C.; López-Cartes, C.; Dornheim, M.; Klassen, T.; Bormann, R.; Fernández, A., Chemical and microstructural study of the oxygen passivation behaviour of nanocrystalline Mg and MgH2. Applied Surface Science 2006, 252 (6), 2334-2345. 19. Borgschulte, A.; Bielmann, M.; Züttel, A.; Barkhordarian, G.; Dornheim, M.; Bormann, R., Hydrogen dissociation on oxide covered MgH 2 by catalytically active vacancies. Applied Surface Science 2008, 254 (8), 2377-2384. 20. Kooi, B. J.; Palasantzas, G.; De Hosson, J. T. M., Gas-phase synthesis of magnesium nanoparticles: A high-resolution transmission electron microscopy study. Applied Physics Letters 2006, 89 (16), 161914. 21. Dehouche, Z.; Goyette, J.; Bose, T. K.; Liang, G. X.; Huot, J.; Schulz, R., Bimetallic Catalyst Effect on the Sorption Properties of Nanocrystalline MgH₂ Hydride. Journal of Metastable and Nanocrystalline Materials 2001, 11, 77-84. 22. Etiemble, A.; Idrissi, H.; Roué, L., On the decrepitation mechanism of MgNi and LaNi 5-based electrodes studied by in situ acoustic emission. Journal of Power Sources 2011, 196 (11), 5168-5173. 23. Nappa, M. J.; Tolman, C. A., Steric and electronic control of iron porphyrin catalyzed hydrocarbon oxidations. Inorganic chemistry 1985, 24 (26), 4711-4719. 24. Trnka, T. M.; Grubbs, R. H., The development of L2X2Ru CHR olefin metathesis catalysts: an organometallic success story. Accounts of Chemical Research 2001, 34 (1), 18-29. 25. Popeney, C.; Guan, Z., Ligand electronic effects on late transition metal polymerization catalysts. Organometallics 2005, 24 (6), 1145-1155. 26. Lanza, G.; Fragalà, I. L.; Marks, T. J., Ligand substituent, anion, and solvation effects on ion pair structure, thermodynamic stability, and structural mobility in “constrained geometry” olefin polymerization catalysts: An ab initio quantum chemical investigation. Journal of the American Chemical Society 2000, 122 (51), 12764-12777. 27. Jovanovic, S. V.; Tosic, M.; Simic, M. G., Use of the Hammett correlation and. delta.+ for calculation of one-electron redox potentials of antioxidants. The Journal of Physical Chemistry 1991, 95 (26), 10824-10827. 28. Cárdenas‐Jirón, G. I.; Gulppi, M. A.; Caro, C. A.; del Rıó , R.; Páez, M.; Zagal, J. H., Reactivity of electrodes modified with substituted metallophthalocyanines. Correlations with redox potentials, Hammett parameters and donor–acceptor intermolecular hardness. Electrochimica acta 2001, 46 (20), 3227-3235. 29. Streeky, J. A.; Pillsbury, D. G.; Busch, D. H., Substituent effects in the control of the oxidation-reduction properties of metal ions in complexes with macrocyclic ligands. Inorganic chemistry 1980, 19 (10), 3148-3159. 30. Zagal, J. H.; Gulppi, M. A.; Cárdenas-Jirón, G., Metal-centered redox chemistry of substituted cobalt phthalocyanines adsorbed on graphite and correlations with MO calculations and Hammett parameters. Electrocatalytic reduction of a disulfide. Polyhedron 2000, 19 (22), 2255-2260.

180

Hybrid Polystyrene-MgH2 Nanoparticles

31. Grochala, W.; Edwards, P. P., Thermal decomposition of the non-interstitial hydrides for the storage and production of hydrogen. Chemical reviews 2004, 104 (3), 1283-1316. 32. Novaković, N.; Grbović Novaković, J.; Matović, L.; Manasijević, M.; Radisavljević, I.; Paskaš Mamula, B.; Ivanović, N., Ab initio calculations of MgH 2, MgH 2: Ti and MgH 2: Co compounds. International Journal of Hydrogen Energy 2010, 35 (2), 598-608. 33. Saillard, J. Y.; Hoffmann, R., Carbon-hydrogen and hydrogen-hydrogen activation in transition metal complexes and on surfaces. Journal of the American Chemical Society 1984, 106 (7), 2006-2026. 34. Johnson, C. E.; Eisenberg, R., Stereoselective oxidative addition of hydrogen to iridium (I) complexes. Kinetic control based on ligand electronic effects. Journal of the American Chemical Society 1985, 107 (11), 3148-3160. 35. Hoyle, C. E.; Bowman, C. N., Thiol–ene click chemistry. Angewandte Chemie International Edition 2010, 49 (9), 1540-1573. 36. Kade, M. J.; Burke, D. J.; Hawker, C. J., The power of thiol‐ene chemistry. Journal of Polymer Science Part A: Polymer Chemistry 2010, 48 (4), 743-750. 37. Lowe, A. B., Thiol-ene “click” reactions and recent applications in polymer and materials synthesis. Polymer Chemistry 2010, 1 (1), 17-36. 38. Roth, P. J.; Boyer, C.; Lowe, A. B.; Davis, T. P., RAFT polymerization and thiol chemistry: a complementary pairing for implementing modern macromolecular design. Macromolecular rapid communications 2011, 32 (15), 1123-43. 39. Hansch, C.; Leo, A.; Taft, R., A survey of Hammett substituent constants and resonance and field parameters. Chemical reviews 1991, 91 (2), 165-195. 40. Marriott, S.; Reynolds, W. F.; Taft, R. W.; Topsom, R. D., Substituent electronegativity parameters. The Journal of Organic Chemistry 1984, 49 (6), 959-965. 41. McDANIEL, D. H.; Brown, H. C., An extended table of hammett substitutent constants based on the ionization of substituted benzoic acids. The Journal of Organic Chemistry 1958, 23 (3), 420-427. 42. Lewis, E. S.; Johnson, M. D., The Substituent Constants of the Diazonium Ion Group. Journal of the American Chemical Society 1959, 81 (9), 2070-2072. 43. Boyer, C.; Granville, A.; Davis, T. P.; Bulmus, V., Modification of RAFT‐polymers via thiol‐ene reactions: A general route to functional polymers and new architectures. Journal of Polymer Science Part A: Polymer Chemistry 2009, 47 (15), 3773-3794. 44. Aakeröy, C. B.; Seddon, K. R., The hydrogen bond and crystal engineering. Chemical Society Reviews 1993, 22 (6), 397-407. 45. Etter, M. C., Hydrogen bonds as design elements in organic chemistry. The Journal of Physical Chemistry 1991, 95 (12), 4601-4610. 46. Setijadi, E. J.; Boyer, C.; Aguey-Zinsou, K.-F., MgH 2 with different morphologies synthesized by thermal hydrogenolysis method for enhanced hydrogen sorption. International Journal of Hydrogen Energy 2013, 38 (14), 5746-5757. 47. Baldi, A.; Gonzalez-Silveira, M.; Palmisano, V.; Dam, B.; Griessen, R., Destabilization of the Mg-H system through elastic constraints. Physical review letters 2009, 102 (22), 226102.

181

Hybrid Polystyrene-MgH2 Nanoparticles

48. Pasquini, L.; Sacchi, M.; Brighi, M.; Boelsma, C.; Bals, S.; Perkisas, T.; Dam, B., Hydride destabilization in core–shell nanoparticles. International Journal of Hydrogen Energy 2014, 39 (5), 2115-2123. 49. Chung, C.-J.; Lee, S.-C.; Groves, J. R.; Brower, E. N.; Sinclair, R.; Clemens, B. M., Interfacial alloy hydride destabilization in Mg/Pd thin films. Physical review letters 2012, 108 (10), 106102. 50. Lu, Y.; Proch, S.; Schrinner, M.; Drechsler, M.; Kempe, R.; Ballauff, M., Thermosensitive core-shell microgel as a “nanoreactor” for catalytic active metal nanoparticles. Journal of Materials Chemistry 2009, 19 (23), 3955. 51. Artuso, F.; D'Archivio, A. A.; Lora, S.; Jerabek, K.; Kralik, M.; Corain, B., Nanomorphology of polymer frameworks and their role as templates for generating size-controlled metal nanoclusters. Chemistry 2003, 9 (21), 5292-6. 52. de Jongh, P. E.; Adelhelm, P., Nanosizing and nanoconfinement: new strategies towards meeting hydrogen storage goals. ChemSusChem 2010, 3 (12), 1332-1348. 53. Jia, Y.; Sun, C.; Cheng, L.; Abdul Wahab, M.; Cui, J.; Zou, J.; Zhu, M.; Yao, X., Destabilization of Mg-H bonding through nano-interfacial confinement by unsaturated carbon for hydrogen desorption from MgH2. Physical chemistry chemical physics : PCCP 2013, 15 (16), 5814-20. 54. Doluda, V. Y.; Sulman, E.; Matveeva, V.; Sulman, M.; Lakina, N.; Sidorov, A.; Valetsky, P.; Bronstein, L., Kinetics of phenol oxidation over hypercrosslinked polystyrene impregnated with Pt nanoparticles. Chemical Engineering Journal 2007, 134 (1), 256-261. 55. Sidorov, S.; Bronstein, L.; Davankov, V.; Tsyurupa, M.; Solodovnikov, S.; Valetsky, P.; Wilder, E.; Spontak, R., Cobalt nanoparticle formation in the pores of hyper-cross-linked polystyrene: control of nanoparticle growth and morphology. Chemistry of materials 1999, 11 (11), 3210-3215. 56. Saliba, S.; Valverde Serrano, C.; Keilitz, J.; Kahn, M. L.; Mingotaud, C.; Haag, R.; Marty, J.-D., Hyperbranched polymers for the formation and stabilization of ZnO nanoparticles. Chemistry of Materials 2010, 22 (23), 6301-6309. 57. Ottaviani, M. F.; Valluzzi, R.; Balogh, L., Internal structure of silver-poly (amidoamine) dendrimer complexes and nanocomposites. Macromolecules 2002, 35 (13), 5105-5115. 58. Lee, J. Y.; Wood, C. D.; Bradshaw, D.; Rosseinsky, M. J.; Cooper, A. I., Hydrogen adsorption in microporous hypercrosslinked polymers. Chemical communications 2006, (25), 2670-2. 59. Germain, J.; Frechet, J. M.; Svec, F., Nanoporous polymers for hydrogen storage. Small 2009, 5 (10), 1098-111. 60. Sulman, E.; Doluda, V.; Dzwigaj, S.; Marceau, E.; Kustov, L.; Tkachenko, O.; Bykov, A.; Matveeva, V.; Sulman, M.; Lakina, N., Catalytic properties of Ru nanoparticles introduced in a matrix of hypercrosslinked polystyrene toward the low-temperature oxidation of d- glucose. Journal of Molecular Catalysis A: Chemical 2007, 278 (1), 112-119. 61. Demirocak, D. E.; Ram, M. K.; Srinivasan, S. S.; Kumar, A.; Goswami, D. Y.; Stefanakos, E. K., Spillover enhancement for hydrogen storage by Pt doped hypercrosslinked polystyrene. International Journal of Hydrogen Energy 2012, 37 (17), 12402-12410. 62. Li, A.; Lu, R. F.; Wang, Y.; Wang, X.; Han, K. L.; Deng, W. Q., Lithium‐Doped Conjugated Microporous Polymers for Reversible Hydrogen Storage. Angewandte Chemie International Edition 2010, 49 (19), 3330-3333.

182

Hybrid Polystyrene-MgH2 Nanoparticles

63. Miyaoka, H.; Ishida, W.; Ichikawa, T.; Kojima, Y., Synthesis and characterization of lithium– carbon compounds for hydrogen storage. Journal of Alloys and Compounds 2011, 509 (3), 719-723. 64. Manguian, M.; Save, M.; Charleux, B., Batch Emulsion Polymerization of Styrene Stabilized by a Hydrophilic Macro‐RAFT Agent. Macromolecular rapid communications 2006, 27 (6), 399-404. 65. Boyer, C.; Stenzel, M. H.; Davis, T. P., Building nanostructures using RAFT polymerization. Journal of Polymer Science Part A: Polymer Chemistry 2011, 49 (3), 551-595. 66. Goldmann, A. S.; Quémener, D.; Millard, P.-E.; Davis, T. P.; Stenzel, M. H.; Barner-Kowollik, C.; Müller, A. H. E., Access to cyclic polystyrenes via a combination of reversible addition fragmentation chain transfer (RAFT) polymerization and click chemistry. Polymer 2008, 49 (9), 2274-2281. 67. Tsitsilianis, C.; Chaumont, P.; Rempp, P., Synthesis and characterization of hetero‐arm star copolymers. Die Makromolekulare Chemie 1990, 191 (10), 2319-2328. 68. Zheng, G.; Pan, C., Preparation of star polymers based on polystyrene or poly (styrene- b- N-isopropyl acrylamide) and divinylbenzene via reversible addition- fragmentation chain transfer polymerization. Polymer 2005, 46 (8), 2802-2810. 69. Wang, X.; Hall, J. E.; Warren, S.; Krom, J.; Magistrelli, J. M.; Rackaitis, M.; Bohm, G. G., Synthesis, characterization, and application of novel polymeric nanoparticles. Macromolecules 2007, 40 (3), 499-508. 70. Hou, Y.; Kondoh, H.; Ohta, T.; Gao, S., Size-controlled synthesis of nickel nanoparticles. Applied Surface Science 2005, 241 (1-2), 218-222. 71. Suntivich, R.; Choi, I.; Gupta, M. K.; Tsitsilianis, C.; Tsukruk, V. V., Gold nanoparticles grown on star-shaped block copolymer monolayers. Langmuir : the ACS journal of surfaces and colloids 2011, 27 (17), 10730-10738. 72. Shao, H., et al., Hydrogen storage properties of magnesium ultrafine particles prepared by hydrogen plasma-metal reaction. Materials Science and Engineering: B, 2004. 110(2): p. 221-226.

183

6 SUMMARY AND RECOMMENDATIONS

The overall aim of this thesis is to investigate novel wet synthesis methods to synthesise

Mg/MgH2 nanoparticles (smaller than 5 nm) particularly through advanced hybrid

inorganic-organic chemistry to enhance its hydrogen storage properties. The structures,

reaction mechanism, kinetics, and thermodynamics were studied in detail to explore the

potential of these hybrid materials.

In Chapter 4, we fully studied the potential of several wet synthesis methods before

combining and modifying their chemical structures. In particular, the thermal

decomposition of organomagnesium compounds have the advantage over the yield,

efficiency and avoidance of by-products in the overall process. Different organomagnesium

compounds including Grignard reagents in addition to di-n-butylmagnesium are capable of

yielding MgH2 upon hydrogenolysis. Interestingly, we found di-n-butylmagnesium could

yield MgH2 even without the presence of any H2 pressure due to -elimination mechanism.

Meanwhile, hydrogenolysis of Grignard reagents yielded magnesium halides which masked

the overall materials and proved to deteriorate their hydrogen sorption properties.

Summary and Recommendations

Nevertheless, materials produced from di-n-butylmagnesium upon hydrogenolysis proved to be in the nanosize scale and to have superior hydrogen sorption properties. It has fast hydrogen desorption kinetics (10 minutes at 300 °C at 0.1 bar) which is comparable to the nano MgH2 produced through mechanical milling with catalyst. However, this material has more superior H2 storage capacity (7.1 wt%) and even more remarkably, the material was stable upon more than 20 H2 cycles unlike the ball milled MgH2 which deteriorated their sorption kinetics after a few cycles. This stability originates in partial stabilisation from the carbon contaminants as the by products of the overall hydrogenolysis process. These carbon contaminants proved to be crucial in determining the overall morphologies of the materials. This is further investigated in the series of hydrogenolysis experiments conducted in solvents as media and comparing the resulting materials with those hydrogenolysed under dried conditions (either under Argon or H2 pressure). In the end, all materials have different morphologies and thus also the overall H2 storage properties. The polarity of the solvent determined the amount of carbon contaminants. It was found that a highly polar solvent such as diethyl ether is highly reactive with Mg surface hence creating a thick layer of carbon contaminants. On the other hand, a non-polar solvent like cyclohexane reacted less with Mg surface hence creating a thin layer of carbon and led to the formation of stable and uniform Mg nanoparticles with diameters ~15 nm. This material proved to have lower temperature for H2 (330 °C) and also fast desorption kinetics, although it lost some H2 capacity (5.1 wt%) due to carbon contaminants. Such unique and remarkable properties caused this material to become the main material used for further modifications that were discussed in Chapter 5. Another precursor was considered, which is MgA.(THF)3 complex where it could generate monoatomic Mg from several different processes. Overall, the catalytic hydrogenation of MgA.(THF)3 alone led to the formation of stabilised MgH2 nanoparticles with 15 nm diameters in size and requiring less harsh temperature conditions (60 °C). It seems that the stability of these nanoparticles also originated from the carbon contaminations as a result of THF solvents and anthracene

185

Summary and Recommendations

compounds which deteriorated the overall H2 capacity to as low as 3.8 wt%. However, it is possible to increase the capacity by recycling and minimising the uses of anthracene and

THF in the system for producing MgH2 in substantial amounts, in this study an increased capacity to around 5 wt% was observed when the concentration of Mg was 10 times the amount of anthracene being used. Overall, all of these materials revealed the potential for large scale MgH2 production with enough versatility for further modifications into hybrid nanostructures.

In Chapter 5, we investigated the uses of polymer, particularly polystyrene because of its attractive properties such as providing a barrier to oxygen and moisture, high thermal stability (up to 400 °C), and neutral without reactive functional groups. The first part of the investigation explored the capability of polystyrene to protect the materials synthesised from hydrogenolysed di-n-butylmagnesium in cyclohexane from oxidation. Linear polystyrene with a low number of chains (MW~5000) was used and as a result surprisingly the nanoparticles increased their sizes quite significantly to 100 nm in average. Hence the properties of the hydrogen sorption were lowered but it indeed gave protection towards oxidation to some extent. Furthermore, thermodynamic properties of the materials after being oxidised interestingly improved upon several hydrogen cycles. The second part of this chapter focused on the effects of different functional groups towards the hydrogen sorption properties of MgH2 produced from hydrogenolysis of di-n-butylmagnesium in cyclohexane. It was found that particle sizes increase depending on the concentration of polystyrene used during the process. It suggested that the side reactions might take place at such high temperature conditions (180 °C). However, a smaller amount of polystyrenes were used to investigate the effects of functional groups attached on the polystyrene alone.

It was found that there were indications that electronic effects of the functional groups have some influence on the hydrogen storage properties. The most interesting result was that more electron withdrawing groups have the tendency to decrease both enthalpy and entropy. However, the enthalpy-entropy compensation effect also took place resulting in

186

Summary and Recommendations

minimum changes across the hydrogen sorption properties of these materials. Nonetheless, this result is novel as such correlation has never been reported anywhere else. The last part of this chapter focused on the uses of polystyrene structures as support for the nanoparticles. In this study, star and hyperbranched polystyrene were synthesised and utilised as nanotemplates . Initially, hydrogenolysed di-n-butylmagnesium was the chosen method however the results confirmed the reaction of the polystyrene with either Mg or a precursor at high temperatures which resulted in large MgH2 particles. Catalytic hydrogenation of MgA.(THF)3 was then selected as the preferred method instead. The results show that this method worked as expected and formed the MgH2 clusters according to the templates. However, the cluster sizes are too large (100 nm) despite the small size porosity these polystyrenes have before the synthetic process. This result suggests the growth of the pores when MgH2 underwent the growth process too and structures within the polymers were too rigid to inhibit those growths. Further modification was done to improve this rigidness, by incorporating the ammonium functionality within the star structures. As a result, the clusters were more uniform and more polymer bindings were observed. However, the clusters are still a similar size suggesting the structures are still too rigid. Another modification attempt was made by utilising the decorated hyperbranched polystyrene with nickel nanoparticles. This led to the formation of new clusters like micelle with nickel nanoparticles as cores and MgH2 nanoparticles attached to them. This material showed the improved hydrogen storage properties only to some extent because the particle sizes are still too large for displaying significant improvement in the sorption behaviours.

However, overall all of these attempts have explored and shown the potential of hybrid magnesium based materials. Obviously there is still much room for improvements to be made especially in designing polymer nanostructures with properties more suitable and fitted to help synthesise MgH2 nanoparticles with properties closer to DOE targets. Indeed, the future directions of these researches could be to further identify other methods to synthesise MgH2 nanoparticles and combine them with these polymers we have used. This

187

Summary and Recommendations

would help in understanding more of the Mg nucleation and growth process with these organic substances. Furthermore, modifications on the polymers should be performed in order to boost their role in helping to synthesise the materials. For example, by improving the rigidness of the pores within the polymer structures and using them as nanotemplates to synthesise the nanoparticles with exact and controlled sizes.

188