<%BANNER%>

Hydrogenation and Dehydrogenation Characteristics of Electrodeposited Mg-Al Alloys

Permanent Link: http://ufdc.ufl.edu/UFE0041120/00001

Material Information

Title: Hydrogenation and Dehydrogenation Characteristics of Electrodeposited Mg-Al Alloys
Physical Description: 1 online resource (198 p.)
Language: english
Creator: Tanniru, Mahesh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: dehydrogenation, electrodeposition, enthalpy, entropy, hydrogenation, magnesium, magnesiumalloys, microstructure, pctcurves, thermodynamics
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Magnesium hydride with theoretical hydrogen capacity of 7.6wt% is one of the promising solid-state hydrogen storage materials. However, the main shortcoming of this hydride is its high hydrogen release temperature owing to its large negative energy of formation. Alloying of magnesium to form either solid solutions or intermetallic compounds is one of the solutions for reduction of the thermodynamic stability of MgH2. In this study, we have fabricated Mg-Al alloy powders with low Al content via electrodeposition route and coated the powders with Ni as catalyst. To investigate the effect of Al addition on de/hydrogenation characteristics of electrodeposited powders, experiments were carried out in the temperature range of 180-400masculine ordinalC. The results of pressure composition isotherms developed under equilibrium conditions and temperatures at and above 300masculine ordinalC elucidated that the enthalpy of formation/dissociation is not significantly changed by the addition of Al. Detailed microstructural and chemical analyses revealed that Al is rejected by the MgH2 upon its formation from hcp-Mg during hydrogenation at high temperatures. The lack of influence of alloying on the stability of MgH2 formed under equilibrium conditions is attributed to the absence of Al in its structure. Hydrogen absorption tests under 1 MPa pressure at low temperatures in the range of 180-280masculine ordinalC illustrated that under non-equilibrium conditions Al is trapped in the MgH2 phase. The desorption of hydrogen in Al containing hydride was found to take place in 2 stages; in the temperature ranges of 90-150masculine ordinalC and 250-320masculine ordinalC, respectively. These results suggest that the entrapment of Al destabilizes MgH2 and hence hydrogen can be released at much low temperatures. Detailed analysis of microstructure suggests that the release of hydrogen in 2 stages is associated with the inhomogeneous distribution of Ni catalyst on the surface of the particles. Addition of Al to hcp-Mg reduced the total capacity of the hydrogen absorbed in the powder. Furthermore, the kinetics of hydride formation is high initially but slowed down due to the diffusion of Al in hcp-Mg. Interestingly, a phenomenon of hydrogen absorption at low temperatures (40-100masculine ordinalC) is noticed in Mg-Al alloy powders and is attributed to the porosity present in the powder.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mahesh Tanniru.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Ebrahimi, Fereshteh.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041120:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041120/00001

Material Information

Title: Hydrogenation and Dehydrogenation Characteristics of Electrodeposited Mg-Al Alloys
Physical Description: 1 online resource (198 p.)
Language: english
Creator: Tanniru, Mahesh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: dehydrogenation, electrodeposition, enthalpy, entropy, hydrogenation, magnesium, magnesiumalloys, microstructure, pctcurves, thermodynamics
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Magnesium hydride with theoretical hydrogen capacity of 7.6wt% is one of the promising solid-state hydrogen storage materials. However, the main shortcoming of this hydride is its high hydrogen release temperature owing to its large negative energy of formation. Alloying of magnesium to form either solid solutions or intermetallic compounds is one of the solutions for reduction of the thermodynamic stability of MgH2. In this study, we have fabricated Mg-Al alloy powders with low Al content via electrodeposition route and coated the powders with Ni as catalyst. To investigate the effect of Al addition on de/hydrogenation characteristics of electrodeposited powders, experiments were carried out in the temperature range of 180-400masculine ordinalC. The results of pressure composition isotherms developed under equilibrium conditions and temperatures at and above 300masculine ordinalC elucidated that the enthalpy of formation/dissociation is not significantly changed by the addition of Al. Detailed microstructural and chemical analyses revealed that Al is rejected by the MgH2 upon its formation from hcp-Mg during hydrogenation at high temperatures. The lack of influence of alloying on the stability of MgH2 formed under equilibrium conditions is attributed to the absence of Al in its structure. Hydrogen absorption tests under 1 MPa pressure at low temperatures in the range of 180-280masculine ordinalC illustrated that under non-equilibrium conditions Al is trapped in the MgH2 phase. The desorption of hydrogen in Al containing hydride was found to take place in 2 stages; in the temperature ranges of 90-150masculine ordinalC and 250-320masculine ordinalC, respectively. These results suggest that the entrapment of Al destabilizes MgH2 and hence hydrogen can be released at much low temperatures. Detailed analysis of microstructure suggests that the release of hydrogen in 2 stages is associated with the inhomogeneous distribution of Ni catalyst on the surface of the particles. Addition of Al to hcp-Mg reduced the total capacity of the hydrogen absorbed in the powder. Furthermore, the kinetics of hydride formation is high initially but slowed down due to the diffusion of Al in hcp-Mg. Interestingly, a phenomenon of hydrogen absorption at low temperatures (40-100masculine ordinalC) is noticed in Mg-Al alloy powders and is attributed to the porosity present in the powder.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mahesh Tanniru.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Ebrahimi, Fereshteh.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041120:00001


This item has the following downloads:


Full Text

PAGE 1

1 HYDROGENATION AND DEHYDROGENATION CHARACTERISTICS OF ELECTRODEPOSITED MG -AL ALLOYS By MAHESH TANNIRU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Mahesh Tanniru

PAGE 3

3 To Family

PAGE 4

4 ACKNOWLEDGMENTS I would like to sincerely express my gratitude and appreciation for my advisor, Dr. Fereshteh Ebrahimi for h er constant encouragement and thought -provoking ideas that helped me in the development of this dissertation. Her enthusiasm towards research, the time and energy she spends on each student to develop their project was always a source of motivation for me to progress in my research. I thank all my committee members, Dr. Sinnott, Dr. Bourne, Dr. Singh, Dr. Slattery and Dr. Wu for their interest and e xpert advice towards my research. Special thanks are due to Dr. Slattery for conducting part of the experiments carried out in this dissertation at Florida Solar Energy Center. I am grateful to Dr. Wu for providing me the necessary equipment to build an ex perimental set up at University of Florida. The support provided by Dr. Bourne for carrying out TEM characterization, sample preparation using FIB and discussions about the research is invaluable. I express my sincere thanks to Dr. Bourne for his time and co operation. Meetings with Dr. Craciun, Dr. Jacob Jones, Dr. Nino to discuss the XRD results were very helpful in carrying out this research and I thank them for their co -operation. The generosity of Dr. Jacob Jones to help me conducting experiments on th e Insitu -XRD in his laboratory is appreciated. Many thanks are due to Wayne, Kerry and Dr. Dempere for their sincere suggestions regarding the characterization techniques and their timely help in carrying out the analysis using EPMA, HRTEM. The characterization facilities provided by MAIC at University of Florida are greatly appreciated. I am grateful to my masters advisor Dr. Devesh Misra, at University of Louisiana for his continued support and motivation during my graduate study. Particularly his strong will towards research and welfare of the stude nts working in his group is incredible. I take this opportunity to thank him for introducing me to research in materials science and providing me an opportunity to learn various new techniques.

PAGE 5

5 I would like to thank Dr. Ebrahimis former and current group members: Dr. Tatiparti, Dr. Wang, Dr. Rios, Dr. Damian, Mike, Ian, Sonalika and Daniel for being helpful and providing a wonderful working environment. I take this moment to thank all the friends I have made during the graduate study at University of Flori da and University of Louisiana. Dr. Ganesan, Dr. Behera, Dr. Nerikar, Dr. Katira, Dr. Omar, Harish, Badri, Dr. Pramanick, Dr. Aidhy, Dr. Ravinuthala, Dr. Kannan, and Pravin to name a few. Special thanks to Soujanya Ponnada (sony), BV Srikanth, Janardhan Si ngaraju and all other friends in India and other parts of the world for their support throughout these years. Lastly but not the least I would like to express my sincere love and gratitude to my parents, brother, sisters sister in law, and brother in law for their blessings, inspiration and continued support, without which this work would be impossible. This project is funded by NSF under the contract number: DMR 0605406.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 9 LIST OF FIGURES ............................................................................................................................ 11 ABSTRACT ........................................................................................................................................ 17 CH A P T E R 1 INTRODUCTION ....................................................................................................................... 19 2 BACKGROUND ......................................................................................................................... 24 2.1 Characteristics of MgH2 ....................................................................................................... 24 2.2 Fabrication of Mg-Al Alloys ................................................................................................ 25 2.3 Characteristics of Mg -Al Alloys .......................................................................................... 26 2.4 Pressure Comp osition Isotherms in Metal Systems ............................................................ 28 2.4.1 Determination of Enthalpy and Entropy from PCT Curves .................................... 30 2.4.2 Effect of Addition of Alloying Elements on PCT Curve ......................................... 31 2.4.3 PCT Curves Developed for Mg-Al Alloys ............................................................... 31 2.5 Hydrogen (de)absorption Characteristics in Mg Based Alloys .......................................... 32 2.5.1 Addition of Catalyst ................................................................................................... 33 2.5.2 Microstructure ............................................................................................................ 35 2.5.3 Composition and Alloying ......................................................................................... 36 2.5.4 Hydrogen Absorption/desorption Behavior of Mg -Al Alloys ................................. 37 3 EXPERIMENTAL PROCEDURES .......................................................................................... 48 3.1 Materials Fabrication ............................................................................................................ 49 3.1.1 Electrodes Preparation ............................................................................................... 49 3.1.2 Electrodeposition of Mg-Al Alloy Powders ............................................................. 50 3.1.3 A ddition of Catalyst ................................................................................................... 53 3.2 Hydrogen Absorption ........................................................................................................... 54 3.2.1 Hydrogenation Setup .................................................................................................. 54 3.2.2 Hydrogenation Procedure .......................................................................................... 55 3.3 Hydrogen Release Ex periments ........................................................................................... 57 3.4 Annealing of Mg-Al Powders .............................................................................................. 59 3.5 Procedure for Conducting a PCT Experiment .................................................................... 60 3.6 Analysis Methods and Characterization Techniques .......................................................... 62 3 .6.1 Compositional Analysis ............................................................................................. 62 3.6.2 X ray Diffraction ........................................................................................................ 63 3.6.3 Scanning Electron Microscopy.................................................................................. 64 3.6.4 Transmission Electron Microscopy ........................................................................... 65

PAGE 7

7 3.6.5 Insitu X -ray Diffraction ............................................................................................. 65 3.6.6 Phase Fraction Analysis ............................................................................................. 66 4 CHARACTERIZATION OF POWDERS ................................................................................. 74 4.1. Characterization of Electrodeposited Mg-Al Powders ...................................................... 74 4.1.1 Morphology and Size of Powders ............................................................................. 74 4.1.2 Addition of Catalyst ................................................................................................... 75 4.1.3 Phases Present in the Alloy Powders ........................................................................ 76 4.1.4 Composition of the Alloy Powders ........................................................................... 77 4.1.5 Microstructural Characterization of Mg -Al Alloy Powders .................................... 79 4.2 Characteristics of Pure Mg Powder ..................................................................................... 80 4.3 Summary and Conclusions ................................................................................................... 81 5 MICROSTRUCTURAL EVOLUTION DURING PRESSURE COMPOSITION ISOTHERMS .............................................................................................................................. 92 5.1 PCT Curves for Pure Mg Powders ....................................................................................... 93 5.2 PCT Curves for Electrodeposited Mg-Al Alloy Powders .................................................. 97 5.2.1 Microstructural Evolution during PCT Test at 350C ............................................. 97 5.2.2 PCT Curves for Mg 8at%Al Powders at Different Temperatures ........................ 106 5.2.3 PCT Curves for Mg 4at%Al Powders at Different Temperatures ........................ 107 5.3 Enthalpy Determination ...................................................................................................... 107 5.4 Discussion ............................................................................................................................ 108 5.4.1 Phase Transformations in PCT Curve of Mg10at%Al Powders .......................... 108 5.4.2 The Effect of Al Content on Different Stages of the PCT Curve .......................... 111 5.4.3 Effect of Temperature on the Different Stag es during the PCT Curve ................. 113 5.4.4 Effect of Al Content on the Enthalpy and Entropy of Hydride ............................. 114 5.5 Summary and Conclusions ................................................................................................. 114 6 EFFECT OF COMPOSITION AND TEMPERATURE ON HYDROGENATION BEHAVIOR OF ELECTRODEPOSITED MG -AL ALLOY POWDERS ........................... 126 6.1 Hydrogen Absorp tion Experiments ................................................................................... 127 6.1.1 Effect of Al addition on Hydrogen Absorption Characteristics of Mg ................. 127 6.1.2 Effect of Amount of Al on the Hydrogenation Characteristics of Mg-Al Powders .......................................................................................................................... 129 6.1.3 Effect of Temperature on Hydrogenation Behavior of Mg-Al Alloy Powders .... 130 6.2 Thermal Stability of Electrodeposited Mg -Al Powders ................................................... 133 6.3 Discussion ............................................................................................................................ 134 6.3.1 Effect of Al Addition ............................................................................................... 134 6.3.2 Effect of Temperature .............................................................................................. 136 6.4 Summary and Conclusions ................................................................................................. 138 7 DEHYDROGENATION CHARACTERISTICS of MgH2 PRODUCED FROM ELECTRODEPOSITED MG -AL ALLOY POWDERS ........................................................ 153 7.1 Hydrogen Desorption Experiments .................................................................................... 154

PAGE 8

8 7.1.1 Effect of A l Addition on Hydrogen Desorption ..................................................... 154 7.1.2 Effect of Hydrogenation Temperature on Hydrogen Desorption of the Mg-Al Alloy Powders ................................................................................................................ 155 7.1.3 Phase Evolution during Desorption of Mg-Al Powders ........................................ 156 7.1.4 Microstructural Analysis of Hydrogenated Mg-Al Powders ................................. 157 7.1.5 The Effect of Catalyst on Desorption of Mg-Al Powders ..................................... 158 7.2 Microstructural Evolution during Desorption of Hydride ................................................ 160 7.3 Discussion ............................................................................................................................ 161 7.4 Summary and Conclusions ................................................................................................. 164 8 LOW TEMPERATURE HYDROGEN ABSORPTION PHOENOMENON in ELECTRODEPOSITED POWDERS ...................................................................................... 174 8.1 Hydrogen Absorption in Electrodeposited Powders ......................................................... 175 8.2 Hydrogen Absorption in Ni coated Electrodeposited Powders ........................................ 177 8.3 Discussion ............................................................................................................................ 177 8.4 Summary.............................................................................................................................. 181 9 CONCLUSIONS AND FUTURE WORK .............................................................................. 187 REFERENCES ................................................................................................................................. 191 BIOGRAPHICAL SKETCH ........................................................................................................... 198

PAGE 9

9 LIST OF TABLES Table page 2 1 Adva ntages and disadvantages of alloying elements on stability of MgH2. ...................... 39 3 1 Experimental conditions for during electrodeposition of Mg -Al alloy powders ............... 67 5 1 Hydrogen solubility and equilibrium plateau pressure of hydrides in pure Mg powders at different tempera tures. ...................................................................................... 115 5 2 Nomenclature of different samples during the development of PCT curve. .................... 115 5 3 EDS compositional analysis of %Al in various phases in different samples mentioned in Table5 1 ......................................................................................................... 115 5 4 Phase fraction analysis on the samples mentioned in Table 5 1 ....................................... 116 5 5 Hydrogen solubility and equilibrium plateau pressure of hydrides in electrodeposited Mg 8at%Al alloy powders at different temperatures ......................................................... 116 5 6 Hydrogen solubility and equilibrium plateau pressure of hydrides in electrodeposited Mg 4at%Al alloy powders at different temperatures ......................................................... 116 5 7 Enthalpy and entropy values calculated from the Vant Hoff plot for the materials studied ................................................................................................................................... 116 6 1 Hydrogen capacity of the materials studied at different temperatures for both the runs. ....................................................................................................................................... 140 6 2 Comparison of the XRD (110) and (101) peak positions of MgH2 produced at different hydrogenation temperatures.. ............................................................................... 140 6 3 The lattice parameters of MgH2 produced in the Mg 8at%Al powder at different hydrogenation temperatures along with the standard values. ............................................ 140 6 4 Compositional analysis of various regions of the hydrogenated Mg8at%Al particles shown in Figure 6 12 ........................................................................................................... 141 7 1 Hydrogen release temperatures for pure Mg and Mg-Al alloy powders hydrogenated at 210C. ............................................................................................................................... 165 7 2 The hydrogenation contents of pure Mg and Mg -Al alloy powder observed during the absorption and desorption. ............................................................................................. 165 7 3 Hydrogen release temperatures for Mg -Al alloy powders hydrogenated at 180, 210, and 280C. ............................................................................................................................ 165

PAGE 10

10 7 4 Hydrogen release temperatures for Mg -Al alloy powders hydrogenated at 180C with different amounts of Ni. .............................................................................................. 165

PAGE 11

11 LIST OF FIGURES Figure page 2 1 Schematic diagram of crystal structure of MgH2. ................................................................ 40 2 2 Binary Phase diagram of Mg H system calculated at 1 bar pressure of hydrogen. ........... 40 2 3 Binary phase diagram of Mg -Al system indicating the cari ous phases present in the syste ......................................................................................................................................... 41 2 4 Schematic representation of a Pressure-CompositionTemeprature (PCT) curve. ........... 41 2 5 Schematic representation of PCT curve at different temperatures showing the effect of temperature on equilibrium plateau pressure and the development of vant hoff plot from different plateau pressures. ................................................................................... 42 2 6 PCT curves of milled and unmilled Mg powders developed at 350C demonstrating the differences in hysteresis observed in the material [reproduced from [98]] ................. 42 2 7 PCT curves of Mg -Fe alloy powders elucidating the effect of temperature on the shape of the PCT curve. ......................................................................................................... 43 2 8 PCT curve of Mg -Co alloy powder at 100C, indicating the multiple plateaus corresponding to the various phases present in the powder. ............................................... 43 2 9 Vant hoff plot developed for Mg-Al alloys from various experiments. ............................ 44 2 10 PCT curves of Al3Mg2 powder developed at different temperatures illustrating the change in enthalpy of MgH2. ................................................................................................. 44 2 11 PCT curves of pure Mg and Mg 10at%Al (%Al= 50%, 42%) developed at 400C revealing the rise in plateau pressur e of Mg with the addition of Al .................................. 45 2 12 Schematic diagram of various stages that occur during the absorption of hydrogen in the metal and hydride formation .......................................................................................... 45 2 13 DSC curves of Mg 5 mol% X (X= Nb2O5, Fe3O4,ZrO2) illustrating the reduction in hydrogen release temperature with the addition of oxide catalysts. ................................... 46 2 14 Hydrogen absorption curves of pure Mg with different grain size created by ball milling showing the faster absorption in n anograined Mg. ................................................. 46 2 15 Hydrogen absorption curves of Mg-Al alloy powders at 400C demonstrating the effect of Al addition on the ki netics of hydride formation. ................................................. 47 2 16 Hydrogen desorption curves of pure Mg and Mg8 mol%Al alloy powders in TGA at 300C. Hydrogen is released in a short time when compared to pure Mg powder. .......... 47

PAGE 12

12 3 1 Photographs showing the shapes and sizes of the electrodes used in electrodeposition. ................................................................................................................... 68 3 2 Schematic representation of rotating cylinder electrodeposition setup .............................. 69 3 3 Schematic of the modified setup used for Ni coating inside the glove box. ...................... 69 3 4 Schematic representation of the hydrogenation setup used in this study. .......................... 70 3 5 Pressure and temperature vs. time plot obtained by a leak test indicating that there is no leak during the test. ........................................................................................................... 70 3 6 Wt loss vs. Temperature (TGA plot) curve obtained during the calibration of TGA/DSC using CuSO4.5H2O. ............................................................................................. 71 3 7 Photograph of the HTP1 volumetric absorption unit used for PCT development. ........... 72 3 8 Photograph of the Inel insitu X ray diffractometer used for studying the phase evolution during desorption of hydrogenated powders. ...................................................... 73 4 1 SEM micrographs revealing the dominant morphology of powders present in electrodeposited hcprich Mg particles ................................................................................ 82 4 2 SEM micrographs of the Mg -Al particles showing different morphology at the root of the dendrite. rite. ................................................................................................................ 82 4 3 SEM micrograph of the morphology present in lower amounts. Higher magnification image showing the globular shape of the branches. ............................................................. 83 4 4 SEM micrographs illustrating the breakage of electrodeposited Mg-Al particles after coating with Ni. ...................................................................................................................... 83 4 5 Higher magnification SEM image of a Ni -coated Mg -Al particle, the EDS Map of Ni for the micrograph shown in (a). ........................................................................................... 84 4 6 Energy Dispersove Spectra of Ni in two Mg-Al particles demonstrating the differences in Ni content. ....................................................................................................... 84 4 7 XRD profiles of electrodeposited Mg4at%Al, Mg 8at%Al and Mg 10at%Al powders after Ni coating illustrating the various phases present in the material befor e hydrogenation. ........................................................................................................................ 85 4 8 XRD profiles of Mg 8at%Al alloy powders deposited for different time intervals and after Ni -coating procedu re showing the variation of intermetallic content at each stage ......................................................................................................................................... 86 4 9 SEM Micrograph of a Mg 8at%Al particle and the corresponding EDS map of Al depicting its distribution in the particle. ............................................................................... 86

PAGE 13

13 4 10 EPMA composition analysis of a Mg-Al powder fabricated using t he 100%Mg sheet as anode demonstrating the Al distribution in the powder. ................................................. 87 4 11 EPMA composition analysis of a Mg-Al powder fabricated using the 80%Mg + 20%Mg sheet as anode along with the measured Al content at various points on the powder. .................................................................................................................................... 87 4 12 A SE M/BSE micrograph of Mg -Al particle illustrating the distribution of different phases in the Mg 8 at%Al powder. ....................................................................................... 88 4 13 Bright field and dark field TEM micrographs of the Mg8at%Al powder revealing the grain size of the mater ial. ............................................................................................... 88 4 14 TEM Micrographs of Mg8a t%Al alloy powder revealing grain size in Bright field, and Dark field after the coating with Ni. .............................................................................. 89 4 15 Bright -field micrograph of an Mg8at%Al alloy powder at high magnifications using HRTEM. ................................................................................................................................. 89 4 16 SEM micrograph of Pure Mg powder illustrating the morpholo gies and particle sizes. ... 90 4 17 XRD profile of pure Mg powder after Ni -coating. .............................................................. 90 4 18 Bright field and dark field TEM micrographs illustrating the microstruc ture of the pure Mg powder. ................................................................................................................... 91 5 1 Pressure -composition isotherms developed for pure Mg powders at different temperatures normal pressure axis and logarithmic pressure axis.. .................................. 117 5 2 Pressure -composition isotherms developed for electrodeposited Mg10at%Al powders at 350C. The vertical dashed lines represent the division of stages. ................ 117 5 3 A comparison of PCT curve for pure Mg and Mg10at%Al developed at 350C. .......... 118 5 4 The PCT slope vs. Hydrogen wt% curves for hydrogenation and dehydrogenation part of the PCT curve developed for Mg10at%Al alloy powder at 350C. .................... 118 5 5 The PCT curve of sample S2 developed at 350C, XRD patterns of the sample S2. Backscattered electron (BSE) micrographs of different types of particles in sample S2 revealing MgH2 phase .................................................................................................... 119 5 6 Higher magnification images of sample S2 showing the nucleation of hydride on the surface. .................................................................................................................................. 119 5 7 Higher magnification SEM/BSE micrograph of a particle in sample S2 showing the microstructure of MgH2. ...................................................................................................... 120 5 8 The PCT curve of sample S3 developed at 350C .. .......................................................... 120

PAGE 14

14 5 9 The PCT curve of sample S4 developed at 350C. ............................................................ 121 5 10 Higher magnification SEM/BSE micrographs of the partially desorbed powders illustrating ............................................................................................................................ 122 5 1 1 XRD patterns of the sample S5 BSE micrographs of particles in sample S5 revealing the presence of hcp -Mg and Mg17Al12 phases. ................................................................... 123 5 12 Pressure -composition isotherms developed for electrodeposited Mg8at%Al alloy powders at different temperatures. The filled symbols represent the absorption curves and the unfilled symbols represent desorption. .................................................................. 123 5 13 Pressure -composition isotherms developed for electrodeposited Mg4at%Al alloy powders at different temperatures. ...................................................................................... 124 5 14 Vant Hoff plot obtained for pure Mg and Mg 8at%Al alloy powders.. .......................... 124 5 15 Comparison of the pressure -composition isotherms for electrodeposited Mg10at%Al, Mg8at%Al and Mg 4at%Al powders developed at 350 C ............................ 125 5 16 Comparison of the pressure -composition isotherms for pure Mg and Mg 8at% Al powders developed at 350C. .............................................................................................. 125 6 1 Hydrogen absorption curves for pure Mg and Mg8at%Al powders developed at 280C and 1 MPa pressure. ................................................................................................. 142 6 2 The XRD patterns of the hydrogena ted and un-hydrogenated powders for Pure Mg, and Mg 8at%Al powder. ...................................................................................................... 142 6 3 Peak fits of (110) XRD peak corresponding to Pure Mg, (b) Mg 8at%Al powder hydrogenated at 280C ......................................................................................................... 143 6 4 Back scattered electron (BSE) micrographs of the polished pure Mg, and Mg8at%Al powders hydrogenated at 280C. ........................................................................................ 143 6 5 SEM/BSE micrograph of Mg8at%Al powder hydrogenated at 280C indicating the %Al in various phases. ......................................................................................................... 144 6 6 Hydrogen absorption curves for Mg8at%Al and Mg 4at%Al alloy powders developed at 210C, and the corresponding XRD patterns of the hydrogenated powders. ................................................................................................................................ 144 6 7 BSE micrographs of Mg4at%Al, and Mg 8at%Al alloy powders hydrogenated at 210C. Note the presence of unhydrogenated regions, as marked by the dotted circles, on the surface of the Mg4at%Al powder. ............................................................. 145

PAGE 15

15 6 8 BSE micrograph of the Mg 8at%Al alloy powder hydrogenated at 210C along with the compositional analysis along the line AB showing the accumulation of Al at the -MgH2/hcp -Mg interface. .................................................................................................. 145 6 9 Hydrogen absorption curves for Mg8at%Al powder developed at different temperatures and at 1 MPa pressure. .................................................................................. 146 6 10 The XRD patterns of the hydrogenated Mg 8at%Al powder at different temperatures. 146 6 11 Peak fits of (110) XRD peak corresponding to Mg8at%Al alloy powder hydrogenated at different temperatures. ............................................................................. 147 6 12 BSE micrographs of the Mg8at%Al powder hydrogenated at 280C 210C and 180C, revealing MgH2 phase as dark, hcp -Mg phase as light and Mg17Al12 phase as bright regions.. ................................................................................................... 148 6 13 The variation of Al in the dark regions of the micrographs shown in Figure 612. ......... 148 6 14 Back scattered electron (BSE) micrographs of the polished Mg 8at%Al powder hydrogenated at 210C and 280C. Higher magnification images and the corresponding variations of Al content. .............................................................................. 149 6 15 SEM/BSE micrographs of Mg8at%Al annealed for 20 minutes at different temperatures. ......................................................................................................................... 150 6 16 BSE micrographs of Mg8at%Al powders annealed for 5 hours at 180C 210C 280C. ................................................................................................................................... 150 6 17 A comparison of the hydrogen absorption curves developed at 280C for the pure Mg and Mg 8at%Al alloy powder after correction for the influence of the intermetallic phase formation. ............................................................................................. 151 6 18 Diffusion length of Al in hcp-Mg as a function of temperature. ....................................... 151 6 19 SEM/BSE micrograph of a Mg8at%Al powder hydrogenated at 180C showing the submicrometer precipitates in the MgH2 region. ............................................................... 152 7 1 Hydrogen release curves (TGA) for pure Mg and Mg 8at%Al powders hydrogenated at 210C, represents the % weight loss, and fraction of hydrogen released as a function of temperature. ....................................................................................................... 166 7 2 The XRD patterns of the desorbed powders for pure Mg and Mg 8at%Al powder revealing that the phas es are similar to that of initial powders. ........................................ 166 7 3 Hydrogen release curves (TGA) for Mg 8at%Al powders hydrogenated at 180, 210, and 280C. represents the % weight loss, and fraction of hydrogen released as a function of temperature. ....................................................................................................... 167

PAGE 16

16 7 4 Phase evolution during desorption in Mg -Al alloy powders performed in high temperature XRD.. ............................................................................................................... 167 7 5 Peak intensity ratios of Mg:MgH2 at different temperatures. ............................................ 168 7 6 SEM/BSE micrographs of Mg8at%Al allo y powder hydrogenated at 180C. .............. 168 7 7 Distribution of different types of particles in Mg 8at%Al powder hydrogenated at180C ................................................................................................................................. 169 7 8 SEM/BSE micrographs of Mg8at%Al alloy powder after the release of hydrogen in 1st stage. ............................................................................................................................... 169 7 9 Comparison of the distribution of hydride phase between the powders in hydrogenated condition and after the 1st stage of hydrogen release ................................. 170 7 10 Hydrogen release curves (TGA) for Mg 8at%Al powders hydrogenated at 180C wit h different amounts of catalyst. ...................................................................................... 170 7 11 SEM/BSE micrographs of the p artially dehydrogenated powders .................................. 171 7 12 XRD profile of an Mg8at%Al powder after dehydrogenation in TGA, and SEM/BSE micrograph of the unhydrogenated particle. ................................................... 172 7 13 SEM/BSE micrographs and the corresponding EDS maps of Ni of Mg8at%Al powders hydrogenated at 180C revealing the differences in the Ni coating (amount of Ni) on different particles. ................................................................................................ 173 8 1 PCT curves developed at 40C and 60C for electrodeposited Mg8at%Al alloy powder without Ni coating .................................................................................................. 183 8 2 PCT curves for electrodeposited Mg 8at%Al alloy powder developed at 40C 60C from another batch of electrodeposited powders without Ni coating. .............................. 183 8 3 PCT curves of Ni coated Mg 8at%Al alloy powder at 40 and 60C. ............................... 184 8 4 BET curve developed for measuring the surface area of the Mg8at%Al alloy powder. .................................................................................................................................. 184 8 5 TEM micrographs of Mg -Al powder revealing its porosity, underfocussed image, focused image, and over focused image. ............................................................................ 185 8 6 Hydrogen absorption curves of Al -Mg powder at 100C. ................................................. 186 8 7 Hydrogen absorption curves of Al -Mg powder at 150C.. ................................................ 186

PAGE 17

17 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HYDROGENATION AND DE HYDROGENATION CHARACTERISTICS OF ELECTRODEPOSITED MG-AL ALLOYS By Mahesh Tanniru December 2009 Chair: Fereshteh Ebrahimi Major: Materials Scie nce and Engineering Magnesium hydride with theoretic al hydrogen capacity of 7.6wt % is one of the promising solid-state hydrogen storage materials. However, the main shortcoming of this hydride is its high hydrogen release temperature owi ng to its large negative energy of formation. Alloying of magnesium to form either solid solutions or intermetallic compounds is one of the solutions for reduction of the thermodynamic stability of MgH2. In this study, we have fabricated Mg-Al alloy powders with low Al content via electrodepositi on route and coated the powders with Ni as catalyst. To investigate the effect of Al addition on de/hydrogenati on characteristics of electrodeposited powders, experiments were carrie d out in the temperat ure range of 180-400C. The results of pressure composition isotherms developed under equilibrium conditions and temperatures at and above 300C elucidated that the enthalpy of formation/dissociation is not significantly changed by the additi on of Al. Detailed microstructural and chemical analyses revealed that Al is rejected by the MgH2 upon its formation from hcp-Mg during hydrogenation at high temperatures. The lack of influe nce of alloying on the stability of MgH2 formed under equilibrium conditions is attributed to the absence of Al in its structure. Hydrogen absorption tests under 1 MPa pressure at low temperatures in the range of 180280C illustrated that under non-equilibrium conditions Al is trapped in the MgH2 phase. The

PAGE 18

18 desorption of hydrogen in Al containing hydride was found to take place in 2 stages; in the temperature ranges of 90-150C a nd 250-320C, respectively. Thes e results suggest that the entrapment of Al destabilizes MgH2 and hence hydrogen can be released at much low temperatures. Detailed analysis of microstruc ture suggests that the release of hydrogen in 2 stages is associated with the inhomogeneous dist ribution of Ni catalyst on the surface of the particles. Addition of Al to hcp-Mg reduced th e total capacity of the hydrogen absorbed in the powder. Furthermore, the kinetics of hydride form ation is high initially but slowed down due to the diffusion of Al in hcp-Mg. Interesti ngly, a phenomenon of hydroge n absorption at low temperatures (40-100C) is noticed in Mg-Al all oy powders and is attrib uted to the porosity present in the powder.

PAGE 19

19 CHAPTER 1 INTRODUCTION Over the past two centuries the major source of energy for the humankind has been based on fossil fuels like coal, crude oil, and natural gas. The continuous increase in demand for fuels along with a limited reserve of fossil fuels has led researchers to focus on alternative sources of energy [1]. Transportation sector is a major consumer of fossil fuels [2]. Fuel cells are one of the promising technologies that can be used in the transportation sector [3]. Among the various fuel cells available, the proton exchange membrane (PEM) fuel cells based on hydrogen oxidation are ideal for the on-board vehicular applications due to their operational conditions [2]. The major issue faced by this technology is to find a means of storing hydrogen and si gnificant research is being carried out to address this issue [4]. Hydrogen can be stored on-board of vehicles eith er in gaseous, liquid or solid form [2, 5]. However, when stored in gaseous form its volumet ric capacity is very low and safety is also an important issue as hydrogen should be compre ssed to higher pressures to store enough volume [6, 7]. Liqudification of hydrogen is not economically feasible as it needs to be cooled below 251C [6]. Hence, hydrogen absorbed in a solid materi al is considered to be a viable solution for storing hydrogen on-board of vehicles [8]. Am ong the various materials available for hydrogen storage, metal hydrides or complex metal hydrides offer several advantages like high gravimetric and volumetric capacities [9]. Hydrogen reacts with almost all metals and forms metal hydrides [10]. However, alkali metals and alkaline earth metals like alumin um and magnesium offer high gravimetric capacity of hydrogen due to their low density [11]. Ma gnesium hydride with theoretical gravimetric capacity of 7.6 wt% is one of the ideal hydrid es that can be used for hydrogen storage application. The major drawback of MgH2 is its high thermodynamic stability (enthalpy of

PAGE 20

20 formation of -76kJ/molH2) and hence requires high temperature of about 350C to release hydrogen [12-14]. In addition, the hydrogen absorpti on and desorption kinetics are also slow for MgH2 [14, 15]. Several solutions have been prop osed to surmount these challenges [9, 16]. Alloying has been identified as a potentia l mechanism for improvi ng the hydrogen release temperature of magnesium hydride. For example, an experimental study on Mg-Ni alloys has revealed that the dehydrogenation temperatur e is decreased from 320C to 250C owing to alloying with Ni [17]. Theoretical calcul ations have shown that alloying of MgH2 with certain metals can reduce the Mg-H bond strength and lowe rs its enthalpy of formation and dissociation [18]. The kinetics of hydrogen absorption and deso rption in Mg-based alloys are improved by adding catalysts, fabricating materials with na nostructure, and/or creat ing high surface area to volume ratio powders [19-21]. It is shown that addition of catalysts helps in the dissociation of hydrogen molecule, which is one of the rate lim iting steps during the hydrogenation process [22]. Nanocrystalline materials offer faster paths fo r diffusion and it is repor ted that a Mg-5mol%Pd material with 30nm grain size absorbed ~ 6 wt% hydrogen while the same material with 1 m grain size absorbed only 0.2 wt% of hydrogen in 120 minutes [23]. Thus, from the above discussion it can be concluded that nanocrystalline MgH2 with a suitable alloying element that can reduce its stability plus some catalyst for faster hydrogen absorption/desorption kinetics will be an ideal candidate for storage of hydrogen. Aluminum when added as an alloying element to MgH2 is calculated to have significant destabilizing effect on the Mg-H bond strength [24]. The light weight nature of Al also helps in retaining the gravimetric capacity of hydrogen in MgH2 [25]. Additionally the high thermal conductivity of Al is advantageous during the desorption of hydrogen as it involves heat transfer

PAGE 21

21 and its affinity towards oxygen also protects Mg from forming MgO, which is very detrimental to the dissociation/recombination of hydrogen [26]. Due to the low solubility of Al in hcp-Mg phase below 300C, alloys of Mg-Al with high concentration of Al form the intermetallic compound Mg17Al12 which is very stable. This intermetalli c phase requires high temperatures (> 300C) to absorb hydrogen [27]. In addition, the total capacity of hydrogen reduces with increasing the Al content as it does not form the hydride to store hydroge n. Therefore, with all these advantages, nanocrystalline powder of Mg w ith small amounts of Al is a potential material for hydrogen storage in PEM Fuel cells. Despite th e apparent benefits of Al addition to Mg, limited work has been reported on this system. Most of the hydrogenati on studies conducted on Mg-Al powders are for the compositions close to that of intermetallic compounds of Al3Mg2 and Mg17Al12, which require high temperatures for hydr ogenation due to their high stability [25, 2830]. Although theoretical calculations suggest that the incorporation of Al in MgH2 structure should significantly reduce the stab ility of this hydride [30], no e xperimental verification of this phenomenon has been conducted. In addition, th ere have been no reported studies on the microstructural evolution during the hydrogenati on and dehydrogenation processes. Based on the discussion so far, there is a need for a funda mental understanding of the hydrogenation and dehydrogenation characteristic s of Mg-Al alloy powders. The objective of this study was to investigate the eff ect of Al addition on the hydrogenation and dehydrogenation characteristics of hcp-Mg. The Mg-Al alloy powders can be synthesized using various techniques like ball mi lling and rapid solidifi cation [31-33]. In our group we have been able to fabricate nanocry stalline Mg-Al alloys. This method can produce relatively pure alloy powders when compared to other processes such as ball milling [34]. In addition, it has been shown that the morphology, composition and microstructure of the alloys

PAGE 22

22 can also be controlled using th is technique [34]. A commercial ly available pure Mg powder was also studied to understand the effect of Al addition. The hydrogen absorption and desorption charact eristics of electrodeposited Mg-Al and pure Mg powders were evaluate d using two methods. The therm odynamic properties of enthalpy and entropy were measured by developing Pre ssure Composition Temperature (PCT) curves under equilibrium conditions at different te mperatures in the range of 300-400C. High temperatures were chosen to dissolve Al in the hcp-Mg phase and thereby understand the effect of Al addition on the enthalpy and entropy of hydride formation/dissociation. The hydrogen absorption behavior of both the Mg-Al and pur e Mg powders below 300C were evaluated using an in house built absorption unit. The hydrogen release temperature for the powders was established using a Thermogravimetric analyz er (TGA). X-ray diffrac tion (XRD), scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS) were employed to investigate the various phase transformations, to understa nd microstructural e volution and to conduct compositional analysis at different stages of hydrogenation tests, respectively. Precise compositional measurements were made using Electron Probe Micro Analyzer (EPMA) to identify the concentration of Al in different pha ses. The initial microstructure of both the pure Mg and Mg-Al powders was characterized using the transmission electron microscopy (TEM). In this dissertation, Chapter 2 presents the relevant background and includes a brief overview of the fundamentals of hydrogenation and dehydrogenation experiments employed to study materials along with the fabrication techni ques and the characteristics of Mg-Al alloy powders before and after hydrogenation tests. The e xperimental details of the fabrication of alloy powders and the different hydrogenation tests carrie d out in this study are reported in Chapter 3. Chapter 4 discusses the characte ristics of the initial powders before hydrogenation. PCT curves

PAGE 23

23 and the effect of Al addition on the thermodynamic stability of the MgH2 are presented in the Chapter 5. Microstructural analyses, along with compositions of various phases at different stages of PCT curves for Mg-Al powders are di scussed in this chapte r. The hydrogenation and dehydrogenation studies of Mg-Al alloy powders as functions of time, temperature and composition along with the details of the microstr uctural evolution are presented in Chapters 6 and 7. An interesting phenomenon of hydrogen absorption at low temperatures and high pressures observed in the electrodeposited na nocrystalline Mg-Al powders is illustrated in Chapter 8. Finally, general conclusions that can be made out of this study along with the future outlook are made in Chapter 9.

PAGE 24

24 CHAPTER 2 BACKGROUND Among the various materials available, magnesium hydride (MgH2) is considered to be one of the most attractive material for hydrogen st orage for on-board vehicular applications due to its high hydrogen capacity and cheap cost [3, 9, 35]. However, the practical application of MgH2 is mainly limited due to its high hydrogen release temperature [8]. Various studies have been carried out to improve the hydrogen absorp tion and desorption properties of magnesium [14, 21, 36-38]. In particular, the alloying of MgH2 with different metals is explored as a solution to reduce its thermodynamic stabili ty [38]. Furthermore, new fa brication techniques like high energy ball milling have been developed to synthesize nanocrystalline structures that improve the kinetics [39, 40]. Even though, e fforts have been carried out in improving the properties of MgH2, the practical realization of this material is still not achieved. Therefore, the hydrogenation and dehydrogenation characteristics of magnesi um hydride are studied in this dissertation. The goal of this chapter is to provide brief knowledge regarding the hydrogen absorption and desorption characteristics of magnesium hydride. Additionally, the properties of MgH2, the fundamentals of various fabri cation processes of materials and the details of various experimental techniques for assessing hydroge n storage behavior ar e also presented. 2.1 Characteristics of MgH2 MgH2 is an ionic compound and its structure und er ambient conditions is tetragonal with a space group P42/mnm. A schematic diagram of the MgH2 crystal along with its lattice parameters is illustrated in Figure 2-1 [41]. There exists a polymorph of MgH2 at higher pressures (80 kbar) which is orthorhombic [42]. The calculated bina ry phase diagram for Mg-H system at 1 bar pressure of hydrogen is shown in Figure 2-2 [12]. According to the phase diagram, MgH2 is stable up to a temperature of 288C, which is very high for transportation application. Alloying

PAGE 25

25 of magnesium hydride with various elements like Ni, Cu, Al, Ti, Fe etc. are reported to reduce the thermodynamic stability of MgH2 and thereby lowering the dehydrogenation temperature [17, 25, 43-49]. First principles calculations of total energy and the heat of formation of the magnesium hydride alloyed with the different meta ls like Al, Ti, Ni, Fe and Cu have indicated that the bonding between the magnesium and hydr ogen is weakened due to alloying [18, 50, 51]. Among all the alloying elements studied, addition of Al exhibited significant reduction in the heat of formation of MgH2 (76 kJ/mol to 28 kJ/mol). In ad dition, Al has other beneficial properties like light weight, oxi dation resistance, better heat c onduction and cheaper compared to other alloying elements [26]. Therefore, the effect of Al a ddition on hydrogenation and dehydrogenation properties of Mg was car ried out in this study. Since MgH2 has very low solubility for many metals under equilibrium conditions, non equilibrium processing methods have been developed to produ ce these alloy powders [27]. 2.2 Fabrication of Mg-Al Alloys Earlier techniques of prepara tion of Mg based alloys include arc melting or induction melting of the alloying elements [8, 29, 52-54]. Th e Mg-30at%Al alloys were prepared by arc melting of the pure metals. Homogeneous alloys are prepared by using this method but it was observed that the composition of the alloy was difficu lt to control This was attributed to the high volatility of Mg at those temper atures [30]. In addition, milli ng of the final cast product was required as high surface area enhances the hydrog enation processes. Rapid solidification is another technique that was employed to produce high quality alloy powders but this technique also require a final step of ball milling to produce th e powder [32]. Ball milling is a very common technique used to fabricate the magnesium based alloy powders [55-58]. Nanocrystalline phases along with enhanced solubility of alloying elements in Mg are created using this tec hnique [19, 39]. These properties have been observed to have

PAGE 26

26 significant impact on the hydrogenation of magn esium powders [19]. For example, it was observed that nanostructured pure Mg powder produced by ball milling showed significant improvement in hydrogen absorption kinetics th an the microcrystalline powder as shown in Figure 2-3 [19]. Similar improvements are report ed in alloy systems like Mg-Ni, Mg-Ti where supersaturated solid solutions or intermetallic compounds with nanostructure [45, 59]. These improvements in hydrogenation ch aracteristics is attributed to the higher amount of grain boundaries and defects present in nanocrystalline materials which help in faster diffusion of hydrogen atoms [40]. Mg-Al alloy powders are also produced by ball milling. Previous studies on this system reported an increase in the solubili ty of Al in Mg [28, 60]. In addition, the kinetics of absorption was also observed to be fast er than pure Mg powder [28]. Even though, ball milling offers significant advant ages, it suffers with a few drawbacks. The powders produced by ball milling are contaminated by either oxides or milling materials. The process requires comparatively longer time, and the size of the po wder is limited by the ball milling media [61]. From the above discussions it can be conclude d that innovative tech niques are required to produce the alloy powders with be tter hydrogenation characteristics. Electrodeposition is another te chnique extensively used to synthesize nanomaterials [33, 62]. It offers several advantages over other met hods of fabrication like production of high purity alloys, nanocrystalline powders and the compos ition of the alloy powders can be controlled effectively by changing the electrodeposition parame ters [34, 63] It is recently shown that MgAl alloy powders can be produced using electrode position with varied compositions [33]. This technique is applied in this research to synthesize the powders for hydrogenation studies. 2.3 Characteristics of Mg-Al Alloys The equilibrium phase diagram of Mg-Al syst em is shown in Figure 2-3 [64]. Along with the terminal solid solutions of fcc-Al, hcp-Mg, intermetallic compounds of -Al3Mg2, -

PAGE 27

27 Mg17Al12, the line compound R (composition 42at%Mg) are the phases present in the Mg-Al alloy system. The solubility of Al in hcp-Mg and Mg in fcc-Al under equilibrium conditions is very less. In addition both the intermetallic compounds present in Mg-Al system exhibit a solubility range for Al. As most of the proces sing techniques employ non-equilibrium conditions, the characteristics of the alloys vary in each sample. Ball milling of Mg and Al powders in the proportion of 90:10 indicated that Mg17Al12 phase forms after short time of ball milling. Further increase in milling time increased the ratio of the Mg17Al12 phase to hcp-Mg phase up to 3 hours of milling and remained constant above that time period [28]. The lattice parameter of hcp-Mg decreased with increase in milling time demonstrating increased Al solubility. This increase d Al content in the initial raw materials also increased the Mg17Al12 phase content [28]. An arc melt sample with 30at% Al exhibited both the hcp-Mg and -Mg17Al12 intermetallic phases in the fabric ated powders [65]. Bulk metallic alloying is another technique which was employ ed to produce the Mg-Al powders [66-68]. The hcp-Mg and the Mg17Al12 intermetallic phases were identified in the fabricated powders with low amount of Al. However, these phases were produced at a shorter processing time when compared to the other studies [68]. A detailed investigation of microstructural anal ysis of the hcp-Mg rich alloy powders using electron microscopy indicates that the -Mg17Al12 precipitate can form either continuously inside the grains or discontinuously along the defects like grain boundaries [69]. Several orientation relationships were predicted to exist between the -Mg17Al12 and hcp-Mg [70, 71]. However, only a few diffraction patterns of the phases showi ng the orientation relationship were presented [71]. In one particular study the morphology of Mg17Al12 phase was shown to be either rod or lath in shape using the HRTEM [69]. It was observed that the morphology was dependent on the

PAGE 28

28 orientation of the phase. High resolution TEM st udies along with micro diffraction carried out on the alloy AZ91 (composition Mg-9Al-1-Zn-0.3 Mn) illustrated that the primary orientation relationship between Mg17Al12 phase and hcp-Mg phase is (011)//(0001), [1-11]//[2-110]which corresponds to the burger s relationship [69]. These prec ipitates formed after 8 hours of aging treatment at 200C were observed to be extremely fine of about 25-50 nm in width to 100-150 nm in length. Convergent Beam Electron Diffraction (CBED) or the microdiffraction was used to identify the orientation relationships due to the large lattice parameters of the Mg17Al12 phase. From the above studies it can be concluded that due to low solubility of Al in hcp-Mg, the stable -Mg17Al12 phase is formed during the processing of Mg-Al alloy powders. However, by employing the non-equilibrium processes the solubility of Al in hcp-Mg can be increased to an extent higher than that of predicted by the equilibrium phase diagram. 2.4 Pressure Composition Isotherms in Metal Systems Pressure Composition Temperature (PCT) curv es are important sources of fundamental information related to thermodynamic prope rties of metal hydrid es. The thermodynamic properties, namely enthalpy and entropy of form ation/dissociation, can be calculated using the PCT curves developed at different temperatures. Sufficient time is provided during the development of PCT curve to achieve near equilibrium conditions. A schematic diagram of a general PCT cu rve is shown in Figure 2-4. The line AB represents the maximum solubility of hydrogen in the material. Af ter the material is saturated with hydrogen, increase in pressure results in nucleation of hydr ide. In an ideal case, at a given temperature and for a pure metal-hydrogen system the formation/dissociation of metal hydride should occur at a constant pressure according to the Gibbs phase ru le [6]. This region is observed

PAGE 29

29 as the plateau BC in Figure 2-4. This constant pressure of hydride formation is defined as equilibrium plateau pressure as it is denoted by a plateau region on the PCT curve. Furthermore, both the equilibrium plateau pressure of formati on/dissociation of metal hydride should be same theoretically, but almost all PC T curves reported in the literatur e show a hysteresis, indicating that the plateau pressures are di fferent [47, 72]. At the end of the plateau region, increasing the pressure during experiment raises the pressure of th e system without any further or little absorption of hydrogen into the material, as the material is completely saturated with hydrogen. The metal hydride phase generally being ionic in nature, it exhibits very low solubility for hydrogen. Figure 2-5 shows the effect of temperatur e on the PCT curves. With the increase in temperature the solubility and the equilibrium pl ateau pressure of hydride formation increases. After a certain temperature, defined as critical temperature, increase in temperature causes the plateau region to disappear and the solid solution of hydrogen in metal continuously transform into MgH2 phase. Hysteresis in the PCT curves is not well unde rstood. Several models have been developed previously to explain the hysteresi s in the powders [73-76]. They attr ibute it to either the defects, in homogenities, compositional differences present in the material or to the elastic-plastic constraints that occur during the hydride formation. The larg e volume expansion during the hydride formation in metal-hydrogen systems cause the elastic or plastic constrains on the unreacted metal. These volume expansions frequen tly cause an irreversible process of plastic deformation and dislocation generation during the hydride formation. Th erefore, the hydride fraction in all the particles is not the same during the absorption of hydrogen and depends on the particle size, defects and inhomogenities present in each particle. This causes pseudo equilibrium

PAGE 30

30 during the experiment rather th an a true equilibrium and ther eby the absorption curve shows a sloping plateau. The opposite is tr ue in desorption as the volume is contracted and hence it almost occurs at a constant pressure. Furtherm ore, the nucleation of metal hydride/metal during hydrogenation/dehydrogenation respec tively may also cause the hysteresis in the PCT curve. As larger driving force is required during the nucleation of metal hydride, the plateau region starts at a slightly higher pressure than that of the equilibrium plateau pressure and hence this causes a hysteresis in the material. Similar is true when nucleation of metal is required during desorption. The hysteresis was observed to be lower in the materials with nanocryst alline grains and is attributed to the significant defects and grain boundaries that are pr esent in material that aid in the nucleation of phases [77]. However, most of the PCT curves are conducted at higher temperatures and under equilibriu m conditions. Under these conditions the nanograins grow to a large extent and the effect of low hysteresis cann ot be explained with respect to the interfaces. For example, PCT curves of pure Mg powder in milled and unmilled condition at 350C are presented in Figure 2-6. It is observed that the eq uilibrium plateau pressure of formation is not affected significantly by milling while it is lower for unmilled powder during desorption. This signifies further the fact that th e nanograins in the material does not contribute for the observed low hysteresis as the equilibrium plateau pressure of hydride formation is similar in both the materials. The other possible reason for the low hyste resis is the size of the particles rather than the grain size of the material used in this study. The higher surface area/volume ration in the milled particles when compared to the unmilled particles may be the reason for nucleation of Mg. 2.4.1 Determination of Enthalpy and Entropy from PCT Curves The fundamental use of the PCT curve in hydr ogen storage studies is to determine the enthalpy and entropy of the hydride formation/dissociation. The equi librium plateau pressures of

PAGE 31

31 hydride formation/dissociation are used to calc ulate these thermodynamic parameters (Figure 25). The pressure corresponding to the mid-plateau region is consider ed as the equilibrium plateau pressure. The relation between the pressure a nd the temperature is given by the Vant hoff equation: R S RT H P P 0ln (2-1) where P is the plateau pressure of formation/dissociation, P0 is the atmospheric pressure, H is the enthalpy of hydrid ing/dehydring reaction, S is the entropy of hydriding/dehydring reaction, T is the absolute temper ature and R is the gas constant. 2.4.2 Effect of Addition of All oying Elements on PCT Curve Addition of alloying elements to pure metals increases the number components of the system. Due to this addition the PCT curve of the a lloys is affected significa ntly as the number of phases and the components participating in the transformation change [16, 43, 78]. For example, PCT curves of Mg-35wt%FeTi1.2 are shown in Figure 2-7. The plateau region which corresponds to the metal hydride formation was observed to show a significant slope wh en compared to that of pure metal [40]. This behavior of the hydride fo rmation in metal alloys was attributed to the different phase distribution and composition of the alloys [40]. Furthermore, multiple plateaus were observed in some alloys with more than one phase as shown in Figure 2-8 [79]. These multiple plateau regions were explained in te rms of hydride formation from various phases present in the initial al loy powders [25, 28, 79]. 2.4.3 PCT Curves Developed for Mg-Al Alloys A very few PCT curves of Mg-Al system are re ported in the literatu re [21,27-29]. Most of the isotherms developed are for the compositions close to that of the intermetallic compounds Mg17Al12 and Al3Mg2. The isotherms developed for the Mg -Al alloys close to that of the

PAGE 32

32 intermetallic compound Mg17Al12 exhibited either a sloping plateau or multiple plateaus depending on the initial composition of the material. The PCT curves developed for Mg17Al12 at high temperatures like 400C di d not illustrate a plateau re gion and the amount of hydrogen absorbed is also very low [28]. Although no dire ct Vant Hoff plot has been developed for a given Mg-Al alloy, the limited av ailable equilibrium pl ateau pressures for various Mg-Al alloy compositions have been collected and plotted in reference [80]. The Vant hoff plot developed is shown in Figure 2-9 and the enthalpy calculated i ndicated similar values to that of the pure Mg powder. In another study, PCT curves were developed for the intermetallic compound Al3Mg2 at different temperature as shown in Figure 2-10 [81]. Only one plateau region was observed in the PCT curve and the analysis after the hydrogenation indicated the presence of -MgH2 and fcc-Al in the alloy powders. The enthalpy of hydride fo rmation calculated from the vant hoff plot developed from these PCT curves was about 62kJ/m ol which was lower th an the pure Mg [81]. Only one study reported a PCT curve for hcpMg alloy powder. The PCT curve developed at 400C for Mg-10at%Al alloy powder is shown in Figure 2-11. A plateau region around 0.7 MPa pressure was observed in this alloy system and this region was followed by a slope with significant amount of hydrogen abso rption. The equilibrium plateau pressure of this alloy was slightly higher than that of pure Mg powder studied under similar conditions. This alloy was not studied at different temperatures and hence no data about the enthalpy of hydride formation and dissociation were calculated. 2.5 Hydrogen (de)absorption Charac teristics in Mg Based Alloys Sorption studies of hydrogen in magnesium and its alloys were star ted as early as 1960 [11]. Since magnesium is light in weight, cheap in cost and posses high capacity for hydrogen,

PAGE 33

33 tremendous research have been carried out to use it as a hydrogen storage ma terial. In addition to high thermodynamic stability, the kinetics of absorption/desorption were the main concerns to employ it in fuel cell applications [82]. Seve ral advancements and modifications to the microstructure were made to overcome these difficulties. The major breakthroughs that occurred to overcome these problems were addition of catalysts, creation of nanostructured material, and design of new alloys with Mg as primary c onstituent [21, 36]. Furthermore, new processing techniques were also developed to achieve one or more of th ese characteristics [19, 33, 83]. 2.5.1 Addition of Catalyst The sequence of steps taking place during hydr ogenation is shown in Figure 2-12 [3]. Hydrogen molecule reaches the surface of the me tal and breaks into two hydrogen atoms. These atoms chemisorb into the surface and diffuse into the bulk of the metal. When the metal reaches saturation at that temperature a nd pressure, the hydrid e nuclei are formed and grow. In case of Mg, the rate limiting step was observed to be the dissociation of hydr ogen on the surface [21, 52]. Due to the high affinity for oxygen, Mg transf orms to MgO even at low partial pressures of oxygen. The major processing techniques like ball milling, which produce the bulk Mg powders, contaminate the surface by poisoning with oxyge n [61]. MgO is a stable compound and have high activation energy for breaking the hydrogen molecule into atom s [61]. Hence to enhance the dissociation of hydrogen molecule catalysts were added to Mg [ 20]. d-block transition metals and their compounds were considered as effec tive catalysts to break hydrogen molecule [21, 84, 85]. Since these metals are heavy and are added to aid the reaction, the amount of the catalysts is kept low so that the total gravimetric capacity of the material is not lowered significantly. Some of the primary metals which are used as catalysts are Ni, Fe, Mn, Nb, V, Pd, Ti [20, 48, 86, 87]. It was observed that addition of 1at% Ni to Mg decreased the onset temperature of hydrogenation from 275 to 175C and the hydrogena tion capacity of the material was increased

PAGE 34

34 by 50% in the same amount of time without catalyst [88]. Ball milling of Mg2Ni, FeTi and LaNi5 with Pd catalyst was reported to improve th e kinetics of absorption even at much lower temperatures and maintain less sensitivity to air exposures [48]. A Mg-5at%V powder was observed to absorb 4.1wt% hydrogen at 200C in 10 minutes while only 1wt% hydrogen was absorbed by pure Mg powder under the same conditions [86]. The effectiveness of the catalyst was observed to depend on its distribution on the surface, the surface area of the catalyst, amount of catalyst, and its chemical affinity towa rds hydrogen. These properties were optimized by changing the processing conditions of the powder. In addition, na no-catalysts of various metals were developed and the absorption kinetics of Mg powders was observed to increase ~ 2 fold by using these catalysts [40, 89]. Recently it was obs erved that other compounds of d-block metals are much better and effective catalysts than the pu re metals [90-92]. In particular these catalysts were observed to help better during desorp tion of hydride. About 6 wt% of hydrogen was released from MgH2 in 10 minutes at 300C with the presence of Cr2O3 or Fe3O4 on the surface while conventional ball milled MgH2 took about 50 minutes to release hydrogen [91]. DSC studies carried out on MgH2-0.5mol% Nb2O5, Fe3O4, and ZrO2 indicated that the dehydrogenation temperature is reduced by the additi on of oxide catalysts as illustrated in Figure 2-13 [61]. In this study the onset temperatur e of hydrogen was observed to be the same irrespective of the catalyst, but the ki netics of desorption was highest when Nb2O5 was added. The advantages and disadvatages of the various alloying elements during hydrogenation/dehydrogenation processe s were listed in Table 2-1. It was shown theoretically that the d-block metals help in the dissociation of hydrogen molecule due to the interactio ns of d-orbital with the elect rons in hydrogen molecule. This orbital interacts with the s-s electrons in th e hydrogen molecule and forms an intermediate

PAGE 35

35 hydride [85, 93]. For example, NiH is formed when hydrogen molecule was split and the hydrogen diffuses through Ni and interacts with Mg near the surface and gets chemisorbed. This is known as spill over mechanism [85]. Ab-initio cal culations have shown that effectiveness of a catalyst on the MgH2 surface was dependent on the d-orbital occupancy of the d-block metal added. If the d-orbital is unoccupi ed or highly occupied, the electr ons are either donated from the bonding orbitals of Mg-H or received from the d-block metal and thereby weakening the Mg-H bond strength [22, 94]. In case of oxi de catalysts a different theory is proposed regarding their catalytic effect. Since Mg has higher affinity towards oxygen, the catalytic oxides are partially reduced to lower stoichiometry oxides and a co mplex oxide of d-block metal, magnesium and oxygen is formed near the interface. The particip ation of Mg in forming the oxide creates a decrease in concentration of MgH2 and several defects are fo rmed making the hydride non stoichiometric. This reduces th e bond-strength of Mg-H and improves the kinetics of desorption [61]. 2.5.2 Microstructure Increasing the surface area to volume ratio of th e material, refining the grain structure into nanoscale and presence of second phase particles are some of the microstructural changes that lead to faster kinetics of hydrogen absorption [ 19, 40, 95]. Ball milling is generally employed to reduce the particle size [17, 49]. It is observed that milling the Mg particles with catalyst helps in better dispersion of the catalyst and lowers the particle size [49]. As illustrated in Figure 2-14, the kinetics of Mg were improved by reducing the grain size from micron size to 50 nm and further down to 30 nm [19]. In another study it was illustrated that the full capacity of 7.6wt% hydrogen can be achieved at 1 MPa pressure in ball milled powders hydrogenated at 300C while unmilled powders absorb only about 6.1 wt% ev en at high temperatures like 400C [2]. The faster kinetics was attributed to the easie r diffusion of hydrogen atoms through the grain

PAGE 36

36 boundaries and the interfaces produced during the milling. It was also shown that the strain created in the material may accommodate the vol ume expansion that was associated with the hydride formation and thereby increase the ra te of hydrogen absorption in the material. The diffusion of hydrogen through MgH2 is very slow and hence the presence of precipitates inside the material may help in providing the interfaces that may act as path for diffusion of hydrogen [95]. For example, the ki netics of hydrogen absorp tion in Mg-Cu alloy was observed to be higher than that of the ball milled pure Mg. The particles in the Mg-Cu alloy consist of the intermetallic phase Mg2Cu and are predicted to provi de the required interfaces for hydrogen diffusion [95]. In addition, the intermetallic particles may also act as nucleation sites and can improve the rate of hydride formation. 2.5.3 Composition and Alloying Alloying of magnesium with vari ous metals like Ni, Fe, Al, Cu, Ti etc have been observed to modify the kinetics of hydrogen absorption and desorption [43, 72]. Several non-equilibrium processes have been developed to prepare the alloys of Mg as it has very little solubility for the most of the alloying elements under equilibrium c onditions [27]. Another as pect that was studied during alloying is the hydrogena tion behavior of intermetalli c compounds that are formed between Mg and the other metals. Among the va rious intermetallic co mpounds of Mg alloy systems, Mg2Ni is the most studied due to its better hydrogenation properties. It was also shown that alloying of Mg with small amounts Al, In and Ga can decrease the activation energy of hydrogenation and improve the sorption kinetics Theoretical studies carried out on MgH2 with different alloying systems have shown that th e enthalpy of formati on and dissociation was reduced [18]. This reduction in enthalpy was attr ibuted to the interact ion of alloying element with the Mg-H bond strength. Due to the different chemical affinity of the alloying element the interaction between the alloyi ng element and hydrogen changes a nd hence the stability of the

PAGE 37

37 crystal is changed. In addition, it is predicted that the differe nce in valency of the alloying element and Mg creates defects in the material which will help in the diffusion of the hydrogen and thereby the kinetics of hydrogen absorption and desorption of the material. Al is one of the alloying elements that are predicted to reduce the enthalpy of hydride dissociation and improve the kine tics of hydrogen absorption and deso rption. It is also proposed that Al can act as heat transfer medium duri ng desorption due to its high thermal conductivity. In addition, the higher affinity towards oxygen rend ers the formation of MgO which hampers the kinetics of hydrogen absorpti on and desorption significantly. 2.5.4 Hydrogen Absorption/desorption Behavior of Mg-Al Alloys The hydrogenation of Mg-Al powders was conducted in both bulk alloys and in thin films [31, 96]. Most of the experiments of hydrogenati on in alloy powders of Mg-Al alloys were primarily conducted at high temperatures in the range of 300-400C. The initial studies carried out on Mg-1at%Al and 4at%Al have suggested that the activation energy of the hydride formation was lowered at lower concentrations of Al while at higher concentrations they approached to that of pure Mg. The kinetics of hydrogen absorption in Mg-X at%Al (where, X= 10,75,42) at 400C are shown in Figure 2-15 [28]. It can be noticed that the kinetics did not change significantly with addition of Al in thes e alloy powders at 400C. However, the activated samples of Mg-8mol%Al have shown significant improvement in the desorption kinetics as presented in Figure 2-16 [97]. At 300C, the total hydrogen of about 5 wt% present in the Mg8mol%Al was completely released within 300 seconds, while the hydrogen from pure MgH2 released only about 2.5 wt% in 1500 sec onds. Air exposed samples of pure MgH2 and MgH2 alloyed with Al (42at%Al) were studied for their dehydrogenation kinetics using insitu-XRD. The kinetics measured from the calculations of area under the XRD curves indicated that the

PAGE 38

38 activation energy for the release of hydrogen is lowered in case of MgH2-Al powders. This was attributed to the higher affin ity of aluminum towards oxygen. Hydrogenation studies of Mg-Al thin f ilms indicated the formation of Mg(AlH4)2 a complex magnesium aluminum hydride when the composition of the film was about 33at%Mg. In this study the thin films in the whole range of compositions present in the phase diagram of Mg-Al system were deposited and capped with Pd. The hydrogenation results at 110C under 1 bar pressure of hydrogen signified that MgH2 was formed in the alloys with rich Mg contents and Mg(AlH4)2 was observed when the Al c ontent was about 33at%.

PAGE 39

39 Table 2-1. Advantages and disadvantages of alloying elements on stability of MgH2. Alloying element Advantages Disadvantages Al Light weight Good conductor of heat Less oxide formation Theoretical calculations predicted low enthalpy of hydride formation Low solubility Forms intermetallic compounds Mg17Al12 and Al3Mg2 Disproportionate reaction takes place. Cu Can act as catalyst Good conductor of heat Forms intermetallic that enhances diffusion Theoretical calculations predicted low enthalpy of hydride formation Heavy metal and reduces the capacity Forms intermetallic in the 1st cycle and hence loses the catalyst activity Ti Excellent catalyst Destabilizes the Mg-H bond if present in the MgH2 crystal lattice. Forms TiH2 which requires higher temperature for hydrogen release Low solubility in Mg Forms intermetallic compounds. Fe Good catalyst Forms complex hydrides Reduces the dehydrogenation temperature Heavy metal and reduces the capacity Forms Mg2FeH6 Ni Excellent catalyst Helps in hydride nucleation. Forms Mg2NiH4 Reduces dehydrogenation temperature Heavy in weight Los solubility Forms intermetallic compounds of Mg2Ni and MgNi2 V Excellent catalyst Helps in hydride nucleation Reduces dehydrogenation temperature Heavy in weight and reduces capacity Low solubility Forms VH2 which is very stable Nb Excellent catalyst Helps in hydride nucleation Changes the electronic structure in Mg-H bonding Heavy in weight and reduces capacity Forms NbH1.5 and reduces the catalytic activity

PAGE 40

40 Figure 2-1. Schematic diagram of crystal structure of MgH2. Figure 2-2. Binary Phase diagram of Mg-H system calculated at 1 bar pressure of hydrogen [reproduced from [12]].

PAGE 41

41 Figure 2-3. Binary phase diagram of Mg-Al system indicating the carious phases present in the system [reproduced from [64]]. H / M P T H / M P T Figure 2-4. Schematic representation of a Pr essure-CompositionTemeprature (PCT) curve.

PAGE 42

42 H / M P 1 / T ln(P/P0) T1T2T3 H / M P 1 / T ln(P/P0) T1T2T3 Figure 2-5. Schematic representation of PCT curve at different temperatures showing the effect of temperature on equilibrium plateau pressu re and the development of vant hoff plot from different plateau pressures. Figure 2-6. PCT curves of milled and unmilled Mg powders developed at 350C demonstrating the differences in hysteresis observed in the material [reprodu ced from [98]].

PAGE 43

43 Figure 2-7. PCT curves of Mg-Fe alloy powders elucidating the effect of temperature on the shape of the PCT curve [reproduced from [16]]. Figure 2-8. PCT curve of Mg-Co alloy powder at 100C, indicating the multiple plateaus corresponding to the various phases presen t in the powder [rep roduced from [79]].

PAGE 44

44 Figure 2-9. Vant hoff plot developed for Mg-Al alloys from various experiments [reproduced from [80]]. Figure 2-10. PCT curves of Al3Mg2 powder developed at different temperatures illustrating the change in enthalpy of MgH2. The corresponding Vant Hoff plot can be seen in the inset [reproduced from [81]]

PAGE 45

45 Figure 2-11. PCT curves of pure Mg and Mg -10at%Al (%Al= 50%, 42 %) developed at 400C revealing the rise in plateau pressure of Mg with the addi tion of Al [reproduced from [28]]. Figure 2-12. Schematic diagram of various stag es that occur during th e absorption of hydrogen in the metal and hydride form ation [reproduced from [6]].

PAGE 46

46 Figure 2-13. DSC curves of Mg-5 mol% X (X= Nb2O5, Fe3O4,ZrO2) illustrating the reduction in hydrogen release temperature with the additi on of oxide catalysts [reproduced from [61]]. Figure 2-14. Hydrogen absorption cu rves of pure Mg with differe nt grain size created by ball milling showing the faster absorption in nanograined Mg [rep roduced from [19]].

PAGE 47

47 Figure 2-15. Hydrogen absorption curves of Mg-Al alloy powder s at 400C demonstrating the effect of Al addition on the kinetics of hydride formati on [reproduced from [28]]. Figure 2-16. Hydrogen desorption curves of pure Mg and Mg-8 mol%Al alloy powders in TGA at 300C. Hydrogen is released in a shor t time when compared to pure Mg powder [reproduced from [97]].

PAGE 48

48 CHAPTER 3 EXPERIMENTAL PROCEDURES Mg-Al alloy powders being developed in this study are anticipated to be used as hydrogen storage materials. In the past, researchers have demonstrated th at specific characteristics of materials like nanocrystalline grains, higher surface area/volume ratio, oxide free surfaces have improved the kinetics of hydrogen absorpti on and desorption of hydrogen [19, 87, 92]. Furthermore, it was shown that alloying of pure Mg affect the thermodynamic properties of hydride formation and dissociation [43, 47, 99]. Diffe rent processing techniques like ball milling, mechanochemical synthesis, induction melting, rapi d solidification, electr odeposition have been developed to achieve the required characterist ics in Mg-Al alloys [28, 30, 31, 33]. Among those techniques electrodeposition ha s several advantages for fabr icating alloy powders. It was elucidated that nanocrystalline metal alloy powders with high purity along with a good control on morphologies and composition can be synthesi zed using electrodeposition [34]. Hence this technique is employed for processing th e Mg-Al alloy powders in this study. The raw materials employed in the synthesis of Mg-Al alloy powders via electrodeposition were extremely sensitive to moisture and oxygen, therefore all the handlin g and fabrication was conducted in an argon filled glove box. The glove box employed was maintained at a very low O2 (< 1 ppm) and H2O (<5 ppm) levels to avoid any cont amination during the production and handling of powders. Nickel was added as catalyst to enhance the kinetics of hydrogen absorption and desorption. The Ni coated electrodeposited Mg-Al alloy powders were hydrogenated using a house built absorption unit. Additionally, the Mg-Al alloys were also tested for their thermodynamic properties at Fl orida Solar Energy Cent er (Dr. Slatterys laboratory) located in Cocoa, Fl. Desorption of hydrogenate d powders was conducted using a Differential Scanning Calorimetry/ Thermogr avimetric Analyzer (DSC/TGA). The Mg-Al

PAGE 49

49 powders at various stages of hydrogenation/de hydrogenation processes and PCT experiments were analyzed using various analytical char acterization techniques like SEM, XRD, EPMA, TEM. The material processing procedures, analy tical testing and the sample preparation for various characterization t echniques applied in this research are explained in this chapter. Moreover, the procedure applied during the deve lopment of Pressure Composition Temperature (PCT) curves is also described in detail. 3.1 Materials Fabrication 3.1.1 Electrodes Preparation The electrodes used during the electrodeposition were prepared outside the glove box. Pure copper (99.99%) and graphite r ods (99.9%) in cylindrical shap e were employed as cathodes (working electrodes). The dimensions and shapes of the cathodes are shown in Figure 3-1(a). The electrodeposition was conducte d on graphite electrode because the powders deposited can be scraped of easily from the graphite surface a nd if any graphite part icles are included it is beneficial during hydrogenation. It was shown in earlier studies that the graphite pa rticles can act as catalyst during the hydrogena tion [100, 101]. The Cu electrode was electrolytically polished using a solution composed of 82.5 vol% orthophos phoric acid in deionized water [102]. A potential of 1.1V was applied betw een the working electrode and th e reference electrode made of aqueous KCl solution using a Princeton Applied Research (PAR) potentiostat. The Cu cathode was polished for 5 minutes and it was rotated at a speed of 200 RPM to polish the surface uniformly. A steel sheet was employed as the second electrode during the electropolishing of copper. The graphite electrodes were machined using the lathe mill to remove the previously electrodeposited layer and were ultrasonically cleaned in toluene. A set of 2 copper electrodes and 5 graphite electrodes were used in each expe riment. The diameters and the weights of all the electrodes (copper and graph ite) were measured and taken into the glove box.

PAGE 50

50 A pure Mg sheet (Fine Metals Inc, 99.999%) or a combination of pure Mg and pure Al (Fine Metals Inc, 99.993%) sheet were used as anodes during the electrodeposition processes. The metal sheets were 1mm in thickness. The anodes used for the de position processes are presented in Figure 3-1(b) and (c). These metal sheets were polis hed using the 800 grit metallographic paper on the surface and were u ltrasonicated in toluene for 15 minutes before taking them into the glov e box for electrodeposition. 3.1.2 Electrodeposition of Mg-Al Alloy Powders The aim of the present study was to pr epare Mg-Al alloy powders using the electrodeposition process. The reduction potentials of Mg and Al (-1.706V for Al and -2.375V for Mg) being lower than the hydrolysis of wate r, electrolytes based on aqueous solutions will release hydrogen at the cathode in stead of metal deposition. Hence, from the previous studies conducted in our laboratory, an electrolyte base d on organometallics was us ed for the synthesis of Mg-Al alloy powders [34]. The electrolyte employed duri ng the electrodeposition processe s was prepared using the similar principles that were de veloped earlier for fa brication of Al-Mg alloy powders [33]. It should be noted that the previous study was conduc ted using a 30 mL electrolytic setup while in this study a 100 mL electrolytic se tup was employed. The amount of electrolyte was increased by 3.3 times and bigger electrodes were used to pr oduce sufficient amount of alloy powders for hydrogenation studies. A rotating cylinder electrode cell setup was employed in this study. A schematic representation of the experimental se tup is shown in Figure 3-2. The anodes employed in these experiments were annular in shape as shown in Figure 3-1(b) & (c). A 250 mL glass beaker was used as the electrolytic cell. A PAR 263/273 potentiostat/galvanostat system with a maximum output of 2 amps was employed to ap ply the required current for electrodeposition. The PAR system was controlled via a computer written program while the rotator of cathode was

PAGE 51

51 controlled externally. The final electrolyte com position after mixing the va rious chemicals in the proportions described in earlier studies was as follows [33]. 1 mol Na[AlEt4] + 2 mol Na[Et3Al-H-AlE t3] + 2.5 mol AlEt3 + 6 mol toluene. The electrodeposition process of Mg-Al alloy powders develope d comprised of two steps. i) Pre-electrodeposition process. ii) Electrodeposition process. The first step was termed as pre-electrodepositio n process in which a copper electrode (working electrode or cathode) was used along with the pure Mg sheet as anode. From the initial composition of the electrolyte mentioned above, it can be noticed that it does not contain any magnesium components. Hence to produce Mg-Al alloy powders, Mg was introduced into the electrolyte using the pre-elec trodeposition process [34]. The aim of this process was to incorporate significant amount of Mg into the electrolyte which can be subsequently used to produce the alloy powders that will be employed in the hydrogenation studies. Depending on the required composition of the Mg-Al powders after electr odeposition, a galvanostatic electrodeposition was conducted for a specific amount of time at a pa rticular temperature as the pre-electrodeposition process. From the knowledge of previo us studies conducted in our laboratory on electrodeposition of Mg-Al all oy powders, a current density of 60mA/cm2 and a temperature of 90C or 60C were employed dur ing pre-electrodeposition process in this study [34]. Toluene was added during the electrodeposition to k eep the electrolyte level at 100 mL as it evaporates during the experiment. For production of alloy powder with high Mg content, longer pre-electrodeposition times were employed. In these experiments, the total pre-electrodeposition time was divided into 2 equal parts and conducte d on different copper elect rodes while the anode remained the same. The change of cathode enhances the Al deposition from electrolyte as it was shown that the initial layers on the cathode during the electro deposition of Mg-Al powders on

PAGE 52

52 copper substrate comprises of fccAl rich solid solution [33]. This preferential deposition of Al enriches the electrolyte with Mg ions which help in producing hcp-Mg rich powders during the electrodeposition. After the pre-el ectrodeposition process, the c opper cathodes were cleaned in toluene which was maintained at 60C for 10 minutes. The cleaning process was repeated 3 times in new toluene to completely elim inate the organometallics from the powder. After the pre-electrodeposition process, the copper cathodes were replaced by the graphite cathodes and the Mg anode with either a pure Mg sheet (new) or a combination of Mg and Al sheets. The type of anode is decided depending on the composition of the powder to be produced. In order to produce powders by electrod eposition, higher deposition rates are required during the electrodeposition processes [ 103]. So, a current density of 150mA/cm2 was employed during this process. Furthermore, all the elect rodeposition processes were conducted at 60C, and 200 RPM. The graphite electrodes were imme rsed into the electrolytic solution up to a length of 2 cm from the bottom. The area of the electrode inside the elec trolyte was calculated using the initial diameter and the total current ob tained was incorporated into the program. One of the problems encountered during the electrodeposition process was the evaporation of the solution due to the rise in temp erature during the deposition. It was shown that if a Mg anode is employed during the electrodeposition of Mg-Al a lloy powders, an exothermic reaction occurs near the surface of the anode which causes the temp erature of the electrolyte solution to increase. The temperature of the electrolyte during the experiment increased from 60-85C and got stabilized around 85C. To maintain the electrolyte level and the te mperature, toluene is added at regular intervals. The other issue faced during th e initial experiments was to retain the produced powders on the surface of the electrode instead of falling into the electrolyte due to their weight. Hence, to overcome these problems the time of the electrodeposition was optimized for the

PAGE 53

53 maximum production of powders without losing the electrolyte (due to evaporation) and the powders by falling into the electrolyte. The el ectrodeposition was carri ed for 45 minutes and toluene was added during the experiment to keep the level of the electrolyte at 100 mL and also to improve the conductivity. Similar to pre-elec trodeposition the powders after electrodeposition, on the graphite electrode were also cleaned using toluene at 60C for 10 minutes (3 times). Finally, the powder on the electrode was scraped using a scalpel into a clean beaker with toluene and preserved. During the cleaning process the el ectrolyte was placed on a second heater that was maintained at 60C to avoid the crystallizat ion and thereby preserving the same electrolytic chemistry for further depositions. The electrodeposition process was repeated 5 times on different graphite electrodes to produce sufficient amount of alloy powders fo r hydrogenation studies. To produce the powders of similar compositions in all the electrodeposition the anodes we re cleaned inside the glove box using a spatula and a pair of tweezers. The s ubsequent electrodeposition was conducted on a new graphite rod along with the cl eaned anodes. The powders from all the electrodeposits were collected in a beaker with clean toluene. The pre-electrodeposition and electrodeposition conditions us ed to fabricate the Mg-Al alloys in the present study are shown in Table 3-1. The ratio of anode is changed during the electrodeposition to produce di fferent composition powders. A pure Mg sheet and a 80%Mg + 20%Al sheets were used in this study. 3.1.3 Addition of Catalyst The powders collected from the electrodeposition were filtered and weighed using a balance with an accuracy of 0.005 g. To enha nce the kinetics of hydrogen absorption and desorption they were coated with a catalyst. Many d-block based transition elements are used as catalyst during hydrogenation/dehydrogenation of Mg powders [20, 49, 87, 104]. Ni is one of the

PAGE 54

54 effective catalysts that were show n to improve these processes and hence it is used as catalyst in this study [21]. The Ni coating procedure wa s conducted inside the glove box to avoid the exposure of powders to oxygen and moisture. An organometalic based Ni compound was used and the procedure described in the Patent: 4,554, 152 was used to coat the powders [105]. Bis(1,5) cyclooctadiene Ni (Strem Chemicals, 98.3% pure) was used as the source of Ni. A schematic representation of the Ni-c oating setup is shown in Figure 3-3 The organometallic Ni and the Mg-Al powders were taken in 0.12:1 proportion and were refl uxed in toluene at 110C for 6 hours. During refluxing, it was difficult to circulate cooling solvent inside the glove box and hence a long jacket was used as shown in the Figure 3-3 so that the toluene vapors are condensed on the walls and drop back into the flask. After 6 hours, the powders were filtered and finally rinsed with anhydrous hexane (Fisher brand, 99.99999%). 3.2 Hydrogen Absorption 3.2.1 Hydrogenation Setup The schematic of the hydrogenation setup used in this study is shown and labeled in Figure 3-4. This setup was built to conduct hydrogena tion experiments based on volumetric method (sieverts type) [106, 107]. The hydr ogenation setup can be divided into two parts viz. reaction chamber and reservoir chamber. As the names indicate the reaction cham ber was used for the reaction of hydrogen with the powders and the re servoir chamber was used for constant supply of hydrogen into the reaction chamber. The two ch ambers were separated using a regulator that controls the flow of hydrogen from reservoir ch amber to the hydrogenation chamber. A pressure sensor was located such that the pressure in eith er of the chambers can be measured at a given point of time. Two J-type thermocouples (OmegaTJ36-ICSS-180-6) were used in the setup to measure the temperatures in the reaction chambe r and the reservoir chamber. The thermocouple in the reaction chamber was inserted such that it was located close to the sample surface to

PAGE 55

55 indicate the temperature of th e sample while the thermocouple on the reservoir chamber was attached on the surface of the rese rvoir and this was maintained at room temperature. The whole setup was connected to high pur ity Ar (Airgas, 99.99999 %) gas, high purity hydrogen (Airgas, 99.99999%) gas cylinders and the vent-hood via a roughing pump. The temperatures, pressure and the duration of the experiments were contro lled using the computer controlled program developed in the Lab-view software from nationa l instruments. The data was recorded using the DAQ system (Omega DAQ 55) and simultaneously a real time pressure was also monitored using the pressure indicator (Omega DP25B-E ; volts: 115). A pressure sensor from omega (PX880-1KGi) that can operate between 1-1000 psi was employed in this study and has the capability of recording the pressure changes of 0.01 psi. All the tubing used in building the system was from Swagelok and were fixed using the corresponding pressure fittings to avoid any leakages. 3.2.2 Hydrogenation Procedure The Ni coated powders after filtration we re weighed precisely using the balance and loaded into the hydrogenation chamber inside the glove box. After loading the sample the hydrogenation chamber was sealed using wrenches and the valve on the other end was also closed. The sealed hydrogenati on chamber was brought out of th e glove box and connected to the hydrogenation setup. The air present in the tu bing was evacuated using the vacuum pump and then back filed with high purity argon. The w hole system including the tubing was flushed with Ar gas 3 times and finally with high pressure of hydrogen. After flushing, the valve near to the hydrogenation chamber was closed and a heating tape was wound around the chamber. The chamber was heated using the resistance heater which was connected and controlled by the DAQ system through a relay. The regulator that was co nnecting the reservoir chamber and the reaction chamber was set to a pre decided pressure valu e. The hydrogenation chamber was heated to the

PAGE 56

56 temperature of the experiment and then the valve near the chamber was opened. Immediately, after opening the valve, the valve near the pressure sensor was rotated to read the pressure on the reservoir chamber side and the valve between th e reservoir and the hydrogen gas cylinder was closed. The reservoir was always maintained at a higher pressure (approx. 2 times) than the experimental pressure, so that the gas flow was always from re servoir to the reaction chamber. As the pressure in the reaction chamber drops du e to the absorption of gas into the material, enough gas enters through the regu lator to compensate the pressu re loss and a drop in pressure on the reservoir side was measured. The pressure in the reservoir was reco rded as a function of time and this drop represents the amount of gas that was absorbed into the material. Using the Ideal gas equation, the drop in pressu re in the reservoir wa s converted into the number of moles of hydrogen absorbed. RT PV n (3-1) where P is the drop in pressure n is the no of moles of hydrogen absorbed T is the reservoir temperature R is the universal gas constant V is the volume of the reser voir chamber along with the tubing The number of moles of hydrogen calculated was converted into the weight of hydrogen as WH = n 2 gms (3-2) and the weight percent of hydrogen was calculated by using the initia l weight of the powder used in the experiment. Approximately 0.3-0.5 gms of pow der was used in each experiment to obtain reasonable amount of pressure drop. The compressibili ty of the gas is ignored in the calculation.

PAGE 57

57 To identify if the system was leaking, be fore conducting the hydr ogenation experiments the whole hydrogen setup was checked for leaks. Di fferent types of leak te sts were performed on the hydrogenation setup. In each te st different sections (reacti on chamber and reservoir chamber) were isolated and then pressurized separately. Th e drop in pressure is recorded as a function of time and temperature in that section was also reco rded to understand the pressure differences due to fluctuations in temperature. For example, the curve in Fi gure 3-5 was generated by running the test by following similar procedure as that of hydrogenation at 250C. The straight line in the curve represents that the pressure was not ch anged with time which de notes no leak in the chamber. Leak tests were conducted at all th e temperatures and pr essures of hydrogenation studied in this research. The pr essure sensor and the thermocoupl es were calibrated at regular intervals. 3.3 Hydrogen Release Experiments The hydrogen release experiments were conducted in a Setaram setsys evolution TGA/DSC. During the thermogravimetric analysis th e changes in the weight of the sample were observed as a function of time and temperature. The weight loss or gain by the sample during the experiment is representative of the reaction taking place in th e material. For example, in the present study the weight loss obser ved during the heating of a sample is indicative of the amount of hydrogen released. The powders after hydrog enation were unloaded from the hydrogen reaction chamber in the glove box without exposing to atmosphere and stored. To conduct the desorption experiments the powders were br ought out of the glove box using a vacuum container. Partial vacuum was created in the ch amber inside the glove bo x so that the powders were not exposed to air. Just before runni ng the hydrogen release expe riment in TGA, the powder was removed from the vacuum chamber a nd was weighed precisely using a balance that

PAGE 58

58 has an accuracy of 0.0005 g. After re cording the initial weight, the sample was loaded into the furnace of TGA. The TGA along with the furnace is flushe d with He and enough time was provided to equilibrate the balance under the He flow. Then th e sample was heated from room temperature to 400C at the rate of 5C/min. The weight loss along with the heat ab sorbed by the sample (because release of hydrogen is endothermic reac tion) during the heating of the sample was recorded as a function of time and temperat ure. The weight loss during the experiment corresponds to the amount of hydrogen release from the material. The wt% of hydrogen was calculated by considering the initial weight of th e sample. To remove the instrumental error from the observed weight loss, same crucible was run in the similar conditions with out the hydrogenated powder and the weight loss observed dur ing this experiment was subtracted from the weight loss in the previous experiment. The final weight loss observed after the empty crucible run denotes the actual amount of hydrogen presen t in the powder. To evaluate the microstructure and phases present at different stages of dehydrogenation, the samples were heated to the te mperatures of interest at the ra te of 5C/min in TGA and were cooled rapidly to room temperature at the rate of 50C/min to freeze the microstructure present at that temperature. The correspond ing structural analysis was c onducted on the powders obtained from these experiments. The TGA was calibrated for the temperature and the weight loss using the standard copper sulphate penta hydrate (CuSO4.5H2O). One of the curve generated using the standard material is shown in Figure 3-6(a) and the we ight losses and transformation temperatures at different stages were compared to the standa rds provided by the company (Figure 3-6(b))and it was observed that they correspond to each other within the e xperimental error.

PAGE 59

59 3.4 Annealing of Mg-Al Powders It has been shown that the Mg-Al alloy powders produced by electrodeposition comprise of supersaturated solid solutions of hcp-Mg and fcc-Al [33]. In addition, it was also demonstrated that very fine nanocrystalline structure of alloy powders can be produced by electrodeposition. The microstructure produced in these experiments is metastable in nature. The rigorous conditions (high current density, high speed of rotation and low temperature) employed to fabricate the alloy powders during electrodepo sition produces powders with the metastable phases. Hydrogen absorption being a high te mperature and time taking process, the microstructure of electrodeposited Mg-Al powders will be affected due to annealing in those conditions. Hence to understand the differences in microstructural and phase changes between hydrogen absorption and the expe rimental conditions like temp erature and time, annealing experiments were conducted on the Mg-Al powde rs at the hydrogenation temperatures and times. To obtain the same conditions during th at of hydrogenation, the annealing experiments were conducted in the hydrogenation chamber its elf but with out pressurizing with hydrogen. Powders were loaded into the hydrogenation cham ber as described in the previous section, and the chamber was connected to the hydrogenation sy stem. After flushing the system with Ar, the reaction chamber was heated to the temperature of the experiment and held at that temperature for specific period of time. Annealing experiment s were conducted for two different periods of time at each hydrogenation temperature. The shorter time experiment was designed to understand the phase changes that take place in Mg-Al powders before the reaction chamber was pressurized with hydrogen in th e hydrogenation experiment. In th is experiment, the powder was heated to the hydrogenation temperature in the same heating rate as that of hydrogenation experiment and once the temperature was reached, the powders were quenched immediately to preserve the phases. In, the second experiment the powder was heated to the temperature of

PAGE 60

60 interest and was held for 5 hours at that temperature and then quenched to room temperature. The annealed powders were characterized for th e phases, microstructure and compositions using XRD, and SEM. 3.5 Procedure for Conducting a PCT Experiment The PCT experiments can be conducted on both volumetric and gravimetric absorption units. A system which works on the volumetric princi ple (sieverts principle) was used in this study [106]. The Pressure Composition Temperatur e (PCT) experiments were carried out on a Hiden Isochema (HTP1) instrument located in Dr Slatterys facility at Florida Solar Energy Center (FSEC), Cocoa, Fl. This unit is shown in Figure 7-2(a) along with the schematic (Figure 7-2b) of the various chambers present inside the equipment. The Ni-coated electrodeposited powders were thoroughly rinsed in ultra high purity hexane (99.99999%) and were stored in small bottles. These bottles were filled with ultr a high pure hexane before enclosing in a vacuum jar. The vacuum jar was sealed inside the glove box and then shipped to the facility in Cocoa. The sealed vacuum jar was taken into a glove box and the powders were filtered from the hexane and weighed precisely before loading them into the reaction chamber. Approximately 0.3-0.4 gm of powder was used in each experiment. The reaction chamber shown in the schematic was detached from the system and taken into the glove box along with a transfer cylinder. Th e reaction chamber has a provision to insert a thermocouple from the bottom of the chamber to measure the temperature of the sample during the experiment and another hole on the top for pressurizing the chamber. The powders were loaded into the reaction chamber and it was sealed in a bigger transfer cylin der. A piece of glass wool was used to cover the hole before loadi ng the sample. The loaded transfer chamber was placed on the top of the absorption unit. A removable glove compartment was then attached to the absorption unit and was evacuated. The glove co mpartment has a provision to back fill with

PAGE 61

61 argon. The evacuated glove compartment was backfilled with argon and the procedure was repeated. Finally, after backfill ing the glove compartment with argon, the reaction chamber from the transfer cylinder was taken out and loaded into the furnace of the absorption unit using the gloves of the compartment. After loading the chamber successfully the glove compartment was detached from the absorption unit before running the experiment. By usin g the detachable glove compartment the samples were loaded into the ab sorption unit without exposing them to air and also the temperature of the sample was also measured accurately by placing a thermocouple very close to the sample surface. After loading the sample, a calibration of volume of the reaction chamber and the piping involved was carried out using a high purity heli um gas. This calibration was conducted at room temperature and at different pressures of helium. Finally the reaction ch amber was evacuated and the sample was heated to the temperature of the experiment. After attaining the reaction temperature the sample was exposed to a meas ured volume of hydrogen pressure. To achieve this, the piping between the valve FCV1 and the valve PCV4 (Figure 7-2(b)) was pressurized to a known value and then the valve PCV4 was opened to expose the sample to hydrogen. The software then waits till the pressure in the chamber reaches an equilibrium value before shutting the PCV4 valve. Based on the volume calibratio n and the observed pre ssure the amount of absorbed hydrogen is calculated using the ideal gas equation. In the next step the piping between the FCV1 and the PCV4 was pressurized to a hi gher pressure and the PCV4 was opened again and allows the pressure in the chamber to reach equilibrium. The same procedure was repeated in incremental steps of pressure till the maximum limit of the equipment. The absorbed hydrogen gas is calculated at each step and a PCT curve is developed from the data.

PAGE 62

62 After reaching the maximum pressure during th e desorption part of the PCT curve, the piping between the FCV1 and PCV4 was evacuated and backfilled with hydrogen to a lower pressure than the reaction chamber. Then the op ening of the PCV4 valve results in flow of hydrogen gas from the reaction ch amber and the system was allo wed to reach equilibrium. The difference in pressure during th e step was calculated as the am ount of hydrogen gas released. The compressibility of the gas at various pressu res was considered during the calculation of the amount of hydrogen absorbed into the material. An upper time limit is provided in the software to move onto the next step if equilibrium is not reached in that specified time. 3.6 Analysis Methods and Ch aracterization Techniques 3.6.1 Compositional Analysis The compositional analysis was carried out using a JEOL 733 micro probe (EPMA) and a JEOL JSM 6400 (EDS) Scanning Electron Microscope (SEM). The error in compositional analysis using EPMA being very low (0.5%), it was used to evaluate the precise composition of components in particular phases while SEM-EDS wa s used to identify the compositions in bulk regions of the samples. The samples for the compositional analysis were prepared by mounting the powders in a cold mount made of epoxy (Buehler). After the epoxy was hardened the samples were polished using the metallographic paper (Buehler 400 grit pa per) until the powders in the epoxy were exposed. Sequentially, after th e 400 grit paper the block was polished using a 800 and 1200 grit papers in that order for about 10-15 minutes. After each stage of polishing the surface of the block was cleaned using a cotton ball dipped in ethyl alcohol and finally ultrasonicating the block in ethyl alcohol for 10 mi nutes. The cotton ball was used to remove the particles which are firmly stuck on the surface. Following the polishing on the grit papers, the sample was polished using 9,6, and 1 micron di amond pastes and the Leco microid diamond compound extender from Leco was used as lubricant. About 15-20 minutes was spent on each

PAGE 63

63 polishing stage to remove the scratches and the sa mple is cleaned with the cotton balls dipped in ethyl alcohol after each stage. In addition ultrasonication in ethyl alcohol was also carried out to remove the particles effectively. After the final stage of polishing and ul trasonication the samples were coated with carbon to improve the conduc tivity during the SEM. Additional carbon paint was also applied in case of EPMA sa mples for better focus of the beam. 3.6.2 X-ray Diffraction A Phillips APD 3720 powder X-ra y diffractometer with Cu-K raditation was primarily employed in this study to identify the phases an d their lattice parameters The alloy powders at various stages of experiments were analyzed using the X-ray diffractometer. The sample was prepared by sticking a double stick tape on the st andard glass slide and spreading the powder on the surface of the tape. The height of the edge of the slide was adjusted to the powder level by applying more tape. To correct the x-ray diffrac tion patterns for the instru mental error, tungsten (W) powder was mixed in the samples during the analysis. The experimental peak positions (2theta value) of the phases were identified by fitting the peaks using the software Profile Fit and the final 2 theta positions were identified by adjusting them using the tungsten peaks in the pattern. All the scans were performed betw een the 2 theta ranges of 20-90 degrees. The lattice parameters of the hydride phase fr om different samples were identified using the (110) and the (101) peaks of magnesium hy dride. After correcti ng the X-ray diffraction patterns for the instrumental error the 2 theta pos itions were used in th e following equations to identify the lattice parameter. Braggs law which is described in the equation was used to calculate the d-spacing of the particular peak Sindn 2 (3-3) where is the wavelength of the radiation used ( 1.54056)

PAGE 64

64 d is the d-spacing of the corresponding plane is the half of the angle measured from the x-ray diffraction pattern The (110) peak was used for identifying the a lattice parameter and using this value and d-spacing of the (101) peak the c lattice parameter was calculated. The equations used for the calculations of lattice parameters were shown. 2 22 2 1101 a kh d (3-4) and 2 2 2 2 2 1011 c l a h d (3-5) It should be noted in some of the samples the amount of the powder and the standard material was kept constant to identify the proporti on of different phases at various stages of the experiments [108]. 3.6.3 Scanning Electron Microscopy A JEOL JSM 6400 SEM was employed to observe the morphology of the alloy powders. The initial powders were analyzed by mounting the scraped powders on an aluminum stub using the carbon paint. The distribution of different phases in the powd ers was also analyzed using the SEM. The average atomic numbers of different phases were exploited in the backscattered electron imaging (BSE) mode to identify the phases. The different phase fractions of the powders were also analyzed by using the BSE images in the software Image J. Multiple images of each sample was taken and the volume fraction of the different phases in the microstructures was calculated.

PAGE 65

653.6.4 Transmission Electron Microscopy A JEOL 200CX Transmission electron micros cope (TEM) was used to identify the microstructure of the Ni-coated electrodeposit ed powders. The TEM samples were prepared using the Leica Ultracut UCT ultramicrotome. Th e powders were embedded in a mixture of resins made up of SPI-PON 812, NMA and DMP 30 from SPI supplies in the proportions described in the reference. The resins were cu red at 50C for 24 hours and these capsules were used to cut the TEM samples using the ultram icrotome. The face of the capsules was trimmed using the razor blade before loading them into the ultramicrotome. The samples were loaded with the cutting face parallel to the edge of the knife. A diam ond knife was employed to cut the sample and the capsule was oscilated with differe nt speeds to obtain the samples. A boat attached to the diamond knife was filled with a mixture of deionized water and ethyl alcohol. The TEM slices that were cut float on the surface of the boat and they were sepa rated using an eyebrow before fishing onto the carbon co ated copper grid of 3 mm size. These grids were used in the TEM to identify the grain size, micros tructure and different phases of the powders. The TEM was operated both in bright field and dark field mode along with the selected area diffraction to iden tify the different phases in the material. 3.6.5 Insitu X-ray Diffraction The insitu X-ray diffraction technique was em ployed to identify the phases that were evolved during the release of hydrogen. The In el insitu XRD presen t in Dr. Jacob Jones laboratory at university of Florid a is employed to identify the ph ases during the desorption. The hydrogenated powders were transferred to the fu rnace of the XRD instrument and were heated from room temperature to 450C in air at the rate of 5C/min. During the heating of the sample the copper Kradiation was used to obtain the X-Ray di ffraction patterns from the powder as a function of temperature and time. To understand the evolution of different phases and to interpret

PAGE 66

66 the hydrogen release temperature, the collected diffraction profiles were plotted in 3-dimensions using Matlab. The differences in the intensity of th e peaks as the sample is heated were identified and were interpreted for the evolution of phases. 3.6.6 Phase Fraction Analysis The microstructure of the Mg-Al powders c onsisted of various phases before and after hydrogenation experiments. The quantification of these phases was necessary to compare various effects like temperature and am ount of hydrogen absorption in different samples. The phase fraction analysis on Mg-Al powders was carried out using the Image J software. Backscattered electron mode was employed to identify the different phases as the average atomic number of all the phases was different. Multiple images of each sample were taken from different particles randomly and these images were used in the Im age J software to identify the fractions of different phases. The brightness in the image mode was adjust ed such that the contrast corresponding to one phase was highlighted and the area under the distribution curve was calculated. By normalizing this area with the tota l area of the particle will give us the phase fraction. This method was employed on both the hydr ogenated and annealed powders to interpret the evolution of phases at differe nt temperatures and pressures.

PAGE 67

67 Table 3-1. Experimental condi tions for during electrodeposi tion of Mg-Al alloy powders Parameter Pre-electrodepositionElectrodeposition Current Density (mA/cm2) 60 150 Cathode rotation speed (RPM) 200 200 Cathode Copper Graphite Anode 100% Mg Temperature (C) 60 60 Time (mins) 180 45

PAGE 68

68 = 12.6 mm = 12.6 mm Al anode Mg anode Mg anodea b c = 12.6 mm = 12.6 mm Al anode Mg anode Mg anode = 12.6 mm = 12.6 mm Al anode Mg anode Al anode Mg anode Mg anode Mg anodea b c Figure 3-1. Photographs showing th e shapes and sizes of the electr odes used in electrodeposition. (a) side view of the Copper and graphite cathodes. Top view of the (b) Pure Mg anode (c) Mg + Al anode.

PAGE 69

69 Heater Rotator Potentiostat Computer control Rotator controller Graphite Cathode Mg anode Electrolytic cell Electrolyte Electrolyte level Heater Rotator Potentiostat Computer control Rotator controller Graphite Cathode Mg anode Electrolytic cell Electrolyte Electrolyte level Figure 3-2. Schematic representation of rotating cylinder electrodeposition setup Heater/stirrer Cooling jacket Toluene level Rubber cork Stirrer Heater/stirrer Cooling jacket Toluene level Rubber cork Stirrer Figure 3-3. Schematic of the modified setup used for Ni-coating inside the glove box.

PAGE 70

70 Tape Tape Tape Figure 3-4. Schematic representation of the hydrogenation setup used in this study. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0100200300400 Time (mins)Pressure (MPa)0 50 100 150 200 250 300 350Temperature (C) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0100200300400 Time (mins)Pressure (MPa)0 50 100 150 200 250 300 350Temperature (C) Figure 3-5. Pressure and temperatur e vs. time plot obtained by a leak test indicating that there is no leak during the test.

PAGE 71

71 20 30 40 50 60 70 80 90 100 110 020040060080010001200 Temperature (C)wt loss (wt%) 28.821 % 7.219 % 32.103 %a 20 30 40 50 60 70 80 90 100 110 020040060080010001200 Temperature (C)wt loss (wt%) 28.821 % 7.219 % 32.103 % 20 30 40 50 60 70 80 90 100 110 020040060080010001200 Temperature (C)wt loss (wt%) 28.821 % 7.219 % 32.103 %a b b Figure 3-6. (a) Wt loss vs. Temperature (TGA plot) curve obtained during the calibration of TGA/DSC using CuSO4.5H2O. (b) Standard Wt loss vs. Temperature (TGA plot) curve provided by the Setaram Co mpany for TGA/DSC calibration.

PAGE 72

72 a a b b Figure 3-7. (a) Photograph of the HTP1 volumet ric absorption unit used for PCT development. (b) Schematic diagram of the reaction chamber and the piping involved for the development of PCT curve.

PAGE 73

73 Figure 3-8. Photograph of the Inel insitu X-ray diffractomet er used for studying the phase evolution during desorption of hydrogenated powders.

PAGE 74

74 CHAPTER 4 CHARACTERIZATION OF POWDERS This chapter details about the initial charac teristics of the electr odeposited Mg-Al alloy powders before the hydrogenation processes. Th e alloy powders were characterized for their morphology, phases, composition and microstructu re. The Mg-Al alloy powders were coated with the catalyst Ni and its distribution on the surface was established using the EDS mapping technique in SEM. Also a pure Mg powder wa s employed in the hydrogenation studies and its initial characteristics were also stud ied and described in this chapter. 4.1. Characterization of Electrodeposited Mg-Al Powders 4.1.1 Morphology and Size of Powders The morphology of the majority of Mg-Al pow ders fabricated using electrodeposition is shown in Figure 4-1. While the high Al conten t electrodeposited powders have been shown to exhibit different morphologies [33],only a single morphology (globular morphology), was observed in the high Mg (>88 at%) powders fabricat ed in this study. As shown in Figure 4-1(a) the particles, due to the high cu rrent density employed in this study, are observed to be 0.5 mm to 1 mm long. The particles were dendritic in na ture as is evident from Figure 4-1(b). These dendrites have many branches that grow in multip le directions. Figure 4-1(c) demonstrates that they exhibit a hierarchical structure consisting of aggregates of approximately 20 m in size. These aggregates were com posed of individual 2-0.5 m size units, which were somewhat flat and faceted (see Figure 4-1(d)). The globular morphology observed in the electrodeposited MgAl powder was reported previous ly, except the roughness of the mo rphology was different [33]. The morphology of Mg-Al powder in this study wa s characterized with lower roughness than the previously observed morphologies in electrodeposite d Al-Mg powders.

PAGE 75

75 The root of the dendrites showed a smooth globular morphology as shown in Figure 4-2(a). This morphology was observed exclusively at th e beginning of the dendrites (the root) and counted for a small fraction of the total weight of the powders fabri cated (Figure 4-2(b)). In addition to the morphology shown in Figure 4-1(c), occasionally, a less rough globular morphology was also observed as illustrated in Figure 4-3(a). At higher magnifications as shown in Figure 4-3(b), it is observed that this mor phology also exhibited a hier archical structure with much finer units which were round (globular) in sh ape. It is to be noted that this type of morphology was observed in very less number of particles and the reasons for its formation could be due to the localized differences in the composition of the electrolyte [4]. 4.1.2 Addition of Catalyst The methodology and materials used to coat th e powders with Ni are discussed in Chapter 3. After Ni coating, it was observed that the branches of the dendrites fell apart during the coating procedure and the particles became smalle r as shown in Figure 4-4 (compare to Figure 41(a)).The distribution of Ni on the surface of the powders was characterized using the EDS mapping in SEM. A secondary electron micrograph along with the corresponding EDS-map of Ni is shown in Figure 4-5. It can be seen from Figure 4-5(b) that Ni is distributed finely on the surface. Further analysis on different powders indi cated that the Ni content was not similar on all the particles. The EDS spectra of different part icles taken for the same amount of time are shown in Figure 4-6. The EDS peak corresponding to th at of the Ni was shown in the spectra. The intensity of the peak varied from one particle to the other as shown in Figure 4-6(a) and (b), which demonstrates that the total amount of Ni on each particle was different. It is anticipated that the particles with variation of Ni cont ent on the surface may cause different hydrogenation behavior in the particle under the same conditions.

PAGE 76

764.1.3 Phases Present in the Alloy Powders The XRD profiles of all the powde rs fabricated in this study are shown in Figure 4-7. The Mg-Al powders primarily consis ted of the hcp-Mg phase, Mg17Al12 intermetallic compound and a small quantity of fcc-Al phase. The amount of the intermetallic phase increased with the Al content of the electrodeposited pow ders. The addition of 4at%Al sh ifted the peak positions of the hcp-Mg to higher values (e.g. (101) peak shif ted from 36.582 for pure Mg (from std to 36.693 for the Mg-4at%Al alloy). This observation indicates a decrease in the lattice parameter of the hcp-Mg phase with Al addition, which is in acco rdance with the previously reported results [43]. No significant change in the 2 theta positions of the peaks for hcp-Mg was found with increasing the Al content of the alloy from 4at% to 8at%, suggesting that the amount of Al dissolved in Mg was similar in these two powders but the additio nal Al in the latter powder was precipitated as the intermetallic phase. However, the hcp-Mg peak in the powder with 10at%Al shifted to a higher 2 theta position when compared to 4 and 8at%al alloy powders. Furthermore, the Mg-Al alloy powders showed wider Mg peaks, elucidating their finer grain size. Previous studies on electrodeposition of the Mg-Al powders using similar electrodeposition conditions indicate d that the formation of intermetallic was suppressed during the synthesis and only supersaturated solid solutions of fcc-Al and hcp-Mg rich phases were observed [33]. In contrast to those studies it was noticed that the in termetallic compound was also present in the electrodeposits fabricated in this study. This intermetallic content increased after the Ni coating. To iden tify the reasons behind the forma tion of the intermetallic phase, electrodeposition was conducted for different time intervals and th e XRD analysis was carried out. The XRD profiles from these powders along with the powders af ter Ni coating are shown in Figure 4-8. This Figure indicate s that for shorter time interval s of electrodeposition, only supersaturated solid solutions of Mg in fcc-Al and Al in hcp-Mg were observed while for

PAGE 77

77 depositions conducted for longer time pe riods, the intermetallic compound of Mg17Al12 was also present in the powders. It was noticed during th e experiments that the temperature during the electrodeposition was increased from 60C to 85C. From the XRD analysis it can be concluded that the longer time periods at high temperatures involved dur ing the processing of the alloy powders by electrodeposition causes th e in-situ annealing and leads to the precipitation of the intermetallic compound in hcp phase. Furthermore, coating of Ni was conducted at 110C for 6 hours which precipitated more intermetallic compound. 4.1.4 Composition of the Alloy Powders The composition of the alloy powders was ev aluated on mechanically polished samples after electrodeposition. An illustra tion of a polished particle is presented in Figure 4-9(a). The corresponding EDS map of Al in th e particles is illustrated in Fi gure 4-9(b). It can be noticed from the EDS map that the root of the particle wa s rich in Al when compar ed to the remainder of the particle. The precise at% of Al in the powder was anal yzed using the EPMA technique. At least 6 particles and 10-15 points on each particle were considered for EPMA composition analysis. Amount of Al at multiple points on the particle was measured and the average value along with standard deviation is calculated. The Mg-Al alloy powders fabricated in th is study can be divided into 2 categories depending on the type of anode used during the electrodeposition process. A 100% Mg sheet is used for the production of Mg-Al powders with low contents of Al in hcp-Mg phase. The compositional analysis of the electrodeposited powder on hcp-Mg phase from these particles indicated the presence of 4 0.4 at%Al. A par ticle used for EPMA compositional analysis is shown in Figure 4-10. The numbers indicated on the powder represent the %Al at that particular location.

PAGE 78

78 For fabricating a powder with relatively highe r amount of Al in hcp-Mg, an anode with 80%area of Mg and 20%area of Al was employed during the electrodeposition process. The EPMA compositional analysis of the powder fabric ated from this alloy indicated that 8 0.1 at%Al was observed in hcp-Mg phase. A SEM/BSE image of a dendritic pa rticle illustrating the different amounts of Al from this experiment is shown in Figure 4-11. The different phases in the alloy powders were identified by using the backscattered imaging. The SEM/BSE images from an Mg-8at %Al powder are presented in Figure 4-12. The root of the dendrite appears bright (higher average atomic number) and its compositional analysis indicated the presence of 18at%Mg. The grey contrast encompassed the whole dendrite. A relatively uniform distribution of Al was found in most dendrit es. The formation of the fcc-Al phase at the root of the Mg-rich dendrites ha s been observed previously and shown to be associated with the high nucleati on barrier required for the forma tion of hcp-Mg on the graphite substrate [34]. The XRD results of both the po wders have indicated the presence of the intermetallic compound Mg17Al12 inside the hcp-Mg phase. The contrast or the composition corresponding to the intermetallic phase is no t observed during the SEM analysis. The SEM analysis suggest that this intermetallic compound is present as very fine phase inside the hcp-Mg phase and could not be observed in the powder particles even at high magnifications due to low resolution of BSE imaging. Because of the low volume fraction of the f cc-Al phase, the Al content of the hcp-phase was found to be similar to the bulk composition of the alloy powders as confirmed by the ICP technique. For example, the apparent average Al content of 8.1 0.2 at% of the hcp-Mg phase as evaluated by EPMA for deposit 2 was very close to the bulk composition of 8.7 at% Al as measured by the ICP technique. The compositi on of the powders fabricated using the

PAGE 79

79 electrodeposition is very sensitive to the initial conditions of the electrodes and raw materials. It was observed that in one of electrodeposited Mg -Al powder fabricated using the 80% Mg and 20%Al anodes, the composition of the powder as analyzed by EPMA was about to 10at% Al in hcp-Mg phase. This deviation in the composition can be attributed to the conditions of the electrode during the deposition. 4.1.5 Microstructural Characteriza tion of Mg-Al Alloy Powders The microstructure of the powder after co ating with Ni was evaluated using the transmission electron microscopy technique. Th e TEM samples were prepared by embedding the powders in the polymer and cutting thin slices using the ultramicrotome. The bright field and dark field TEM micrographs of the Mg-8at%Al alloy powder are presente d in Figure 4-13(a) & (b) respectively. The corresponding diffraction patte rn of the TEM micrographs is presented in Figure 4-13(c). The diffraction s pot corresponding to the plane (002) was used to image the micrograph in dark field mode. In the diffraction pattern other than the hcp-Mg phase, diffused rings corresponding to the MgO were identified and indexed (Figure 4-13(c)). It is believed that the MgO was formed during the sample preparation of TEM as it involved cutting the slices using ultramicrotome and fishing the 50 nm slices from the water in the boat of microtome. The average grain size of the Mg -8at%Al alloy powder calculate d from multiple images was measured to be 44 5 nm. Since the electrodepo sition conditions were same in all the powders fabricated, similar grain sizes were anticipat ed in all the electrode posited powders. The nonuniformity of the diffraction spots in the ring patterns and their cluste ring suggested a strong microtexture in the Mg(Al) alloy powders. This behavior is believed to be associated with the sub-micrometer size faceted units shown in Figure 4-1(d). The intermetallic phase could not be detected by the conventional TEM techniques probabl y owing to their small size. Several slices from different samples were cut using the ultram icrotome and were analyzed in the TEM. Figure

PAGE 80

80 4-14(a) shows the TEM micrograph imaged from a different sample. The microstructure indicated that the grains were s lightly elongated which also repres ented that these materials were highly textured. The corresponding da rk field micrograph is shown in the Figure 4-14(b). Figure 4-15 shows the microstructural analysis carried out in the HRTEM on Mg -8at%Al alloy powder. The d-spacing measured from th e micrograph shown in Figure 4-15 was close to that of (0001)Mg basal plane and no phases corresponding to intermetallic were identified in the sample. One of the problems aroused during the TEM analysis was the samples used to peel off from the polymer when the beam was focused on the sample. The structure of the intermetallic compound is complex and the lattice parameter is very high (about 10.56 ). Therefore, th e diffraction spots from this cr ystal form very close to the transmitted beam due to large d-spacing involved and were unable to get resolved during the TEM analysis. Furthermore, previous studies on bulk Mg-9at%Al alloys illustrated that the Mg17Al12 phase was formed as fine precipitates in the order of 100 nm size laths after carrying out annealing treatment for 8 hour s at 200C [69]. However, the Mg-Al alloys employed in this study were in powder form with nano size gr ains and were processed at much lower temperatures. Therefore, the intermetallic phase formed in the Mg-Al powder fabricated using electrodeposition was expected to be very fine and was unable to get detected even at high magnifications. 4.2 Characteristics of Pure Mg Powder To compare the effect of Al addition to Mg on hydrogenation properties, pure Mg powder from Alfa aesar was bought and tested unde r the same hydrogenation conditions. The pure magnesium powder exhibited a wide range of particle sizes and shap es as seen in Figure 4-16. The mean particle size of the pure magnesi um powder based on the volume fraction was measured to be 43 m. These powders were also coated with Ni using the same procedure

PAGE 81

81 mentioned in Chapter 3. The XRD profile after th e coating of Mg with Ni is shown in Figure 417. Other than the peaks corresponding to hcp-Mg phase, the peaks corresponding to the internal standard tungsten and the catalys t Ni phases were observed in th e powder. The evaluation of the microstructure in TEM indicated that the grain size of this powder was in micrometer regime. The TEM samples were prepared using the ultr amicrotomy, similar to that of Mg-Al alloy powder. The bright field and dark field TEM micrographs of th e pure Mg powder after coating with Ni are shown in Figure 4-18 (a) & (b) respectively. The average grain size of the particles was about 1.5.3 m. The diffraction pattern from the grai ns presented in Figure 4-18(c) shown a spot pattern with the spots corresponding to the hcp Mg. 4.3 Summary and Conclusions Hcp-rich Mg-Al alloy powders were fabr icated with 4, 8 an d 10 at%Al using the electrodeposition technique. The el ectrodeposition parameters were chosen such that there was extensive growth of dendrites and the grain size of the pow ders fabricated was in the nanoregime. Ni was added as catalyst to prom ote the hydrogen absorption and desorption. The Ni coating procedure rendered particles with di fferent amounts of Ni on the surface. The primary phases present in these powders were hcp-Mg, the intermetallic compound Mg17Al12 and small amounts of fcc-Al. The compositional analysis indicat ed that the fcc-Al was found at the root of the dendrite while the hcp-Mg was formed on top of fcc-Al. The intermetallic compound observed in the powders was formed due to th e insitu annealing of the powder during the electrodeposition and its content in creased after the Ni coating. The microstructure of the electrodeposited powders after Ni co ating revealed the presence of nanocrystalline grains in the hcp-Mg phase. However, the presence and distri bution of the intermetallic compound in hcp-Mg was not observed in the microstructural analysis due to their fine size.

PAGE 82

82 1 mm a 10 m d 50 m c 500 m b 1 mm a 10 m d 50 m c 500 m b Figure 4-1. SEM micrographs revealing th e dominant morphology of powders present in electrodeposited hcp-rich Mg particles, (a) low magnification micrograph illustrating the particle size, (b) the dendritic nature of the electrodeposited Mg-Al particles, (c) higher magnification micrograph showing the branches of the dendrite, (d) finer structural units present in the branches of the powders. 200 m a 50 m Different morphology at the root b 200 m a 200 m 200 m a 50 m Different morphology at the root b 50 m 50 m Different morphology at the root b Figure 4-2. (a) SEM micrographs of the Mg-A l particles showing different morphology at the root of the dendrite. (b) Higher magnifi cation image showing the different globular morphology at the root of dendrite.

PAGE 83

83 50 m a 50 m a 10 m b 10 m b Figure 4-3. (a) SEM micrograph of the mo rphology present in lower amounts. (b) Higher magnification image showing the glo bular shape of the branches. 200 m 200 m Figure 4-4. SEM micrographs i llustrating the breakage of electro deposited Mg-Al particles after coating with Ni.

PAGE 84

84 5 m a b 5 m a 5 m a b b Figure 4-5. (a) Higher magnifi cation SEM image of a Ni-coate d Mg-Al particle, (b) the EDS Map of Ni for the micrograph shown in (a). Figure 4-6. Energy Dispersove Spectra of Ni in two Mg-Al particles dem onstrating the differences in Ni content.

PAGE 85

85 3035404550 2 theta (deg)Intensity (AU).hcp-MgMg17Al12 fcc-AlW Mg-4at%Al Mg-8at%Al Mg-10at%Al 3035404550 2 theta (deg)Intensity (AU).hcp-MgMg17Al12 fcc-AlW Mg-4at%Al Mg-8at%Al Mg-10at%Al Figure 4-7. XRD profiles of electrodeposit ed Mg-4at%Al, Mg-8at%Al and Mg-10at%Al powders after Ni coating illustrating the various phases present in the material before hydrogenation.

PAGE 86

86 30 40 50 60 2 theta (deg)Intensity (AU)hcp-MgMg17Al12 fcc-Al 15 minutes 45 minutes After Ni coating 30 40 50 60 2 theta (deg)Intensity (AU)hcp-MgMg17Al12 fcc-Al 15 minutes 45 minutes After Ni coating Figure 4-8. XRD profiles of Mg-8at%Al alloy po wders deposited for different time intervals and after Ni-coating procedure showing the varia tion of intermetallic content at each stage 100 m a b 100 m a 100 m a b b Figure 4-9. (a) SEM Microgra ph of a Mg-8at%Al particle a nd (b) the corres ponding EDS map of Al depicting its dist ribution in the particle.

PAGE 87

87 50 m 3.6 4.2 4.1 3.5 3.4 4.3 4.2 %Al 50 m 3.6 4.2 4.1 3.5 3.4 4.3 4.2 %Al Figure 4-10. EPMA composition analysis of a Mg-Al powder fabricated using the 100%Mg sheet as anode demonstrating the Al distribution in the powder. 7.8 100 m 8.1 8.2 7.6 7.8 82.4 %Al 7.8 100 m 8.1 8.2 7.6 7.8 82.4 %Al Figure 4-11. EPMA composition an alysis of a Mg-Al powder fabricated using the 80%Mg + 20%Mg sheet as anode along with the measured Al content at various points on the powder.

PAGE 88

88 50 m fcc-Al hcp-Mg + Mg17Al12 50 m fcc-Al hcp-Mg + Mg17Al12 Figure 4-12. A SEM/BSE mi crograph of Mg-Al particle illustra ting the distribution of different phases in the Mg-8 at%Al powder. 200 nm a 200 nm b Mg{10-10} Mg{0002} Mg{10-11} MgO {111} c 200 nm a 200 nm a 200 nm b 200 nm b Mg{10-10} Mg{0002} Mg{10-11} MgO {111} c Mg{10-10} Mg{0002} Mg{10-11} MgO {111} c Figure 4-13. (a) Bright field and (b) dark fi eld TEM micrographs of the Mg-8at%Al powder revealing the grain size of the material. (c ) The selected area diffraction pattern from the region shown in (a).

PAGE 89

89 0.5 m 0.5 m a b 0.5 m 0.5 m a b Figure 4-14. TEM Micrographs of Mg-8at%Al all oy powder revealing grain size in (a) Bright field, and (b) Dark field after the coating with Ni. 10 nm 10 nm Figure 4-15. Bright-field mi crograph of an Mg-8at%Al allo y powder at high magnifications using HRTEM.

PAGE 90

90 200 m 200 m Figure 4-16. SEM micrograph of Pure Mg powder illustrating the morphologies and particle sizes. 253035404550 2 theta (deg)Intensity (AU).hcp-MgW Ni 253035404550 2 theta (deg)Intensity (AU).hcp-MgW Ni Figure 4-17. XRD profile of pur e Mg powder after Ni-coating.

PAGE 91

91 a 0.5 m b 0.5 m cB= 2110 0112 0122 1110 a 0.5 m a 0.5 m b 0.5 m b 0.5 m c cB= 2110 0112 0122 1110 Figure 4-18. (a) Bright field and (b) dark field TEM micrographs illustrating the microstructure of the pure Mg powder. (c) The selected area diffraction pattern from the region shown in (a).

PAGE 92

92 CHAPTER 5 MICROSTRUCTURAL EVOLUTION DURING PR ESSURE COMPOSITION ISOTHERMS One of the methods to evaluate hydrogena tion and dehydrogenation processes is to establish the pressure-composition isotherms. These isotherms are developed under equilibrium conditions. As explained in Secti on 2.4, PCT tests are conducted at relatively high temperatures and sufficient time is given at each step to obtain equilibrium conditions. These isotherms are usually carried out to identif y the thermodynamic properties of enthalpy and entropy of hydride formation and decomposition. The major limitation in using magnesium hydrid e as a hydrogen storage system is its high enthalpy of formation [8]. Theoretical calculations have reported that the enthalpy of formation was reduced from 76 to 28 kJ/mol with Al inco rporation in magnesium hydride [18]. Therefore, there has been significant interest to investigate the effect of Al on the stability of magnesium hydride. Depending on the composition of the all oy several phases may exist in the starting microstructure, namely hcp-Mg, Mg17Al12, and Al3Mg2 and the equilibrium pl ateau pressure of hydride formation from these phases is different. For example, the magnesium hydride formation from Mg17Al12 phase was observed to take place around 0.8-0.9 MPa pressure of hydrogen at 350C [28]. In case of the Al3Mg2 phase, at the same temperat ure, the equilibrium plateau pressure of hydride formation was reported to be significantly higher (1.4 MPa) [28]. An equilibrium plateau pressure of hydride formatio n of about 1.6 MPa had been observed for an Mg-10at%Al at 400C [28]. Most of the PCT studies reported for Mg ba sed alloy powders focused on measuring the enthalpy and entropy of formation/dissociation of metal hydride [43, 49, 109]. However, no direct Vant Hoff plot had been developed for a given Mg-Al alloy. But the sparsely available equilibrium plateau pressures for various Mg-Al alloy compositions have been collected and the

PAGE 93

93 Vnt Hoff plot was developed as was shown in Chapter 2 (See Figure 2-10) [80]. This curve indicates that the enthalpy of formation/dissociation of MgH2 does not change significantly by the addition of aluminum. It shoul d be noted that the equilibriu m plateau pressure data were compiled from different experiments and a consiste nt set of data for a particular Mg-Al alloy powder has not been developed so far. Since the PCT curves in Mg-Al alloy powders with more than one phase are complex, it is necessary to investigate the e volution of different phases to measure the thermodynamic properties precisely. One of the aims of this study has been to ev aluate the enthalpy a nd entropy of magnesium hydride formation in Al-rich hcp-Mg solid soluti ons. A systematic approach was considered to evaluate the phase transformati ons during the development of the PCT curves for Mg-Al alloy powders. In addition, PCT curves for a commer cially available pure Mg powder were also developed to understand the effect of Al add ition to hcp-Mg. An Mg-10at%Al alloy powder was employed to investigate the phase transformations that occur during the development of a PCT curve at 350C. Using the knowledge about the ph ase transformations in Mg-10at%Al powder, the effect of temperature, and Al content in hcp-Mg are demonstrated by establishing PCT curves for Mg-8at%Al and Mg-4 at%Al alloy powders. Finally, the effect of Al addition on enthalpy and entropy of formation/dissociation of magnesium hydride are evaluated using the Vant hoff plots developed from these PCT curves. 5.1 PCT Curves for Pure Mg Powders The pressure-composition isotherms for the co mmercial pure Mg powder, obtained within 275-350C temperature range, are illustrated in Figur e 5-1(a). Figure 5-1(b) represents the plot with a logarithmic scale for the pressure axis for additional clarity. The pressure of hydrogen during the initial stages of the experiment was increased by 0.025 MPa at each step and the step size was increased to 0.05, 0.1, 0.25 MPa at higher pressures. The small step size used during the

PAGE 94

94 initial stages was to provide e nough data points to observe the so lubility of hydrogen in the hcpMg phase. To achieve equilibrium conditions duri ng the experiment, a time period of 30 minutes was provided at each step. The total time to develop each PCT curve was approximately 46 hours. The maximum solubility of hydrogen in the Mg po wder at different temperatures tested is provided in Table 5-1, which eluc idates an increase in the solub ility limit with an increase in temperature. Furthermore, at a constant pressure the solubility of hydroge n at higher temperature is observed to increase with temperature. It has to be recognized that two methods have been used in the litera ture to identify the solubility of hydrogen in magnesium. One method (method-1) involves the volumetric measurements of hydrogen absorption in Mg-H system at a constant pressure and temperature. The experiments using this method were genera lly conducted at low pressures (P<1.013 bar) to avoid hydride formation [12]. The other method (method-2) evaluates the pressure composition isotherms at high pressures and hi gh temperatures [11]. The solubil ity calculated in the present study is based on method-2. The solubility data reported vari ed significantly based on the method of determination. The results obtained us ing method-2 for hydrogen solubility are always one order higher than that of method-1. For exam ple, the solubility of hydrogen was calculated to be 0.001 wt% at 450C at 0.4 MPa using method-1 while about 0.02 wt% of hydrogen was reported to be soluble in Mg at 440C and 0. 45MPa when measured with method-2 [11, 12, 110]. From Table 5-1 it can be noticed that the solubility of hydrogen in pure Mg from this study were a lot higher (0.02wt% at 440C to 0.09wt% at 350C) than the reported data using both the methods. The higher solubility values in this study can be attributed to the experimental

PAGE 95

95 approach of the PCT curve, as the solubility of hydrogen in this method was determined near the three-phase equilibria of hcp-Mg, H2 gas and MgH2. The solubility values represent the maximum hydrogen (saturated) solubility in hcp-Mg at that temperature and pressure. Furthermore, the solubility data reported in the literature was for bulk Mg, while the Mg considered in this study is in powder form The higher surface ar ea of the powder also contributes to the higher solubil ity value as the solubility of hydrogen in magnesium involves a gas-solid reaction which increase s with increase in surface area. When maximum solubility of hydrogen is reached, a plateau region follows, which represents the formation of MgH2 (Figure 5-1). The equilibrium plateau pressures of hydride formation and dissociation at different temperatures are presented in Table 5-1. Decreasing temperature resulted in a reduc tion of the equilibrium plateau pressure from 0.48 MPa at 350C to 0.08 MPa at 275C. The plateau pressure of 0.67 MPa during absorption was reported for unmilled Mg powder at 350C which is close to the value observed in this study [98]. However, MgH2 powders fabricated by reactive mechanical milling showed higher plateau pressures than the pure Mg powders [2]. Consistent with previously reported data [2], the PCT curves exhibited hysteresis, i.e., the plateau pr essures of absorption and desorp tion at a given temperature are different. The difference in the equilibrium plat eau pressure between absorption and desorption at 350C for the unmilled Mg powder was reported as 0.25 MPa which is slightly higher (0.22 MPa) than for pure Mg powder used in this stud y [98]. The hysteresis observed in this study is due to the difficulty in nucleation of hc p-Mg during desorption of hydrogen from -MgH2. As the temperature of the PCT is reduced the hyste resis in pure Mg powde r also reduced and the total amount of hydride is also lower. This observation suggests that the particles were hydrogenated completely to the theoretical cap acity during absorption and upon desorption the

PAGE 96

96 nucleation of hcp-Mg is hindered at equilibrium plateau pressure. However, further decrease in pressure helps in breaking the Mg-H bond and nucleates hcp-Mg. At the end of plateau region the pressure increases steeply as a function of the absorbed hydrogen. The length of the plat eau region was found to decrease with decreasing the test temperature. At 350C the total amount of hydrogen absorbed was close to the theoretical capacity of 7.6 wt% and it decreased to as low as 6.4 wt% at 275C. These results are consistent with the reported values in th e literature, where a complete transformation of hcp-Mg into MgH2 was observed in the unmilled and milled Mg powders at 350C and 10MPa pressure [98]. Furthermore, the hydrogen content previously re ported was ~ 6.8wt% at 300C for mechanically milled hcp-Mg under high pressure [2]. Several factors may contribute to the apparent loss of capacity with reducing temperature. The transformation of hcp-Mg to -MgH2 involves a significant change in volume (approximately 32%). At lower temperatures the st rain created due to hydride formation is not relieved and further transformation of hcp-Mg into magnesium hydride is inhibited[75]. In addition, the nucleation of magnesi um hydride is fast while the grow th is slow at relatively low temperatures like 275C. These conditions of hydride transformation forms a layer of MgH2 around the surface and further hydrogenation of hcp-Mg is restricted as hydrogen has to diffuse through the hydride layer which is very slow. Therefore, the tota l amount of hydrogen stored is less at low temperatures. From Figure 5-1(b), it can be no ticed that the dehydrogenation part of the curve is not complete. i.e., not all hydrogen is released. This is due to the lack of enough assigned data points towards the end of the experiment. Due to the sl ow kinetics at low temperatures the total number of steps required to complete the dehydrogenation process is high. However, when the PCT

PAGE 97

97 system moves to the next experiment, the reactio n chamber is evacuated till the pressure in the system reaches a minimum value as mentioned in Chapter 3. Even though the data is not acquired during this step, the sample is allowed to dehydrogenate completely before the start of next experiment. 5.2 PCT Curves for Electrodeposited Mg-Al Alloy Powders Three electrodeposited Mg-Al al loy powders with 4at%, 8at% and 10at% Al content were characterized in this study. The Mg-10at%Al alloy powder was used to identify the phase evolution during the development of PCT curve at 350C. This goal was achieved by partially hydrogenating or dehydrogenating along the path of the PCT curve at 350C. Initially a complete absorption part of the PCT was carried out on a sample and based on the obtained curve new tests were designed, where the te st was stopped at vari ous points along the curve. The absorption part of the Mg-Al powder during PCT was c onducted to design further experiments for identifying the phase transformations. The samp les from each experiment were evaluated for phases and microstructures. The details of th e experiments are described in the following sections. 5.2.1 Microstructural Evolutio n during PCT Test at 350C The PCT curve for Mg-10at%Al alloy powder at 350C is shown in Figure 5-2. A comparison of PCT curves for pure Mg powder and Mg-10at%Al powder (Figure 5-3) depicts that the PCT curve of Mg-Al powder exhibits a continuous increase in slope during hydrogen absorption in contrast to relatively flat regi on for pure Mg powder. The solubility limit of hydrogen near the saturation point of the PCT cu rve in Mg-10at%Al was noticed to be lower than that of pure Mg powder. For the Mg-Al powd er the hydride formation demonstrated a short plateau followed by a sloping curv e rather than a flat region as was observed for the pure Mg powder. Furthermore, the PCT curve indicates a second plateau region at a higher pressure. To

PAGE 98

98 understand the reasons behind the differences, the PCT curve for the electrodeposited Mg-10at% Al alloy powder was divided into several stages as shown in Figure 5-2 a nd separate experiments were carried out to evaluate the phase transforma tions in these stages. This division was based on the change in the transformation reaction, which was predicted by plotting the slope of the PCT curve versus the hydrogen wt% for both the hyd rogenation and dehydrogen ation parts of the PCT experiment as shown in Figures 5-4(a) and (b). A significant change in the slope of PCT curve indicates a change in phase transformati on or a combination of phase transformations. From the Figure 5-4(a) it can be noticed that during the hydrogenation process the first significant change in slope occurs at 0.01 wt% of hydrogen. The absorption up to solubility limit is defined as stage 1. Immediately after stage 1, the slope of the PCT curve drops to almost zero (flat region), which indicates the formation of a hydride phase. This region is defined as stage 2 and the slope of the curve in Figure 5-4(a) re mained same till the hydrogen concentration in MgAl alloy powders reaches approximately 2.5 wt%. After this stage the slope increases till the hydrogen content of ~ 4.0 wt% (stage 3). Further in crease of the pressure resulted in a drop of the slope of the PCT but does not reach zero. Th is region corresponds to the stage 4 of the PCT curve. Beyond this stage, with increase in pressure, no significant absorption of hydrogen was observed (stage 5). The change in slope during desorption part of the PCT curve is shown in Figure 5-4(b). The slope of the desorption part of the curve is very high initially till the hydrogen wt% in the powder reached 4.6wt% where its value drops clos e to zero. The region ti ll the hydrogen wt% in the powder is 4.6wt% is defined as stage 6 while the region where the slope is almost equal to zero is defined as stage 7 as indicated on the Figure 5-4(b).

PAGE 99

99 Several experiments were conducted, where the powders were pressurized to various points along the PCT curve for the Mg-10at%Al powder and were analyzed for their phases and microstructure. These samples are identified in Table 5-2. Quantitative metallography was employed to calculate the phase fractions in th e alloy powder at the e nd of each experiment. After each experiment the samples were cooled u nder hydrogen pressure. It should be noted that the cooling of the sample takes time and during th is time period it is anticipated that slightly more amount of hydrogen is absorbed into the material. Hence, the phases that are analyzed do not correspond exactly to the points on the PC T curve, but they will be close enough for evaluating the transformations occurring during th e PCT curve. Similar to that of pure Mg the pressure of hydrogen during these experiments was increased by 0.025 MPa at each step initially and then increased to 0.05, 0.1, 0.25 MPa at hi gher pressures. However, the time provided for reaching equilibrium at each step was selected to be longer (60 minutes) than the experiments conducted for pure Mg in-order to achieve true e quilibrium as our previous experience on similar powders indicated that complete equilibrium could not be achieved in 30 minutes. As described in Chapter 4 the phases present in the initial powder (sample S1) before the PCT experiment were hcp-Mg solid solution wi th small amounts of the intermetallic compound Mg17Al12 and fcc-Al phases (Figure 4-7). Figure 5-5 presents the results of characterization of the Mg-10%Al powder after 4.2w t% hydrogen absorption (sample S2), which correspond to the middle of stage 4 (see Figure 5-2). The resul ting PCT curve from this experiment is shown Figure 5-5(a). The XRD profile of sample S2 is presented in Figur e 5-5(b). A comparison of this profile with that of sample S1 reveals that significant amount of -MgH2 has formed in sample S2. In addition, the relative intensities of the peaks corresponding to the intermetallic compound Mg17Al12, and fcc-Al were reduced. A new peak corresponding to the Al3Mg2 intermetallic

PAGE 100

100 compound was also observed in the XR D profile of sample S2. It ha s to be noted that the peak corresponding to the hcp-Mg was not present, indicating its complete transformation into magnesium hydride and incorporatio n in other phase. A comparison of the relative intensities of Mg17Al12 reveal that part of the intermetallic compound Mg17Al12 was transformed into MgH2 as the relative intensity was decreased. In addi tion, the peak corresponding to the fcc-Al phase present in the sample S1 (See Figure 4-7) wa s not observed in sample S2. This observation suggest that fcc-Al phase absorbs more Mg at 35 0C (solubility is higher) and upon cooling it is precipitated as part of Al3Mg2 at room temperature. Representative SEM/BSE micrographs of sample S2 are shown in Figure 5-5(c) and (d). The phases were identified by thei r composition analysis (Table 5-3) and are labeled on the micrographs. Consistent with the XRD results, th e grey contrast corresponding to hcp-Mg phase, which was observed in sample S1, was not found in these particles. In general, two types of particles were observed in sample S2 based on the amount of hydride. Some particles were hydrogenated to a signifi cant extent while the others were partially hydrogenated. The dark phase, which corresponds to the MgH2 phase, formed like a network inside the powder in some particles (Figure 5-5(c)) while it was restricted only near the su rface in other particles as shown in Figure 5-5(d). The fcc-Al phase observed in these particles was disc onnected as shown in Figure 5-5(c). From the microstructural analysis it can be concluded that Al was rejected from hcp-Mg phase during the hydride fo rmation and precipitated as fccAl phase. Similar theories have been proposed in the litera ture that during the hydrogenation of Mg-Al alloys [28, 80], the Mg-Al alloy powder dis-proportionates during hy drogen absorption into fcc-Al and hcp-Mg phases. The hcp-Mg phase formed during this reaction absorbed more hydrogen to form the -

PAGE 101

101 MgH2 while the fcc-Al does not transform [26, 28, 80, 97, 111]. This observation is attributed to the low solubility of Al in MgH2 [112]. Figure 5-6 presents high magni fication SEM/BSE micrographs of sample S2. It appears that MgH2 nucleates at specific locations on the surf ace and grows into particle as indicated by the white dashed lines on the micrograph. The va rious nucleation points on the surface can be the triple junction of catalyst, hcp-Mg and hydroge n or any defects on the surface. In addition, cracks were observed in MgH2 grains due to the stress in hydride because of the excessive volume expansion during the hydride formation. Higher magnification image of a particle from sample S2 (Figure 5-7), which is hydrogenated to a significant extent, illustrated that the MgH2 was under significant stress and hence extensive cracking was observed in the hydride grains. The fractions of different phases for sample S2 are presented in Table 5-4. As was mentioned previously two types of particles were found in this sample (See Figure 5-5). In Table 5-4, the values in the first row correspond to the particles that were hydrog enated to a significant extent (Figure 5-5(c)) while the second row represents for particles that were partially hydrogenated (Figure 5-5(d)). The phase fraction analysis illustrated that during hydrogenation all the particles did not absorb hydrogen equally. The differences in the absorption behavior of the particles can be attributed to the catalyst distribution, and the ther mal and elastic stresses created during the transformations. Analytical model developed for hydrogenation of metal particles also predicts that at any given instance, the fraction of metal transformed into hydride in the particles is not same for all the particles but it follows a log-normal distribution [2, 76]. The origin of the two kinds of particles can be attributed to the different sizes of the particles or the distribution of catalyst on the surf ace of the particle. As noticed from the results in the present study, stress was developed in the particles during the hydrogen ation and inhibited the hydride

PAGE 102

102 growth. The amount of stress in th e particle was dependent on the size of the particle and hence the total hydride formed in each particle was affected. Sample S3 was hydrogenated to the maximum pressure (9MPa) that was possible in the PCT system to understand the phase transformati ons after the stage 4 a nd to characterize the powders after the completion of hydrogen absorption part in the PCT experiment. The slope of the PCT curve rose significantly as is evident in Figure 5-4(a) after the stage 4. The PCT curve and the XRD profile of the sample S3 are presented in Figure 5-8(a) and (b). The -MgH2 phase along with the fcc-Al phases were observed in this sample. Very small peak corresponding to the intermetallic compound Al3Mg2 was also present in the XRD profile. By comparing the XRD profile of sample S3 with S1 and S2, it can be concluded that the intermetallic compound Mg17Al12 was hydrogenated completely to form the Al3Mg2 and this intermetallic further hydrogenated to form MgH2 and fcc-Al. The microstructure of sample S3 as observed in SEM/BSE mode is shown in Figure 5-8(c) and (d). Other than the -MgH2 phase, the only other contrast (bright) that was observed correspond to fcc-Al as indicated by the compositional evaluation (Table 5-3). The microstructure of the particles in this sample was noticed to be similar to that of the particles that are completely hydrogenated in sample S2. The MgH2 in sample S3 formed as a network all around and inside the particle. As demonstrated in Figure 5-8(c), f cc-Al phase was present inside the particle and a sign ificant amount in chunks near the r oot. The amount of Al in the hydride was about 0.8at% owing to its lo w solubility and confirming that Al was being rejected during the hydrogenation of the hcp-Mg. A higher magni fication image of the hydrogenated powder is shown in Figure 5-8(d). Similar to that of sample S2 extensive cracking was also observed in this sample.

PAGE 103

103 The phase fractions found in sample S3 indica ted that the hydride content was increased and the amount of fcc-Al phase was present similar to that of the particles that were completely hydrogenated in sample S2. Even though a sma ll peak corresponding to the intermetallic compound Al3Mg2 was seen in the XRD profiles of sample S3, the contrast corresponding to that phase was not detected in the SEM/BSE images because of its very low volume fraction. From the above results it can be concluded that the intermetallic compounds of Mg17Al12 and Al3Mg2 hydrogenate at higher pressures and form the -MgH2 and fcc-Al. Stage 6 corresponds to the initial change in sl ope during the start of desorption as shown in Figure 5-4(b). With the drop in pressure, th e concentration of hydrogen was not reduced significantly until the pressure of the system reac hed 2 MPa. Hence no samples were analyzed in this stage as the phases would be sim ilar to that of the hydrogenated powder. Stage 7, which was defined as the release of hydrogen from the magnesium hydride, was studied to understand the evoluti on of different phases during the desorption process. Sample S4 was produced by dehydrogenating the powder partia lly till the hydrogen c ontent in the powder was about 3 wt%. The PCT curve for this sample is shown in Figure 5-9(a). After cooling the sample S4 to room temperature, the XRD profile (Figure 5-9(b)) indicated the presence of the hcp-Mg, Al12Mg17 phases in the powder. Other than these phases, peaks of remaining -MgH2 were also observed in the XRD profile. Furtherm ore, the fcc-Al phase present in sample S3 was not observed in this sample. These results suggest that upon desorption of hydrogen from the MgH2 phase, hcp-Mg was formed and it reacted with the fcc-Al present in the powder to form the intermetallic compound Mg17Al12. Representative SEM/BSE micrographs of sample S4 are presented in Figures 5-9(c) and (d). Similar to sample S2, two different particle s were observed in sample S4, namely, particles

PAGE 104

104 with low amount of hydrogen released (Figure 5-9(c)) and particles with significant amount of hydrogen released (Figure 5-9(d)). In addition, two other contrasts were observed in the SEM/BSE micrographs. The compositional analysis on these contrasts revealed that the grey contrast corresponds to the hcp-Mg phase while th e bright contrast corresponds to that of the intermetallic compound Mg17Al12. These phases are labeled on the micrographs presented in Figure 5-9. SEM/BSE micrographs of sample S4 at higher magnifi cations is presented in Figure 5-10. Detailed analysis of the microstructural features reveal ed that desorption of hydrogen started at the surface of the particle. As hydrogen was released from the powder, the hcp-Mg nucleated and reacted with the fcc-Al to form the Mg17Al12 as shown in Figure 5-10(a) near the surface. Furthermore, it can be seen in Figure 5-9( d) that the bright and grey phases are always connected to the surface with no di sconnected bright phase inside the dark phase, as was seen in sample S3, were identified. Therefore, it can be concluded that desorp tion of hydrogen started on the surface by the nucleation of th e hcp-Mg, which grew into the pa rticle. During this process the nucleated hcp-Mg phase dissolved Al until it was saturated and formed the intermetallic compound Mg17Al12. It is to be noted th at during the cooling of the sample S4, hydrogen absorption took place as the sample was cooled unde r high pressure. This pressure was sufficient enough to form the MgH2 at lower temperatures as the eq uilibrium pressure of hydrogenation was lower at these temperatures. The newly form ed hydride during cooling rejected the Al back into Mg and contributed to the formation of Mg17Al12. The volume fractions of different phases in sample S4 is presented in Table 5-4. These values suggest that the volume fractions of the hcp-Mg and the intermetallic compound Mg17Al12 are higher in some particles (F igure 5-9(d)) when compared to the other particles. Based on the microstructures and the volume fraction cal culations it can be suggested that the

PAGE 105

105 dehydrogenation takes place to a different extent in various particles of the powder which is similar to the observation made for hydrogenation. Sample S5 represents the powder that has undergone a complete hydrogenation and dehydrogenation cycle. The XRD profile (Figure 511(a)) of this powder illustrates the presence of hcp-Mg and the intermetallic Mg17Al12 phases which were similar to that of initial powders. It is to be noted that the peak corresponding to the fcc-Al phase was absent. Th e distribution of the phases present in the sample was analyzed by SEM and the corresponding micrographs are presented in Figure 5-11(b). The compositional analysis of the powders (Table 5-3) indicated that the bright phase in the microstructure corresponds to the in termetallic compound Mg17Al12 phase, while the grey phase was hcp-Mg phase. A comparison of this sample with the sample S1 represented that the intermetallic compound Mg17Al12 had grown after 1 complete cycle of hydrogen absorption/desorption at 350C while th e XRD analysis suggested that the volume fraction of the intermetallic increased slightly In sample S1, as shown in Figure 4-6, the intermetallic compound was present as fine precip itates in the hcp-Mg, while in sample S5 the size of the intermetallic preci pitates was as large as 10 m. In addition, the fcc-Al phase, which was present in the initial powder, was not observed in sample S5. This shows that that the high temperature and longer times invol ved during the desorption stage of the PCT curve, dissolves the Al in hcp-Mg and forms the intermetallic compound Mg17Al12 upon cooling. The phase fractions calculated revealed that about 80% of the microstructure consisted of the hcp-Mg phase, which is lower than the initial content. The lower content of hcp-Mg in sample S5 could be due to the transformati on of the hcp-Mg into Mg17Al12 by absorbing the Al which was present at the root of the initial powder.

PAGE 106

1065.2.2 PCT Curves for Mg-8at%Al Powders at Different Temperatures Figure 5-12 illustrates the pressure-composition isotherms developed for the Mg-8at%Al alloy powders. The first PCT curves ware deve loped sequentially at 350C, 365C, 380C and 395C. The solubility of hydrogen in hcp-Mg is pr esented in Table 5-5. Addition of Al to Mg reduced the solubility of hydrogen in the hcp-Mg phase significantly. By comparing the data in Tables 5-1 and 5-5, it can be noticed that the solubility of hydrogen is much higher in pure Mg even at lower temperatures. This phenomenon was al so observed previously and is attributed to the reduction in unit cell volume of hcp-Mg with addition of Al [113]. The equilibrium plateau pressures of magne sium hydride formation and decomposition at various temperatures are listed in Table 5-5. The plateau pressure s indicated in the table are for the formation/dissociation of hydride from hcpMg. These values are obtained from the mid plateau regions of the most fl at part of PCT curve. Increasing temperature rendered a higher equilibrium pressure of hydride formation. Furthe rmore, the extent of the plateau region, which corresponds to the hydride formati on, also increased with temperature, but at a given temperature the extent was reduced significantly when comp ared to the results for pure Mg due to the addition of Al. Similar to the Mg-10at%Al a lloy powders, rather than a sharp bend in the isotherm, the slope of the pre ssure-composition curves varied gradually with composition. This behavior was found to be temperature dependent such that the steepest change in slope was found at the highest temperature. In addition, at all temperatures it was observed that during the dehydrogenation part of the PCT curve the hydrogen content was initially increased and then decreased with a reduction in pr essure. This observation suggests that the time limit and the pressure step size chosen during hydrogenation ar e not sufficient to reach equilibrium, hence instead of an equilibrium condition a pseudo e quilibrium condition existed during the test and further hydrogenation took place upon the dehydrogenation part at higher pressures. From Figure

PAGE 107

107 5-12 it can be noticed that the hydrogen content af ter the hydrogenation cycl e at all temperatures did not reach zero, similar to that of pure Mg powders. However, an examination of the powder after a cycle, revealed that the final microstruc ture do not have any magnesium hydride, hence it is assumed that the all of the hydrogen was released from the powder even though the PCT curves indicated the presence of hydrogen in the powder. 5.2.3 PCT Curves for Mg-4at%Al Powders at Different Temperatures The PCT curves of the Mg-4at%A l alloy powder at different te mperatures are presented in Figure 5-13. Similar to that of Mg-8at%Al powde r, the PCT curves were developed at 350C, 365C, 380C and 395C in that order. As indicated in Table 5-6 the solu bility of hydrogen did not change significantly with a decrease in Al content Mg-4at% Al powder. The equilibrium plateau pressures for hydrid e formation and decomposition along with the extent of hydride formation are presented in Tabl e 5-6. With the rise in temperature of the PCT experiment, the equilibrium plateau pressure of the hydride formation also increased while the extent of plateau region (MgH2 formation) remained almost c onstant in all the PCT curves developed at higher temperatures after the 1st experiment at 350C. This behavior was also observed in pure Mg powders. 5.3 Enthalpy Determination The plateau pressures given in Tables 5-1, 5-5 and 5-6 for pure Mg, Mg-8at%Al and Mg4at%Al were used to determine the enthalpy of formation/dissociation of magnesium hydride. The enthalpy and entropy of hydrogenation or dehydrogenation can be calculated according to the Vant Hoff equation: R S RT H P P 0ln (5-1)

PAGE 108

108 where P0 is the atmospheric pressure, H is the enthalpy of hydrid ing/dehydring reaction, S is the entropy of hydriding/dehydring reaction, T is the absolute temperature and R is the gas constant. The resulting Vant Hoff plots for ma gnesium hydride formation and dissociation are shown in Figure 5-14. The enthalpy and entropy valu es calculated from the slopes and intercepts of the linear fits to the Vant Hoff plots are repr esented in Table 5-7. The enthalpy of formation was reduced from 79 KJ/mol for pure Mg to 72.6 KJ/mol for Mg-Al powder, however, the enthalpy of dissociation almost remained the same for both the powders. The calculated enthalpy of hydride formation for both Mg-8at%Al and Mg-4 at%Al powder were slightly lower than the enthalpy of formation of hydride reported for Mg17Al12 but higher than the enthalpy value reported for Al3Mg2 [80]. Even though, the enthalpy valu es of hydride formation/dissociation were not changed significantly from that of pure Mg powder, the plateau pressures of formation/dissociation were in creased for Mg-Al alloy powder. 5.4 Discussion 5.4.1 Phase Transformations in PCT Curve of Mg-10at%Al Powders As indicated in Figure 5-2, a hydrogenation/dehydrogenation cycle of Mg-Al alloy powder can be divided into 7 stages. The division of va rious stages was based on the significant change in slope of the PCT curve as it indicated a chan ge in transformation reaction. These changes in slopes are presented in the Figures 5-4 (a) and (b) for hydrogenation and dehydrogenation parts of the PCT curve respectively. The different st ages are marked on these curves to show the difference in the slope, which are related to a change in transformation reaction. Stage 1 indicated on Figure 5-4 corresponds to the formation of -solid solution of hydrogen in Mg-Al alloy powder. A comparison of solubility values for pure Mg and Mg10at%Al alloy powder revealed that the solubility of hydrogen decreased with the addition of Al to hcp-Mg. These results are consistent with the calculated Al-Mg-H phase diagrams based on

PAGE 109

109 the available experimental data, which suggest th at the hydrogen solubility is very low in the hcp-Mg phase, and the addition of Al results in a decrease in solubility of hydrogen [12]. After stage 1, the slope of the PCT curve decr eased significantly to almost zero (stage 2) and continues to be zero till the hydrogen con centration in the powder reached about 2.5 wt%. This stage indicated the forma tion of magnesium hydride from the hcp-Mg phase. During this transformation, Al was not dissolved in the magne sium hydride phase and the rejected Al reacted with the hcp-Mg to form the intermetallic compound Mg17Al12. In previous studies, it has been shown that under the conditions of P= 0.6 MPa, and T= 350C during stage 2, the intermetallic compound Mg17Al12 is stable and does not form th e magnesium hydride [53]. The phase transformations taking place during the stage 2 can be written as follows: MgAlMgHHAlMg 2 2 (5-2) 1217AlMgAlMgMgAl (5-3) The rise in the slope of PCT curve and absorp tion of hydrogen after stage 2 indicated that the hydride formation was occurring from differen t phases. According to the Gibbs Phase rule, the formation of MgH2 occurs at constant pressure in a PCT curve when 3 phases and 2 components are present in the system. However, in a multi component system with more phases participating in the transformation, it allows more degrees of freedom and hence the pressure of the system also changes with composition during the transforma tion. Therefore a sloping curve rather than a flat plateau re gion was observed during the hydr ide formation in Mg-Al alloy powders.

PAGE 110

110 A comparison of XRD curves of sample S1 a nd S2 it can be concluded that the magnesium hydride was being formed from both the hcp-Mg and the intermetallic compound Mg17Al12 as relative intensities of their peaks was reduced (compare Figure 4-9 and Figure 5-5(b)). Stage 3 was observed to continue till the concentration in the powder reaches 4.1 wt% of hydrogen. The drop in the slope at the end of stage 3 illustrated that all the hcp-Mg phase present in the powder was converted into MgH2. The slope of the PCT curve after stage 3 was observed to drop continuously till the hydrogen concentration in the powder reaches 4.3 wt %. This region was defined as stage 4 of the PCT curve. The drop of the slope indicated th at the hydride formation was occurring from a single phase. Based on the XRD analysis of sample S1 and sample S2 it can be concluded that the Mg17Al12 content decreased while the Al3Mg2 phase content increased. This observation suggested that the MgH2 was being formed from the intermetallic compound Mg17Al12. The amount of hydrogen absorbed during the stage 4 was very low due to the low amount of total intermetallic present in the powder. The transf ormation reaction occurring during the stages 3 and 4 is given as 23 2 2 1217MgAlMgHHAlMg (5-4) During stage 5 the slope of the PCT curve increased continuously till the maximum pressure available in the system was reached By comparing the microstructural and phase analyses of the sample S2 and S3 it can be concluded that the intermetallic compound Al3Mg2 was hydrogenating to form MgH2 and fcc-Al phases in stage 5. The transformation reaction taking place during the stage 5 can be represented as follows: MgAlMgHHMgAl 2 223 (5-5)

PAGE 111

111 Sample S4 indicated that the phases present in the powder after par tial desorption of MgH2 were the hcp-Mg phase, the intermetallic Mg17Al12 phase along with the remaining MgH2. The absence of fcc-Al phase in sample S4 indicate d that the hcp-Mg phase dissolved the Al and transformed into the intermetallic phase of Mg17Al12. The transformation reactions involved during stage 6 can be written as follows: 21217 2HAlMgAlMgMgAlMgH (5-6) Comparison of XRD profiles of sample S1 and S6 indicated that the primary phases present in the initial powder and final powder were hcp-Mg and Mg17Al12. The small amount of fcc-Al present in the initial powder was not pres ent. This proved that the electrodeposited Mg-Al powders can be recycled for further hydroge nations as Al did not absorb hydrogen. 5.4.2 The Effect of Al Content on Different Stages of the PCT Curve The PCT curves of Mg-10at%Al, Mg-8at %Al and Mg-4at%Al de veloped at 350C are compared together in Figure 5-15 to understand the effect of Al cont ent on the different stages of PCT curves. It is clear that increasing the Al content affected the PCT curve. Figure 5-15 illustrates that, in the case of the Mg-10at%Al powder, stages 2 and 3 are well distinguished when compared to PCT curve developed for Mg -8at%Al powder. In case of Mg-4at%Al alloy powder the slopes corresponding to the stag e 2 and 3 were not distinguishable. The shape of the PCT curve depends on the co mposition and the time allowed at each step to reach equilibrium. Based on the results in this study, more intermetallics (Mg17Al12) are formed with higher amount of Al in Mg. When the intermetallic phase hydrogenates along with the hcp-Mg, a slope was observed on the PCT curve (stage 3 in Figure 5-3). However, after the

PAGE 112

112 hcp-Mg was completely consumed, a plateau region corresponding to the formation of MgH2 from Mg17Al12 was observed (stage 4 of Figure 5-3). Another variable during the experiment, the time given to reach equilibr ium, also affected the shape of the curve. If the ki netics of transformation is slow and sufficient time is not given to reach equilibrium, then the points on the PCT curv e represents pseudo equilibrium rather than a true equilibrium. In the present study this can be observed in the PCT curves developed for Mg8at%Al alloy powders. The contin uous increase in slope of th e PCT curve at 350C ( Figure 515) when compared to Mg-10at%Al powder was due to lower time applied to reach equilibrium (30 minutes for Mg-8at%Al and 60 minutes for Mg-10at%Al). The sloping nature of the PCT curve during th e absorption cycle was also observed to be less during the PCT curves of Mg-4at%Al alloy po wders. This suggested that decreasing the Al content the PCT curve tends to reach the shape of the PCT cu rve of pure Mg (Figure 5-1). A comparison of the PCT curves for pure Mg and Mg-8at%Al powders revealed that the addition of Al decreased the plat eau region (Figure 5-16). The shorte r plateau region is attributed to the formation of the Mg17Al12 intermetallic phase upon hydrogenation, which consumes the hcp-Mg. The higher plateau pressu re observed in the Mg-8at%Al sy stem can be explained by the presence of Al in the hcp-Mg, which changes th e equilibrium conditions for hydride formation as well as the solubility of hydrogen in Mg. Hysteresis in PCT curves was observed fo r both pure Mg and Mg-Al alloy powders. In a pure metal-hydrogen system the hysteresis is de fined as difference between the equilibrium plateau pressures of hydride formation to that of hydride dissociation. This difference was observed to be smaller in Mg-Al alloy powders wh en compared to that of pure Mg powders at 350C. As discussed in chapter 2, previous stud ies have suggested that the hysteresis in metal

PAGE 113

113 hydrogen systems is due to the elas to-plastic constraints in the material [76]. Similar to the model predicted in the literature, the particles in Mg-Al alloy powders also showed a variation in hydride fraction and hence exhibited the hysteresis. The smaller difference between the equilibrium plateau pressures of hydride forma tion and dissociation in case of Mg-Al alloy powders can be explained by considering the driv ing force required for nucleation of Mg from the Mg hydride. From the PCT curves shown in Fi gure 5-16 it can be noticed that the desorption of hydrogen started at a relatively higher pres sure in the case of Mg-Al alloy powder when compared to that of pure Mg powder. The nucleat ion of hcp-Mg was assisted by the presence of fcc-Al in case of Mg-Al powder. However, the pure Mg powder was completely hydrogenated close to its capacity (7.6wt%) and hence the nuc leation of hcp-Mg required higher driving force and this was achieved when the pressure in the system was significantly lower than the equilibrium plateau pressure. This causes a higher hysteresis in case of pure Mg powders. 5.4.3 Effect of Temperature on the Di fferent Stages during the PCT Curve From Figure 5-12 it can be seen that increasing the temperatur e of the PCT curve increases the extent of the plateau region in stage 2. Furthermore, the slope of the PCT curve also increases with temperature after stage 2. These phenomena are attributed to the solubility of Al in the hcpMg phase. As illustrated in Secti on 5.2.1, Al was rejected out of the hydride and into the hcp-Mg phase during the formation of magnesium hydride in Mg-Al powders. The rejected Al dissolved in the remaining hcp-Mg and precip itated as the inte rmetallic compound Mg17Al12 when the maximum solubility of Al in the hcp-Mg phase at the temperature of the experiment was reached. The solubility of Al in the hcp-Mg pha se increased with temperature in the range of 350-395C (solubility increases fr om 8.5at% to 12at%) [27] and hence less amount of intermetallic compound formed at higher temperat ures. The higher amount of the hcp-Mg phase

PAGE 114

114 available at higher temperatures ca used the increase in the extent of the plateau during the stage 2 as well as the rise in the sl ope at higher temperatures. 5.4.4 Effect of Al Content on the Enthalpy and Entropy of Hydride The Vant Hoff plots indicated that the addition of Al to Mg did not change the enthalpy of magnesium hydride formation/dissoci ation (Table 5-7) significan tly while theoretical studies have suggested that the incorporation of 8 at % of Al in the magnesium hydride reduces the enthalpy to as low as 28.36 KJmol-1 [18]. These contrasting results between the theoretical and experimental studies can be explained by considering the absence of Al in the MgH2 phase at high hydrogenation temperatures. Based on the results shown in Table 5-3, the amount of Al in the magnesium hydride at 350C was negligible The long duration involved in achieving the equilibrium conditions during the development of the PCT curves provided ample time for Al to diffuse out of the hydride phase. Thereby the magnesium hydride produced during PCT experiments contained only the equilibrium level of Al, which is quite low [112]. Hence the bond strength between Mg-H was not affected due to the absence of Al and thereby the enthalpy of formation/dissociation was not changed signi ficantly when compared to pure Mg powders. 5.5 Summary and Conclusions Pressure Composition isotherms of Pure Mg, Mg-10at%Al, Mg-8at %Al, and Mg-4at%Al were developed at different temperatures. The ev olution of different pha ses and their fractions during the development of PCT curve of the Mg -10at%Al alloy powder was established. The results of this study reveal that the initial plateau region obser ved in the PCT curve corresponds to the formation of magnesium hydride. During this transformation the Al is rejected out of the hydride and into the hcp-Mg phase, which results in the formation of the intermetallic compound Mg17Al12. This intertmetallic compound hydrogenates further at a higher pressure to form a different intermetallic compound, i.e. Al3Mg2, and magnesium hydride. At much higher

PAGE 115

115 pressures the Al3Mg2 phase also reacts with hydrogen to form magnesium hydride and fcc-Al. The enthalpy of formation/dissociation is not a ffected significantly when compared to the pure Mg powders as the magnesium hydride produced fo rm Mg-Al alloy powders is depleted of Al and hence the magnesium-hydrogen bond strength is not affected. Table 5-1. Hydrogen solubility and equilibrium pl ateau pressure of hydrides in pure Mg powders at different temperatures. Temperature(C) Solubility (wt%) Equilibrium plateau pressure (MPa) Absorption Desorption 275 0.01 0.09 0.08 300 0.02 0.18 0.16 325 0.04 0.37 0.29 350 0.09 0.7 0.48 Table 5-2. Nomenclature of different samp les during the development of PCT curve. Sample code Condition S1 Initial Powders S2 Hydrogenated partially upto 2.5 MPa S3 Hydrogenated upto maximum pressure (9 MPa) S4 Hydrogenated to maximum pressu re and dehydrogenated partially S5 Complete 1 cycle of hydrogenation/dehydrogenation Table 5-3. EDS compositional analysis of %Al in various phases in different samples mentioned in Table5-1 Sample Dark Grey Bright S1 9.9 2.0 S2 0.8 0.4 62.8 1.03 0.8 0.4 94.0 1.2 S3 0.8 0.5 93.0 1.2 S4 0.5 0.2 10.2 0.7 39.2 2.7 S5 5.5 1.3 41.1 1.2

PAGE 116

116 Table 5-4. Phase fraction analysis on the samples mentioned in Table 5-1 Sample Dark Grey Bright S1 0.9.02 0.1.01 S2 0.82 0.07 0.15 0.08 0.6 .02 0.35.1 S3 0.92 0.02 0.07 0.03 S4 0.77 0.05 0.13 0.04 0.09 0.04 0.6 0.06 0.2 0.03 0.16.09 S5 0.8 0.05 0.2 0.1 Table 5-5. Hydrogen solubility and equilibrium pl ateau pressure of hydrides in electrodeposited Mg-8at%Al alloy powders at different temperatures Temperature(C) Solubility (wt%) Equilibrium plateau pressure (MPa) Absorption Desorption 350 0.02 0.73 0.63 365 0.01 0.97 0.89 380 0.03 1.33 1.23 395 0.04 1.89 1.78 Table 5-6. Hydrogen solubility and equilibrium pl ateau pressure of hydrides in electrodeposited Mg-4at%Al alloy powders at different temperatures Temperature(C) Solubility (wt%) Equilibrium plateau pressure (MPa) Absorption desorption 350 0.02 0.67 0.57 365 0.01 0.88 0.83 380 0.02 1.22 1.16 395 0.03 1.7 1.6 Table 5-7. Enthalpy and entropy va lues calculated from the Van t Hoff plot for the materials studied Material Process H(KJ/mol H) S(J/mol H K) Pure Mg absorption -79.07.13 -143.12.8 desorption -79.91.3 -123.45.5 Mg-8at%Al absorption -72.69.8 -142.01.2 desorption -78.97.2 -133.04.9 Mg-4at%Al absorption -71.88.2 -111.85.46 desorption -79.17.77 -122.47.86

PAGE 117

117 0.001 0.01 0.1 1 10 100 012345678 Hydrogen (wt%)ln P (MPa) 350C 325C 300C 275Cb 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350C 325C 300C 275Ca 0.001 0.01 0.1 1 10 100 012345678 Hydrogen (wt%)ln P (MPa) 350C 325C 300C 275Cb 0.001 0.01 0.1 1 10 100 012345678 Hydrogen (wt%)ln P (MPa) 350C 325C 300C 275Cb 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350C 325C 300C 275Ca 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350C 325C 300C 275Ca Figure 5-1. Pressure-composition isotherms deve loped for pure Mg powders at different temperatures (a) normal pressure axis (b) logarithmic pressure axis. The filled symbols represent the absorption curves and the unfilled symbols represent desorption. 0 2 4 6 8 10 0246 Hydrogen Content (wt%)Pressure (MPa) 2 3 4 5 7 1 350C 6 0 2 4 6 8 10 0246 Hydrogen Content (wt%)Pressure (MPa) 2 3 4 5 7 1 350C 6 Figure 5-2. Pressure-composition isotherms developed for electrodeposited Mg-10at%Al powders at 350C. The vertical dashed lines represent the division of stages.

PAGE 118

118350C Mg-10at%Al Pure Mg 0 2 4 6 8 10 012345678 Hydrogen Content (wt%)Pressure (MPa)350C Mg-10at%Al Pure Mg 0 2 4 6 8 10 012345678 Hydrogen Content (wt%)Pressure (MPa) Figure 5-3. A comparison of PCT curve for pur e Mg and Mg-10at%Al developed at 350C. 0 50 100 150 200 250 0123456 Hydrogen (wt%)Slope (MPa/wt%) 2 3 4 1 absorption -2 3 8 13 18 23 2 3 4 5 Hydrogen (Wt%)Slope x 10-2 (MPa/wt%) 6 7 desorption 5b a 0 50 100 150 200 250 0123456 Hydrogen (wt%)Slope (MPa/wt%) 2 3 4 1 absorption -2 3 8 13 18 23 2 3 4 5 Hydrogen (Wt%)Slope x 10-2 (MPa/wt%) 6 7 desorption -2 3 8 13 18 23 2 3 4 5 Hydrogen (Wt%)Slope x 10-2 (MPa/wt%) 6 7 -2 3 8 13 18 23 2 3 4 5 Hydrogen (Wt%)Slope x 10-2 (MPa/wt%) 6 7 desorption 5b a Figure 5-4. The PCT slope vs. Hydrogen wt% curves for (a) hydrogenation (b) dehydrogenation part of the PCT curve developed fo r Mg-10at%Al alloy powder at 350C.

PAGE 119

119 0 2 4 6 8 10 024 Hydrogen Content (wt%)Pressure (MPa)350C 2030405060 2 theta (deg) Intensity (AU) Al3Mg2Mg17Al12fcc-Al W MgH2 2030405060 2 theta (deg) Intensity (AU) Al3Mg2Mg17Al12fcc-Al W MgH2 20 m 20 m a b d cAl3Mg2 -MgH2 Al3Mg2fcc-Al -MgH2 Figure 5-5. (a) The PCT curve of sample S2 deve loped at 350C, (b) XRD patterns of the sample S2. Backscattered electron (BSE) micrographs of different types of particles in sample S2 revealing MgH2 phase as dark and (c) F cc-Al as bright and (d) Al3Mg2 phase as bright regions. 10 m Al3Mg2 MgH2 10 m 10 m Al3Mg2 MgH2 Figure 5-6. Higher magnification images of samp le S2 showing the nucl eation of hydride on the surface.

PAGE 120

120 10 m fcc-Al MgH2 10 m fcc-Al MgH2 Figure 5-7. Higher magnification SEM/BSE micrograph of a particle in sample S2 showing the microstructure of MgH2. 0 2 4 6 8 10 0246 Hydrogen Content (wt%)Pressure (MPa)350C a c 50 m -MgH2 fcc-Al c 50 m -MgH2 fcc-Al 2030405060 2 theta (deg) Intensity (AU) fcc-Al W MgH2 b Al3Mg2 2030405060 2 theta (deg) Intensity (AU) fcc-Al W MgH2 b Al3Mg2 fcc-Al -MgH2 d 10 m fcc-Al -MgH2 d 10 m Figure 5-8. (a) The PCT curve of sample S3 deve loped at 350C (b) XRD patterns of the sample S3. (c) BSE micrographs of particles in sample S3 revealing MgH2 phase as dark and fcc-Al as bright regions. (d) Higher magnification micrograph of the microstructure shown in (c).

PAGE 121

121 0 2 4 6 8 10 0246 Hydrogen Content (wt%)Pressure (MPa)350C a 2030405060 2 theta (deg)Intensity (AU) hcp-Mg fcc-Al W Mg17Al12 MgH2 b d 50 m Mg17Al12hcp-Mg -MgH2 d 50 m Mg17Al12hcp-Mg -MgH2 50 mMg17Al12hcp-Mg -MgH2c 50 mMg17Al12hcp-Mg -MgH2c Figure 5-9. (a) The PCT curve of sample S4 developed at 350C. (b) XRD patterns of the sample S4. BSE micrographs of different types of particles in sample S4 (c) particles with higher amount of MgH2, and (d) particles with low amount of MgH2.

PAGE 122

122 hcp-Mg Mg17Al12 -MgH2a20 m 10 m hcp-Mg Mg17Al12 -MgH2b hcp-Mg Mg17Al12 -MgH2a20 m hcp-Mg Mg17Al12 -MgH2a20 m 10 m hcp-Mg Mg17Al12 -MgH2b 10 m hcp-Mg Mg17Al12 -MgH2b Figure 5-10. Higher magnification SEM/BSE micr ographs of the partially desorbed powders illustrating (a) the phase evolution during desorption, (b) nucleation of different phases.

PAGE 123

123 20304050602 theta (deg)Intensity (AU) hcp-Mg W Mg17Al12 20 m bMg17Al12hcp-Mg Mg17Al12hcp-Mg 20 m c 20304050602 theta (deg)Intensity (AU) hcp-Mg W Mg17Al12 20304050602 theta (deg)Intensity (AU) hcp-Mg W Mg17Al12 20 m bMg17Al12hcp-Mg 20 m bMg17Al12hcp-Mg Mg17Al12hcp-Mg 20 m c Mg17Al12hcp-Mg 20 m c Figure 5-11. (a) XRD patterns of the sample S5. (b) and (c) BSE microgra phs of particles in sample S5 revealing the presence of hcp-Mg and Mg17Al12 phases. 350C 365C 380C 395C 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350C 365C 380C 395C 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) Figure 5-12. Pressure-composition isotherms de veloped for electrodeposited Mg-8at%Al alloy powders at different temperatures. The filled symbols represent the absorption curves and the unfilled symbols represent desorption.

PAGE 124

124 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350 C 365 C 380 C 395 C 0 2 4 6 8 10 012345678 Hydrogen (wt%)Pressure (MPa) 350 C 365 C 380 C 395 C Figure 5-13. Pressure-composition isotherms de veloped for electrodeposited Mg-4at%Al alloy powders at different temperatures. The filled symbols represent the absorption curves and the unfilled symbols represent desorption. -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 14.515.516.517.518.519.5 104/T (1/K )ln (P/P0) Pure Mg Mg-Al -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 14.515.516.517.518.519.5 104/T (1/K )ln (P/P0) Pure Mg Mg-Al Figure 5-14. Vant Hoff plot obtained for pure Mg and Mg-8at%Al alloy powders. The filled symbols represent the plots from absorption data while unfilled symbols represent the plots from desorption data.

PAGE 125

125 0 2 4 6 8 10 02468 Hydrogen Content (wt%)Pressure (MPa)Mg-10at%Al 350C Mg-8at%Al Mg-4at%Al 0 2 4 6 8 10 02468 Hydrogen Content (wt%)Pressure (MPa)Mg-10at%Al 350C Mg-8at%Al Mg-4at%Al Figure 5-15.Comparison of the pressure-com position isotherms for electrodeposited Mg10at%Al, Mg-8at%Al and Mg-4at%Al powders developed at 350 C 0 2 4 6 8 10 02468 Hydrogen content (wt%)Pressure (MPa)350C Mg-8at%Al Pure Mg 0 2 4 6 8 10 02468 Hydrogen content (wt%)Pressure (MPa)350C Mg-8at%Al Pure Mg Figure 5-16. Comparison of the pressure-com position isotherms for pure Mg and Mg-8at%Al powders developed at 350 C.

PAGE 126

126 CHAPTER 6 EFFECT OF COMPOSITION AND TEMPERA TURE ON HYDROGENATION BEHAVIOR OF ELECTRODEPOSITED MG-AL ALLOY POWDERS In addition to the high ther modynamic stability of MgH2, the Mg powders are also limited in usage due to their slow rate of hydrogen abso rption and desorption. As illustrated in Section 25, most of the previous hydrogenation studies on the Mg-Al alloy powde r had been conducted at relatively high temperatures, and have reported the -MgH2 and fcc-Al phases as products upon hydrogenation [28, 80]. For example, hydrogenation of ball-milled hcp-Mg alloy powders with 10at% Al at 400C resulted in the final products consisted only of -MgH2. However, the hydrogenation of a powder with composition close to the Mg17Al12 intermetallic compound resulted in the formation of the fcc-Al phase as well as the magnesium hydride phase [28]. In another study, the hydrogenation of a cast and powdered Mg-30at%Al at 350C had revealed that the amount of the MgAl-phase increased during the in itial period of hydrogenation and upon further hydrogenation for longer periods of tim e, the intermetallic phase transformed to magnesium hydride and fcc-Al [30]. Not only the hydrogenation behavior of Mg-Al alloy powders was conducted at higher temperatures, th e microstructural stud ies of hydrogenated MgAl alloys has been very few [8 114]. The investigation of a Mg alloy with 3at%Al (Alloy AZ31) using optical microscopy showed that the magnesium hydride nucleated at grain boundaries during hydrogenation at 450C and its formation did not result in the development of the intermetallic Mg17Al12 phase [114]. On the other hand the hydrogenation of induction melt samples of Mg-10at%Al at 450C re vealed that the hydride may have nucleated inside the grains of the hcp-Mg [8]. Although a large number of studies have been conducted on the hydrogenation behavior of Mg-Al alloy powders at high temper atures (>300C), very less atte ntion have been given to the

PAGE 127

127 hydrogenation behavior of Mg-Al alloy powders at lower temperatures th at are close to the intended application. In addition, no systematic microstructural and compositional analyses have been carried out for Mg-Al alloy powders. The goal of this chapter is to evaluate the effect of Al addition on the hydrogenation behavior of Mg powde rs at low temperatures. Microstructural and compositional analyses were carried out to ad dress the evolution of magnesium hydride and elucidate the presence of Al in MgH2. To understand the effect of Al addition, pure Mg powders were also hydrogenated under similar conditions. The difference in microstructural changes due to temperature and hydrogen absorption was eval uated by comparing the microstructures of the hydrogenated powders with annealed powders that were heat treated in similar conditions. 6.1 Hydrogen Absorption Experiments The electrodeposited nanocrystalline Mg-8at%Al, Mg-4at%Al powders and the microcrystalline pure Mg powders, after coati ng with Ni, were hydrogenated using the in house built hydrogenation setup. The details of the hydrogenation setup and the hydrogenation procedure are described in Chapter 3. Pressure wa s kept constant at 1 MPa, which is much higher than the equilibrium pressure of magnesiu m hydride formation from magnesium at the temperatures studied here. Each hydrogenation test described belo w was conducted at least twice under the same conditions for investigating the repeatability. 6.1.1 Effect of Al addition on Hydroge n Absorption Characteristics of Mg Figure 6-1 compares the hydrogenation curves at 280C for pure Mg and Mg-8at%Al alloy powders. The pure Mg powder exhibited a fast hydrogenation rate up to approximately 4wt% hydrogen and then the rate decr eased sharply and eventually saturated at about 7.2wt% hydrogen, near to the theoreti cal capacity. The electrodeposited Mg-Al alloy powder initially hydrogenated very fast but the rate decrease d gradually after about 0.35 wt% hydrogen. The hydrogen capacity, as defined by the approximate saturation level after 5 hours of hydrogenation,

PAGE 128

128 was much lower for the Mg-Al al loy powder (Table 6-1). The X-ray diffraction profiles of both powders after hydrogenation are shown in Figure 6-2, which confirmed the formation of the magnesium hydride in both cases. The fcc-Al and the hcp-Mg phases present in the initial Mg8at%Al powders were not observed after hydrogenation, while the content of the Mg17Al12 intermetallic was increased. The (110) XRD peak for the MgH2 phase was fitted using a Peasrons VII function to calcula te the lattice parame ter, a. The fitted peaks along with the original peaks for Mg-Al alloy and pure Mg pow ders hydrogenated at 280C are shown in Figure 6-3. Similarly, the c lattice parameter was calc ulated from the fitted (101) peak. The 2 theta positions and the calculated lattice parameters presented in Tables 6-2 and 6-3, respectively, indicate that there is no significant change w ith the addition of Al when the powders are hydrogenated at 280C. The backscatterd SEM (BSE) micrographs of pure Mg and Mg-8 at%Al hydrogenated at 280C for 5 hours are presented in Figure 6-4 (a) & (b), respectively. In both powders the surface was completely covered with magnesium hydride. Pure Mg powders exhib ited a small core of Mg in the particles, while the hydride in Mg-8at %Al exhibited a network like structure inside the particles. Based on compositional analysis, the bright phase observed in Figure 6-4(b) contained ~ 42 at%Al, which corresponds to the Mg17Al12 intermetallic phase. This phase was observed mostly inside the particles and in between the hydride phases. The distri bution of % Al acquired from EPMA is shown in Figure 6-5. EPMA compositional analysis at multiple spots on 5 particles in this sample indicated that the amount of Al present in the -MgH2 is about 0.1.1 at%, which is very low within the accuracy of measurement. The absence of fcc-Al and Al3Mg2 phases suggested that Mg17Al12 intermetallic was stable under the conditions of 280C temperature and 1MPa pressure and did not react with hydrogen. This observation was consistent

PAGE 129

129 with previous studies that have shown that the Mg17Al12 intermetallic phase did not react with hydrogen below 300C and was very stable even at high pressures of hydrogen. Also, as demonstrated in Chapter 5, the equilibrium pressure for hydrogenation of this phase at 350C was 1.7MPa, higher than the pressure used here. 6.1.2 Effect of Amount of Al on the Hydroge nation Characteristics of Mg-Al Powders The effect of Al content on the hydrogenati on behavior was inves tigated by hydrogenating powders with 8at% and 4at% at 210C and 1MPa pressure. The hydrogen absorption curves for the electrodeposited Mg-8at%Al an d Mg-4at% Al alloy powders are compared in Figure 6-6 (a) and the corresponding XRD profiles are shown in Figure 6-6 (b). The Mg-4at%Al powder exhibited a slower hydrogenation rate in the first 50 minutes, but the rate di d not decrease as fast as it did for the Mg-8at%Al powder. Therefore, higher amount of hydrogen was absorbed in the former powder after 5 hours of hydrogenation (Tab le 6-1). In addition, the hydrogenation curve of Mg-4at%Al indicated that even after 5 hours, the saturation has not been reached. From Figure 6-6(b) it can be noticed that both Mg-8at%Al and Mg-4at% powders contains hcp-Mg, fcc-Al and Mg17Al12 phases after hydrogenati on. The amount of fcc-Al phase present in both the powders after hydrogenation was similar to that of the initial powder (Fi gure 4-9), showing that this phase did not transform during the hydroge nation process at 210C. The XRD peaks of (110) and (101) for MgH2 were fitted using the Pearsons VII function and the peak values are presented in Table 6-2. The corresponding lattice parameters calculated from the peak positions are presented in Table 6-3. These results indica ted that the a parame ter for Mg-8at%Al alloy powder was slightly higher than the value for Mg-4at%Al, while the c parameter was almost similar in both the powders. The SEM/BSE micrographs of the Mg-4at%A l and Mg-8at%Al powders hydrogenated at 210C are compared in Figure 6-7. In both powders the ma gnesium hydride was found only

PAGE 130

130 along the surface but in case of the Mg-4at%Al alloy powder the surface was not completely covered with the hydride phase as indicated by dash ed circles in the micrograph shown in Figure 6-7(a). The Mg-8at%Al alloy powde r showed relatively finer hydrid e colonies in comparison to the Mg-4at%Al alloy powder. Much less Mg17Al12 intermetallic phase was found in the Mg4at%Al powder and it was limited to the locati ons near the hydride phase. However, the Mg8at%Al powder, exhibited precipitation of this in termetallic phase inside the hcp-Mg and away from the hydride phase. Figure 6-8 shows the line compositional analysis using EPMA. The compositional analyses along the line from A to B indicated that the amount of Al was increasing until the hcp-Mg/MgH2 interface. The composition of Al decreased immediately after the interface inside the hydride phase. In addition, the formation of intermetallic compound as indicated in Figure 6-8, near the hydride and the hcp-Mg in terface, suggested that the concentration of Al reached beyond the saturation of hcp-Mg. The Al content in the hydride was about 2.1at%, which was well below the concentration of Al in initial powder and these observations demonstrated that upon the formation of magnesium hydride, Al was rejected by this phase and was accumulated at the hcp-Mg/ -MgH2 interface. Compositional measurements revealed that the Al content of the hydride decreased significantly from 2.2.2at% to 0.1.1at% with reducing the Al conten t of the powders from 8at% to 4at%. 6.1.3 Effect of Temperature on Hydrogenation Behavior of Mg-Al Alloy Powders The effect of temperature on hydrogenati on was evaluated by conducting hydrogenation tests at 280C, 210C, and 180C for the Mg-8at%A l alloy powder. As shown in Figure 6-9, the hydrogenation temperature affected the absorp tion behavior of hydr ogen significantly. The hydrogenation curves obtained at di fferent temperatures revealed that the initial kinetics of hydrogenation was similar at 280C and 210C, while their total capaciti es after 5hrs were

PAGE 131

131 significantly different. Reducing th e temperature to 180C drastical ly reduced the transformation rate initially, however, saturation in hydrogen absorption was not reached even after 5 hours. The total amount of hydrogen present in e ach powder is shown in Table 6-1. The XRD profiles of the Mg-8 at%Al alloy powder hydrogenated at different temperatures are presented in Figure 6-10, wh ich indicated the formation of -MgH2 at all temperatures. Other than the hydride, only Mg17Al12 phase was observed in the powder hydrogenated at 280C. For powders hydrogenated at 210C and 180C, along with these phases, the fcc-Al phase and a significant amount of the hcp-Mg pha se were present in the structure. The peak positions (2 theta values) for -MgH2 was found to depend noticeably on th e hydrogenation temperature of the Mg-8at%Al alloy powder. The XR D (110) and (101) peaks for MgH2 produced at each temperature were fitted with a Pearsons VII function. The peaks fitted for (110) plane are presented in Figure 6-11 and their calculated positions are presented in Table 6-2. The lattice parameters of the tetr agonal structure of MgH2 were calculated from these values and are also included in Table 6-3. Interestingly, the shift in the peak positions increased with a decrease in the hydrogenation temperature. The calculated lattice parameters presented in Table 6-3 demonstrated that the a parameter was notic eably increased by lowering the hydrogenation temperature, while the c parame ter remained almost constant. Backscattered electron micrographs of th e Mg-Al alloy powder hydrogenated at 280C, 210C, and 180C are shown in Figure 6-12. Three different contrasts can be identified in these micrographs. The darker regions correspond to the magnesium hydride phase, the brighter regions represent the Mg17Al12 and the grey region corresponds to the hcp-Mg phase. The average compositions of different contrast s are given in Table 64. The amount of Al present in the dark regions as a function of hydr ogenation temperature is given in Table 6-4 and

PAGE 132

132 plotted in Figure 6-13. The results indicated that the amount of Al in the dark region increases with reducing the hydrogena tion temperature. It is interesti ng to note that parallel to this increase, the standard deviation of the measured compositions increases significantly. The lighter region in the core of the powders hydrogenate d at 210C and 180C had a composition close to the alloy composition and is believ ed to consist of hcp-Mg and Mg17Al12 intermetallic phases due to the limited solubility of Al in Mg. Consistent with the hydrogenation curves, a smaller fraction of hydride (dark phase) was found in the 210C and 180C hydrogenated samples. -MgH2 formed as a shell on the surface with a network of hydride extending through th e particles hydrogenated at 280C (Figure 6-12 (a)). However, after hydrogenation at 210C, the hydride formation was limited to the surface layer and a non-hydrogenated core remained in the particles (Figure 6-12(b)). Furthermore, the powder hydrogenated at 180C exhibited an incomplete shell of hydride, in addition to a core as shown in Figure 6-12(c). The brighter phase wa s observed in large amounts inside the powder hydrogenated at 280C, while small precipitates of bright phase were observed at 210C. In order to identify the phases observed on the microstructure and investigate the partitioning of Al among various phases, compos itional line analyses were conducted on several particles. The analyses for two su ch particles are presented in Fi gures 6-14. Figures 6-14 (a) and (b) represent the low magnification SEM/BSE mi crographs of particles hydrogenated at 210 and 280C respectively. Figure 6-14(c) represents the distribution of Al along the line AB on a particle that was hydrogenated at 210C. This region covered th e three contrasts that were observed in the powder. The Al content near the MgH2/Mg17Al12 and Mg17Al12/hcp-Mg interface changed sharply showing the accumulation of Al Only two contrasts were observed in the powder hydrogenated at 280C and the line composition analyses of the corresponding particle is

PAGE 133

133 presented in Figure 6-14(d). These results revealed that the bright areas in the microstructure of the hydrogenated powders contained approximately 43 to 46 at% Al, which corresponds closely to the composition of the Mg17Al12 (41 at%) intermetallic com pound. This phase was found in between the hydride colonies or at the interf ace between the hcp-Mg and the hydride in the 210C hydrogenated powder. However, in the powder hydrogenated at 280C, the intermetallic phase had grown significantly larger between the MgH2 colonies due to the high temperature. The microstructure of the powder hydrogenate d at 210C contained a large amount of the Mg(Al)-phase (grey phase) as shown in Figure 612 (b) which is in accord with the XRD results shown in Figure 6-10. On the other hand, the mi crostructure and the XRD results (Figure 6-10) of the 280C hydrogenated powder did not show det ectable amount of the grey phase, i.e. the hcp-Mg phase. 6.2 Thermal Stability of Electrodeposited Mg-Al Powders To identify the phase changes in Mg-8at%Al alloy powder due to the hydrogenation and the temperature, annealing tests were conducte d on these powders. Th e Mg-8at%Al powder was annealed at 180, 210 and 280C fo r two different time periods of 20 minutes and 5 hours. The annealing test for 20 minutes corresponds to the in itial time where the powder was heated before insertion of hydrogen and the 5 hours experime nt corresponds to the hydrogenation time. The SEM/BSE micrograph of the powders heat treated for 20 minutes at 180C, 210C, and 280C are presented in Figure 6-15. Other th an hcp-Mg, the contrast corresponding to the intermetallic compound was not identified in powders heat treated for shorter times. At higher magnifications, very fine particles with bright contrast appeared but the composition of those particles could not be identified due to their size. The SEM/BSE micrographs of the powders heat treated for 5 hours are s hown in Figure 6-16. The interm etallic compound has grown to a certain extent at 180 and 210C while it has grow n significantly at 280C after 5 hours. Small

PAGE 134

134 precipitates of the intermeta llic compound were observed in case of the powder hydrogenated at low temperatures while large intermetallic pha se was observed when annealed at 280C. The intermetallic compound formed at 280C was obser ved primarily near the surface. The larger particles and the formation near the surface indica tes that the diffusion of Al at 280C is much higher than at 180 and 210 C. The presence of Al12Mg17 inside the particle and near the defects like surfaces illustrates that the nucleation of in termetallic takes place both in continuous and discontinuous form. Previous work on Mg-9Al all oy also indicated the formation of intermetallic particles of Mg17Al12 both continuously and discontinuously inside hcp-Mg [69]. Furthermore, even in a conventional bulk alloy used in this st udy, the precipitate sizes after aging at 200C for 8 hours were very fine (approximately 25 nm wide and 50 nm long). Therefore, it is predicted that the intermetallic phase present in the pow der annealed at 280C af ter 20 minutes is grown when compared to the powders at 180 and 210C. However, it is not observed in the SEM/BSE micrographs due to its poor reso lution at higher magnifications. 6.3 Discussion 6.3.1 Effect of Al Addition The results of this study i ndicated that under relatively lo w temperatures the hydrogenation of Mg-Al alloy powders resu lted in the formation of -MgH2 and Mg17Al12 phases. However, previous studies at higher hydrogena tion temperatures have revealed the formation of the fcc-Al phase in addition to the hydride [28]. This difference can be a ttributed to the hydrogenation of the Mg17Al12 intermetallic compound at higher temperatur es which resulted in the formation of magnesium hydride and fcc-Al based on equation pres ented in section 5.4.1. The presence of the intermetallic phase after 5 hours at 280C at 1MPa hydrogen suggested that these conditions were not sufficient for the hydrogenation of this phase. It was also reported that the

PAGE 135

135 hydrogenation of Mg17Al12 takes place at temperatures a bove 300C and high pressures (5MPa) [53]. According to the equilibrium Mg-Al phase diagram [115] the maximum concentration of aluminum soluble in the hcp-Mg at 280C is a pproximately 5.2at%. As shown in Figure 6-8, upon hydrogenation, magnesium hydride rejected Al and consequently Al accumulates at the hydride/hcp-Mg interfac e and resulted in the formation of more Mg17Al12 phase than expected from the binary Mg-Al phase diagram. This was also evident by analyzing the microstructure of the powder after annealing. A comparison of Figure 6-4(b) and Figure 6-16( c) indicated that the intermetallic phase formed during hydrogenation was much higher when compared to that formed in annealing. In addition, the quantit ative metallographic analysis on annealed powder indicated that the volume % of Mg17Al12 was ~13.5% (close to th at of predicted by phase diagram ~ 14.5%) but similar analysis on many pa rticles revealed that the hydrogenated powders contained 28.4% of Mg17Al12 phase. The formation of in termetallic phase upon hydrogenation resulted in a much lower hydrogen capacity in Mg-Al alloy powders in comparison to the pure Mg-powder (Figure 6-1). However, correcting the hydrogenation curv e with the total fraction of Mg available (Figure 6-17), the relative hydrogen capacity was equivalent to that for the pure Mg. Another consequence of th e Al rejection by the hydride pha se was the observation of a lower hydrogenation rate for the Mg -8at%Al alloy powder. This findi ng can be attributed to the required diffusion of Al. The diffusion coeffici ent of Al in hcp-Mg at 280C is 1.869X10-15 m2/s while that of hydrogen in hcp-Mg is 8.1475X10-9 m2//s [116, 117]. From the mentioned coefficients, it can be noticed th at the hydrogen diffusion is much hi gher than that of Al in hcpMg.

PAGE 136

136 The difference observed in the hydrogenati on behaviors of the Mg-8at%Al and Mg4at%Al powders at 210C temperature (Figure 66(a)) can be attributed to the amount of the intermetallic phase available when the hydrogenati on begins. The higher intermetallic content of the former powder resulted in an initially faster hydrogenation rate but the impingement of the hydride colonies rendered a hydride layer. The formation of the hydride layer resulted in a fast reduction in the hydrogenation rate as the diffusion of hydrogen through MgH2 was very slow. Consistently, the microstructural examinations showed that in contrast to the Mg-8at%Al powder, the Mg-4at%Al powder di d not develop a continuous laye r of magnesium hydride (see Figure 6-7). In summary, the addition of Al af fects the hydrogenation ki netics in two opposing manners. The presence of intermetallic particles expedites the nucleation rate of the magnesium hydride, which enhances the kinetics of the hydrid e formation. On the ot her hand, the rejection of Al by the hydride phase slows do wn the growth of the hydride. 6.3.2 Effect of Temperature The kinetics of absorption and the hydroge n capacity of the Mg-8at%Al alloy powder showed a strong dependency on the hydrogenation temperature (Figure 6-10). It was shown recently that the impingement of hydride coloni es, which nucleate and gr ow from the surface of powders, causes the saturation of hydrogen abso rption and any parameter that increases the nucleation rate, decreases the hydrogen capacity by reducing the thickness of the hydride [118]. The Vant Hoff plot (Figure 5-14) shown in the previous chapter suggest s that the equilibrium temperature of hydride formation at 1 MPa from the Mg-8at%Al powder is 370C and hydrogenation below this temperature and pressure will result in a hi gher driving force for nucleation. In addition, intermetallic particles can also act as the nucleation sites for magnesium hydride [95, 119]. The high hydride nuc leation rate encouraged by the larger driving force due to low temperature cause the hydrogen saturation to be reached at a smaller hydride thickness at

PAGE 137

137 210C in comparison to the hydrogenation at 280 C. A comparison of Figures 6-12 (a) & (b) elucidated that entire surface of the Mg-8at%Al part icles hydrogenated at 210 C was covered with the magnesium hydride phase. Similarly, microstructural evaluati on indicated (Figure 612(c)) that the particles hydrogena ted at 180C were not completely covered and the absence of impingement of the hydride colonies was consistent with the lack of sa turation indicated by the hydrogenation curve. The hydrogenation curve at 180C indicated that the hydrogen absorption kinetics was slower and this can be associated with the diffusion of Al, which was rejected by the hydride upon its formation [119]. The diffusion length of Al as a function of temperature in 5 hours is shown in Figure 6-18 [116]. It can be noticed that the diffusion le ngth of Al started to decrease significantly in the te mperature range of 180C. The fine precipitate particles observed in the powder hydrogenated at 180C also confirme d that the diffusion of Al was significantly slow at this temperature. The compositional analysis of the dark regions formed at different temperatures indicated that the amount of Al reduces from 6.8% to 0.1% with increasing the hydrogenation temperature from 180C to 280C (Figure 6-14). While the accommodation of 6.8 at% of Al in MgH2 was not realistic as the isotherms of Al -Mg-H developed at different temp eratures suggest that there was no solubility of Al in MgH2[112]. However, the detailed an alysis of the XRD results and compositional analyses (Tables 6-2 & 6-3) s howed that indeed re ducing the hydrogenation temperature resulted in the entrapment of Al in the lattice and caused lattice expansion. The observation of MgH2 lattice expansion due to the presen ce of Al was contradictory to the theoretical calculations performed on this system [24, 50]. On the other hand, these calculations did not consider the formation of defects and since the radius of Al cation is smaller than that of Mg cation, the lattice naturally contracts. In reality, magnesium hydr ide is an ionic compound

PAGE 138

138 and the lattice parameter is influenced by the do pants ionic radius as well as the amount and type of defects created in the crystal for maintaining the charge neutrality. When Mg+2 ions are replaced by Al+3 ions the charge neutrality is maintain ed by either creating interstitial hydrogen ions or magnesium vacancies. Both of these de fect formations are an ticipated to expand the lattice. However, the H-1 ion being very large (ionic radius = 1.53 [120]), makes the formation of interstitial hydrogen atom improbable. The la ttice expansion owing to the addition of the higher valance cations has also be en reported in oxides with the flourite structure [121, 122]. The high level of Al content found in the dark region is believed to be associ ated with the inclusion of the Mg17Al12 precipitates. Higher magnification imag es of the powder hydrogenated at 180C are shown in Figure 6-19. Figure 6-19 illustrates the submicrometer intermetallic particles present in the MgH2. During the composition analysis the sm allest beam size used in EPMA was about 1 m in diameter. Since the intermetallic part icles were smaller than the beam size and were present everywhere inside the hydride re gion, it was impossible to eliminate them and measure the Al content in the MgH2. The larger standard deviation in compositional analysis (Figure 6-13) at lower hydrogenation temperatures may be attributed to the statistics of the inclusion of the sparsely separated and larger intermetallic particles within the region of analysis. However, at high temperatures (280C) the diffusion of Al was high and the Al diffused away from hydride to form the Mg17Al12. These results suggest that the -MgH2 phase has very low solubility for Al under the hydrogenation conditions of this study, but under non-equilibrium conditions Al may be incor porated in this hydride. 6.4 Summary and Conclusions In this chapter, the hydrogena tion behavior in pure Mg powder, and electrodeposited hcp-Mg rich alloy powders were studied. Based on the microstructural and compositional analyses it can

PAGE 139

139 be concluded that under nonequilibrium hydrogenation conditions Al can be incorporated in the MgH2 crystal. Due to the slower diffusion of Al, its presence in the hcp-Mg reduces the hydrogenation kinetics at a given temperature in comparison to pure Mg. This phenomenon is attributed to the rejection of Al by the ma gnesium hydride, which slows the growth of the hydride. The presence of the Mg17Al12 intermetallic prior to the hydrogenation increases the hydride nucleation rate, which causes an initially fast hydrogenation rate but reduces the hydrogen capacity owing to the coverage of th e powder surfaces with magnesium hydride in a short time. Raising the hydrogenation temperature in creases the solubility of Al in the hcp-Mg phase. Thereby, the intermetallic content prior to hydrogenation at low temperatures is high. The higher intermetallic content at low temperatur es of hydrogenation decreases the energy barrier for hydride nucleation significantly and enhances the hydride nucleation rate. This reduces the hydrogen capacity at low temperatures. Thermal stab ility studies on Mg-Al powders indicate that the intermetallic compound Mg17Al12 grows to a large extent when annealed for 5 hours and forms as relatively small precipitates inside the particle.

PAGE 140

140 Table 6-1. Hydrogen capacity of the materials studie d at different temperatures for both the runs. Temperature (C) Material Run 1 (wt%) Run 2 (wt%) 280 Pure Mg 7.17 7.20 Mg-8at%Al 4.84 4.81 210 Mg-4at%Al 4.16 4.12 Mg-8at%Al 2.78 2.80 180 Mg-8at%Al 3.60 Table 6-2. Comparison of the XRD (110) and (101) peak positions of MgH2 produced at different hydrogenation temp eratures. Note that the standard value for the -MgH2 [123] is similar to the position found in the hydrogenated pure Mg. Material 2 (110) 2 (101) Std MgH2 27.942 35.743 Pure Mg, 280C 27.942 35.742 Mg-8at%Al, 280C 27.941 35.742 Mg-8at%Al, 210C 27.862 35.682 Mg-8at%Al, 180C 27.842 35.722 Mg-4at%Al, 210C 27.914 35.713 Table 6-3. The lattice parameters of MgH2 produced in the Mg-8at%Al powder at different hydrogenation temperatures along with the standard values. Material a () c() Standard MgH2 4.514 3.022 Pure Mg, 280C 4.514 3.022 Mg-8at%Al, 280C 4.515 3.022 Mg-8at%Al, 210C 4.527 3.025 Mg-8at%Al, 180C 4.529 3.019 Mg-4at%Al, 210C 4.517 3.024

PAGE 141

141 Table 6-4. Compositional analysis of various regions of the hydrogenated Mg-8at%Al particles shown in Figure 6-12 Temperature Dark at%Al Light at%Al Bright at%Al 280C 0.1.1 44.2 210C 2.2.2 7.3 41.5 180C 6.8.66 8.2 42.1.3

PAGE 142

142 0 1 2 3 4 5 6 7 8 050100150200250300350 Time (mins)Hydrogen (wt%)Pure Mg Mg-8at%Al 280C, 1 MPa 0 1 2 3 4 5 6 7 8 050100150200250300350 Time (mins)Hydrogen (wt%)Pure Mg Mg-8at%Al 280C, 1 MPa Figure 6-1. Hydrogen absorption curves for pur e Mg and Mg-8at%Al powders developed at 280C and 1 MPa pressure. 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-Mg W Ni un-hydrogenated hydrogenated Pure Mga 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-MgMg17Al12 fcc-AlW Ni Mg-8at%Al hydrogenated un-hydrogenated b 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-Mg W Ni un-hydrogenated hydrogenated Pure Mga 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-Mg W Ni un-hydrogenated hydrogenated Pure Mga 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-MgMg17Al12 fcc-AlW Ni Mg-8at%Al hydrogenated un-hydrogenated b 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-MgMg17Al12 fcc-AlW Ni Mg-8at%Al hydrogenated un-hydrogenated b Figure 6-2. The XRD patterns of the hydrogenate d and un-hydrogenated powders, (a) for Pure Mg, and (b) Mg-8at%Al powder.

PAGE 143

143 2727.52828.529 2 theta (deg)Intensity (AU)a b 2727.52828.529 2 theta (deg)Intensity (AU) 2727.52828.529 2 theta (deg)Intensity (AU)a 2727.52828.529 2 theta (deg)Intensity (AU)a b 2727.52828.529 2 theta (deg)Intensity (AU)b 2727.52828.529 2 theta (deg)Intensity (AU) Figure 6-3. Peak fits of (110) XRD peak co rresponding to (a) Pure Mg, and (b) Mg-8at%Al powder hydrogenated at 280C 20 m a 20 m a 50 m b 50 m bMg17Al12 -MgH2hcp-Mg -MgH2 Figure 6-4. Back scattered elec tron (BSE) micrographs of the po lished (a) pure Mg, and (b) Mg8at%Al powders hydrogenated at 280C.

PAGE 144

144 44.15 0.12 0.15 0.08 0.18 0.01 0.02 42.65 42.50 0.13 5 mMg17Al12 -MgH2 44.15 0.12 0.15 0.08 0.18 0.01 0.02 42.65 42.50 0.13 5 mMg17Al12 -MgH2 Figure 6-5. SEM/BSE microgra ph of Mg-8at%Al powder hydrogenated at 280C indicating the %Al in various phases. 0 1 2 3 4 5 0100200300 Time (mins)Hydrogen (wt%)Mg-8%Al Mg-4%Al a 1 MPa, 210C 0 1 2 3 4 5 0100200300 Time (mins)Hydrogen (wt%)Mg-8%Al Mg-4%Al a 1 MPa, 210C 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-MgMg17Al12 fcc-AlW Ni b Mg-4%Al Mg-8%Al 25303540455055 2 theta (deg)Intensity (AU)MgH2 hcp-MgMg17Al12 fcc-AlW Ni b Mg-4%Al Mg-8%Al Figure 6-6. (a) Hydrogen absorption curves for Mg-8at%Al and Mg-4 at%Al alloy powders developed at 210C, and (b) the corr esponding XRD patterns of the hydrogenated powders.

PAGE 145

145 Figure 6-7. BSE micrographs of (a) Mg-4 at%Al, and (b) Mg-8at%Al alloy powders hydrogenated at 210C. Note the presence of un-hydrogenated regions, as marked by the dotted circles, on the surf ace of the Mg-4at%Al powder. 10 m 2.1 7.4 10.9 AA BB -MgH2hcp-Mg Mg17Al12% Al 10 m 2.1 7.4 10.9 AA BB -MgH2hcp-Mg Mg17Al12% Al Figure 6-8. BSE micrograph of the Mg-8at%A l alloy powder hydrogenate d at 210C along with the compositional analysis al ong the line AB showing the acc umulation of Al at the -MgH2/hcp-Mg interface.

PAGE 146

146 0 1 2 3 4 5 6050100150200250300350 Time (mins)Hydrogen (wt%)1 MPa, Mg-8at%Al 280C 180C 210C 0 1 2 3 4 5 6050100150200250300350 Time (mins)Hydrogen (wt%)1 MPa, Mg-8at%Al 280C 180C 210C Figure 6-9. Hydrogen absorption curves for Mg-8at%Al powder developed at different temperatures and at 1 MPa pressure. MgH2 hcp-Mg Mg17Al12 fcc-Al W 25303540455055 2 theta (deg)Intensity (AU) 210C 180C 280C MgH2 hcp-Mg Mg17Al12 fcc-Al W 25303540455055 2 theta (deg)Intensity (AU) 210C 180C 280C Figure 6-10. The XRD patterns of the hydr ogenated Mg-8at%Al powder at different temperatures.

PAGE 147

147 2727.52828.5 2 theta (deg)Intensity (AU) 2727.52828.5 2 theta (deg)Intensity (AU) 2727.52828.5 2 theta (deg)Intensity (AU)280 C 180 C 210 C 2727.52828.5 2 theta (deg)Intensity (AU) 2727.52828.5 2 theta (deg)Intensity (AU) 2727.52828.5 2 theta (deg)Intensity (AU)280 C 180 C 210 C Figure 6-11. Peak fits of (110) XRD peak corresponding to Mg-8at%Al alloy powder hydrogenated at different temperatures.

PAGE 148

148 50 m Mg17Al12 MgH2 280C a 50 m Mg17Al12 MgH2 280C a hcp-Mg +Mg17Al12 -MgH2 20 m Mg17Al12 210C b hcp-Mg +Mg17Al12 -MgH2 20 m Mg17Al12 210C b 20 m 180C hcp-Mg + Mg17Al12 -MgH2 c 20 m 180C hcp-Mg + Mg17Al12 -MgH2 c Figure 6-12. BSE micrographs of the Mg-8at %Al powder hydrogenated at (a) 280C (b) 210C and (c) 180C, revealing MgH2 phase as dark, hcp-Mg phase as light and Mg17Al12 phase as bright regions. The arro ws indicate the locations on the surface where no hydride had formed. 0 1 2 3 4 5 6 7 8 150180210240270300 Hydrogenation Temperature (C)at%Al Figure 6-13. The variation of Al in the dark regions of the micrographs shown in Figure 6-12.

PAGE 149

1497.6 50 m a 10 m % Al 2.1 43.2 A B A B c 50 m b 10 m A B A B 0.1 46.2 0.3 d % Al 210C 280C hcp-Mg hcp-Mg Mg17Al12 -MgH2 Mg17Al12 -MgH2 -MgH2 Mg17Al127.6 50 m a 10 m % Al 2.1 43.2 A B A B c 50 m a 10 m % Al 2.1 43.2 A B A B c 50 m b 10 m A B A B 0.1 46.2 0.3 d % Al 50 m b 10 m A B A B 0.1 46.2 0.3 d % Al 210C 280C hcp-Mg hcp-Mg Mg17Al12 -MgH2 Mg17Al12 -MgH2 -MgH2 Mg17Al12 Figure 6-14. Back scattered el ectron (BSE) micrographs of the polished Mg-8at%Al powder hydrogenated at (a) 210C and (b) 280C. Higher magnification images and the corresponding variations of Al content along the AB lines developed by the EPMA method for powders hydrogenated at (c) 210C and (d) 280C.

PAGE 150

150 50 m hcp-Mg + Mg17Al12180C 50 m hcp-Mg + Mg17Al12180C 210C hcp-Mg + Mg17Al12 20 m 210C hcp-Mg + Mg17Al12 20 m 280C hcp-Mg + Mg17Al12 10 m 280C hcp-Mg + Mg17Al12 10 m Figure 6-15. SEM/BSE micrographs of Mg-8at %Al annealed for 20 minutes at different temperatures. 20 m hcp-Mg + Mg17Al12a Mg17Al12 20 m hcp-Mg + Mg17Al12a Mg17Al12 20 m hcp-Mg c Mg17Al12 20 m hcp-Mg c Mg17Al12 b 20 m hcp-Mg + Mg17Al12Mg17Al12 b 20 m hcp-Mg + Mg17Al12Mg17Al12 Figure 6-16. BSE micrographs of Mg-8at%Al powd ers annealed for 5 ho urs at (a) 180C (b) 210C (c) 280C.

PAGE 151

151 0 1 2 3 4 5 6 7 8 0100200300 Time (mins)Hydrogen (wt%)Pure Mg Mg-8at%Al 280C, 1 MPa 0 1 2 3 4 5 6 7 8 0100200300 Time (mins)Hydrogen (wt%)Pure Mg Mg-8at%Al 280C, 1 MPa Figure 6-17. A comparison of the hydrogen absorp tion curves developed at 280C for the pure Mg and Mg-8at%Al alloy powder after correction for the influence of the intermetallic phase formation. 0 100200300 Temperature (C)Diffusion Length Figure 6-18. Diffusion length of Al in hc p-Mg as a function of temperature.

PAGE 152

152 MgH2+ Mg17Al12 10 m MgH2+ Mg17Al12 10 m Figure 6-19. SEM/BSE micrograph of a Mg-8at %Al powder hydrogenated at 180C showing the submicrometer precipitates in the MgH2 region.

PAGE 153

153 CHAPTER 7 DEHYDROGENATION CHARACTERISTICS OF MGH2 PRODUCED FROM ELECTRODEPOSITED MG -AL ALLOY POWDERS Different methods have been employed to reduce the thermodynamic stability of MgH2. The addition of Al to MgH2 has been predicted to decrease the stability of the magnesium hydride by reducing its enthalpy of formation. Theoretical calculations have proved that the addition of Al weakens the magnesium hydrogen bond [18, 50]. Very few experimental studies have been carried out on the desorption behavior MgH2 produced from Mg-Al alloy powders [97, 111]. A desorption study conducted on the MgH2-Al composite produced by ball milling of 8 mol% of Al and -MgH2, revealed that the kinetics of hydrogen desorption increased significantly at 300C when compared to pure MgH2 [97]. However, the eff ect of Al addition on desorption temperature of MgH2 is not investigated and the attr ibution of faster kinetics of hydrogen release due to additi on of Al was not very clear [97]. Other dehydrogenation experiments that have been carried out previously on hydrogenated Mg-Al powders were under equilibrium conditions as part of the PCT curve development [31]. The absorption tests described in Chapter 6 on Mg-Al powder suggested that different amounts of Al is trapped in MgH2 crystal lattice depending on th e hydrogenation conditions. The goal of the present study was to i nvestigate the effect of the tra pped Al on the stability of the MgH2. To achieve this goal, desorption tests on MgH2 synthesized at different temperatures from electrodeposited Mg-Al alloy powders were conducted. To unde rstand the effect of Al addition, dehydrogenation of hydrogenated pure Mg powder wa s also carried out and compared with the hydrogenated Mg-Al alloy powder. Finally, the microstructural ev olution during desorption was understood by analyzing the samples th at were dehydrogenated partially.

PAGE 154

1547.1 Hydrogen Desorption Experiments The desorption tests of hydrogenated electrodeposited Mg-8at%Al pow ders and the pure Mg powders were conducted in TGA (details of the dehydrogenation proc edure were described in Chapter 3). The transformation temperatures were identified from the DSC/TGA curves by using the first deviation and extrapolated tange nt method as described in ASTM E967-03 [124]. In this method the temperature where the weight loss curve deviates from the base line for the first time was reported as onset temperature while peak temperature is ca lculated by drawing the tangents across the base line and the heat flow curve. Each de hydrogenation test described below was conducted at least twice under the same c onditions for investigating the repeatability. 7.1.1 Effect of Al Addition on Hydrogen Desorption Figure 7-1 compares the hydrogen desorption curves (TGA) for pure Mg and Mg-8at%Al alloy powders that were hydrogenated at 210C. Figure 7-1(a) reveals that the hydrogen was released in two stages from hydrogenated Mg-Al pow ders while it was released in a single stage from hydrogenated pure Mg powders. The first st age of hydrogen release was observed to occur in the temperature range of 100150C for Mg-Al alloy powders while the second stage occurred in the temperature range of 200350C. The transformation temperatures of both the pure Mg powder and the Mg-Al alloy powders are pres ented in Table 7-1. The hydrogen release temperatures (peak) from hydrogenated Mg-Al powders were about 115 and 327C while for hydrogenated pure Mg powder was about 350C. From these results it can be noticed that the hydrogen release temperature was lo w in hydrogenated Mg-Al powders. The fraction of hydrogen released from both the hydrogenated Mg-Al and pure Mg samples as a function of temperature is plotted in Figure 7-1(b), which showed ~ 12% of the total hydrogen was released in the first stage for hydrogenated Mg-Al alloy powder. However, the total hydrogen present was releas ed in 1 stage for hydrogenated pure Mg powder. The amount of

PAGE 155

155 the hydrogen content in the hydrogenated samp les and the hydrogen release during the dehydrogenation tests are shown in Table 7-2. Thes e values in the table suggest that the total hydrogen present in the pow ders was released. The XRD profiles of both powders after the de hydrogenation are shown in Figure 7-2. The dehydrogenation products of both the powders s howed the presence of hcp-Mg phase. In addition to the hcp-Mg phase, Mg17Al12 phase was also observed in the Mg-Al alloy powders. A comparison of these XRD profiles with that of the in itial powders before hydrogenation (Figure 4-6 and 4-10) revealed that the fcc-Al present in the initial powders was completely transformed into Mg17Al12 or dissolved in hcp-Mg and the 2 position of the hcp-Mg increased (for Mg-Al, (initial 2 = 36.662 and after 1 cycle 2 =36.850). This increase in 2 position for Mg-8at%Al indicated that the fcc-Al was dissolved in the hcp-Mg lattice and caused the reduction in the lattice parameter. Since the initial phases of the Mg-Al powder were not altered significantly, this powder can be cycled for more number of absorption and desorption experiments, similar to the pure Mg. The above results indicated that the addition of Al to hcp-Mg reduced the hydrogen release temperature from MgH2. However, the release of hydrogen in two stages from the Mg-Al alloy powder was not clearly understood and further in vestigations were car ried out on the Mg-Al alloy powders. 7.1.2 Effect of Hydrogenation Temperature on Hydrogen Desorption of the Mg-Al Alloy Powders Figure 7-3 shows the hydrogen release curv es of the Mg-8at%Al alloy powder hydrogenated at 180, 210 and 280C. The release of hydrogen was observed to take place in 2 stages when the Mg-Al samples were hydrogena ted at 180 and 210C, but not at 280C (Figure 7-3(a)). The temperatures of hydrogen release for these samples are presented in Table 7-3.

PAGE 156

156 These temperatures indicated that the samp le hydrogenated at 280C released hydrogen at a much higher temperature than the samples hydrogenated at lower temperatures. The transformation temperatures corresponding to th e 2 stages were observed to be similar in powders hydrogenated at 180 and 210 C. Figure 7-3 (b) presents th e fraction of hydrogen release in the samples hydrogenated at different temperat ures. It can be noticed that the fraction of hydrogen released in the 1st stage increased with decreasing the hydrogenation temperature. Approximately, 35% of the total hydrogen was released in the first stage for sample hydrogenated at 180C while only 10% was releas ed when hydrogenated at 210C. However, it was not confirmed whether the hydro gen release in the first stage was due to the dissociation of MgH2. 7.1.3 Phase Evolution during Desorption of Mg-Al Powders The evolution of phases during desorption of powders was evaluated using the High temperature X-ray diffraction unit. The Mg-8 at%Al powder hydrogenated at 180C was placed in the sample holder of the XRD and was heated in air at 5K/min. The XRD profile of the sample was collected as a function of temperature (and tim e) and is shown in Figure 7-4. As marked by dashed lines on this figure, it can be noticed that there is a drop in intensity of the (110) peak of MgH2 (2 theta = 28.59) with respect to the other peaks in the temperature range of 100-150C. The drop in intensity of MgH2 peak indicated that hydrogen was released from the MgH2, but upon further heating it was observed th at the intensity of this peak increased. This analysis was not very conclusive for the release of hydrogen from MgH2 as an analysis of peak ratios of Mg to MgH2 varied with temperature (Figure 7-5). Figure 7-5 indicates that the p eak intensity ratio of Mg to MgH2 decreased initially with the increase in temperature till the temperature reaches 200C and remained constant.. It can be noticed that the intensity of MgH2 peak varied with temperature and this can be attributed to the displacement of sample holder during the heating

PAGE 157

157 process. This displacement affects the sampling vol ume of the grains present in the surface of the powder. It was reported earlier that XRD peak in tensities do change because of the displacement of sample holder at high temperatures [125]. Ther efore, the intensity of the peak depends on both the phase transformation and the di splacement of sample holder. Consistent with the TGA results, the hydroge n release from the powder was observed to start around 250C. The intensity of the MgH2 peak decreased from 250C until the hydrogen in the powder was completely released and at this point the peak diminished. Further increase in temperature caused the appearance of a peak co rresponding to that of Mg O, indicating that the hcp-Mg was oxidizing under these conditions. In addition, the peak corresponding to the Mg17Al12 phase was observed to decr ease in intensity in the te mperature range of 300-400C. From this analysis it may be suggested that the hydrogen was released from MgH2 in stage 1 as well as in stage 2. However, the reasons behind the two distinct stages of hydrogen release need further investigation. 7.1.4 Microstructural Analysis of Hydrogenated Mg-Al Powders The SEM/BSE micrographs of the Mg-8at %Al powder hydrogenated at 180C are presented in Figure 7-6. Two ki nds of particles were observed in the hydrogenated condition with different amounts of hydride present in them. These particle s were categorized as Type I and Type II. The Type I particles showed significant amount of MgH2 as shown in Figure 7-6(a) and (b). In comparison to the other particles, T ype II particles exhibited a lower amount of MgH2 as presented in Figures 7-6(c) & (d). As indi cated by the arrows on the micrographs, the surface of the particles was not completely covered with MgH2 and there by the volume fraction of the hydride in the powders was observed to be lowe r. A volume fraction analysis on many particles in the powder revealed that the hydride content in the particles followed a bi-modal distribution

PAGE 158

158 as shown in Figure 7-7. Majority of the particles exhibited lower am ount of magnesium hydride (20-40%) content. In order to understand the reasons behind the 2 stages, a sample hydrogenated at 180C was desorbed partially till the end of the first stage (see Figure 7-3, till 200C) and its microstructure was analyzed. The SEM/BSE microgr aphs of the particles from this powder are shown in Figure 7-8. Most of the part icles exhibited similar amounts of MgH2 and the volume fraction of the MgH2 in these particles was about 20-45% (F igures 7-8(a) and (b)). Furthermore, it is to be noted that some particles, as illustra ted in Figure 7-8(c), did not show the presence of MgH2 at all. A volume fraction analysis on 45 particles of this sample indicated that the hydride fraction in this sample followed a normal dist ribution (Figure 7-9). Figure 7-9 compares the volume fraction of magnesium hydride in the particle s between the Based on the volume fraction analysis it is conclu ded that two kinds of particles existed in the hydrogenated powder which were termed as Type I and Type II. Upon desorption the Type I particles released hydrogen in the temperature range of 100150C (stage 1) while Type II particles released the hydrogen at higher temperatures. The primary difference in Type I and Type II was the amount of hydride present in the particles. The volume fraction analysis of the hydride predicted that the weight loss in the temperature range of 100-150C could be due to release of hydrogen from MgH2. 7.1.5 The Effect of Catalyst on Desorption of Mg-Al Powders The effect of catalyst on de sorption process was investigat ed by coating the Mg-8at%Al powder with two levels of Ni content. Diffe rent amounts of the catal yst on the powder was achieved by changing the proportion of the organometallic Ni to the Mg-Al alloy powders during the Ni coating procedure described in Chapter 3. A Mg-8at%Al powder is coated with half the amount of catalyst was synthesized (0.295g of orga nometallic Ni was used instead of 0. 59 g for

PAGE 159

159 5 gms of Mg-Al powder) and hydrogenated at 180C under 1 MPa pressure of hydrogen for 5 hours. The hydrogen desorption curves for samples w ith two different amounts of Ni coating are presented in Figure 7-10. The desorption curves indicated that hydrogen was released in 2 stages from both samples (Figure 7-10(a)). However, the total amount of hydrogen released in the sample with less content of catalyst Ni was interestingly much lower. The hydrogen release temperatures of these samples are presented in Table 7-4. It can be noticed that the transformation temperatures did not ch ange by reducing the amount of Ni. Figure 7-10(b) illustrates the fr action of hydrogen released in different stages. The amount of hydrogen released in each stag e was changed significantly by va rying the content of catalyst. Only 12% of the total hydrogen present in the powder was observed to release in stage 1 (Temperature range 100-150C) when the catalyst content was reduced to half. However, Mg-Al alloy powder with the high amount of Ni on the su rface of the particles ha ve illustrated that approximately 33% of the total hydrogen is releas ed at low temperature. From these observations it can be concluded that the amount of Ni covering the particle played an important role in the dehydrogenation behavior of the Mg-Al powder. Th ese results suggested that the high amount of Ni on the surface of the powders may result in highe r fraction of particles w ith Al trapped in the MgH2. Hence, larger fraction of MgH2 was destabilized due to the presence of Al trapped and released hydrogen at low temperatures. It was also observed previously that the increase in the amount of catalyst led to higher amount of hydrogen release in less time at the same temperature. However, the faster kinetics in those cases was attributed to the lower diffusion distances of hydrogen on the surface to recombine into hydrogen molecule [86].

PAGE 160

1607.2 Microstructural Evolution during Desorption of Hydride To understand the microstructu ral evolution during desorp tion of hydrogen from the hydrides, samples were partiall y dehydrogenated and analyzed using the SEM and XRD. The 210C hydrogenated Mg-8at%Al powder was heated at 5 K/min in the TGA till the temperature of 290C was reached and then the sample was cooled rapidly at 50 K/min to freeze the microstructure present at 290C. Figure 7-11(a) illustra tes that the hydrogen was released from the surface as evidenced by the el imination of the dark hydride phase and the formation of the grey hcp-Mg. A higher magnifi cation micrograph shown in Fi gure 7-11(b) revealed the formation of the hcp-Mg phase on the surface of a particle. The intermetallic phase, similar to the microstructure of the hydrogenated phase (see Figure 7-11(b)), was found next to the magnesium hydride. Figure 7-11(c) shows that the hydrogen was not released uniformly from the whole surface in some particles. As demonstrated in the high magnification pict ure presented in Figure 7-11(c), hydride grains/colonies st ill existed un-hydrogenated on the surface of the particles. The inhomogenity of the dehydrogenation process may be hypothesized to be associated with the nucleation of the hcp-Mg phase, however this phas e was present next to the hydride phase in all the particles and hence the nucleation will not e xplain the observed phenomena Therefore, the inhomogenity of the dehydrogenation process can be attributed to the non-uniformity in the Ni distribution on the surface of particles. The de hydrogenation process requ ires the formation of hydrogen molecules from the hydrogen dissolved in the hcp-Mg, which, similar to dissociation process, was enhanced by the presence of a catalyst. It should be noted that during hydrogenation, Mg can diffuse into Ni (or vise-versa) and any change in the interfacial structure between Ni and hcp-Mg can possibly affect the dehydrogenation kinetics. Figure 7-12 illustrates the phases and the microstructure of the powders after complete dehydrogenation. The sample was cooled from the dehydrogenation temperature to room

PAGE 161

161 temperature at 10 K/min. The XRD profile of the dehydrogenated powder is shown in Figure 712(a). Hcp-Mg along with the Mg17Al12 phase was observed in the powder while the peak corresponding to the fcc-Al presen t in the initial powder was abse nt. It can be concluded from this analysis that duri ng the dehydrogenation processes the fc c-Al was dissolved in the hcp-Mg phase at high temperatures and precipitated as Mg17Al12 during cooling to room temperature. The microstructure after complete dehydrogenation pr esented in Figure 7-12(b) is similar to the initial microstructure of the powders with hcp-Mg. The contrast corresponding to the Mg17Al12 phase was not observed indicating it s presence as very fine particles. These results suggest that the microstructure of the powder was not changed significantly and can be recycled. 7.3 Discussion The results of this study indicated that MgH2 produced from the electrodeposited Mg-Al powders released hydrogen at much lower temperatures when compared to the MgH2 produced from pure Mg powder. In addition, hydrogenati on temperature played a significant role in desorption of hydrogen from Mg-Al alloy powders. MgH2 fabricated from Mg-Al alloys at low hydrogenation temperatures was observed to re lease hydrogen at low temperatures. During hydrogenation, higher diffusion of Al in hcp-Mg at high temperatures caused Al to diffuse away from the MgH2. Hence the MgH2 synthesized at high temperatur es like 280C was depleted of Al and the hydrogen release temp erature was observed to be very high in this case. So, the decrease in desorption temperature can be attr ibuted to the presence of Al in the MgH2 crystal lattice. From the results presented in the Chapter 6, it was concluded that Al was trapped in MgH2 crystal when the Mg-Al alloy powder wa s hydrogenated at low temperatures. The presence of Al in the MgH2 lattice reduced the magnesium hydr ogen bond strength. In addition, it should be noted that for every two Al atoms present in the latt ice, a vacancy is created to maintain the charge neutrality. Thes e defects created destabilize the MgH2 to a significant extent.

PAGE 162

162 Earlier studies on defects in the MgH2 have shown that desorption temperature of MgH2 is decreased by the increase of these defects [ 61]. The reduction in the bond strength and the formation of defects leads to the release of hydrogen at lowe r temperatures. These results confirmed the theoretical ca lculations performed on MgH2-Al system regarding the destabilization of Mg-H bond [18, 50 97]. It has been predicted that when 8 mol% of Al is added to MgH2, the bond energy between the magnesi um and hydrogen is lowered and the enthalpy of dissociation was reduced from -76 kJ/molH2 to -28kJ/mol H2 [97]. The equilibrium temperature of hydrogen release from MgH2 at 1 atm pressure of hydrogen is 288C. From the results observed in this study it can be noticed that the desorption temperature was reduced significantly to 115C in some particles and 260 C in the other particles when hydrogenated at low temperatures. However, the MgH2 produced at 280C from Mg-Al alloy powders released hydrogen at very high temperatures when compared to the pure Mg powder. The reasons behind this phenomenon are not very clear, but can be attributed to the distri bution of catalyst. Microstructural anal ysis of the Mg-8at %Al powder hydrogenated at 180C for 5 hours indicated that particle s were hydrogenated in homogeneously and exhibited different amounts of hydride. The variation in the amount of hydride can be attributed to the amount of Ni present on the surface of different particles. The studies carried out on the Ni distribution on the initial particles before hydrogenation revealed that th e Ni distribution was not homogeneous and the total content on the particles was also not simila r. Figure 7-13 presents SEM/BSE micrographs (a and c) of 2 Mg-8at%Al particle s along with their correspondi ng SEM/BSE micrographs (b and d) after hydrogenation at 180C. It can be noticed that from the EDS maps that the Ni distribution (green dots) was not same in both the particles. ED S Map of Ni in Figure 7-13(d) elucidates that big part icles of Ni were present on the pow der. A comparison of the EDS maps

PAGE 163

163 illustrated that the amount and the distribution of Ni in the particles were not the same. Hence, it can be predicted that the amount of hydrogenati on in both the particles will be different. As mentioned in Chapter 2, the role of the catalys t during the hydrogenation pr ocess is primarily to break the hydrogen molecule into two hydrogen atoms [85]. Therefor e, the particles with higher content of Ni absorb hydrogen at a higher rate and the nucleat ion of magnesium hydride takes place at a faster rate. The fast transformation ra te provides less time for diffusion of Al in the material and therefore more Al is anticipated to be trapped in the hydride phase of the particle. Thus, it can be suggested that th e hydride in the part icles with higher amount of Ni might have higher amount of Al in MgH2 and it is expected that the enthalpy of dissociation of MgH2 in these particles is decreased to a significant exte nt. This leads to two st ages of hydrogen release observed in the hydrogenated Mg-Al alloy powders. Similar release of hydrogen in two different stages was observed in hydrogenated samples of Mg produced by equi-channel angular pressing. However, the temperature range observed in this study was higher and the two stage release was attributed to the differences in the distribution of grain size in the material [126]. Results of the present study indicate that changing the amount of catalyst affects the amount of hydrogen released due to larger fractio n of particles with hi gher amount of Ni. These results are in agreement with the published data for MgH2 with different amount of transition metal catalysts. In a study with 0-10at% of Ni ball milled with on MgH2, it was shown that the 1wt% of Ni in contact with the MgH2 surface releases less hydrogen from MgH2 when compared to the powder with 5wt% Ni on MgH2 surface [21]. From the above discussions it can be c oncluded that the destabilization of MgH2 can be achieved by creating defects via alloying with Al and reduce th e bond strength between Mg and hydrogen. The findings in the present study suggest that hydrogenation at low temperatures and

PAGE 164

164 under the conditions with high growth rate of hydride can lead to the formation of MgH2 with Al trapped inside the crystal lattice. The trapped Al inside the MgH2 crystal creates the defects that weaken the bond strength between magnesium and hydrogen. The trapping of Al in MgH2 contribute for the destabilization of MgH2 and hence reducing the desorption temperature. 7.4 Summary and Conclusions Hydrogen desorption behavior of both pure Mg powders and elec trodeposited Mg-Al powders was studied using TGA and detailed mi crostructural analysis. The major conclusions that can be drawn from this study are as follows: When Al is trapped in the crystal structur e, desorption temperature can be reduced to temperatures as low as 115C. The release of hydrogen in the Ni coated electrodeposited Mg-Al alloy powders hydrogenated at low temperatur es takes place in two stag es. This phenomenon can be attributed to the existence of a bimoda l distribution of hydride contents of the hydrogenated powders. The presence of this bimodal distribution is believed to be associated with the inhomogeneous distribution of Ni catalyst on the su rface of particles. Decreasing the hydrogenation temperature incr eases the amount of hydrogen released in the 1st stage of hydrogen release. This is attributed to the larger entrapment of Al in MgH2 at lower temperatures due to the slower diffusion of Al in the hcp-Mg phase. Reduction in the catalyst content decreases the total amount of hydrogen in the powder and also the amount of hydrogen released in the 1st stage. Microstructural analysis of partially dehydrogenated particle s indicates that the hydrogen is released from the surface of the powders.

PAGE 165

165 Table 7-1. Hydrogen release temperatures for pure Mg and Mg-Al alloy powders hydrogenated at 210C. Stage 1 Stage 2 On-Set Temperature (C) Peak Temperature (C) On-Set Temperature (C) Peak Temperature (C) Pure Mg 296.75 349.91 Mg-Al 102.13 114.35 265.95 327.46 Table 7-2. The hydrogenation contents of pure Mg and Mg-Al alloy powder observed during the absorption and desorption. Hydrogen Capacity Absorption (wt%) Desorption (wt%) Pure Mg 6.17 6.09 Mg-Al 2.93 2.90 Table 7-3. Hydrogen release temperatures for Mg-Al alloy powders hydrogenated at 180, 210, and 280C. Stage 1 Stage 2 On-Set Temperature (C) Peak Temperature (C) On-Set Temperature (C) Peak Temperature (C) 180 93.75 108.06 213.25 258.85 210 102.13 114.35 265.95 327.46 280 430.09 433.46 Table7-4. Hydrogen release temperatures for Mg -Al alloy powders hydrogenated at 180C with different amounts of Ni. Stage 1 Stage 2 On-Set Temperature (C) Peak Temperature (C) On-Set Temperature (C) Peak Temperature (C) Low Ni 93.41 107.05 215.65 258.85 High Ni 93.75 108.06 213.25 265.92

PAGE 166

166 93 94 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)210C Mg-Al Pure Mg a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (fraction)210C Pure Mg Mg-Al b 5K/min 93 94 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)210C Mg-Al Pure Mg a 5K/min 93 94 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)210C Mg-Al Pure Mg a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (fraction)210C Pure Mg Mg-Al b 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (fraction)210C Pure Mg Mg-Al b 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (fraction)210C Pure Mg Mg-Al b 5K/min Figure 7-1. Hydrogen release cu rves (TGA) for pure Mg and Mg-8at%Al powders hydrogenated at 210C, (a) represents the % weight loss, and (b) fraction of hydrogen released as a function of temperature. hcp-Mg Mg17Al12W 3035404550 2 theta (deg)Intensity (AU)Pure Mg Mg-Al hcp-Mg Mg17Al12W 3035404550 2 theta (deg)Intensity (AU)Pure Mg Mg-Al Figure 7-2. The XRD patterns of the desorbed powders for pure Mg and Mg-8at%Al powder revealing that the phases are similar to that of initial powders.

PAGE 167

167 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)180C 210C 280C a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (Fraction)180C 210C 280C b 5K/min 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)180C 210C 280C a 5K/min 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)180C 210C 280C a 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)180C 210C 280C a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (Fraction)180C 210C 280C b 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (Fraction)180C 210C 280C b 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500 Temperature (C)Hydrogen release (Fraction)180C 210C 280C b 5K/min Figure 7-3. Hydrogen release cu rves (TGA) for Mg-8at%Al pow ders hydrogenated at 180, 210, and 280C. (a) represents the % weight loss, and (b) fraction of hydrogen released as a function of temperature. MgH2 MgOMg MgOMgH2 MgH2 Mg Mg Mg MgH2+ Al12Mg17 T e m p e r a t u r e ( C ) 2 t h e t a ( d e g )5K/min MgH2 MgOMg MgOMgH2 MgH2 Mg Mg Mg MgH2+ Al12Mg17 T e m p e r a t u r e ( C ) 2 t h e t a ( d e g ) MgH2 MgOMg MgOMgH2 MgH2 Mg Mg Mg MgH2+ Al12Mg17 T e m p e r a t u r e ( C ) 2 t h e t a ( d e g )5K/min Figure 7-4. Phase evolution dur ing desorption in Mg-Al alloy powders performed in high temperature XRD. The marked circle indicate s the drop in intensit y of the (110) peak of MgH2.

PAGE 168

168 0 0.5 1 1.5 2 0100200300400500 Temperature (C)Peak Ratios (Mg/MgH2) Figure 7-5. Peak intensity ratios of Mg:MgH2 at different temperatures. 20 m b hcp-Mg + Mg17Al12 MgH2 20 m d hcp-Mg + Mg17Al12 MgH2 20 m c hcp-Mg + Mg17Al12MgH2 20 m a hcp-Mg + Mg17Al12 MgH2 20 m b hcp-Mg + Mg17Al12 MgH2 20 m b hcp-Mg + Mg17Al12 MgH2 20 m d hcp-Mg + Mg17Al12 MgH2 20 m d hcp-Mg + Mg17Al12 MgH2 20 m d hcp-Mg + Mg17Al12 MgH2 20 m c hcp-Mg + Mg17Al12MgH2 20 m c hcp-Mg + Mg17Al12MgH2 20 m a hcp-Mg + Mg17Al12 20 m a hcp-Mg + Mg17Al12 MgH2 Figure 7-6. SEM/BSE micrographs of Mg-8at%Al alloy powder hydrogenated at 180C. (a,b) Type I particles revealing higher amount of hydrides (c,d) Type II particles with less amount of hydride.

PAGE 169

169 0 0.2 0.4 0.6 00.20.40.60.8 Hydride (fraction)Particles (fraction) Figure 7-7. Distribution of di fferent types of particles in Mg-8at%Al powder hydrogenated at180C 50 m a hcp-Mg + Mg17Al12 MgH2 50 m a hcp-Mg + Mg17Al12 MgH2hcp-Mg + Mg17Al12 c hcp-Mg + Mg17Al1220 m c hcp-Mg + Mg17Al1220 m 10 m MgH2b Mg17Al12 10 m MgH2b Mg17Al12 Figure 7-8. SEM/BSE micrographs of Mg-8at%Al alloy powder af ter the release of hydrogen in 1st stage. (a,b) particles revealing amount of hydride present. (c) particles showing the absence of hydride.

PAGE 170

170 0 0.2 0.4 0.6 00.20.40.60.8 Hydride (fraction)Particles (fraction) Figure 7-9. Comparison of the distribution of hydride phase between the powders in hydrogenated condition and after the 1st stage of hydrogen release 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)Low Ni High Ni a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500Temperature (C)Hydrogen Release (fraction)Low Ni High Ni b 5K/min 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)Low Ni High Ni a 5K/min 95 96 97 98 99 100 101 0100200300400500 Temperature (C)Wt Loss (%)Low Ni High Ni a 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500Temperature (C)Hydrogen Release (fraction)Low Ni High Ni b 5K/min 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0100200300400500Temperature (C)Hydrogen Release (fraction)Low Ni High Ni b 5K/min Figure 7-10. Hydrogen release curves (TGA) for Mg-8at%Al powders hydrogenated at 180C with different amounts of catalyst, (a) represents the % weight loss, and (b) fraction of hydrogen released as a function of temperature.

PAGE 171

171 MgH2 20 m hcp-Mg Mg17Al12 a MgH2 10 m hcp-Mg Mg17Al12 b hcp-Mg 5 m d MgH2Mg17Al12Ni hcp-Mg 10 m c MgH2 Mg17Al12Ni MgH2 20 m hcp-Mg Mg17Al12 a MgH2 20 m hcp-Mg Mg17Al12 a MgH2 10 m hcp-Mg Mg17Al12 b MgH2 10 m hcp-Mg Mg17Al12 b hcp-Mg 5 m d MgH2Mg17Al12Ni hcp-Mg 5 m d MgH2Mg17Al12Ni hcp-Mg 10 m c MgH2 Mg17Al12Ni hcp-Mg 10 m c MgH2 Mg17Al12Ni Figure 7-11. SEM/BSE micrographs of the partially dehydrogenated powders (a,c) revealing that dehydrogenation starts on the surface. (b,d) higher magnification micrographs indicating the nucle ation of hcp-Mg.

PAGE 172

172 20304050 2 theta (deg)Intensity (A U)hcp-Mg Mg17Al12a 20304050 2 theta (deg)Intensity (A U)hcp-Mg Mg17Al12a hcp-Mg + Mg17Al1250 m Figure 7-12. (a) XRD profile of an Mg-8at%A l powder after dehydrogenation in TGA, and (b) SEM/BSE micrograph of th e un-hydrogenated particle. b

PAGE 173

173 10 m 10 m Ni a b d c 10 m 10 m 10 m Ni 10 m Ni a b d c Figure 7-13. SEM/BSE micrographs and the co rresponding EDS maps of Ni of Mg-8at%Al powders hydrogenated at 180C revealing the differences in the Ni coating (amount of Ni) on different particles.

PAGE 174

174 CHAPTER 8 LOW TEMPERATURE HYDROGEN ABSORPTION PHOENOMENON IN ELECTRODEPOSITED POWDERS According to the DOE targets a suitable hydr ogen storage material for PEM fuel cell technology require a gravimetric capacity of 2kWh/kg, a volumetric capacity of 1.5kWh/L and the hydrogen release temperature around 6080C [ 127]. To achieve these requirements another class of materials, which are porous in nature, is being developed. These materials can adsorb or absorb hydrogen under high pressure and release at relatively lo w temperatures [128, 129]. The materials under consideration for this type of hydrogen storage are generally carbon based porous materials like carbon nanotubes, carbon nanofibers, fullerenes, gra phite nanostrutctures [130]. In addition to these materials recent studie s have shown that hydrogen can be physisorbed or chemisorbed into various porous metallic na nostructures like Ni nanoclusters and Pd nanostructures [131, 132]. The hydrogen absorption onto the carbon based materials is primarily attributed to their high surface area and the active bonds on the surface [130]. However, due to their low heat of adsorption, typically around 4-10kJ/mol, the adsorption can take place only at sub-zero temperatures and under hi gh pressure of hydrogen. For example, an activated carbon is reported to absorb 7 wt% of hydrogen at 77K under a pressure of 20 MPa [129]. The maximum hydrogen absorption in carbon based materials re ported at room temperature is about 2 wt% under 70 MPa pressure [133]. The major drawback of the carbon based materials is to produce ultra clean nanostructures with high surface area as the fabrication processes involves chemical routes and the absorption of hydrogen is a su rface sensitive property. Furthermore, the absorption of hydrogen is possible only at subzero temperatures which is not very economic and feasible for practical applications. Recently, metallic nanostructures have also been investigated for their hydrogen absorption properties. Nanostructures of dblock metals either supported by an organic frame work or

PAGE 175

175 activated carbon are reported to adsorb 0.5-1. 0 wt% of hydrogen at room temperature [134]. The mechanism for the adsorption on the metal surface is not clear. However, it is predicted that the spill over effect which describes the mechanis m for catalysts could be the reason behind the absorption [134]. Similar studies have been repo rted in literature wher e the hydrogen is trapped inside the metal lattices [135]. Furthermore, in metals, it has been noticed that hydrogen can be chemisorbed into the meal lattice and can be stored near the vacanci es, dislocations, grain boundaries, and at the interface of se cond phase particles [135, 136]. In this study, interesting observations were made duri ng the hydrogen absorption of electrodeposited Mg-Al a lloy powders at low temperatures and high pressures. Hydrogenation tests conducted at low temperat ures around 40-100C and up to hydrogen pressure of 9MPa indicate that both the Ni coat ed and as deposited powders absorb hydrogen. The hydrogenation results and the possible reasons for the observe d phenomenon are presented in this chapter. 8.1 Hydrogen Absorption in Electrodeposited Powders Figure 8-1 presents the PCT curves of electrodeposited Mg-8 at%Al alloy powder developed at 40 and 60C in the as deposited form without coating with Ni. 0.02wt% hydrogen was absorbed into the material during the absorption part and even upon desorption, the material continued to absorb hydrogen till the pressure in the system reached 2MPa and the total amount of hydrogen absorbed was about 0.035wt%. This observation suggests that the Mg-Al powder was not saturated at each step during the absorp tion part. Below the pressure of 2 MPa, partial amount of hydrogen was released and after th e PCT experiment about 0.028wt% of hydrogen was remained in the material. The PCT curve at 60C developed after the 40C illustrated that the hydrogen absorption increased from 0.03 wt% to 0.037 wt% and approximately 0.026wt% of hydrogen was remained in the powder.

PAGE 176

176 However, in a different batch of Mg-8at%Al powders, which were prepared under similar conditions, a high amount of hydrogen (2.7 wt%) was absorbed into the material at low temperatures. Figure 8-2 illustrates the PCT curves at 40 and 60C of the alloy powder fabricated in. Figure 8-2(a) reveals that about 2.6 wt% of hydrogen was absorbed by the powder at 40C and 9 MPa pressure. A continuous increase of hydr ogen absorption was noticed as the pressure of hydrogen increased in steps. During the initi al part of desorption, 0.4wt% of hydrogen was released till the pressure in the system reached 8.5 MPa and upon further decrease in pressure hydrogen was not released signifi cantly. Figure 8-2 (b) illustrates a PCT curve developed at 60C. Interestingly during the absorption part of the PCT cu rve at 60C, the alloy powder released hydrogen even as pressure of the syst em was increased. This phenomenon occurred till the pressure in the system reached 2 MPa. Approximately 1.3wt% of hydrogen was released during this part. After further increase in pre ssure above 2 MPa, the Mg-Al powder started to absorb hydrogen and finally the total content of hydrogen in the powder reached about 2 wt% at the maximum pressure of 9 MPa. The desorption part of the PCT curve followed a similar trend to that of the PCT curve at 40C, where noticeab le amount of hydrogen (0.3wt%) was released at high pressures followed by almost no hydrogen desorp tion. The results of the PCT curves at 40 and 60C suggest that about 2-3 wt% hydrogen is absorbed into the material under high pressures and it is not released under these conditions. It should be noticed that the amount of hydrogen absorbed in a material fabricated under similar conditions exhibited much less hydrogen capacity (Compare Figures 8-1 and 8-2). Similar observat ions have been reported in the literature for carbon based materials. For example, in a particular study reported 65.55 wt% of hydrogen was adsorbed at room temperat ure and under 12 MPa pre ssure in graphite nanofibers [137]. However, these experiments have not been reproducible and the differences

PAGE 177

177 were attributed to the vari ed surface conditions produced in different experiments. The inconsistency in the hydrogen absorption results of simila r composition powders at low temperatures can be attributed to the surface condition of the powders. 8.2 Hydrogen Absorption in Ni coated Electrodeposited Powders The PCT curves developed for the Ni coat ed Mg-8at%Al powders at 40 and 60C are shown in Figure 8-3. The powder employed for the Ni coating was taken from the same batch of powders fabricated to conduct the low temperature hydrogenation tests that were shown in Figure 8-1. The amount of hydrogen absorbed was approximately 0.07 wt% at 40C and 0.075 wt% at 60C. In contrast, to the uncoated powders, the hydrogen is released during the desorption part of the PCT curve at these te mperatures. However only 0.03 wt% of the total hydrogen was released and about 0.04wt% of hydrogen remained in the material. It is noticed that same amount of hydrogen was present afte r the PCT curve at both the temperatures. 8.3 Discussion The results observed in the present study suggest that electrodeposited Mg-Al alloy powders can absorb hydrogen in the temperature range of 40-100C. A comparison of solubility of hydrogen at high temperatures (350-400C, Table 5-5) and low temperatures (40-60C) in Nicoated electrodeposited Mg-8at%Al powder reveals that higher amount of hydrogen is soluble in the powder at low temperatures and high pressu res. However, at a constant pressure the solubility of hydrogen increased w ith increase in temperature. Fo r example, at 1.89 MPa pressure (equilibrium plateau pressure of hydride forma tion at 395C), the solubility of hydrogen was about 0.02 wt% at 60C and 0.04wt% at 395C. Under the condition of low temperatures, the formation of magnesium hydride is hindered due to the high nuclea tion barrier and therefore, the Mg-Al alloy powder absorb continuous ly hydrogen at higher pressures.

PAGE 178

178 However, the high amount of hydrogen absorpti on observed in the case presented in Figure 8-2 cannot be explained by the solubility of hydr ogen. Specifically, at lo w temperatures, it has been reported that the hydrogen can be physisorbed on the surfaces at diffe rent active sites [134]. Further more, defects present in crystalline materials can act as hydrogen trapping sites [134]. Hydrogen atoms, due to their small size, occupy interstitial sites in the metal lattice and the diffusion of hydrogen in metals is reported to be very high [138]. For example, the diffusivity of hydrogen in vanadium is 2 x 1012 jumps per second which exceeds 15-20 orders of magnitude for interstitials like oxygen, nitr ogen and hence can diffuse easily in the metal lattice [138]. In addition, the small size of hydrogen atom permits dense packing of hydrogen atoms in metal host lattices [134]. The above studies suggest that the hydrogen can be st ored in various defects of the materials like dislocations, vacancies and gr ain boundaries. Therefore the amount of hydrogen trapped in a material depends on the microstruc ture, surface area, and th e affinity of the host lattice towards hydrogen. Materials with large amount of defects and interfaces have been reported to absorb significant amount of hydr ogen [131, 134, 139, 140]. For example, the solubility of hydrogen in the ball milled samp les with high surface area, more of grain boundary area, dislocations and point defects has been re ported to be higher than the conventional grain size materials [131]. In a niobium sample the solubility of hydrogen increased from 0.06H/Nb atom to 0.37H/Nb atoms when the grain size of the Nb was decreased from 200 nm to 10 nm. Therefore, the solubility of hydrogen observe d in the present study can be explained by considering the microstructure of the electrodeposited material. The hydrogen can be either physisorbed onto the surface or absorbed into the material. To verify the physisorption mechanism, the surface area of the powder was measured using the BET technique. The BET curve obtained is shown in Figure 8-4. This curve illu strates that the amount

PAGE 179

179 of nitrogen absorbed at various pressures under standard conditions. By employing the molecule length of nitrogen, the total surface area is calculated. The BET surface area measurements indicated that the total connected surface area of the powder was about 1.1m2/g. Assuming that a monolayer of hydrogen molecules was physisorbed on the surface of the powders, the hydrogen content absorbed in the experi ments was calculated using diameter of the hydrogen molecule (289 X10-12m) [141]. The amount of powde r used during the PCT experiments was about 0.34 g. Using the BET surface area, the total amount of h ydrogen absorbed on the powder is obtained as 0.2 wt%. The hydrogen calculated using this met hod is higher than the value observed in the PCT experiments for Ni coated powders. Howe ver, if the hydrogen is physisorbed on the surface, upon desorption the powder should rele ase complete hydrogen as the activation energy for breaking the bonds in physisorption is very low. Hence, it can be concluded that the total hydrogen absorbed in the Mg-Al po wder may not be due to physisor ption. Furthermore, the high amount (2.5wt%) of hydrogen absorbed in uncoa ted powders can not be explained by this phenomenon. Previously it was shown that the electrodepos ited Al-Mg powders fabricated in our group can contain significant amount of porosity due to the deposition c onditions [34]. Microstructural analysis of the electrodeposited powders conducte d using TEM revealed that the Mg-rich Mg-Al powders also contain significant amount of porosity. The TEM micrographs of the Mg-Al powder investigated for these experiments are shown in Figure 8-5. The under focused and over focused micrographs in bright field mode are also presented along with the focused image to reveal the porosity of the electrodeposited powde r. It can be noticed th at significant amount of pores of different sizes are pres ent in the material. The diffrac tion patterns corresponding to the image indicated that the material is hcp-Mg The BET surface tests performed on the Mg-Al

PAGE 180

180 powders will not include the surface area of th e pores inside the material. Therefore, the physisorption of hydrogen inside the Mg-Al alloy powder may not be explained using the surface area obtained from the BET surface area calcul ations. The amount of closed porosity was investigated using the density measurements. Th e density of the Mg-Al powder was measured using the Pyknometer. The density value measured for uncoated Mg-Al powder was 1.7958 gm/cc. The theoretical density was calculated using the lattice parame ter of Mg-8at%Al powder measured from XRD. The theoretical density of Mg-Al powder calculated was 1.8074 gm/cc. Even though the density values are not significan tly different, the values suggest that closed pores are present in th e electrodeposited powder and the volum e fraction of pores was estimated to be 0.65%. The amount of hydrogen absorbed (physisorbed) for the pore sizes noticed in TEM and in the range of observed porosity volume (0.5-1.0 % of volume por osity) are calculated. It is assumed that a monolayer of hydrogen is formed on the surface of the pore. The total surface area of the available pores in the powder wa s calculated and the diameter of the hydrogen molecule, as mentioned in the literature, was used to calculate the to tal amount of hydrogen absorbed [141]. From the TEM micrographs it can be noticed that different sizes of pores are present. The calculations carried out suggest that the maximum amount of hydrogen that can be physisorbed into the pores is about 0.06wt % for an average pore size of 30 nm with 0.65 volume % of porosity. Initial studies on electrodeposit ed fcc-Al rich powders indicated that hydrogen can also be absorbed in the material. These alloy powders were prepared under similar electrodeposition conditions as that of the hcp-Mg rich all oy powders and showed extensive porosity. The hydrogenation curve of an fcc-Al rich powder at 100C is presented in Figure 8-6. The total

PAGE 181

181 amount of hydrogen absorbed into the material was observed to be very low around 50 micro moles. The weight percent of the hydrogen was not calculated as the weight of the powder used was not known but based on our previous experime nts it can be predicted that 0.4-0.5 wt% of hydrogen was absorbed in this experiment. When the temperature of the PCT experiment was increased from 100C to 150C under a hydrogen pressu re of 4.6 MPa, a release of hydrogen was observed as shown in Figure 8-7. This observation, which is similar to the finding for the Mgrich powders, suggests that higher activation en ergy is required to release hydrogen from the Mg-Al powder. These results suggest that the absorption of hydrogen is independent of the composition of the powders but de pends on the microstructure. Based on the results observed in this study and the calculations carried out, the unusual high hydrogen absorption capacity of the powders produced in this study at low temperatures can partially be attributed to th e physisorption of hydrogen in th e porous structure and the high surface to volume ratio of the elec trodeposited powders. However, the further work is needed to address the large amount of hydrogen absorption and the fact that it is not released when the pressure is reduced. 8.4 Summary The results of this study reveal that el ectrodeposited nanoporous nanocrystalline Mg-Al powders can potentially absorb large quantity of hydrogen (2.5wt%), i.e. beyond the solubility limit, at low temperatures and high pressures wi thout formation of a hydride phase. The lack of repeatability of this phenome non is consistent with previously reported studies on hydrogen absorption in high surface area materials. Such inconsistencies suggest th at the surface condition plays an important role in the hydr ogen capacity at low temperatures Coating of Ni is associated with the treatment of the electrodeposited powde rs with chemicals and the low absorption of hydrogen may suggest that the surface of the po wder is changed during this procedure. The

PAGE 182

182 calculations based on the physisorption of the material did not explain the observed high capacity. The fact that hydrogen rele ase requires thermal activation suggests that it is trapped in the material rather than physisorped.

PAGE 183

183 0 2 4 6 8 10 00.010.020.030.04 Hydrogen Content (Wt%)Pressure (MPa)40C 60C 0 2 4 6 8 10 00.010.020.030.04 Hydrogen Content (Wt%)Pressure (MPa)40C 60C Figure 8-1. PCT curves developed at (a) 40 C (b) 60C for electrodeposited Mg-8at%Al alloy powder without Ni coating 0 2 4 6 8 10 -1.5-1-0.500.51 Hydrogen content (wt%)Pressure (MPa)b60C 0 2 4 6 8 10 00.511.522.53 Hydrogen Content (wt%)Pressure (MPa)a40C 0 2 4 6 8 10 -1.5-1-0.500.51 Hydrogen content (wt%)Pressure (MPa)b60C 0 2 4 6 8 10 -1.5-1-0.500.51 Hydrogen content (wt%)Pressure (MPa)b60C 0 2 4 6 8 10 00.511.522.53 Hydrogen Content (wt%)Pressure (MPa)a40C 0 2 4 6 8 10 00.511.522.53 Hydrogen Content (wt%)Pressure (MPa)a40C Figure 8-2. PCT curves for electrodeposited Mg-8 at%Al alloy powder developed at (a) 40C (b) 60C from another batch of electrode posited powders without Ni coating.

PAGE 184

184 0 2 4 6 8 10 00.020.040.060.080.1Pressure (MPa)40C 60CHydrogen content (Wt%) 0 2 4 6 8 10 00.020.040.060.080.1Pressure (MPa)40C 60CHydrogen content (Wt%) Figure 8-3. PCT curves of Ni coated Mg-8at%Al alloy powder at 40 and 60C. 0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 0.00.10.20.30.4Pressure (P/P0)1/(W((P0/P)-1))Slope = 2904, Intercept = 35.6 Co-relation Coefficient= 0.986Surface area = 1.19 m2/g 0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 0.00.10.20.30.4Pressure (P/P0)1/(W((P0/P)-1))Slope = 2904, Intercept = 35.6 Co-relation Coefficient= 0.986Surface area = 1.19 m2/g Figure 8-4. BET curve developed for measur ing the surface area of the Mg-8at%Al alloy powder.

PAGE 185

185 0.2 m c 0.2 m 0.2 m b a 0.2 m c 0.2 m 0.2 m b a Figure 8-5. TEM micrographs of Mg-Al powder re vealing its porosity, (a) underfocussed image, (b) focused image, and (c) over focused image.

PAGE 186

186 0 1 2 3 4 5 6 7 8 9 10 0204060 Hydrogen uptake ( moles)Pressure (MPa)100C 0 1 2 3 4 5 6 7 8 9 10 0204060 Hydrogen uptake ( moles)Pressure (MPa)100C Figure 8-6. Hydrogen absorption curv es of Al-Mg powder at 100C. -50 -40 -30 -20 -10 0 10 020406080100 Time (mins)Hydrogen Uptake ( moles)4.6MPa 150C -50 -40 -30 -20 -10 0 10 020406080100 Time (mins)Hydrogen Uptake ( moles)4.6MPa 150C Figure 8-7.Hydrogen absorption curves of Al-M g powder at 150C. Hydrogen uptake vs time curve indicating the releas e of hydrogen at 4.6MPa.

PAGE 187

187 CHAPTER 9 CONCLUSIONS AND FUTURE WORK The hydrogenation and dehydrogenation characteris tics of Mg-Al all oy powders with 010at% Al in hcp-Mg in the te mperature range of 180-400C were established. The Mg-Al alloy powders were fabricated using the electrodeposition technique and the amount of Al in hcp-Mg phase was varied from 4 to 12 at%. The processed Mg-Al powders were coated with Ni as catalyst to enhance the kinetics of hydroge n absorption and desorption. Even though the electrodeposition parameters were se lected to produce supersaturated solid solution of Al in hcpMg, the formation of intermetallic compound Mg17Al12 was observed to be inevitable due to the conditions present during the powder processi ng and Ni coating. A commercial powder was employed for studying pure Mg. The de/hydrogenation characteristics of the Mg-Al and pure Mg powders under equilibrium conditions were es tablished by developing the Pre ssure Composition Temperature (PCT) curves in the temperature range of 275-400C. A plateau region was observed during the absorption part of the PCT curve for pure Mg powder signifying the formation of magnesium hydride. However, instead of a plateau region an increasing slope of hydrogen absorption with pressure was found in Mg-Al alloy powders at 350C. The sloping cu rve during the hydrogen absorption of Mg-Al alloy powder was due to in volvement of more than 2 phases during the hydrogen absorption. The slope of the PCT curves was observed to depend on the composition and the time allowed to the reach th e equilibrium during the experiments. The PCT curves for the electrodeposited Mg-Al powders were divided into various stages based on the changes in the slope of the curves The extent of each stage was found to depend on the composition of the powder studied. Microstruc tural and compositional analyses revealed that Al is rejected from MgH2 during its formation. The rejected Al dissolves in the available hcp-Mg

PAGE 188

188 and precipitates as Mg17Al12 and Al3Mg2 intermetallic compounds, which were observed to hydrogenate at higher pressures. The high diffusi vity of Al at these temperatures and the equilibrium conditions provided dur ing the experiment causes Al to diffuse away from the MgH2 and precipitate as fcc-Al phase. These results show that under equilibrium conditions, MgH2 has very low solubility for Al even at high temper atures, consistent with phase diagram predictions [112]. The equilibrium plateau pressures of magnesium hydride formation for hcp-Mg, Mg17Al12 and Al3Mg2 increase in that order. Interestingly it was noticed, that th e hydrogenation of the powders did not take place homogeneous ly during the abso rption process. Desorption of hydrogen from the hydrogenated Mg-Al powders took place at a relatively constant pressure, which is consistent with th e absence of Al in the magnesium hydride. The plateau pressure of dissociation of MgH2 was lower than that of the plateau pressure of hydride formation and caused the hysteresis in the PCT curv e. At a given temperature, the hysteresis was higher for pure Mg powders than that of Mg-Al powders. The observed hyst eresis is believed to be due to the high energy barrier for nucleati on of the hcp-Mg phase, which requires a lower pressure than the equilibrium pressure. The di fference in the hydrogena tion and dehydrogenation pressures at a given temperature, i.e. the de gree of hysteresis, was smaller for the Mg-Al powders than that of the pure Mg powder. Although the equilibrium pressure of formation increased at a given temperature with the addition of Al, the enthalpy of formation/dissoci ation of magnesium hydride, calculated from Vant Hoff plots, was found not to be affected significantly. This lack of influence, despite the theoretical prediction that Al should reduce this enthalpy by a factor of 3, is attributed to the absence of Al in the MgH2 crystal lattice under equilibrium conditions that exist during the development of PCT curves.

PAGE 189

189 Dehydrogenation of the Mg-Al powders was found to start on the surface of the powders producing hcp-Mg phase. This phase dissolved the available fcc-Al phase and formed the intermetallic compound Mg17Al12. After complete desorption, th e phases and the microstructure of the powders were similar to that of the in itial powders and therefore the powders could be recycled for further hydrogenation experiments. Hydrogenation studies accompanied by deta iled microstructural and compositional evaluations of Mg-Al powders in the temperat ure range of180C to 280C at 1MPa pressure demonstrated that the addition of Al may e nhance the kinetics of hydrogenation due to the presence of Mg17Al12 precipitates which act as nucleation si te for the hydride phase. On the other hand, the rejection of Al that l eads to the formation of more intermetallic phase limits the total hydrogen capacity of these alloy powders. The compositional analysis using EPMA and lattice parameter calculations using XRD showed that the MgH2 produced at 210C and 180C at 1 MPa pressure contained Al, whose amount increased with a decrease in the hydroge nation temperature. This observation suggests that under non-equilibrium conditions Al can be trapped in the MgH2 lattice. The hydrogen release in Mg-Al pow ders was observed to take place in two stages in the temperature ranges of 90-150 C and 250-330C when the MgH2 was developed at 210C and 180C hydrogenation temper atures. However, MgH2 produced at 280C from Mg-Al alloy powders released hydrogen at 430C. The change in the release temperature is attributed to the trapping of Al in the MgH2 phase when produced under non-equi librium conditions, consistent with the theoretical predictions. The two stage rel ease behavioris believed to be associated with the inhomogeneous distribution of Ni in a batch of powders.

PAGE 190

190Suggested Future Work: In this dissertation we have shown that the hydrogen release temperature of MgH2 can be reduced significantly by additi on of Al. The results of this study demonstrate that Al can be incorporated into MgH2 by fabricating th e hydride under nonequilibrium conditions. However, several phen omenons were not completely understood to design these materials for the inte nded application. In order to fully understand this phenomenon, further investigations to understand the microstr ucture and phase transf ormations are required. For example, formation of intermetallic compound Mg17Al12 is inevitable during the hydrogenation/dehydrogenation processes. Hence an understanding of this formation will help in the design of microstructure. Base d on the results in Chapter 7 we have shed some light on the effect of amount of catalyst a nd the mechanism that helps in dehydrogenation. It is worth studying the effect of different amounts of catalyst as its content should be optimized for the optimum performance. The desorption results carried out in this study indicate that the dehydrogenation temperature is redu ced but no kinetic analysis was carried out. Hence, a kinetic study on desorption of the hydrogen from these mate rials is required to understand the various rate limiting mechanisms.

PAGE 191

191 REFERENCES [1] Gross KJ, Chartouni D, Leroy E, Zuttel A, Schlapbach L. J Alloys Comp 1998;269:259. [2] Varin RA, T C, Z.S W. Nanomaterials for solid stat e hydrogen storage, 2009. [3] Schlapbach L, Zuttel A. Nature 2001;414:353. [4] Bowman RC, Fultz B. MRS Bull. 2002;27:688. [5] Sandi. G. The electrochem soc Inter 2004;13:40. [6] Dornheim. M, Klassen. T, Borm ann. R. Jap Soc Prom Sci 2004:1. [7] Chandra. D, Reilly JJ, Chellappa R. J Metals 2006. [8] Douglass DL. Metall Trans A 1975;6:2179. [9] Sakintuna B, Lamari-Darkrim F, Hirs cher M. Int. J. Hydrogen Energy 2007;32:1121. [10] Fukai Y, Metal-Hydrogen system, Germa ny, Springer series in Materials Science 2005. [11] Stampfer JF, Holley CE, Suttle JF. J Amer Chem Soc 1960;82:3504. [12] Zeng K, Klassen T, Oelerich W, Bormann R. Int. J. Hydrogen Energy 1999;24:989. [13] Bogdanovic B, Bohmhammel K, Christ B, Reiser A, Schlicht e K, Vehlen R, Wolf U. J Alloys Comp 1999;282:84. [14] Klose W, Stuke V. Chemie Ingenieur Technik 1992;64:360. [15] Stander CM. J Inorg Nuc Chem 1977;39:221. [16] Liang GX, Wang ED, Fang SS. J Alloys Comp 1995;223:111. [17] Liang G, Boily S, Huot J, Van Neste A, Schulz R. J Alloys Comp 1998;267:302. [18] Song Y, Guo ZX, Yang R. Phys Rev B 2004;69:094205. [19] Zaluska A, Zaluski L, StromOlsen JO. J Alloys Comp 1999;288:217. [20] Liang G, Huot J, Boily S, Van Neste A, Schulz R. J Alloys Comp 1999;292:247. [21] Eisenberg FG, Zagnoli DA, Sheridan JJ. J Less-Common Metals 1980;74:323. [22] Li S, Jena P, Ahuja R. Phys Rev B 2006;74.

PAGE 192

192 [23] Zaluska A, Zaluski L, StromOlsen JO. J Alloys Comp 1999;289:197. [24] Song Y, Guo ZX, Yang R. Mater Sci Engg a-Struct Mater Props Microstruct Proces 2004;365:73. [25] Mintz MH, Gavra Z, Kimmel G, Hada ri Z. J Less-Comm on Metals 1980;74:263. [26] Andreasen A, Sorensen MB, Burkarl R, Moller B, Molenbroek AM, Pedersen AS, Andreasen JW, Nielsen MM, Jensen TR. J Alloys Comp 2005;404:323. [27] Binary Alloy Phase Diagrams, ASM Handbook; 3, [28] Bouaricha S, Dodelet JP, Guay D, Hu ot J, Boily S, Schulz R. J Alloys Comp 2000;297:282. [29] Gavra Z, Hadari Z, Mintz MH. J Inorg Nuc Chem 1981;43:1763. [30] El-Amoush AS. J A lloys Comp 2007;441:278. [31] Bououdina M, Guo ZX. J Alloys Comp 2002;336:222. [32] Urgnani J, Di Chio M, Palumbo M, Feue rbacher M, Fernandez JF, Leardini F, Baricco M. J Phys 2009;144. [33] Sankara Sarma V. Tatiparti, Ebrahi mi F. J. Electrochem. Soc. 2008;155:D363. [34] Tatiparti SSV. Fabrication Of Na nocrystalline Al-Mg Alloy Powders By Electrodeposition And Their Characterization, PhD University of Florida, 2008. p.166. [35] Yao XD, Wu CZ, Du AJ, Lu GQ, Cheng H M, Smith SC, Zou J, He YH. J Phys Chem B 2006;110:11697. [36] Schlapbach L, Seiler A, Stucki F, Siegmann HC. J Less-Common Metals 1980;73:145. [37] Suda S, Kobayashi N, Yoshid a K. J Less-Common Metals 1980;73:119. [38] Bogdanovic B, Claus KH, Gurtzgen S, Sp liethoff B, Wilczok U. J Less-Common Metals 1987;131:163. [39] Zaluski L, Zaluska A, StromOlsen JO. J Alloys Comp 1997;253:70. [40] Zaluska A, Zaluski L, Strom-Olsen JO. Appl Phys A 2001;72:157. [41] Noritake T, Towata S, Aoki M, Seno Y, Hirose Y, Nishibori E, Takata M, Sakata M. J Alloys Comp 2003;356:84.

PAGE 193

193 [42] Bastide JP, Bonnetot B, Letoffe J M, Claudy P. Mater Res Bull 1980;15:1779. [43] Liang G. J Alloys Comp 2004;370:123. [44] Oskarsson F, Stier W, Camargo LG, Jonsson H. Abs Papers Amer Chem Soc 2005;229:U852. [45] Zaluski L, Zaluska A, Stromolsen JO. J Alloys Comp 1995;217:245. [46] Vegge T, Hedegaard-Jensen LS, Bonde J, Munter TR, Norskov JK. J Alloys Comp 2005;386:1. [47] Huot J, Liang G, Schulz R. Appl Phys A 2001;72:187. [48] Zaluski L, Zaluska A, Tessier P, Stro molsen JO, Schulz R. J Alloys Comp 1995;217:295. [49] Liang G, Huot J, Boily S, Schulz R. J Alloys Comp 2000;305:239. [50] Kelkar T, Pal S, Kanhere DG. Chem Phys Chem 2008;9:928. [51] Bououdina M, Grant D, Walker G. Int. J. Hydrogen Energy 2006;31:177. [52] Stander CM. Zeitschrift Fur Phys Chem 1977;104:229. [53] Zhang QA, Wu HY. Mater. Chem. Phys. 2005;94:69. [54] Vigeholm B, Kjoller J, Larsen B, Pede rsen AS. Int. J. Hydr ogen Energy 1983;8:809. [55] Tessier P, Enoki H, Bououdina M, Akiba E. J Alloys Comp 1998;268:285. [56] Zaluski L, Zaluska A, Tessier P, Stro molsen JO, Schulz R. J Alloys Comp 1995;227:53. [57] Imamura H, Masanari K, Kusuhara M, Katsumoto H, Sumi T, Sakata Y. J Alloys Comp 2005;386:211. [58] Charbonnier J, de Rango P, Fr uchart D, Miraglia S, Pontonni er L, Rivoirard S, Skryabina N, Vulliet P. J Alloys Comp 2004;383:205. [59] Asano K, Akiba E. J Alloys Comp 2009;481:L8. [60] Gubicza J, Kassem M, Ungar T. Euro Pow Diff Epdic 8 2004;443-4:103. [61] Borgschulte A, Bosenberg U, Barkhordari an G, Dornheim M, Bormann R. Cata Today 2007;120:262. [62] Ebrahimi F, Zhai Q, Kong D. Scrip Mater 1998;39:315.

PAGE 194

194 [63] Jovic VD, Jovic BM, Pavlovi c MG. Acta Electrochim 2006;51:5468. [64] Murray JL. Bulletin of A lloy Phase Diagrams 1982;3. [65] El-Amoush AS. J A lloys Comp 2008;463:475. [66] Crivello JC, Nobuki T, Kato S, Abe M, Kuji T. J Alloys Comp 2007;446:157. [67] Yabe H, Kuji T. J Alloys Comp 2007;433:241. [68] Gubicza J, Kassem M, Ribarik G, Ungar T. Mater Sci Engg 2004;372:115. [69] Nie JF, Xiao XL, Luo CP, Muddle BC. Micron 2001;32:857. [70] Crawley AF, Lagowski B. Metall Trans 1974;5:949. [71] Duly D, Zhang WZ, Audi er M. Philos Mag A 1995;71:187. [72] Liang G, Schulz R. J Mater Sci 2003;38:1179. [73] Qian S, Northwood DO. Int. J. Hydrogen Energy 1988;13:25. [74] Schwarz RB, Khachaturyan AG. Acta Mater 2006;54:313. [75] Schwarz RB, Khachaturyan AG. Phys Rev Lett 1995;74:2523. [76] Rabkin E, Skripnyuk VM. Sc ripta Materialia 2003;49:477. [77] Huot J, Liang G, Boily S, Van Neste A, Schulz R. J Alloys Comp 1999;295:495. [78] Gross KJ, Spatz P, Zuttel A, Sc hlapbach L. J Alloys Comp 1996;240:206. [79] Zhang Y, Tsushio Y, Enoki H, Akiba E. J Alloys Comp 2005;393:147. [80] Andreasen A. Int. J. Hydrogen Energy 2008;33:7489. [81] Fernandez JF, Leardini F, Bodega J, Sanc hez C, Joubert JM, Cuevas F, Leroy E, Baricco M, Feuerbacher M. J Alloys Comp 2009;472:565. [82] Sandrock G. J Alloys Comp 1999;295:877. [83] Schulz R, Huot J, Liang G, Boily S, Lalande G, Denis MC, Dodelet JP. Mater Sci Engg A 1999;267:240. [84] Pozzo M, Alfe D. Int. J. Hydrogen Energy 2009;34:1922.

PAGE 195

195 [85] Fleisch T, Abermann R. J. Catal. 1977;50:268. [86] Rivoirard S, de Rango P, Fruchart D, Charbonnier J, Vempaire D. J Alloys Comp 2003;356:622. [87] Bazzanella N, Checchetto R, Miot ello A. Appl. Phys. Lett. 2004;85:5212. [88] Holtz RL, Imam MA J Mater Sci 1999;34:2655. [89] Bhat V, Rougier A, Aymard L, Nazr i GA, Tarascon JM. Int. J. Hydrogen Energy 2007;32:4900. [90] Recham N, Bhat VV, Kanda vel A, Aymard L, Tarascon J M, Rougier A. J Alloys Comp 2008;464:377. [91] Oelerich W, Klassen T, Borma nn R. Adv. Eng. Mater. 2001;3:487. [92] Barkhordarian G, Klassen T, Bo rmann R. J Phys Chem B 2006;110:11020. [93] Norskov JK, Houmoller A, Johansson PK, Lundqvist BI. Ph ys Rev Lett 1981;46:257. [94] Tsuda M, Dino WA, Kasai H, Nakanishi H, Aikawa H. Thin Solid Films 2006;509:157. [95] Au M, Wu J, Wang QD. Int. J. Hydrogen Energy 1995;20:141. [96] Gremaud R, Borgschulte A, Lohstroh W, Schreuders H, Zu ttel A, Dam B, Griessen R. J Alloys Comp 2005;404:775. [97] Shang CX, Bououdina M, Song Y, Guo ZX. Int. J. Hydrogen Energy 2004;29:73. [98] Huot J, Liang G, Boily S, Van Neste A, Schulz R. International Symposium on MetalHydrogen System, p.135 1998. [99] Gennari FC, Castro FJ, Urretavi zcaya G. J Alloys Comp 2001;321:46. [100] Huang ZG, Guo ZP, Calka A, Wexler D, Wu J, Notten PHL, Liu HK. Mater Sci Engg A 2007;447:180. [101] Imamura H, Tabata S, Shigetomi N, Ta kesue Y, Sakata Y. J Alloys Comp 2002;330:579. [102] Li.H, PhD. Gainesville 2004;University of Florida. [103] Walsh FC, Herron ME. J Phys D 1991;24:217.

PAGE 196

196 [104] Bazzanella N, Checchetto R, Miotello A, Sada C, Mazzoldi P, Mengucci P. Appl. Phys. Lett. 2006;89. [105] Bogdanovic B. Method of preparing activ e magnesium-hydride or magnesium hydrogen storer systems vol. US Patent 4,554,152, 1985. [106] Checchetto R, Trettel G, Miote llo A. Meas. Sci. Technol. 2004;15:127. [107] Blach TP, Gray EM. J Alloys Comp 2007;446:692. [108] Cullity BD, Stock SR. Elements of X-Ra y Diffraction: Prentice Hall, Newyork, 2001. [109] Klose W, Stuke V. Int. J. Hydrogen Energy 1995;20:309. [110] Fromageau R, Hillairet J, Ligeon E, Mair y C, Revel G, Tzanetakis P. J Appl Phys 1981;52:7191. 111] Shang CX, Bououdina M, Guo ZX. Mater Trans 2003;44:2356. [112] Palumbo A, Torres FJ, Ares JR, Pisani C, Fernandez JF, Baricco A. Calphad-Comp Coupling of Phase Diagrams and Thermochemistry 2007;31:457. [113] Huang YC, Watanabe T, Komatsu R. Proceedings of 4th International Conference in Vacuum Metallurgy 1973:176. [114] Takamura H, Miyashita T, Kamegawa A, Okada M. J Alloys Comp 2003;356:804. [115] ASM Handbook Alloy Phase Diag rams' vol. 3, online resource. [116] Braszczynska-Malik KN. J Alloys Comp 2009;477:870. [117] Schimmel HG, Kearley GJ, Huot J, Mulder FM. J Alloys Comp 2005;404-406:235. [118] Tien HY, Tanniru M, Wu CY, Ebrahimi F. Int. J. Hydrogen Energy 2009;34:6343. [119] Tanniru M, Ebrahimi F. Int J Hydrogen Energy 2009;Submitted. [120] Shannon RD. Acta Crystall Section A 1976;32:751. [121] Kim DJ. J Amer Cer Soc 1989;72:1415. [122] Hong SJ, Virkar AV. J Amer Cer Soc 1995;78:433. [123] American Society for Testing Materials, 35-1185 F. ASTM 1996. [124] American Society for Testing Ma terials, 967-03 E. ASTM 1996.

PAGE 197

197 [125] Shobit O, Eric DW, Jacob LJ, Juan CN. J Amer Cer Soc 2009;9999. [126] Skripnyuk V, Buchman E, Rabkin E, Estrin Y, Popov M, Jorgensen S. J Alloys Comp 2007;436:99. [127] Lennie Klebanoff, Materials Go/No-Go Decisions Made Within the Department of Energy Metal Hydride Center of Excellence (MHCoE), Oct 2007, http://www1.eere.energy.gov/hydrogenandfuelcells/news_detail.html?news_id=11721 Aug 2008. [128] Benard P, Chahine R. Scrip Mater 2007;56:803. [129] Jorda-Beneyto M, Suarez-G arcia F, Lozano-Castello D, Cazorla-Amoros D, LinaresSolano A. Carbon 2007;45:293. [130] Sahaym U, Norton MG J Mater Sci 2008;43:5395. [131] Andrievski RA. P hys Uspekhi 2007;50:691. [132] Zielinski M, Wojcieszak R, Mont everdi S, Mercy M, Bettahar MM. Cata Comm 2005;6:777. [133] de la Casa-Lillo MA, Lama ri-Darkrim F, Cazorla-Amoros D, Linares-Solano A. The J Phys Chem B 2002;106:10930. [134] Pundt A, Kirchheim R. Ann Rev Mater Res 2006;36:555. [135] Turnbull A. Corro Sci 1993;34:921. [136] Krom A, Bakker A. Metall Mater Trans B 2000;31:1475. [137] Chambers A, Park C, Baker RTK, R odriguez NM. The J Phys Chem B 1998;102:4253. [138] Volkl J, Schauman.G, Alefeld G. J Phys Chem Solids 1970;31:1805. [139] Makenas BJ, Birnbaum HK. Acta Metall 1980;28:979. [140] Schober T. Metall Trans 1981;12 A:951. [141] Cheng JR, Yuan XH, Zhao L, Huang DC, Zhao M, Dai L, Ding R. Carbon 2004;42:2019.

PAGE 198

198 BIOGRAPHICAL SKETCH Mahesh Tanniru was born in 1982, at Hydera bad, Andhra Pradesh, India. He did his bachelors degree in metallurgic al engineering from Jawaharlal Nehru Technological University at Hyderabad in 2003. He received his maste rs degree in chemical engineering from the University of Louisiana at Lafayette in 2005. Mahesh joined the Materials Science and Engineering Department at University of Fl orida in Fall 2005 for his PhD and joined Dr. Fereshteh Ebrahimis group in Summer 2006.