<%BANNER%>

Fabrication of Nanocrystalline Al-Mg Alloy Powders by Electrodeposition and Their Characterization

Permanent Link: http://ufdc.ufl.edu/UFE0022697/00001

Material Information

Title: Fabrication of Nanocrystalline Al-Mg Alloy Powders by Electrodeposition and Their Characterization
Physical Description: 1 online resource (166 p.)
Language: english
Creator: Tatiparti, Sankara
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: aluminum, anisotropy, composition, crystallography, dendrites, electrodeposition, evolution, fcc, galvanostatic, hcp, hydrogen, magnesium, microstructure, morphological, morphology, nanocrystalline, nucleation, organometallic, powders, storage, substrate, supersaturated, tem, texture, ultramicrotomy
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: Aluminum-magnesium alloy powders can potentially be used as hydrogen storage materials. In order to enhance the kinetics of hydrogenation it is desirable to have agglomerates of fine powders with very small grain size. In this study, nanocrystalline Al-Mg alloys in the form of powders were successfully fabricated by the electrodeposition technique using an organometallic based electrolyte. Mg was introduced into the electrolyte by a process called pre-electrodeposition. The mechanism for Mg accumulation can be explained considering the electrode reactions as well as the chemical changes in the electrolyte. Using a copper cathode, the effects of the electrolyte composition and current density on composition of the deposit, its constituent phases and morphology were investigated. The magnesium content of the deposits improved with increasing Mg concentration in the electrolyte, temperature and current density. Depending on the composition, the deposits consisted of FCC-Al(Mg) and HCP-Mg(Al) phases and no intermetallic phase was found except for long deposition times. Generally, the deposits formed initially on the copper substrate with three dendritic morphologies namely, rod-like, feather-like and small globular, which eventually evolved into the large globular morphology. This observation is attributed to the establishment of spherical diffusion conditions at the sharp dendrite tips. Potentiostatic studies suggested that the appearance of different morphologies is associated with differing rates of deposition. While the initial dendrites consisted of the FCC Al-rich phase, the large globular morphology manifested as both FCC Al-rich and HCP Mg-rich phases, with the latter always forming over the former. The observation of formation of only the FCC phase implies that the nucleation barrier for the HCP phase on the copper substrate is quite high. The investigation of the effect of substrate, namely, Cu, graphite and Mg, revealed that the HCP phase can directly nucleate on an oxide-free Mg surface. This finding can be explained in terms of surface/interfacial energies. Detailed TEM analysis revealed that the observed morphologies consist of randomly distributed nanocrystalline grains except for the feather-like dendrites, which exhibited a strong crystallographic texture.
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 Sankara Tatiparti.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Ebrahimi, Fereshteh.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022697:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022697/00001

Material Information

Title: Fabrication of Nanocrystalline Al-Mg Alloy Powders by Electrodeposition and Their Characterization
Physical Description: 1 online resource (166 p.)
Language: english
Creator: Tatiparti, Sankara
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: aluminum, anisotropy, composition, crystallography, dendrites, electrodeposition, evolution, fcc, galvanostatic, hcp, hydrogen, magnesium, microstructure, morphological, morphology, nanocrystalline, nucleation, organometallic, powders, storage, substrate, supersaturated, tem, texture, ultramicrotomy
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: Aluminum-magnesium alloy powders can potentially be used as hydrogen storage materials. In order to enhance the kinetics of hydrogenation it is desirable to have agglomerates of fine powders with very small grain size. In this study, nanocrystalline Al-Mg alloys in the form of powders were successfully fabricated by the electrodeposition technique using an organometallic based electrolyte. Mg was introduced into the electrolyte by a process called pre-electrodeposition. The mechanism for Mg accumulation can be explained considering the electrode reactions as well as the chemical changes in the electrolyte. Using a copper cathode, the effects of the electrolyte composition and current density on composition of the deposit, its constituent phases and morphology were investigated. The magnesium content of the deposits improved with increasing Mg concentration in the electrolyte, temperature and current density. Depending on the composition, the deposits consisted of FCC-Al(Mg) and HCP-Mg(Al) phases and no intermetallic phase was found except for long deposition times. Generally, the deposits formed initially on the copper substrate with three dendritic morphologies namely, rod-like, feather-like and small globular, which eventually evolved into the large globular morphology. This observation is attributed to the establishment of spherical diffusion conditions at the sharp dendrite tips. Potentiostatic studies suggested that the appearance of different morphologies is associated with differing rates of deposition. While the initial dendrites consisted of the FCC Al-rich phase, the large globular morphology manifested as both FCC Al-rich and HCP Mg-rich phases, with the latter always forming over the former. The observation of formation of only the FCC phase implies that the nucleation barrier for the HCP phase on the copper substrate is quite high. The investigation of the effect of substrate, namely, Cu, graphite and Mg, revealed that the HCP phase can directly nucleate on an oxide-free Mg surface. This finding can be explained in terms of surface/interfacial energies. Detailed TEM analysis revealed that the observed morphologies consist of randomly distributed nanocrystalline grains except for the feather-like dendrites, which exhibited a strong crystallographic texture.
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 Sankara Tatiparti.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Ebrahimi, Fereshteh.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022697:00001


This item has the following downloads:


Full Text

PAGE 1

1 FABRICATION OF NANOCRYSTALLINE AL-MG ALLOY POWDERS BY ELECTRODEPOSITION AND THEIR CHARACTERIZATION By SANKARA SARMA V. TATIPARTI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Sankara Sarma V. Tatiparti

PAGE 3

3 To my parents, my brother, and my advisor

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my deep est gratitude to my advisor Dr. Ebrahimi, for her invaluable guidance. She has been extremely supportive in al most all the facets of my PhD. Through her I have learnt a great deal of science, thinking pr ocess and general approach to any given problem. She has been very helpful in not only identifyi ng my flaws but also inst rumental in correcting them. Throughout my PhD she has been my motivati on and will serve as the same in the rest of my life. I consider myself extremely lucky to ge t Dr. Ebrahimi as my advisor. Finally, as they say in Sanskrit Guru Devo Bhava, I re spect her with my deepest honesty. I thank all of my present a nd past committee members (Dr. Dempere, Dr. Sinnott, Dr. Baney, Dr. White, Dr. Seifert and Dr. Bourne) fo r their valuable suggestions and time without which progress in my PhD would have been ex tremely difficult. I thoroughly enjoyed the long talks with Dr. White and Dr. Baney on the organo metallic chemistry. They have shown me some of the valuable bibliographical sources on the subject through wh ich I have benefited extensively. I thank Dr. Dempere for valuable in puts on the various charact erization techniques such as microscopy and compositional analysis I was deeply motivated by the extreme enthusiasm and efforts taken by Dr. Sinnott in delivering her knowledge in the Structures course. Dr. Bourne has taught me the TEM course in one of the best possible ways anybody can ever teach. The after-class discussi ons with him have been very helpful for me to gain great insight into the subject. I thank my past and present group memb ers Ebony Westbrooke, Yanli Wang, Mike Kessler, Ian Liu, Mahesh Tanniru, Orlando Rios, Sonalika Goyel, Da niel and Chamoria for their help and support. My special thanks go to Ch amoria in helping me make my TEM sample.

PAGE 5

5 I consider myself to be lucky in gaining so me of the very good fr iends, the list of who includes Rakesh Behera, Dr. Shobit Omar, D ilpuneet Singh, Krishna Prakash Ganesan, Pankaj Nerikar, Abhijit Pramanick, etc. Ot hers are not to be left out. I would like to convey my thanks to the sta ff at MAIC particularly, Wayne and Kerry for their valuable advise on both the characterization and sample preparation techniques. My special thanks go to Rosabel who brings chee rful working environment with her. My special thanks go to Bert Caringal at Vacuum Atmospheres in teaching me a great deal about the operation and maintenance of the glove box. He has become my friend through the long phone calls that we had. Finally, I express my deepest gr atitude to my father and my brother without whose support and permission I would not have come this far to learn a great deal abou t everything. They have been extremely supportive in my life.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............15 CHAPTER 1 INTRODUCTION..................................................................................................................17 2 BACKGROUND....................................................................................................................20 2.1 Preparation of Al-Mg Based Materials for Hydrogenation..............................................20 2.1.1 Earlier Methods of Preparation..............................................................................20 2.1.2 Electrodeposition of Powders.................................................................................22 2.1.2 Electrodeposition of Al-Mg Alloys........................................................................23 2.2 Structure of Different Electrolyte Components................................................................28 2.2.1 AlEt3.......................................................................................................................28 2.2.2 Na[Et3Al-H-AlEt3].................................................................................................29 2.2.3 Na[AlEt4]................................................................................................................29 2.3 Morphological and Microstructu ral Analysis of the Deposits..........................................29 2.3.1 Morphological Analysis.........................................................................................29 2.3.2 Microstructural Analysis........................................................................................33 3 EXPERIMENTAL PROCEDURES.......................................................................................42 3.1 Substrate Preparation...................................................................................................... ..42 3.2 Anode and Reference Electrode Preparation....................................................................44 3.3 Electrolyte Preparation.................................................................................................... .45 3.3.1 Pure Al.................................................................................................................. ..45 3.3.2 Al-Mg Alloys..........................................................................................................46 3.4 Electrodeposition.......................................................................................................... ....48 3.4.1 Experimental Setup................................................................................................48 3.4.3 Al-Mg Alloys..........................................................................................................49 3.5 Sample Preparation......................................................................................................... ..50 3.5.1. Electron Microprobe Samples...............................................................................50 3.5.2. Ultramicrotomy Samples.......................................................................................51 3.5.3. TEM Samples........................................................................................................52 3.5.4. Focused Ion Beam Samples...................................................................................52 3.6 Characterization Techniques............................................................................................53 3.6.1. Compositional Analysis.........................................................................................53 3.6.2. Transmission Electron Microscopy.......................................................................54

PAGE 7

7 3.6.3. Scanning Electron Microscopy..............................................................................54 3.6.4. X-Ray Diffraction..................................................................................................55 4 PRE-ELECTRODEPOSITION: A PROCESS FOR ACCUMULATING Mg IN THE ELECTROLYTE....................................................................................................................63 4.1 Selection of Electrolyte................................................................................................... ..63 4.2 Phase Analysis............................................................................................................. .....64 4.3 Compositional Analysis....................................................................................................65 4.3.1 Deposits................................................................................................................. .65 4.3.2 Electrolyte.............................................................................................................. .65 4.4 Anodic and Cathodic Reactions........................................................................................69 4.5 Roles of Electrolyte Components.....................................................................................71 4.5.1 Component Electrolyte (I)......................................................................................71 4.5.2 Component Electrolyte (II).....................................................................................71 4.5.3 Component Electrolyte (III)...................................................................................72 4.6 Al-Mg Alloy Deposition Scheme.....................................................................................74 4.7 Summary.................................................................................................................... .......75 5 ELECTRODEPOSITION OF PURE Al AND Al-Mg ALLOYS..........................................84 5.1 Electrodeposition of Pure Al............................................................................................84 5.1.1 Current Density Vs Time Results...........................................................................84 5.1.2 Morphological Analysis of Pure Al........................................................................85 5.2 Electrodeposition of Al-Mg Alloys..................................................................................86 5.3 Growth and Morphology of the Deposits.........................................................................87 5.4 Effect of Electrodeposition Parameters on Morphology and Composition Phases of the Deposits................................................................................................................... ......88 5.4.1 Effect of Pre-Electrodeposition Time.....................................................................88 5.4.2 Effect of Deposition Time......................................................................................90 5.4.3 Effect of Current Density.......................................................................................90 5.4.4 Effect of Anode Composition.................................................................................92 5.5 Consistency of the El ectrodeposition Process..................................................................93 5.6 Discussion................................................................................................................. ........93 5.7 Summary.................................................................................................................... .......96 6 MORPHOLOGICAL AND MICROSTRUCT URAL EVOLUTION OF Al-Mg DENDRITES...................................................................................................................... ..108 6.1 Morphological Analysis of Dendrites.............................................................................109 6.1.1 Rod-Like Morphology..........................................................................................109 6.1.3 Feather-Like Morphology....................................................................................110 6.1.4 Globular Morphology...........................................................................................111 6.1.5 Morphological Evolution in Dendrites.................................................................112 6.2 Growth Mechanisms of Morphologies...........................................................................113 6.3 Microstructural Anal ysis of Dendrites............................................................................115 6.3.1 Rod-Like Morphology..........................................................................................115

PAGE 8

8 6.3.2 Feather-Like Morphology....................................................................................115 6.3.3 Globular Morphology...........................................................................................116 6.4 Compositional Analysis..................................................................................................118 6.4.1 Feather-like Morphology......................................................................................118 6.4.2 Globular Morphology...........................................................................................118 6.5 Discussion................................................................................................................. ......120 6.6 Summary and Conclusions.............................................................................................123 7 EFFECT OF SUBSTRATE ON NUCL EATION OF HCP Mg PHASE.............................145 7.1 Effect of Electrolyte Composition on the Nucleation of the HCP Phase.......................145 7.2 Effect of Substrate on Nucleation of HCP Phase...........................................................146 7.3 Discussion................................................................................................................. ......149 7.4 Summary.................................................................................................................... .....152 8 CONCLUSIONS..................................................................................................................157 LIST OF REFERENCES.............................................................................................................159 BIOGRAPHICAL SKETCH.......................................................................................................166

PAGE 9

9 LIST OF TABLES Table page 4-1 Calculations from Mg and Al mass ba lance schemes in electrolyte (30 mL) and deposits....................................................................................................................... .......83 4-2 Estimate for the valence of dissolving or depositing Mg..................................................83 5-1 Experimental parameters in th e electrodeposition of Al-Mg alloys................................107 6-1 Experimental conditions of the sa mples used for compositional analysis.......................144

PAGE 10

10 LIST OF FIGURES Figure page 2-1. Appearance of different hydrides and othe r phases at different compositions of AlMg alloys...................................................................................................................... .....34 2-2 Schematic of the three layers (electric double, diffusion and convection layers) that form ahead of the cathode during electrodeposition..........................................................35 2-3 Schematic of cyclic voltammetry curve depicting different regions for forming different morphologies.......................................................................................................35 2-4 Composition of the deposits (Mg at%) as a function of initial Mg/Al ratio in the electrolyte.................................................................................................................... .......36 2-5 The SEM images of the Al-Mg alloy de posit obtained using electrolyte Na[AlEt4] + 2 AlEt3 + 3.3 Toluene........................................................................................................36 2-6 Structure of AlMe3 (where Me = methyl) dimer...............................................................37 2-7 Structure of [Me3Al-H-AlMe3]-1........................................................................................37 2-8 Structure of [AlEt4]-1. ........................................................................................................38 2-9 The SEM images showing different morphol ogies of Zn electrodeposited at different applied potentials showing chunks, di sks,and hexagonal shaped dendrites......................38 2-10 Anisotropic growth of Zn dendrites...................................................................................39 2-11 Anisotropic growth of Ag dendrites..................................................................................39 2-12 The SEM images of Ni-Co alloys for different star ting ratios of Ni+2/Co+2 in the electrolyte showing di ffferent morphologies.....................................................................40 2-13 The SEM images showing morphological evolution in Au-Ag alloys, Co powders.........40 3-1 Current vs time curves generated during el ectropolishing of Cu electrode when in the absence and presence of wobbling.....................................................................................56 3-2 Aluminum-magnesium deposit showing the irregular coverage of substrate. This irregularity was found to be due to wi ping the substrate with a kimwipe after electropolishing............................................................................................................... ...56 3-3 Grey residue from mixing Na and AlEt3............................................................................57 3-4 Experimental setup showing glove box w ith different accessories, PAR system with computer, and electrodeposition cell setup........................................................................58

PAGE 11

11 3-5 Cyclic voltammetry curve in the case of pure Al, and Al-Mg alloys................................59 3-6 Schematic showing the embedding of Al-Mg dendrites parallel to the cutting face of the mold and the obtained capsule after curing.................................................................60 3-7 The TEM grid with slices of polymer on it........................................................................60 3-8 The FIB sample preparation sequence...............................................................................61 3-9 Broken pellet made of used electrolyte..............................................................................62 4-1 The XRD profiles of Al-Mg alloy powder s produced after pre-electrodeposition for different times with major peaks of all the phases indexed...............................................77 4-2 Magnesium and aluminum in the de posits plotted as function of preelectrodeposition time........................................................................................................77 4-3 Concentration variation of Mg in the electrolyte...............................................................78 4-4 Magnesium (at.%) in the deposit plotted as function of Mg/Al ratio in electrolyte..........79 4-5 Efficiency of the pre-electrodeposition pr ocess as a function of Mg/Al ratio in the electrolyte.................................................................................................................... .......79 4-6 The SEM images of the deposit obtained using el ectrolyte (II)........................................80 4-7 The SEM images of the deposit ob tained using elect rolyte (III).......................................80 4-8 Analysis of black film on anode.showing the presence of MgH2......................................81 4-9 Scheme of Al-Mg alloy depos ition using electrolyte (III).................................................82 5-1 Current vs time curves for electrod eposition of pure Al using two different compositions of electrolyte................................................................................................97 5-2 The SEM images of pure Al deposit fr om electrolyte with different compositions..........97 5-3 Values of potential during a ga lvanostatic deposition conducted at 60 oC........................98 5-4 The SEM images showing three stages of deposit growth namely, ground, nodules, and dendrites in deposits #1 and #3...................................................................................98 5-5 The SEM images showing nodules and dendrites from deposit #4...................................99 5-6 The SEM images of the ground obtained at same current density (60 mA/cm2) and different temperatures (60 oC, and 90 oC)..........................................................................99 5-7 The SEM images showing smooth and r ough globules in deposit #1, and very rough globules in deposit #3......................................................................................................100

PAGE 12

12 5-8 Low magnification SEM images showing extensive dendritic growth in deposit #1, and fine nodular growth in deposit #4.............................................................................100 5-9 The XRD profiles of deposits #1-4 showi ng the effect of pre-electrodeposition time on the phases present.......................................................................................................100 5-10 The SEM images showing the growth of dendrites when deposition time is increased from 10 min (deposit #4), and 30 min (deposit #5).........................................................101 5-11 The SEM images showing different mor phologies of the deposits made at 60-150 mA/cm2............................................................................................................................101 5-12 The SEM images showing rod-like morphol ogy at the root of a globular dendrites, and the presence of rough globules over the smooth ones in deposit #8.........................102 5-13 The SEM images of the cross sectio n of a dendrite rev ealing the porosity.....................102 5-14 Low magnification SEM images demonstr ating the refinement of the dendrites...........103 5-15 The XRD profiles of deposits #6-8 showi ng the effect of current density on the phases present................................................................................................................. .103 5-16 The SEM images showing a general representative morphologies in deposits made using anodes having composition of pur e Mg, and 80 area%Al + 20 area% Mg............104 5-17 The XRD profiles of deposits made us ing anode composition of pure Mg, and 80 area% Al + 20 area% Mg.................................................................................................104 5-18 Temperature variation du ring deposition at 150 mA/cm2 and starting temperature of 60 oC.............................................................................................................................. ...105 5-19 The SEM images at different level of cathode for three depositions done for checking consistency.................................................................................................................... ...105 5-20 The XRD profiles of three depositions done for checking consistency...........................106 5-21 General growth mechanism of de posits under galvanostatic technique..........................106 6-1 The SEM images of the rod-like morphology formed at current density of 100 mA/cm2 and temperature of 60 oC...................................................................................125 6-2 The SEM images showing general f eather-like morphology, and hierarchical behaviour exhibited by feather-like morphology.............................................................125 6-3 The SEM images of feather-like mo rphology showing primary dendrite, and secondary dendrite from head-on position.......................................................................126 6-4 The SEM images of feather-like mor phology showing the angles made by primary arms with the dendrite axis..............................................................................................126

PAGE 13

13 6-5 Schematic showing the angular relations hips and crystallographic directions along which the dendrite and primary arms of the feather-like morphology are growing........127 6-6 The SEM images showing the forma tion of dendrite containing globular morphology..................................................................................................................... .128 6-7 The SEM images showing HCP Mg-rich rough globules forming over the FCC Alrich smooth ones, and coarsening of the sm ooth globules and eventual formation of HCP Mg-rich rough globules...........................................................................................128 6-8 The SEM images showing the hierar chy exhibited by smooth globules, and rough globules....................................................................................................................... .....129 6-9 The SEM images showing morphological evol ution from rod-like, and feather-like to large globules................................................................................................................. ..129 6-10 Morphologies as a function of applied potentials............................................................130 6-11 Morphologies as a function of applied potentials for constant deposited charge............130 6-12 Morphological and mictrostructural description of rod-like morphology.......................131 6-13 Morphological and microstructural de scription of feather-like morphology..................132 6-14 Grain size of the feather-like morphology.......................................................................133 6-15 Morphological and microstructural de scription of fine globular morphology................134 6-16 Grain size of the fine globular morphology.....................................................................135 6-17 Microstructural descri ption of large globules..................................................................136 6-18 Microstructural description of la rge-grained region of large globules............................136 6-19 Microstructural description of fine -grained region of large globules..............................137 6-20 The HCP region of large globules...................................................................................138 6-21 Phase boundary in large globular morphology................................................................139 6-22 Compositional analysis along the le ngth of the feather-like morphology.......................139 6-23 Composition along dendrites containing gl obular morphology from deposit made at pre-electrodeposition time of 180 min, current density of 60 mA/cm2, and temperature of 90 oC........................................................................................................140 6-24 Composition along a line in the dendri tes containing globular morphology from the deposit made at pre-electr odeposition time of 180 mi n, current density of 60 mA/cm2, and temperature of 90 oC..................................................................................141

PAGE 14

14 6-25 Compositional analysis in large globular morphology made at pre-electrodeposition time of 60 min, current density of 150 mA/cm2, and temperature of 60 oC....................142 6-26 Compositional analysis between FCC and HCP regions of large globular morphology..................................................................................................................... .143 7-1 Composition along a dendrite from a deposit made at 180 min of preelectrodeposition containing F CC Al-rich phase at the root and HCP Mg-rich phase on Al-rich phase...............................................................................................................153 7-2 The SEM images of deposits on Cu, oxi de-containing Mg and amorphous carbon substrates..................................................................................................................... .....153 7-3 The SEM images deposit on Cu substrat e ground with 800 SiC paper showing Al in majority, and on Mg substrate ground with 800 SiC paper showing Mg in majority.....154 7-4 The SEM images of deposit made on oxide -free Mg substrate mounted on graphite stub........................................................................................................................... ........154 7-5 The XRD profile from deposit made on oxide-free substr ate confirming the formation of HCP Mg-rich phase....................................................................................155 7-6 Schematic showing different interfacial energies at a hemispherical nucleus.................155 7-7 Lattice parameter of HCP Mg-rich phase as a function of at.% Al. From 0.96 at.% Al to 6.89 at.% Al................................................................................................................ .156 7-8 Interatomic distance mismatch between nucleating materials (FCC Al-rich and HCP Mg-rich phases) and different substrates.........................................................................156

PAGE 15

15 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 FABRICATION OF NANOCRYSTALLINE AL-MG ALLOYS POWDERS BY ELECTRODEPOSITION AND THEIR CHARACTERIZATION By Sankara Sarma V Tatiparti December 2008 Chair: Fereshteh Ebrahimi Major: Materials Science and Engineering Aluminum-magnesium alloy powders can pot entially be used as hydrogen storage materials. In order to enhance the kinetics of hydrogenation it is desirable to have agglomerates of fine powders with very small grain size. In this study, nanocrystalline Al-Mg alloys in the form of powders were successfully fabricat ed by the electrodepositi on technique using an organometallic based electrolyte. Mg was introdu ced into the electrolyte by a process called pre-electrodeposition. The mechanism for Mg accumulation can be explained considering the electrode reactions as well as the ch emical changes in the electrolyte. Using a copper cathode, the effects of the electrolyte composition and current density on composition of the deposit, its constituent pha ses and morphology were investigated. The magnesium content of the deposits improved with in creasing Mg concentration in the electrolyte, temperature and current density. Depending on th e composition, the deposits consisted of FCCAl(Mg) and HCP-Mg(Al) phases and no inte rmetallic phase was found except for long deposition times. Generally, the deposits formed initially on the copper substrate with three dendritic morphologies namely, rod-like, feather-like and small globular, which eventually evolved into

PAGE 16

16 the large globular morphology. This observation is attributed to the establishment of spherical diffusion conditions at the sharp dendrite tips. Potentiostatic studies suggested that the appearance of different morphologies is asso ciated with differing rates of deposition. While the initial dendrites consisted of the FCC Al-rich phase, the large globular morphology manifested as both F CC Al-rich and HCP Mg-rich phase s, with the latter always forming over the former. The observation of forma tion of only the FCC phase implies that the nucleation barrier for the HCP phase on the copper substrate is quite high. The investigation of the effect of substrate, namely, Cu, graphite a nd Mg, revealed that the HCP phase can directly nucleate on an oxide-free Mg surface. This finding can be explained in terms of surface/interfacial energies. Detailed TEM analysis revealed that th e observed morphologies consist of randomly distributed nanocrystalline grai ns except for the feather-like de ndrites, which exhibited a strong crystallographic texture.

PAGE 17

17 CHAPTER 1 INTRODUCTION With the rapid consumption of the conventiona l fossil fuels and increasing global warming the trend is shifting to the development of nonconventional sources of energy [1, 2]. Among the non-conventional sources of energy, fuel cells are one of the promising solutions for the mobile applications [3]. Of the available fuel cells Pr oton exchange membrane fuel cells (PEMFC) are ideally suited for on-board vehicular applica tions mainly due to their low operational temperatures (~100 oC). One of the fuels for these PEMF Cs is hydrogen which is non-toxic and has high specific energy delivery capacity [4]. Various means of hydrogen storage such as gaseous state [5], liquid state [6], and solid stat e [7] have been consider ed for the transportation applications. Due to the low gravimetric and volumetric capacities, safety issues and the requirements of the high specific energies yi elded way to solid state storage [7, 8]. Metal hydrides with their larg e gravimetric hydrogen capacitie s are potential candidates for hydrogen storage [8]. Modifications have been carried out on thes e metal hydrides through alloying additions [9, 10] to improve their kine tics [11] and stability [9]. The added alloying elements are shown to assume two different ro les namely metal-hydrogen bond destabilizers [9] or complex hydride formers (a hydride with co mbination of two or more metals) [12]. Among the various modified hydrides Mg-Al base d hydrides are the cons idered to be one of the most promising candidates due to their light weight, high theore tical gravimetric and volumetric capacities [13]. Al, when added as an alloying element to Mg was mentioned to improve heat conduction during dehydrogenation [10]. Moreover, a complex hydride with Mg and Al called magnesium alanate (Mg(AlH4)2) is a potential candidate with its enthalpy of formation close to zero and stoichiometric hydroge n capacity of 9.3% which is the highest among the available complex metal hydrides.[14 -16]. Magnesium alanate can give about 7.6

PAGE 18

18 wt% of hydrogen at approximately 100 oC [12]. So far this compound is produced by various chemical routes, which were shown to suffer from high energy consumption, prolonged process times, and solvents needing tedious purification methods [16-18]. Th is can be avoided if suitable alternative methods of prepar ation of the hydrogen storage ma terials can be brought forth. One possibility is to first fa bricate alloys of aluminum and magnesium and then subject them to hydrogenation. Since gas-solid reactions are surface area dependent best results would be obtained if the alloy is in the form of powder. Moreover, diffusion is enhanced in the presence of nanocrystalline materials which have grai n sizes typically less th an 100 nm [19-21] For example, Mg-8%V with a grain size of 30 nm absorbed 7 % of hydrogen within 16 minutes [22]. Hence, it is advantageous for the aluminum magnesium alloys to be in the form of fine-grained powders for hydrogenation studies. There have been several processes in use to produce nanocrystalline materials in powder form. For example, ball milling can give high yields of materials with nanocrystalline and/or amorphous structure [23], and introduce mechanical alloying [24]. However, this process suffers from long milling times, presence of oxides, contamination from the milling environment, and powder size limitations. Electrodeposition is a well established technique for producing materials. Generally, electrodeposition can produ ce relatively pure nanocrystalline powders in shorter times compared to ball milling. In addition, it is a versatile technique and the characteristics of the powders may be tailored by varying different parameters like current density or applied potential, temperature, el ectrolyte composition, and agitation rate [25-37]. Limited attempts have been made to produ ce Al-Mg alloys using electrodeposition [3841]. Lehmkuhl et al. [40,41] and Mayer [39, 42] used organometallic-based electrolytes to

PAGE 19

19 successfully fabricate Al-Mg alloys in the form of corrosion resistance films rather than powders but the production of Al-Mg alloy powder s has not been previously attempted. The above mentioned requirements in terms of material and its structure serve as motivations to the present research leading to ob jectives. The objective of the present research is to develop a technique for fabricating nanocrysta lline Al-Mg alloys in the form of powders using electrodeposition and to carry out the fundamental investigation of their characteristics such as morphology, microstructure, compos ition and phases present and th eir inter-relationships. The alloy powders were characterized using various characterization techni ques including scanning electron microsopy (SEM), transmission electro n microsopy (TEM), electron probe micro analyzer (EPMA), energy dispersive spectro metry (EDS), and X-ray diffraction (XRD). The research conducted for the present work is distributed into four chapters supplemented by background and literature review in Chapter 2. The experimental procedures that were carried out are presented in Chapter 3. A systematic inves tigation of the electrolyte system is discussed in Chapter 4. The effect of different paramete rs used in the electrode position of Al-Mg alloy powders on their characteristics is presented in Chapter 5. Chapte r 6 is devoted to a detailed analysis of the morphology, microstructure, phases present, and compositional analysis of the electrodeposited Al-Mg alloy powde rs. The effect of substrate on the nucleation of different phases is discussed in Chapter 7. Finally, impor tant conclusions pertaining to the present research are presented in Chapter 8.

PAGE 20

20 CHAPTER 2 BACKGROUND Magnesium hydride is one of the promis ing materials for hydrogen storage with stoichiometric capacity around 7.6 wt%. About 7-7.3 wt% of hydrogen can be released from MgH2 at about 1 bar pressure [43-45]. However, the reported desorption temperatures were around 300-350 oC. Hence there is a need to reduce deso rption temperatures to make it suitable for transportation applications. The Mg-H bond wa s shown to be destabilized by addition of various alloying elements such as Al, Ti, Fe, Ni and Cu [9]. Among these alloying elements aluminum was shown to destabilize the Mg -H bond to the maximum extent. Moreover, improvements in heat conduction during dehydrogen ation by addition of Al were mentioned in literature [10]. With many such advantages Mg-Al based hydrogen storage materials have been developed [13]. Also a complex hydrid e namely magnesium alanate (Mg(AlH4)2) is a potential candidate with its enthalpy of formation close to zero and stoi chiometric hydrogen capacity of 9.3 wt% which is the higher among the availabl e complex metal hydrides [14-16]. Magnesium alanate can give about 7 .6 wt% of hydrogen at around 100 oC along with MgH2 which needs to be heated to around 300 oC for further release of hydrogen [12]. 2.1 Preparation of Al-Mg Based Materials for Hydrogenation The Al-Mg alloys are prepared by various chemical, mechanical, electrochemical techniques. The present secti on introduces all these methods and discusses about their advantages and disadvantages. 2.1.1 Earlier Methods of Preparation The complex hydrogen storage materials are us ually prepared by three methods namely, direct synthesis, mechano-chemical synthesis an d metathesis reaction. Di rect synthesis is the process of melting the elements and allowing them to react with hydrogen to form the complex

PAGE 21

21 hydrides at high pressures [17]. In mechano-chemical synthesis a chemical reaction is forced by applied mechanical treatment (e.g. ball-milling) to the reactants. For exam ple, in the case of sodium alanate, NaH and Al are mixed in the form of powders along with TiCl3 catalyst and ballmilled to obtain the alanate [18]. Generally, metathesis reactions are used to synthesize magnesium alanate by reacting sodium alanate wi th magnesium chloride [16]. The use of high energy (in case of mechano-chemical synthesis) for prolonged process times and solvents (in case of metathesis reaction), which need tedious purification methods, can be avoided if suitable alternative methods of prepar ation of the hydrogen storage ma terials can be brought forth. Preparation of the Al-Mg materials followed by their hydrogenation is one of the options. There are occasions in history where attempts were made on synthesizing Al-Mg alloys and subjecting them to hydrogenation for useful hydr ogen capacity [46, 47]. In their investigation Gremaud et al. [46] prepared Al-Mg alloys by sputtering techniques a nd then subjected to hydrogenation. Among various compositions of Mg and Al studied for hydrogenation, magnesium alanate could be produced at an at omic ratio of Al:Mg of about 2:1. All other compositions yielded MgH2. The formation of magnesium alanat e at about 33 at.% Mg is shown in Figure 2-1 with the equilibrium phase diagram of Al-Mg system. Figure 2-1(B) shows that the composition of 33 at.% Mg is in the two phase region with one phase be ing an intermetallic, Al3Mg2. As it is difficult to hydrogenate the stable phases such as Al3Mg2, production of materials in their supersaturated state is necessary. Ball milling is another technique that ha s been in use to produce materials for hydrogenation. Ball milling can yield materials with nanocrystalline and/or amorphous structure [23, 48] and introduce mechan ical alloying [24] in the material s. However, this process suffers from disadvantages such as long milling times presence of oxides, contamination from the

PAGE 22

22 milling environment and powder size limitations. Hence, better methods of preparing the materials should be sought for. 2.1.2 Electrodeposition of Powders Electrodeposition is one of th e very well established techni ques in producing materials. The success of the electrodepositi on dates back to the middle of the nineteenth century when it was used mainly as an art to coat objects with copper or brass [49]. Howe ver, technical treatise on this subject only appeared at the beginning of twentieth century. Since then, electrodeposition has been gaining extensive attention in producing ma terials with nearly desired characteristics. The experimental setup for the electrodepositi on consists of electroly te, cathode, anode and sometimes reference electrode. The process of electrodeposition can be performed by using two different techniques in general namely, potentio static and galvanostatic. In potentiostatic electrodeposition, a consta nt potential difference is applie d from external source between cathode (where reduction of cations takes place) and anode (anions form due to oxidation). In the case of galvanostatic electrodeposition, potential difference is generated between cathode and anode by application of constant current. When current density or potential difference is applied across the electrodes, the generated cations drift towards cathode and ar range against it forming an elec trical double layer. Due to the fast arrangement of the cations at the surface of the cat hode, there is a depletion of the cations in front of the double layer in the electrolyte side. Since the bulk concentration of the cations is higher than that at the double layer there is di ffusion of the cations from the bulk towards the double layer through this depleted region. Hence, this depleted region is called the diffusion boundary layer. Beyond the diffusion boundary la yer is the convection layer, where the movement of the cations depends on convection in the bulk of the electrolyte The three layers in front of the cathode are shown in Figure 2-2.

PAGE 23

23 Depending on the values of either applied po tential difference or current, the morphology of the deposits can be varied significantly. In order to select appropriate current or potential difference to be applied, a t echnique known as cyclic voltammetry is performed. In cyclic voltammetry, potential is swept b ack and forth between two different values at a constant rate. Based on the values of the selected potential a typical curve between current and applied potential during cyclic voltammetry may look lik e the one shown in Figure 2-3. The dendritic growth and powder formation can be realized when electrodeposition is performed under the conditions, where mass transfer be comes the controlling factor. Recently, both pure metals such as Ni [26, 27], Cu [28], Pd [25], Co [29] and alloys such as Ni-Cu [30], Ni-Fe[31, 32], Au -Pd [33], Zn-Co [34] have been produced successfully using this technique. Electrodeposition being a nonequ ilbrium process, allows the fabrication of metastable phases. For example, the stability of the FCC Ni-Fe alloy can be extended to high concentrations of iron [35]. Similarly, supersat urated solid solutions were obtained in Ag-Ni [36] and Cu-Co [37] systems which show limited solubility at equilibrium. 2.1.2 Electrodeposition of Al-Mg Alloys Electrodeposition of eith er Al or Mg can not be achiev ed in aqueous solutions because their reduction potentials (-1.706V for Al and -2 .375V for Mg) [50] ar e less than that of hydrolysis of water. Hence, the latter takes place when electrodeposition of either Al or Mg is attempted from aqueous solutions. Therefore, there has been search for alte rnative electrolyte to be used successfully in producing Al and Mg either in pure or in alloy form. Several non-aqueous solutions have been used in electrodeposition of Al. Zhao [51] and Great [52] summarized the charac teristics of the electrolyte sy stem for electrodeposition of Al. The electrolyte should consist of a solvent acting as Lewis base and a solute, the source of Al, should act as a Lewis acid. The solute should have high solubility in the solvent. The

PAGE 24

24 coordination of the Al must not be strong in order for easy release of Al. There should be a reservoir of Al during elect rodeposition for its continuous supply. Based on the above characteristics three organic solvent systems ha ve been used for deposition of Al namely, ethereal solvents, dimethyl sulphoxide and aromatic hydrocar bons. The ethereal solutions consisted of either diethyl ether (Et2O) or tetrahydrofuran. AlCl3 has been used in combination with LiH or LiAlH4 as an electroactive component [53, 54]. Using Et2O as a solvent, a cathodic current efficiency of 90-100% wa s obtained at ambient temperat ures. However, the Al anode used for deposition of Al had to be covered in a diaphragm in order to prevent formation of anodic reaction products on it. The formation of anodic reaction products was avoided by using tetrahydrofuran (THF) as a solvent in addition to benzene [55, 56]. Howe ver, the solutions of complexes of organoaluminum compounds are the mo st widely accepted electrolyte systems for deposition of Al. The electrolyte s consist of trialkyaluminum a nd alkali halides dissolved in aromatic solvents such as benzene or tolu ene [57, 58]. The typica l alkyl groups in the organoaluminum include methyl and ethyl, although various highe r groups were also explored [50]. The alkali halides include NaF, KF in general. Due to the melting points of those electrolytes being higher than the room temperature, the de positions were carried out at temperature range typically 90-100 oC. The higher productivities obtained using these electrolytes have lead to deve lopment of an industrial scale pr ocess of electrode position of Al called SIGAL (Siemens Galvano Aluminum) [50]. The electrodeposition of Mg has been mainly done with two different electrolyte systems na mely, Grignard reagents which are basically organomagnesium halides [59] and dialkylma gnesiums [60]. The electrodeposited Mg found applications in batteries which requi red smooth finish of the deposit.

PAGE 25

25 In the literature Grignard reag ents [61] were demonstrated to be one of the potential candidates for electrodeposition of Al-Mg alloys [39]. Using differe nt Grignard reagents in the presence of other trialkylaluminums, Maye r could deposit Al-Mg alloys with various compositions [39]. In Mayers st udy, the electrolytes consisted of alkali metal fluorides (NaF, KF and CsF), triethylaluminum (AlEt3, where Et = C2H5), triisobutylaluminum (iBu3Al, where Bu=C4H9), diethylmagnesium (MgEt2), and toluene as a solvent. Us ing various initial ratios of Mg/Al in the electrolyte deposits with differe nt compositions could be obtained. Figure 2-4 shows the composition of the deposit (at.% Mg) plotte d as a function of diffe rent ratios of Mg/Al in the electrolyte at the beginning of electr odeposition. Figure 2-4 shows that with a proper selection of the initial ratio of the Mg/Al desire d compositions in the depo sits could be obtained [38]. Figure 2-4 also shows that with a very sm all change to the ratio of the Mg/Al in the electrolyte, very high concentra tions of Mg in the deposit can be realized. Although the results are very interesting in production of broad spectrum of concentra tions of Mg in the deposits, the chemical changes occurring in the electrolyte syst em were not clear at that time. For example, Mayer hinted about the formation of some co mplex compounds in the electrolyte during the electrodeposition of Al-Mg alloys, which, however ,was not known at his time [39]. Latter, Lehmkuhl et al. have shown that one of the comple xes that is forming in such kind of electrolyte systems, which contai n NaF or KF and AlEt3, is Na[Et3Al-F-Al-Et3] [41]. Also as the deposition progressed, an insoluble film was observed to be forming over the Mg anode which was unknown at that time. Moreover, in the present el ectrolyte system diethylmagnesium was used as one of the sources of Mg (the other sour ce being Mg anode). Since the shipping of diethylmagnesium is banned in USA, the other way to obtain it is by ch emical synthesis. The synthesis of diethyl magnesium is a very diffi cult process. Typically, diethyl magnesium is

PAGE 26

26 produced by disturbing the chemical equilibriu m known as Schlenk equilibrium [62-65] existing between Grignard reagents (R-Mg-X, where R=alkyl group and X=halide) and diethylmagnesium by adding 1,4 dioxane as show n in equation (2-1) [66, 67]. According to equation (2-1), the reaction is driven towards the products side and more of the R2Mg is produced as more of the MgX2 is precipitated and removed. 2R-Mg-X + 1,4-dioxane R2Mg + MgX2,4dioxane (2-1) Since dioxane is an ether, complexes such as Na[Et3Al-F-Al-Et3] are mentioned to be cleaved by these ether groups to yield etherattached parts of these complex and Na[Et3Al-X]. These Na[Et3Al-X] were shown to hinder the depos ition of Al-Mg alloys due to the codeposition of alkali metal [68]. He nce to prevent the de position of alkali metal, and to alleviate the handling of the diethyl magnesium, the electr odeposition community has been in search of better alternative solutions for the continuous deposition Al-Mg alloys, one of which is by changing the chemistry of the electrodeposition bath. Recently, Lehmkuhl et al. [40, 41, 50] have deve loped a potential electrolyte system for the deposition of Al-Mg alloys. One of their electrolytes is Na[AlEt4] + 2 AlEt3 + 3.3 Toluene. When this electrolyte has been used for deposi tion of Al-Mg alloys using Mg anode, they could obtain only about a maximum of 24 wt%Mg in the alloys. In order to obtain increased amounts of Mg in the deposits, they have modified this electrolyte by the addition of another complex component namely, Na[Et3Al-H-AlEt3], which was mentioned to have improved the conductivity of the electrolyte system. Although, the addition of Na[Et3Al-H-AlEt3] improved the conductivity of the electr olyte, this complex, by itself, faile d to deposit any Mg. However, the reasons for its failure in depositing Mg were not reported in their work. Moreover, the Mg anode was shown to develop an unknown insoluble layers on it during the electr odeposition [41]. Also, when the above mentioned modified electrolyte wa s used for deposition, th e concentration of Mg

PAGE 27

27 in the electrolyte was mentioned to build up befo re reaching a steady state [41]. The reasons for this steady state were not clear from their work. Although Lehmkuhl et al. [40, 41] were successful in producing Al-Mg alloys, alleviating the use of Grignard reagents, the deposits were mainly used as corrosion protective coatings for automobiles. For the present work, since Al-Mg alloys in the form of powders have to be produced, different electrodepositio n parameters than the ones used by Lehmkuhl et al. have to be used. Using current densities of 10, 40-50 mA/cm2 they could obtain deposits having spherical or worm-like morphologi es, the SEM images of which are shown in Figure 2-5. Figure 2-5 shows that the structur e is refined with increased current de nsity values. Hence, the effect of different current density values and other para meters on the morphology of the deposits with a goal of producing dendritic structure should be in vestigated. Moreover, Lehmkuhl et al.s work [40, 41] bears a minimal account of the phase analysis. They mentioned the presence of AlxMgy phase in the Al-Mg deposits and did not e xplore beyond it. Since the properties of the deposits depend very much on their phases, a thorough phase analysis of the Al-Mg alloy deposits is necessary for the pres ent work. The presence of diffe rent phases in the Al-Mg alloy deposits depends on their composition. Since it is shown in Mayers work that the composition of the deposits depends on the concentration of Mg di ssolved in the electrolyte, it is necessary to understand the compositional variations in the el ectrolyte system and the relationship between composition of deposits and that of electrolyte used by Lehmkuhl et al. Although they mentioned about the electrolyte reachi ng a steady state with Mg upon prolonged electr odeposition, the reasons were not explored or at least were not presented. Th e absence of analysis on the compositional changes in the electrolyte in their work necessitate s a thorough study on the compositional analysis both in the electrolyte and in the deposits. Although the electrolyte

PAGE 28

28 system developed by Lehmkuhl et al. [40, 41] is a promising one for the deposition of Al-Mg alloys (eliminating the use of Gri gnard reagents), there are various issues to be explored such as the ones mentioned above. Hence for our study, the electrolyte system developed by Lehmkuhl et al. [40, 41] is considered. 2.2 Structure of Different Electrolyte Components In order to understand the chemical changes that might be occurri ng in the electrolyte during electrodeposition, knowledge about the structure of different components in the electrolyte is necessary. Most of the components in the present electrolyte system are electron deficient compounds [69]. Their structures are discussed in the following sections. 2.2.1 AlEt3 The structure of AlEt3 is similar to that of AlMe3 (where Me = CH3) [70] where Me is replaced by an Et (where Et = C2H5) group. When the three valence el ectrons in Al are shared by three alkyl groups there will be six valence electr ons in the Al. However, for achieving a stable electronic configuration the AlR3 (where R = alkyl) has to acqui re two more electrons. This is done by having these AlR3 in the form of dimers rather than as monomers. In the case of AlMe3, it was suggested by Lewis and Rundle [71] that an electron pair is shared by two aluminum atoms and one carbon atom from Me giving rise to an electron deficient Al-C-Al bridge. These kinds of bridge bonds are normally called 3-cente red-2electron (3c2e) bonds. In these electrondeficient compounds each Al is sp3 hybridized and is bonded to four carbon atoms. Figure 2-6 shows the structure of AlMe3 as suggested by Lewis and Rundle [71]. As seen from Figure 2-6 the bond length of the Al-C bond in the bridge (2 .24 ) is longer than that of the Al-C bond outside the bridge (1.99 ). It is the longer Al-Me that makes the AlMe3 in particular and AlR3 in general vulnerable to cleave at this pos ition. Also, the Al-C-Al bond angle is about 70o. However, when alkyl groups heavier than Me, such as Et are present in the electron deficient

PAGE 29

29 bonds these bond lengths and angles will be differ ent to accommodate for the larger electronic repulsions created by thes e heavier alkyl groups. 2.2.2 Na[Et3Al-H-AlEt3] The structure of Na[Et3Al-H-AlEt3] is similar to that which has its Et group replaced by Me. Similar to the above structure Na[Me3Al-H-AlMe3] also has an electron deficient bond shared by two aluminums and hydrogen [72]. Be cause of the symmetry of the complex ligand the Al-H-Al bond angle in this case is 180o. The Al is sp3 hybridized with the bonds shared by three Et groups and H. Due to the asymmetry th e bonds of Al with the Et and H are not arranged tetrahedrally in space. Figure 2-7 shows the structure of [Me3Al-H-AlMe3]-. 2.2.3 Na[AlEt4] Unlike the above two structures, Na[AlEt4] does not have any electron deficient bond in it. Na[AlEt4] is crystalline in nature with a monoclinic structure having a space group of P21/c [73] .The Al is again sp3 hybridized in this case. It is an ionic compound with the Na+ being closest to carbons of Et group. Figure 2-8 shows the struct ure of the negatively charged complex ligand with Et groups arranged tetrahedrally around Al. 2.3 Morphological and Microstructural Analysis of the Deposits In this section the morphology and microstr ucture of electrode posited materials are discussed. So far, no work has been conducted on the characterization of electrodeposited Al-Mg powders prior to our research for their morphology or the microstructure. Hence, in this section morphological and microstructural features of other metallic systems are discussed. 2.3.1 Morphological Analysis An extensive work has been done on the morphological characterization of electrodeposited materials. The morphology of the electrodeposits in metallic systems depends on various parameters such as current density or applied potential [ 74, 75], composition of the

PAGE 30

30 powders in case of multi component systems [76], pH of the solu tion [75]. Although the electrodeposition of la rge crystals in not the main object ive for our study, a brief literature review about the same is presented here to s how the contrast from the electrodeposition of dendrites. When the conditions are such that the deposit ion rates are lower at current densities or applied potentials lower than limiting diffu sion conditions, the depositing atoms have enough time to arrange themselves to minimize the surfac e or interfacial energy. The fastest growth of deposits is realized in a directi on perpendicular to the plane with the highest surface or interfacial energy [74]. Since the growth rates depend on the interfacial energies of different crystallographic planes [77], it is possible to modify these in terfacial energies and thereby change their growth rates. For example, in the case of the elec trodeposition of the Cu2O the growth morphology showed increased tendency to wards branching with rise in pH of the electrolyte [78]. The different growth rates of different crystallograp hic planes result in anisotropy in the electrodeposits. For example, in the case of th e electrodeposition of Zn, when deposition was done at lower applied potential s (E =-1.4V to -1.63V vs Ag/AgCl reference electrode), isolated crystals or well-faceted hexagonal chunks were observed [79]. The reason was attributed to the lower reduction rates of Zn+2 at lower applied potentials. When the reduction rates of the Zn+2 are lower, the loss of the Zn+2 at the cathode can be compensated by the supply of more Zn+2 by diffusion. Under these conditions th e deposit can assume a shape that minimizes the interfacial energy. The different an isotropic morphologies th at were obtained at different applied potentials in the case of Zn can be seen in Figure 2-9. With further increase in the deposition rates either due to applied potential or the current density, branching can initiate. For example, in the electrodeposition of Zn from the same group

PAGE 31

31 which conducted the above study [79], the increase in the applied poten tial to about E =-3V resulted in the formation of hexagonal shaped re gular dendrites. One such dendrite, which is in the initial stages of formation is shown in Figure 2-9(C). As can be seen in Figure 2-9(C), the branching started from the corners of the he xagonal crystal. In a nother study made on the dendritic growth of Ag, it was shown that the br anching always begins at the vertices or the corners of large crystals as ther e is an increased mass flux in t hose areas in comparison to that near the edges of the crystals [80]. The dendri tic Ag and a schematic showing the mass flux near the corners of the large crysta l are shown in Figure 2-11. Thes e results are similar to the observations made by Pal and Chakravorty [81], w ho demonstrated that the electrodeposition of Zn in the mass transfer controlled conditi on can yield dendrites growing in specific crystallographic directions. The r easons of the anisotropic growth were attributed to the relative differences in the interfacial energies of differe nt growing crystallographi c planes. The dendrites of Zn growing in different crystallographic directions can be seen in Figure 2-10. Various types of morphologies can be realized in metallic systems when electrodeposition is performed at deposition rates higher than that of limiting current densities. For example, in the electrodeposition of Au a dense branched morphology (DBM) was observed [82]. In the electrodeposition of Co, spongy-like agglomerates were obtained at current densities much higher than limiting values [76, 83]. As seen in the above discussion, applied current density or deposition rates relative to the limiting va lues affects the morphology of deposits. In the case of multi-component systems, apar t from the current density, the composition of the deposits or that of the elec trolyte also affects the morphology of the deposits. For example, in the galvanostatic electrode position of Ni-Co alloys by Jovic et al. [76] various in itial ratios of Ni+2/Co+2 in the electrolyte and in turn the compositio ns in the deposits were shown to affect the

PAGE 32

32 morphologies of deposits. Some of the morphologies observed in the electrodeposition of Ni-Co alloys resembled those of the pure Ni or Co el ectrodeposits [83]. The r easons for appearance of different morphologies as functi on of different compositions in the deposits were not clear. It was stated that the possible answer could be found by comparing the limiting current density values in each case with that of applied curren t density. However, this kind of comparison could not be possible in the case of electrodeposition of Ni-Co allo ys as hydrogen evolution also played a significant role in the deposition process. The diffe rent morphologies obtained for different ratios of Ni+2/Co+2 in the electrolyte is shown in Figure 2-12. Although, electrodeposition started by imposing certain values of current density or potential difference has shown to form certain morphologies to begin with, a morphological evolution has been observed as a function of tim e during the electrodeposi tion in the literature. For example, in the potentio static deposition of Au-Ag allo ys, dendritic precursors were observed in the initial stages of deposition under mixed interfacial -diffusion controlled conditions (see Figure 2-13(A)). As they grow, the tips of these dendrites experience a greater flux of depositing ions than the rest of the deposit. Moreover, th e dendrites experience flux of cations along the lateral sides of these growing dendrites. Due to these reasons spherical diffusion conditions are setup (radius of tip << diffusion boundary layer thickness) [84] (see Figure 2-13(C)). Due to the onset of the spheri cal diffusion condition in the later stages of deposition, poorly defined crystallites and gl obular morphologies were mentioned to be developed [85]. In the case of electrodeposit ion of Co powders, the morphological evolution from initially forming disperse deposits to compac t ones is attributed to the onset of spherical diffusion condition [83] (see Figure 2-13(B)).

PAGE 33

33 2.3.2 Microstructural Analysis Recently the microstructural analysis of the electrodeposited dendrites has received much attention. Electrodeposited Cu was investigat ed for its microstructure [82, 86]. The nanocrystalline nature of the Cu dendrites was explained with a concept of oscillating local overpotential [87]. A space charge field including the cations is de veloped near th e cathode [88]. As the space charge is develope d the local overpotential raises. When the value of the local overpotential is more than the critical local overpotential for nucleation of a grain then nucleation of a grain occurs. With the nucleation of the gr ain the local space charge collapses since all the cations are consumed in nucleation of the grain. During this period the local concentration of the cations gradually builds up due to which the local overptential value also rises. During the time when the local overpotential is rising, the growth of the nucleated grain occurs. Again, when the space charge is built up to and extent that th e local overpotential valu e exceeds the critical overpotential for nucleation of a grain anothe r nucleation event occu rs. And this process continues during the macroscopic growth of the dendrites. The Cu dendrite with nanocrystalline grains is shown in Figure 2-14. Although, morphological and microstructural inve stigations of the dendrites have been reported in the literature, an in ter-relationship between them is missing. Hence it is essential for our study to be able to correlate the obse rved morphology of the dendrites with their microstructure.

PAGE 34

34 (A) (B) (A) (B) A B 450 oC 437 oC Al3Mg2 Al12Mg17 650 oC (A) (B) (A) (B) A B 450 oC 437 oC Al3Mg2 Al12Mg17 650 oC Figure 2-1.Appearance of different hydrides and other phases at different compositions of Al-Mg alloys. A) Different hydrides formed as a function of Al:Mg ratio [46, 47], and B) AlMg equilibrium phase diagram with co mposition suitable for magnesium alanate formation indicated in it [89]

PAGE 35

35 Electric double layer Solvated cations Adsorbed cations Electrode Solvated cations Diffusion layer Convection layer Bulk electrolyte Electric double layer Solvated cations Adsorbed cations Electrode Solvated cations Diffusion layer Convection layer Bulk electrolyte Figure 2-2. Schematic of the three layers (elect ric double, diffusion and convection layers) that form ahead of the cathode during electrodeposition (from Walsh [90]). Figure 2-3. Schematic of cyclic voltammetry cu rve depicting different regions for forming different morphologies (from Walsh [90]).

PAGE 36

36 0 20 40 60 80 100 120 0.10.150.20.250.30.35Mg/Al ion Ratio in ElectrolyteMg in Deposit (at.%) Figure 2-4. Composition of the deposits (Mg at%) as a function of initial Mg/Al ratio in the electrolyte [38]. A B A B Figure 2-5. The SEM images of the Al-Mg allo y deposit obtained using electrolyte Na[AlEt4] + 2 AlEt3 + 3.3 Toluene showing A) spherical morphology at 10mA/cm2, and B) wormlike morphology at 40-50 mA/cm2 [41].

PAGE 37

37 70o 110o124o 2.24 1.99 Al Me (Methyl) 70o 110o124o 2.24 1.99 Al Me (Methyl) Figure 2-6. Structure of AlMe3 (where Me = methyl) dimer (from [71]). Al Me (Methyl) H Al Me (Methyl) H Al Me (Methyl) H Figure 2-7. Structure of [Me3Al-H-AlMe3]-1 (from [72]).

PAGE 38

38 Al Et (Ethyl) Al Et (Ethyl) Figure 2-8. Structure of [AlEt4]-1 Figure 2-9. The SEM images showing different mor phologies of Zn electrodeposited at different applied potentials. A) Chunks, E =-1.4V, B) Disks, E =-1.55V, and C) Hexagonal shaped dendrite showing branching in its intial stages of gr owth E =-3V [79].

PAGE 39

39 Figure 2-10.Anisotropic growth of Zn dendrites A) SEM image of Zn dendrites growing in mass transfer condition, and B) Schematic show ing the crystallographic directions along which different arms are growing [81]. A B C A B C Figure 2-11. Anisotropic growth of Ag dendrite s A) and B) TEM images of Ag dendrites growing in mass transfer condition, and C) Schematic showing the mass flux around hexagonal shaped dendrite tip [81].

PAGE 40

40 Figure 2-12. The SEM images of Ni-Co all oys for different st arting ratios of Ni+2/Co+2 in the electrolyte. The morphologi es are A) spongy, B) agglomerate + fern-like, C) agglomerate + fern-like, a nd D)majorly agglomerates [76]. Figure 2-13.The SEM images showing morphological evolution A) Dendrites with sharp tips in Au-Ag deposit [85], B) Evolution from disp erse to compact deposit in Co powders [83], and. C) Schematic showing the lateral flux on the dendrite leading to spherical diffusion around its tip [84].

PAGE 41

41 Figure 2-14. The SEM image of Cu dendrit e showing nanocrystalline grains [86].

PAGE 42

42 CHAPTER 3 EXPERIMENTAL PROCEDURES Preparation methods of various substrat es, electrolytes, electrodeposition and characterization of the deposits are described in the present chapter. Since the chemicals used for preparing the electrolyte are highl y moisture and oxygen sensitive the use of inert atmosphere was inevitable for the present research. All the procedures that deal with the moisture and oxygen sensitive chemicals we re carried out in a glove box maintaining an argon atmosphere. Generally moisture and oxygen levels in the glove box were maintained at <5 ppm and <1 ppm, respec tively. The samples were analyzed using various characterization tec hniques such as SEM, TEM, XRD, EPMA, and EDS. These techniques and the sample prep aration methods used for this research are described in this chapter in a detailed manner. 3.1 Substrate Preparation Three types of materials namely, polyc rystalline Cu (99.99%), Mg (99.9%) and amorphous graphite were used as substrates for electrodeposition. Cu was selected for its good conductivity. The graphite was selected as a substrate because when added to the Mg, it was shown to improve the kinetics of Mg-based materials [91-93]. Mg was selected as a substrate to study the eff ect of substrate on the competition between nucleation of FCC Al-rich and HCP Mg-rich phase s. The substrates were in the form of cylinders of length about 35 mm and a nomin al diameter of 6 mm. A rotating cylinder electrode cell setup was used for electropolis hing. The substrates were connected to a rotator using a steel shaft and rotated at 50 rpm. They were electropolished using a PAR 273 Potentiostat/Galvanost at system. The graphite substrat e was ultrasonicated in toluene

PAGE 43

43 after machining and was applied without any other surface preparation. However, Cu and Mg anodes were electropolishe d prior to electrodeposition. Cu substrates were electr opolished in an electrolyte made of 82.5 vol.% orthophosphoric acid and 17.5 vol.% deio nized water. An applied poten tial of 1.1 V relative to an aqueous KCl reference electrode was used for about 5 minutes to electropolish Cu substrates [94]. In the case of the Mg subs trates the electropolishi ng solution consisted of 37.5 vol% ortho-phosphoric acid and 62.5 vol % ethanol solution was used. For Mg substrates also an applied potential of 1.1 V relative to an aqueous KCl reference electrode was used for about 5 minutes fo r electropolishing. The quality of the electropolishing was judged based upon the ge nerated current vs time curves. The quality of polishing deteriorated in the presence of wobbling of the substrate which was the anode in the present case. Wobbling was identi fied by the irregular spikes in the current vs time curve [95]. The wobbli ng was corrected for by readjus ting the steel shaft and the anode Examples of current vs time curves generated during elec tropolishing of Cu (in the absence and presence of wobbling) and Mg electrode are shown in Figure 3-1. Immediately after electropolishi ng the Cu substrate was removed from the cell and rinsed with deionized water. In the initial stages of research, after rinsing with deionized water the Cu substrate was wiped with a Kimwipe to dry it. However, this method was found to produce irregularities in the c overage of the substrate with the deposit. One such case where the coverage was not uniform is show n in Figure 3-2. Hence, the cleaning process was improved to eliminate th ese irregularities in the mor phologies of the deposits. The Cu substrate was rinsed with ethyl alcohol im mediately after rinsing with deionized water Mg substrate was rinsed with toluene immediatel y after the electropolishing. In the case

PAGE 44

44 of the Mg substrate ethyl alcohol was not used for final rinsing as it oxidizes the surface of the Mg substrate. Although rinsed with to luene, the Mg substrate developed an oxide layer on it. Since it was difficult to obtain oxide free Mg substrate outside the glove box, no further attempts were made towards remova l of oxide layer outside the glove box. The surface of selected Mg substrat es was scratched with 800-grit SiC paper inside the glove box to remove the oxide layer and rinsed with to luene in order to investigate the effect of oxide-layer. The effect of oxidation of Mg cathode on the nucleation process during electrodeposition will be di scussed in Chapter 7. The weights of the substrates were measur ed using a balance with an accuracy of .005 g. The diameters of the substrate were measured with a caliper with an accuracy of .1 mm. After measuring the weights and diameters of these substrates, they were then transferred into a glove box. These subs trates were used as cathodes during the electrodeposition. 3.2 Anode and Reference Electrode Preparation For electrodeposition two different annul ar shaped anodes made of about 0.5 mm thick sheets of namely Mg (Goodfellow, 99.9%) and Al (Goodfellow, 99.99%) were used. The anodes were polished with 800 grit Si C paper and ultrasonica ted in toluene for about 5 min prior to taking into the glove box. During the potentiostatic electrodepo sition, a non-aqueous Ag/AgCl reference electrode filled with 1M LiCl in ethylene glycol monobutyl ether was used. The solution for the reference electrode was prepared by dissolving the required amount of LiCl in ethylene glycol monobut yl ether at around 80 oC.

PAGE 45

45 3.3 Electrolyte Preparation The electrodeposition of pure Al and Al-Mg alloys is described in this section. As the reduction potentials of Al and Mg are less th an that of electrolysi s of water, they can not be deposited using aqueous solutions. Hen ce for deposition of Al and Al-Mg alloys non-aqueous solutions are used. 3.3.1 Pure Al For electrodeposition of pure Al an orga nometallic-based electrolyte was used as suggested by Lehmkuhl et al. [50]. The elec trolyte was composed of sodium fluoride (NaF), triethylaluminum AlEt3 and toluene. According to Lehmkuhl et al. [50] the following reaction takes place upon addition of NaF to AlEt3 NaF + 2 AlEt3 Na[Et3Al-F-AlEt3] (3-1) The existence of the Na[Et3Al-F-AlEt3] was suggested by Lehmkuhl [96]. Since the melting point of Na[Et3Al-F-AlEt3] is mentioned to be about 35 oC [50], a temperature of 60 oC was chosen to make the above complex. In order to understand the function of the complex Na[Et3Al-F-AlEt3] in the electrolyte two differe nt electrodeposition experiments were done by varying the amount of AlEt3 relative to NaF. In the first experiment, 3 mol of toluene (Fisher, laboratory grade) was added to an electrolytic cell. To this stoichiometric amount i.e. 2 mol of AlEt3 (93%, Strem, impurities include mixtures of trin butylaluminum, diethylbutylaluminum a nd ethyldibutylaluminum)) was added and stirred using a glass spin bar. The solution was maintained at 60 oC using a heater. To the solution 1 mol of NaF (Fisher) in the fo rm of powder was added. The stirring was continued at 60 oC until an opaque solution was obtaine d. In the second experiment, the amount of AlEt3 was increased to 4 mol. Also the am ount of toluene was increased in this case to 6 mol in order to maintain homogeneity in the solution [41].

PAGE 46

46 3.3.2 Al-Mg Alloys The raw materials used for the electr odeposition of Al-Mg alloys were AlEt3 (same purity as mentioned in the above section) Na (ACROS), NaH (ACROS, 60% dispersion in mineral oil) and toluene (same quality as mentioned in the above section). The electrolyte for the deposition of Al-Mg all oys typically consiste d of 1 mol Na[AlEt4], 2 mol Na[Et3Al-H-AlEt3], 2.5 mol AlEt3 and 6 mol toluene based on previously reported studies [40, 41]. Preparation of the tw o complex compounds mentioned above was critical for the deposition of Al-Mg alloys. The preparation of Na[AlEt4] was suggested by Zakharkin and Gavrilenko [97].Their studies mentioned that Na[AlEt4] can be prepared with or without addition of any solv ent. When Na in a finely divided form was added to AlEt3 and stirred in the te mperature range of 60-80 oC the colourless Na[AlEt4] crystals should start forming according to following reaction. 3 Na + 4 AlEt3 3 Na[AlEt4] + Al (3-2) In our early attempts to prepare the Na[AlEt4] complex, liquid AlEt3 was taken in a beaker and stoichiometric amount of Na cut into small pieces were added to it in the absence of any solvent. As soon as the Na pieces were added to AlEt3 they turned to grey in colour. This mixture was heated while st irring to different temperatures ranging from 60 oC to 150 oC. All these attempts were futile as no crystallization was observed. Hence, to improve the process, the addition of chem icals was conducted in the presence of small amounts of toluene. The mixture, with gr ay residue in it was heated to around 60 oC for about 45 minutes while stirring it. When this mixture was removed fr om heater and kept for cooling, a colourless crysta lline front was observed in the liquid port ion of this mixture which started to grow with time. When this mixture was left overnight at room temperature, the whole of the liquid turned crystalline. This crystalline portion was

PAGE 47

47 separated carefully from the gray residue. Sin ce this crystalline material was flammable it could not be taken out of glove box for furthe r characterizatio n. In order to check whether this crystalline material is Na[AlEt4] or not, a couple of its ph ysical properties were taken as aid namely, physical appearance and me lting point. According to Zakharkin and Gavrilenko [97] the melting point of Na[AlEt4] is around 122-124 oC.When this colourless crystalline material was heated to temperatures of around 120 oC it started melting. Hence, it was assumed that this crystalline mate rial was Na[AlEt4]. Moreover, a close observation of the gray residue left by Na revealed that extremely fine shiny particles were struck on it. These fine shi ny particles could be Al according to equation (3-2).This gray residue was washed with to luene and taken out of glove box for analysis of composition using EDS in SEM. The EDS an alysis revealed that this gray residue contained Na and Al. Figure 3-3 shows the SE M images of this gray residue along with the EDS spectrum showing the presence of Na and Al. Na[Et3Al-H-AlEt3] was prepared by adding stoichio metric amount of NaH to AlEt3 in presence of toluene at 60 oC according to Ziegler et al [98]. When this mixture was stirred for some time at 60 oC an opaque solution was obtained. The Na[Et3Al-H-AlEt3] was believed to have formed, as suggeste d by Ziegler et al. [98], according to the following equation NaH + 2 AlEt3 Na[Et3Al-H-AlEt3] (3-3) After preparing the above complexes, the el ectrolyte for deposition of Al-Mg alloys was prepared by adding first AlEt3 to the Na[Et3Al-H-AlEt3] and toluene solution at 60 oC and followed by the addition of Na[AlEt4]. This solution was opaque in appearance.

PAGE 48

48 The solution was always maintained at 60 oC. Thus prepared electrolyte was used for electrodeposition of Al-Mg alloys. 3.4 Electrodeposition The present section describes the experi mental setup and th e electrodeposition of pure Al and Al-Mg alloys. 3.4.1 Experimental Setup Since the electrolytes for the deposition of pure Al and Al-Mg alloys were highly moisture and oxygen sensitive, all the experiments were done in Ar atmosphere in a glove box maintained at moisture and oxygen le vels of <5 ppm and <1 ppm respectively. PAR 263/273 Potentiostat/Galvanostat systems interfaced with computer were used for all the experiments. Figure 34(A) shows the glove box with some of its accessories and PAR system. The electrolytic cell setup and its schematic are shown in Figure 3-4(B). Wherever reference electrode was used 100 mL electrolyte was used in a large cell to accommodate the reference electrode.In the cases where reference electrode was not used, smaller cell with 30 mL of electrolyte was used. 3.4.2 Pure Al For pure metals the morphology of the de posits depends on the applied current density (applied potent ial) during the electrodeposition. Wh en the applied current density exceeds its limiting value dendritic growth is r ealized [90]. Hence, in order to find this limiting current density cyclic voltammetry was done prior to the electrodeposition. In the case of pure Al cyclic voltammetry was conducted on Cu substrate at a temperature of 60 oC using pure Al as the anode. Th e applied potential was swept between 0 and -2 V (Ag/AgCl reference elec trode) at a rate of 6.67 mV/s. A rotation speed of 400 rpm was chosen for our study. Th e maximum rotation speed was chosen at

PAGE 49

49 which no vortex formation was observed. Figure 3-5(A) presents the relation between the applied potential and the current density values obtained during the cyclic voltammetry test. The spikes in the curve are due to the wobbling caused by high rotation speeds selected here. The limiting current density is a saturated value of current density, which does not change within a certain range of th e applied potentials. For the present case the saturation in the current density was seen at a value of 0.2 mA/cm2 within -0.5 V to -1.2 V. Hence, this value of the current density wa s taken as the limiting value. In order to be slightly above this limiting value of the curr ent density, an applied potential of -1.5 V was taken for the deposition of pure Al in potentiostatic mode using a pure Al anode. 3.4.3 Al-Mg Alloys The cyclic voltammetry for the case of Al-Mg alloys also was conducted on Cu substrate at a temperature of 60oC using pure Mg as anode. The applied potential was swept between 0 and -2 V (Ag/AgCl refere nce electrode) at a rate of 6.67 mV/s. A rotation speed of 200 rpm was chosen here to reduce the spikes in the voltammetry curve. Figure 3-5(B) shows the cyclic voltammetry curv e for the present case. As can be seen in Figure 3-4(B) the spikes were reduced in magn itude significantly with the decrease in the rotation speed of the electrode. The limiting valu e of the current density in this case was found to be around 0.3 mA/cm2. In order to enhance the dendritic growth of the deposits current density values of 60, 100 and 150 mA/cm2, which are far greater than 0.3 mA/cm2 were chosen for electrodeposition in galvanos tatic mode. Similarly ap plied potentials of 4 and above were chosen for electrode position using potentiostatic mode. For electrodeposition of Al-Mg alloys either pure Mg or Mg and Al anodes in combination were used. When both Mg and Al were used as anodes, they were connected in parallel in

PAGE 50

50 the circuit. The ratio of their areas was kept constant at 80 area% Al to 20 area% Mg when they are used in combination. Since the electrolyte did not contain any Mg in it, it was necessary to dissolve some Mg into it prior to using it for electrod eposition of Al-Mg alloys. Mg was anodically dissolved by using a pure Mg as anode and cy lindrical Cu substrate as cathode. A current density of 60 mA/cm2 was used at a temperature of 90 oC. During this process some Mg dissolved into the electrolyte, part of which was deposited on the cathode. The rest of the Mg remained in the electrolyte. Hence, there was an accumulation of Mg in the electrolyte during this process. The amount of Mg dissolved into the electrolyte depended on the duration of this process. Since this process of introduction of Mg into the electrolyte was conducted before every elect rodeoposition experiment, it was termed Pre-electrodeposition process. Chapter 4 de als in detail about the pre-electrodeposition process. After pre-electrodepos ition, the above mentioned cu rrent densities or applied potentials were used for different durations to produce Al-Mg alloys. 3.5 Sample Preparation After electrodeposition the substrate with either pure Al or Al-Mg deposit on it was removed from the electrolytic cell. The subs trate with the deposit on it was dipped in toluene maintained at a temperature of about 60 oC and rotated at 100 rpm for 10 minutes to clean the electrolyte off the deposit. This cleaning process was repe ated for three times before taking the deposit for characterization. 3.5.1. Electron Microprobe Samples For composition analysis using electr on microprobe (EPMA) the Al-Mg powders were embedded in a two part epoxy resin sy stem from Buehler consisting of epoxy cure resin (5 parts) and epoxy cure hardener (1 part). The resin mixture was allowed to set for

PAGE 51

51 overnight at room temperature. After the resin was hardened, mechanical grinding was done on the hardened block with 400 grit SiC paper until the powders were exposed. When the powders were exposed the block was ground with 800 and 1200 grit SiC papers in that order for about 10 minutes on each paper. For grinding the block on a higher grade grit paper a direction which wa s perpendicular to the previous one was chosen to eliminate the scratches formed on the sample surface from the previous grit paper. After grinding with each paper the bloc k was ultrasonicated in ethyl alcohol for about and 5 minutes. Following the grinding on grit papers, the block was polished with 9, 6 and 1 micron diamond pa stes in that order by using Leco microid diamond compound extender oil from Buehler as a lubricant. Each polishing was done for about 15 to 20 minutes and at the end of each polishing step the block was ultrasonicated in ethyl alcohol for 5 minutes. All the ultrasonica tion steps were done keeping the block in a vertical position in th e beaker containing ethyl alcohol so that the SiC powder or diamond paste that was jammed in the block was knoc ked off into the beaker containing ethyl alcohol. After the final cleaning with ethyl alcochol, the block was carbon painted over the areas where there was no sample. Finally the block was coated with carbon all over the surface and taken to composition analysis. 3.5.2. Ultramicrotomy Samples Ultramicrotome samples were prepared by embedding the Al-Mg dendrites in a three part resin consisting of SPI-PON 812 (24 parts), NMA (20 parts) and DMP 30 (1 part) from SPI supplies such that they were parallel to the cutti ng face on the mold. In order to avoid formation of any bubbles, th e resin was poured gently from the face opposite to the cutting face. The resi n was cured at a temperature of 50 oC for 24 hours to

PAGE 52

52 obtain the samples in the form of capsules. Figure 3-6 shows the schematic of the embedded dendrites in a mold and th e capsule obtained after curing. 3.5.3. TEM Samples TEM samples were prepared using Leica Ultracut UCT ultramicrotome. The capsules were loaded into ultramicrotome su ch that the cutting f ace of the capsules was parallel to the edge of the kni fe. The slices were cut from the capsule at different speeds using a diamond blade with 45o wedge angle. Initially slow cutting speeds were chosen which resulted in pullout of the dendrites from the polymer slices. Hence, cutting speed was increased to about 100 mm/s which im proved the sample preparation by minimizing pullout. As the slices were cut they floate d in the boat attached to the diamond knife which contained deionized water with a fe w drops of ethyl alcohol. The purpose of adding ethyl alcohol was to reduce the surface tension of the water to minimize the wrapping up of slices as soon as they float on water. The slices were fished using a 3mm grid of 200 mesh size which was held by a tweezer s. The grid with the slices is shown in Figure 3-7. 3.5.4. Focused Ion Beam Samples For preparing the Focused ion beam (FIB) samples the dendrites were coated with carbon for three times. Prior to the ion milling with Ga ions a Pt layer of about 1 m thickness was deposited onto the area of inte rest to prevent its damage. Two trenches were made along the sides of this Pt deposited area to prepare the sa mples lift off. After the trenches were made, an in-situ microm anipulator was attached to the sample by welding it onto it using Pt. The bottom of th e sample was cut free from the rest of the dendrite and the sample was then lifted off the dendrite carefully. The lifted-off sample was welded onto an edge of a grid used to hold the sample. The micromanipulator was

PAGE 53

53 then separated from sample using Ga ions The sample, which was on the edge of the grid, was milled initially with a Ga ion current of 300 pA. As the thickness of the sample was reduced considerably, the Ga ion current was reduced to 100 pA to minimize the sample damage. The final thickness of the sa mple prepared thus was about 70 nm. Figure 3-8 shows the SEM images of the sequence of FIB sample preparation. The sample prepared by FIB was analyzed for their poros ity using the scanning electron microscope in the FIB instrument. 3.6 Characterization Techniques For our study characterization for vari ous features such as compositional, morphological, mictrostructural and phase analysis was conducted. The techniques used for these analyses are describe d in he following sections. 3.6.1. Compositional Analysis The compositional analysis was done using a EPMA JEOL 733 Superprobe. The samples were prepared using a method described in Section 3.5.1 The composition of the dendrites was analyzed by selecting some random points on them. Whenever compositional variations in different phase s of the dendrites were sought for, the dendrites were analyzed al ong their length in the interv al range of 1 to 10 m. Inductively coupled plasmaoptical emi ssion spectroscopy (ICP-OES) at Columbia Analytics Company was used to find the overall composition of the alloys. Nominal composition of the deposits was analyzed usi ng energy dispersive spectroscopy (EDS) in a JEOL JSM 6400 scanning electron microscope. It was difficult to find the composition of th e electrolyte after deposition as it could not be taken out of the glove box. Hence, th e used electrolyte wa s quenched by using a mixture of 90 vol%toluene+10 vol% isopropyl alcohol inside the glove box. This

PAGE 54

54 quenched product could be taken out of the glove box safely as it was insensitive to moisture and oxygen. When this oxidized el ectrolyte was taken out of the glove box, white slurry started to appear in the electroly te. This slurry was dried at room temperature for almost 24 hours after which it changed into a colourless powder. The powder was washed with acetone and mixed with polyvinyl alcohol binder and green compacts of pellets were made out of this mixtur e. The pellets were sintered at 1200 oC for 3 hours and cooled in the furnace. The sintered pe llets were broken into four quarters and composition was analyzed on each face and sel ected edges of the pellets using EPMA JEOL 733 Superprobe. One such broken pe llet with the areas indicating where composition was analyzed is shown in Figure 3-9. The composition was found to be uniform throughout the pellets. 3.6.2. Transmission Electron Microscopy The microstructure of dendrites was studied mainly using a JEOL 200CX TEM. For the microstructural analysis both the bright field and dark field imaging were performed along with selected area electron diffraction (SAED). In order to study certain crystallographic growth featur es of the dendrites, rotation calibration was performed on the images to make them superimposable on the SAED patterns. For high resolution analysis a JEOL 2010F was used. 3.6.3. Scanning Electron Microscopy The morphology of the deposits was studi ed using JEOL JSM 6400 and JEOL JSM 6335F scanning electron microscopes. The de posits were analyzed both in the asdeposited condition when on the substrate, a nd also by scraping them and mounting on an aluminum or graphite stub.

PAGE 55

55 3.6.4. X-Ray Diffraction The phases in the deposits were analyzed using Philips APD 3720 X-ray diffractometer (XRD) with Cu-K radiation. The deposits were scraped from the substrate and distributed randomly on a glass slide in or der to eliminate any kind of preferentiality in sampling. Sometimes when the scraping was done hard to remove the deposit from the substrate some scrapings of the substrate also entered the deposits. This happened in the case of Cu substrate.

PAGE 56

56 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0100200300400Time (s)Current (A) Cu wobbling Cu No wobbling Mg Figure 3-1. Current vs time curves generate d during electropolishi ng of Cu electrode when in the absence and presence of wobbling. 20 m 20 m Figure 3-2. Aluminum-Magnesium deposit showi ng the irregular coverage of substrate. This irregularity was found to be due to wiping the substrate with a kimwipe after electropolishing.

PAGE 57

57 Figure 3-3. Grey residue from mixing Na and AlEt3 A) SEM image, B) SEM images of the grey residue at higher magnifica tion, and C) EDS spectrum taken on the gray residue showing the presence of Na and Al

PAGE 58

58 PAR Computer Glove Box O2sensor Moisture sensor Antechamber Argon Purification unit Display unit PAR Computer Glove Box O2sensor Moisture sensor Antechamber Argon Purification unit Display unit A Rotator PAR Computer CathodeRef. ElectrodeAnode Steel Shaft Beaker Annular Anode Cathode with Deposit Heater/Stirrer Electrolyte Rotator PAR Computer CathodeRef. ElectrodeAnode Steel Shaft Beaker Annular Anode Cathode with Deposit Heater/Stirrer Electrolyte Rotator PAR Computer CathodeRef. ElectrodeAnode Steel Shaft Beaker Annular Anode Cathode with Deposit Heater/Stirrer Electrolyte B Figure 3-4. Experimental setup showing A) glove box with different accessories and PAR system with computer, and B) electr odeposition cell setup and its schematic.

PAGE 59

59 -20 -10 0 10 20 30 40 -2.0 -1.5 -1.0 -0.5 0.0Applied Potential (V)Current Density (mA/cm2)60oC RPM 400 Ag/AgCl non-aqueous reference electrode -20 -10 0 10 20 30 40 -2.0 -1.5 -1.0 -0.5 0.0Applied Potential (V)Current Density (mA/cm2)60oC RPM 400 Ag/AgCl non-aqueous reference electrode A -20 -10 0 10 20 30 40 -2.0 -1.5 -1.0 -0.5 0.0Applied Potential (V)Current Density (mA/cm2)60oC RPM 200 Ag/AgCl non-aqueous reference electrode -20 -10 0 10 20 30 40 -2.0 -1.5 -1.0 -0.5 0.0Applied Potential (V)Current Density (mA/cm2)60oC RPM 200 Ag/AgCl non-aqueous reference electrode B Figure 3-5. Cyclic voltammetry curve in the case of A) pure Al, and B) Al-Mg alloys. The arrows indicate the direction of the sweep of the applied potential.

PAGE 60

60 Capsule for ultramicrotomy Al-Mg dendrite 200 m 14 7 3 3All dimensions are in mm Resin Mold Embed dendrites parallel to cutting face Cure at 50oC for 24 hrs Cutting face Al-Mg dendrite Capsule for ultramicrotomy Al-Mg dendrite 200 m 14 7 3 3All dimensions are in mm Resin Mold Embed dendrites parallel to cutting face Cure at 50oC for 24 hrs Cutting face Al-Mg dendrite Figure 3-6. Schematic showing the embedding of Al-Mg dendrites parallel to the cutting face of the mold and the obt ained capsule after curing. 2 mm Grid Polymer slice 2 mm Grid Polymer slice Figure 3-7. The TEM grid with slices of polymer on it.

PAGE 61

61 20 m 10 m 50 m 20 m 10 m 5 m A B D C E F 20 m 10 m 50 m 20 m 10 m 5 m A B D C E F Figure 3-8.The FIB sample preparation show ing A) trenches made on both sides of sample, B) attachment of in-situ micromanipulator to sample by welding it to the latter using Pt, C) sample lifted off the dendrite, D) sample welded onto an edge of a grid and micromanipulator was cut off from the sample, E) edge-on view of sample before milling, and F) edge-on view of sample after final milling (sample thickness was 70 nm)

PAGE 62

62 y Q1Q2Q3 Q4 x 2mm Y y Q1Q2Q3 Q4 x 2mm Y y Q1Q2Q3 Q4 x 2mm Y y Q1Q2Q3 Q4 x 2mm Y Figure 3-9. Broken pellet made of used electrolyte. The circles and arrows indicate where the composition was analyzed.

PAGE 63

63 CHAPTER 4 PRE-ELECTRODEPOSITION: A PROCESS FOR ACCUMULATING MG IN THE ELECTROLYTE Since the electrolyte used fo r our study did not contain a ny Mg in it Mg has to be introduced from external sources such as a node. This chapter describes the methodology of introduction of Mg into the electrolyte. 4.1 Selection of Electrolyte Most of the electrolytes that were used to deposit pure Mg or Al-Mg alloys were made from Grignard reagents (alkylmagnesium halides) [42, 59, 96, 99-101]. However, certain drawbacks exist when Grignard reagents are used for electrodeposition. During the electrodeposition using elect rolytes containing Na[Et3Al-X-AlEt3] (where X = halide) it was mentioned that a continuous addition of Mg in th e form of Grignard reagents was necessary in order to replenish the concentra tion of Mg in the electrolyte [41]. Also, since most of the Grignard reagents are available in the form of their ethereal solutions, complexes like Na[Et3AlX-AlEt3] were shown to be cleaved by ether to form Et3AlO Et2 and Na[Et3Al-X]. These Na[Et3Al-X] were shown to hinder th e deposition of Al-Mg alloys due to the co-deposition of alkali metal [68]. One possible solution for avoiding the Na[Et3Al-X] thereby avoiding codeposition of alkali metal was by using another mean s of incorporating Mg into the electrolyte. This was done by typically using a Mg anode [41]. Although, using of Mg anode obviated the use of Grignard reagents, this Mg anode wa s mentioned to develop insoluble layers of magnesium halides on it during the electrodepos ition. Hence, electrodeposition community has been in search of better alternative solutions fo r the continuous depositio n of Al-Mg alloys, one of which is by changing the chem istry of electrodeposition bath. Recently, Lehmkuhl et al. [40, 41, 50] have deve loped a potential electrolyte system for the deposition of Al-Mg alloys. One of their electrolytes is Na[AlEt4] + 2 AlEt3 + 3.3 Toluene.

PAGE 64

64 When this electrolyte has been used for deposi tion of Al-Mg alloys using Mg anode, they could obtain only about a maximum of 24 wt%Mg in th e alloys. Hence, they have modified this electrolyte by addition of anothe r complex component namely, Na[Et3Al-H-AlEt3] which was mentioned to have improved the conductivity of the electrolyte system. Although, the addition of Na[Et3Al-H-AlEt3] improved the conductivity of the electrolyte, this complex by itself failed to deposit any Mg. However the reasons for its failur e in depositing Mg were not reported in their work. Moreover, when the above mentioned, modi fied electrolyte was used for deposition, the concentration of Mg in the electrolyte was show n to build up before reach ing a steady state [41]. The reasons for this steady state were not clear from their work. For our study a modified electr olyte suggested by Lehmkuhl et al. [41] was used which has a composition of Na[AlEt4] + 2.5 AlEt3 + 3.3 Toluene. A pre-electrodeposition process was carried out to introduce Mg into the el ectrolyte. Current density of 60 mA/cm2, at temperature of 90 oC throughout this research for introducing Mg into the electrolyte. Si nce the electrolytes similar to the present one ha ve melting point of about 40 oC, a temperature of 90 oC was used so that the electrolyte remains in liquid state. When the using such a high temperature the use of current densities higher than 60 mA/cm2 was found to cause evapor ation of the electrolyte. Hence. to minimize the evaporation of the el ectrolyte a current dens ity value of 60 mA/cm2 was used for our study. In order to understand the proc ess of Mg accumulation in the electrolyte a systematic study on the changes in the electrol yte composition with pre-electrodeposition time is necessary. For our study the pre-electrodepositio n time was varied from 20 to 120 minutes. 4.2 Phase Analysis Figure 4-1 shows the XRD profiles of th e powders produced after different preelectrodeposition times. It can be observed from Figure 4-1 that when pre-electrodeposition was done for 20 minutes, peaks corresponding to only F CC Al-rich solid solution are obtained. In the

PAGE 65

65 case of powder produced after 80 minutes of pr e-electrodeposition time peaks corresponding to HCP Mg-rich solid solution and an intermetallic phase (Al12Mg17) just started to appear. Here still FCC Al-rich solid solution phase is in majo rity. After pre-electrode position for 120 minutes HCP Mg-rich solid solution is found in majority. The relative amount of intermetallic phase also increased. The peaks corresponding to copper are due to a small amount of copper entered into the alloy powders while scraping these powders fr om the copper electrode to prepare samples for XRD analysis. 4.3 Compositional Analysis This section describes the compositional analys is conducted in both the electrolyte and the deposits. Various techniques such as mass bala nce calculation and direct measurement of weights were carried on which are de scribed in the following sections. 4.3.1 Deposits The compositions of the deposits at each pre-electrodepositio n time were measured using EPMA. The validity of some of these result s was confirmed by IC P-OES analysis. The composition of the deposits as a function of pr e-electrodeposition time is plotted in Figure 4-2. These results reveal that at low pre-electrodepos ition times the concetrations of Mg and Al stay constant and the level of Mg in the deposits is re latively low. This stage is denoted as stage I. However, at longer pre-electrodeposition times th e concentration of Mg increases rapidly. This stage is denoted as stage II. 4.3.2 Electrolyte To estimate the amounts of Mg and Al contai ned in the electrolyte two different mass balance schemes were used for each. In the case of Mg the anode weight loss was assumed to be equal to sum of the weight of Mg in the deposit and that left in the elect rolyte. For aluminum, the weight difference between the initially added Al (in the form of different chemical components)

PAGE 66

66 and that in the deposit is considered to be left in the electrolyte. The calculated weights of Mg and Al in the electrolyte using these ma ss balance schemes are given in Table 4-1. The relative ratio of the Mg and Al contents in the elec trolyte was estimated by using EPMA technique. Since the electrolyte is moistu re and oxygen sensitive, it was oxidized after conducting deposition using isopropy l alcohol + toluene mixture befo re taking it out of the glove box. The electrolyte after quenching was rendered mo isture and oxygen-safe and was in the form of powder. This powder was cleaned and pellet s were made out of them for compositional analysis. The details of electrolyte treatment af ter pre-electrodeposition for composition analysis are discussed in Chapter 3. The measured Mg/Al rati o in the electrolyte is plotted as a function of pre-electrodeposition time in Figur e 4-3(A). The dashed line should only be taken as a visual aid as it does not represent any curve fitting among the data points The calculated Mg/Al ratios in the electrolyte (from the mass balance scheme s) are also plotted in Figure 4-3(A) for comparison. As shown in Figure 4-3(A) the ca lculated Mg/Al ratios are in a reasonable agreement with the measured values. Figure 4-3(A) demonstrates that the Mg/Al ratio increases sharply in the beginning, during st age I, and reaches almost a sa turation level with the progress of pre-electrodeposition during stag e II. It should be noted here th at this behavior contains two effects namely, one from both th e dissolution and removal (for deposition) of Mg from the electrolyte and the other from the removal of Al from the electrolyte. The concentrations of Mg and Al that are present in the electrolyte as a function of preelectrodeposition time are plotted in Figure 4-3(b). These results reveal that not only the ratio but the c oncentration of Mg and Al also increase and decrease, respectiv ely, with the pre-electrodepositi on time during stage I, before reaching nearly a saturation level in stage II.

PAGE 67

67 Figures 4-2 and 4-3 show that the process of pre-electrodeposition is divided into two stages in terms of the composition of either electrolyte or that of deposits. During the stage I of the pre-electrodeposition there is a rapid increase in the amount of Mg in the electrolyte (see Figure 4-3(A) and (B)). During this stage the concentration of Mg in the deposits is only very slightly increasing and maintained at very low value (see Figure 4-2). The concentration of Mg in the deposits at, for exam ple, pre-electrodeposition times 20 and 80 minutes is very low although there is a gradual increase in Mg in the electrolyte from low to high values. This means that although more Mg is di ssolving in the electrolyte wi th the increase in the preelectrodeposition time, only a small amount of Mg enters the deposits In other words there is an accumulation of Mg in the electrolyte during stage I. Moreover, during this stage the Al content in the electrolyte is reducing as shown from Figure 4-2(B). Consiste nt with the low values of Mg in the deposits, the deposit at 20 minutes of pre-electrodeposition time showed only FCC Al-rich solid solution (see Figure 4-1). Th e Al lost in the electrolyte must be entering the deposits in the form of FCC Al-rich phase. The slight increase in Mg in the deposits from pre-electrodeposition time of 20 minutes to 80 minutes is started to ma nifest as a small fractio n of HCP Mg-rich solid solution in the deposit made at pre-electro deposition time of 80 minut es (see Figure 4-1). However, the FCC Al-rich phase is in majority in this deposit. Apart from the Mg-rich phase the deposit made at 80 minutes pre-electr odeposition time consisted of some Al12Mg17 intermetallic phase also. So it can be said that the stage I of preelectrodeposition is realized when there is an accumulation of Mg and depletion of Al in the el ectrolyte at which time the concentration of Mg in the deposits would be low and that of Al w ould be high, both of which are changed only very slowly with the preelectrodeposition time.

PAGE 68

68 In the stage II of the preelectrodeposition the amount of Mg in the electrolyte is maintained almost constant (see Figure 4-3), wh ereas in the deposit the concentration of Mg is increased rapidly (see Figure 4-2). This indicates that almost all the Mg that dissolved in the electrolyte during this stage is deposited on the cathode main taining a steady state of Mg concentration in the electrolyte. The increased Mg in the deposits during this stage appeared in the form of increased fraction of HCP-Mg-rich phase and intermetallic phase. Figure 4-2 shows that HCP Mg-rich phase is in majority although th ere is some FCC Al rich phase in the deposit made at pre-electrodeposition time of 120 minutes. Th e slight change in the amount of Al in the electrolyte during this stage can be explained w ith only a small concentration of Al that is entering HCP Mg-rich phase in the deposits. When the HCP Mg-rich phase is populating in the deposits the relative concentration of Al in the de posits is becoming low. This is the reason that there is drop in the concentration of Al observed in the deposits during this stage (Figure 4-2). So the stage II of the pre-electrodeposition is identi fied with an establishm ent of steady state in terms of concentration of Mg and slight decrease in the concentrati on of Al in the electrolyte at which time there is a rapid increase (or decrease) in the composition of the deposits in terms of Mg (or Al). Thus there seems to be a critical time fo r pre-electrodeposition fo r given experimental conditions prior to which, called th e stage I, there is an accumula tion of Mg in the electrolyte. After this critical time of preelectrodeposition, a stea dy state in terms of concentration of the Mg in the electrolyte is reached similar to the results of Lehmkuhl et al. [41]. This is called the stage II. Since this critical pre-el ectrodeposition time may depend on th e experimental conditions such as current density, temperature etc. a more precise way of identifyi ng this critical situation is by considering the concentration of Mg in the electrolyte. The composition (in terms of Mg) of

PAGE 69

69 deposits is plotted as a function of the Mg/Al ra tio in the electrolyte in Figure 4-4.From Figure 44 it is seen that a critical Mg/A l ratio in the electrolyte can be identified prior to which the concentration of Mg in the deposits is very low. This is the stage I during which only FCC Alrich solid solution forms in the deposits. When the Mg/Al ratio in the electrolyte exceeds this critical Mg/Al ratio the concentr ation of Mg in the deposits increases rapidly even for a small change in the Mg/Al ratio in the electrolyte. Th is is the stage II during which HCP Mg-rich phase forms and populates in the deposits. This critical Mg/Al ratio in the electrolyte depends on when a second phase (HCP Mg-rich) is forming in the deposits. 4.4 Anodic and Cathodic Reactions The anodic and cathodic reactions were assumed to be oxidation of Mg and reduction of Al and Mg ions respectively according to the fo llowing: These assumptions are based on 100% efficiency. As will be shown later, there may be ot her side reactions that result in a reduction in the observed efficiency. Anodic reaction: Mg Mg+n + ne(4-1) Cathodic:reactions: Mg+n + neMg (4-2) Al+3 + 3eAl (4-3) As shown in equations (4-2) and (4-3), the va lence state of the disso lving or depositing Mg is taken as unknown (n). The valence of the di ssolving Mg was considered as both univalent and/or divalent Mg in either aqueous [102] or non-aqueous [103] electrolyte systems. Irrespective of whether th e electrolyte system is aqueous or non-aqueous, the studies suggested that the presence of molecules wh ich act as electron acceptors in the electrolyte system increase the stability of the univalent Mg+, which eventually gives up it s valence electron to those

PAGE 70

70 electron acceptors before forming Mg+2. For our study the Al cations are assumed to have a constant valence of 3. Since the electr ochemical changes taking place during the electrodeposition are not ye t known, the valence of the depositi ng as well as dissolving Mg is treated as unknown but equal to each other (n). Charge balanc e calculations were performed, which were coupled with values of weight loss of anode and weight s of Al and Mg in the deposit formed at cathode. An estimate of the valen ce of both the dissolving and depositing Mg was found. While calculating the vale nce of the Mg ions the erro rs involved from the equipment (EPMA and balance) were considered. The values of these valences are listed in Table 4-2 for different pre-electrodeposition times Table 4-2 shows that the valu es of the valence for all the pre-electrodeposition times are nearly close to 2 which is a reasonable number. Hence for our study the valence of both dissolving and depositing i ons is taken as 2, (i.e n=2). Rausch et al. [103] mentioned that the molecules which act as electron acceptors (better known as Lewis acids) stabilize the Mg+. According to Rausch et al. [103], Mg dissolves anodically to yield a mixture of both Mg+ and Mg+2. In the absence of electron acceptor molecules the Mg+ is relatively less stable than Mg+2 hence only the latter forms. The value of n=2 in our study suggests that during the electrodeposition also no electron acceptors would be forming. Figure 4-5 shows the efficiency values as a function of different Mg/Al ratios in the electrolyte. The efficiency va lues are plotted both by assuming a valence of 2 for Mg and by considering the calculated values for the same. As shown in Figure 4-5 both sets of the values of efficiencies are very close to each other indicating that the variation in the efficiency values are not resulted from change in the va lence of the Mg. Hence the other possibility of the variation of the efficiencies and their values typically far from 100 may be due to side reactions taking place

PAGE 71

71 in the system (if any).At this moment it is not known about the side reactions that might be present during el ectrodeposition. 4.5 Roles of Electrolyte Components In order to understand the ro les of different components in the electrolyte used, three component-electrolyte systems were considered in our study. They ar e electrolyte (I): AlEt3 + 3.3 Toluene; electrolyte (II): Na[AlEt4] + 2 AlEt3 + 3.3 Toluene; electrolyte (III): Na[Et3Al-HAlEt3]+3 Toluene. These component electrolytes were chosen such that the results from electrolyte (I) and (II) would yield an understanding of the purpose of Na[AlEt4]. Also, the results from electrolyte (III) would yield an understanding of the purpose and possible effects that the complex may have on the deposition. El ectrodeposition was performed with each of these component electrolytes at current density of 60 mA/cm2 and temperature of 90 oC for 15 minutes. 4.5.1 Component Electrolyte (I) The component electrolyte (I) has the composition AlEt3 + 3.3 Toluene. When electrodeposition was attempted with electrolyte (I) no deposit was obtained at the cathode. This suggests that since toluene is not strongly polar AlEt3 could not be dissociated into ionic species. 4.5.2 Component Electrolyte (II) The component electrolyte (II) has the following composition Na[AlEt4] + 2 AlEt3 + 3.3 Toluene. The main differences between this electrolyte and electrolyte (I) are the presence of Na[AlEt4] and increased ratio of AlEt3 to toluene in electrolyte (II). The reason for increased ratio of AlEt3 to toluene is that Lehmkuhl et al. [ 40, 41] suggested that an additional AlEt3 is required to prevent the deposition of alkali meta l in preference to either Al or Mg. When deposition was conducted with electrolyte (II) both Al and Mg were deposited and the Mg anode did not develop any black film on it. Figure 46 shows SEM images of the Al-Mg alloy deposit

PAGE 72

72 in the as-deposited condition obtai ned in this case. The deposit showed a composition range from about 4-25 at%Mg. Sodium tetraethylaluminate (Na[AlEt4]) was mentioned to be i onic in nature maintaining equilibrium with its dissociation products Na+ and [AlEt4][58]. Hence in the case of electrolyte (II), Na[AlEt4] should be imparting polarity to electr olyte (II) by dissociating into ions and reforming Na[AlE4]. However, since the toluene is not st rongly polar, these dissociation products exist as ionic pair Na+//[AlEt4]-. At present the effect of the Na[AlEt4] on electrolyte (II) is not understood. 4.5.3 Component Electrolyte (III) The composition of electrolyte (III) is Na[Et3Al-H-AlEt3] + 3Toluene. When electrodeposition was done with th is electrolyte using Mg anode, an Al-Mg alloy was deposited. Figure 4-7 shows the SEM images of the deposit obtained in this case. Interestingly, this result is in contradiction to that of Lehmkuhl et al. [ 40, 41]. Their studies suggest that when deposition was attempted with only Na[Et3Al-H-AlEt3] complex in the presence of toluene with Al and Mg anodes connected in parallel the deposit contained only Al. There was a weight loss in the Al anode. However, the weight of the Mg anode rema ined the same. One possibility for the absence of Mg in their results could be that the curren t density values used by Lehmkuhl et al. [40, 41] were very low (about 30 mA/cm2 in comparison to 60 mA/cm2 in our study). The low current densities mean that the reacti on rates during the electrodepositi on process are low. Hence the amount of Mg dissolved into the electrolyte and that deposited on the cathode was probably low. It will be shown in the Chapter 5 that the amount of Mg in the deposits is directly proportional to the applied current density value. It should be not ed that if the amount of Mg deposited is below the detectable limits of the characterization tec hniques used for compositional analysis then it could not have been detected at all. In the present work although an Al-Mg alloy was obtained

PAGE 73

73 using electrolyte (III), it is impor tant to note that the composition in terms of Mg in this case is less than that observed in the deposit obtained usi ng electrolyte (II). This could be probably due to the differences in the chemical changes occu rring in both these electro lyte systems during the electrodeposition process. Another interesting observation during the de position of Al-Mg alloys using electrolyte (III) is that a black film was formed on the anode This film was reasonably adherent on the Mg surface. This observation is similar to the studie s of Lehmkuhl et al [41] which reported the appearance of insoluble films of magnesium halide on Mg an ode during the electrodeposition using Na[Et3Al-X-AlEt3] (where X = halide). Figure 4-8(A) s hows the picture of Mg anode with the black film on it. XRD was done on a sample of this black film by scraping it from the anode. Figure 4-8(B) shows the XRD profile and th is black film was identified as MgH2. The peaks corresponding to Mg are due to some Mg that en tered the XRD sample while scraping the black film from the anode. This kind of black film was also observed by others during anodic dissolution of Mg into aqueous elec trolytes. In the case of the aque ous electrolytes the black film was identified as MgO [104]. The formation of a black film during anodic dissolution of Mg into either aqueous or non-aqueous elec trolytes suggests that an anionic species reacts with the positively charged Mg anode. As the results suggest, in the present case th is cationic species is H-. Unlike in the studies of Lehmkuhl et al [40, 41], Al-Mg alloy was deposited from electrolyte (III) in our study. Since, the Na[Et3Al-H-AlEt3] is conductive enough [40, 41], during the electrodeposition Na[Et3Al-H-AlEt3] must be dissociating into Na+ and [Et3Al-H-AlEt3]-. The formation of MgH2 black film during electrodeposition w ith electrolyte (III) indicates that the source of hydrogen is from the breakage of Al-H-Al bond in the [Et3Al-H-AlEt3]liberating

PAGE 74

74 H-.This liberated Hmust be absorbed into Mg anode forming MgH2 black film. When the Al-HAl bond is broken the AlEt3 is released from the complex gr oup and this must be dissociating further to give Al+3 which will be deposited at cathode Now in the present case the AlEt3 can not dissociate in the same way it does in the case of electrolyte (II). Th is is because of the decreased stability of [AlEt2]+due to the presence of Na+. The [AlEt2]+ was shown to be stable only in the presence of electron donors (be tter known as Lewis bases) su ch as THF, pyridine and NEt3 [105107]. Hence, in the presence of Na+ the AlEt3 may dissociate into [AlEt4]and Al+3 according to the following way as suggested by Ziegler et al. [58], 4AlEt3 3[AlEt4]+Al+3 (4-4) Because of the low conductivity of the solv ent (toluene) the ions can not exist by themselves. One possibility is to exist as ion pairs consisting of cation and anion per pair. At present it is not known about the existence of ion pair s in the present electrolyte systems. 4.6 Al-Mg Alloy Deposition Scheme Considering the above mentioned observa tions and dissociati on scheme of AlEt3 in presence of component electrolyte (III) a scheme of deposition of Al-Mg alloy is proposed for from the electrolyte (III) as shown in Figure 4-9. During the electrodeposition from electroly te (III), as shown in Figure 4-9, Al+3 which is liberated according to the above me ntioned equation (4-4) gets deposit ed on the cathode. In order to maintain the charge neutrality in the electrolyte again equi-coulombic amount of Mg+2 dissolves from the anode and ente rs electrolyte (III). This Mg+2 in the electrolyte combines with [AlEt4]to yield Mg[AlEt4]2 and is accumulated in electrolyte (III). This Mg in the form of Mg[AlEt4]2 accounts for the rest of the Mg accumulated in the actual electrolyte that is observed in the stage I of pre-elet rodeposition (see Figure 4-3).

PAGE 75

75 As the deposition progresses with electrol yte (III), more Mg is removed from the accumulated Mg in the electrolyte and deposited the cathode. Hence again an equi-coulombic amount of Mg is dissolved from anode to maintain the charge neutrality of the electrolyte at any time. Thus at any instance, the concentration of Mg in the electrolyte remains constant. This explains the steady state observed in terms of the Mg in the electrolyte during the stage II of preelectrodeposition (see Figure 4-3).It should be no ted here that the concentration of Mg accumulated in the case of electrolyte (II) is mo re (1.5 mole) than that during deposition from electrolyte (III) (0.5 mole). This difference in th e concentration of accumulated Mg in the case of electrolyte (II) in comparison to electrolyte (II I) is the reason for observing higher compositions (in terms of Mg) in the de posit obtained from electrolyte (II) than those from electrolyte (III) (see Figures 4-6 and 4-7). The amount of Mg that is entering the deposits at any time depends on the accumulated Mg in the electrolyte and thus de pend on the pre-electrodeposition time. Moreover, the concentration of Mg in the deposits might also depend on th e other experimental conditions such current density, temperature etc. 4.7 Summary Mg can be incorporated into the electrolyte during the electro deposition using the modified electrolyte suggested by Lehmkuhl et al. [40, 41]. The process of incorporation of Mg in the electrolyte is termed pre-electrodeposition. Pre-electrodeposition process can be divided into two stages. In the stage I, the rate of accumulation of Mg in the el ectrolyte is higher than the rate at which Mg enters the deposits. The low concentrat ions of Mg in the deposit during the stage I are due to the presence of FCC Al-rich phase in majority in the deposits. During the stage II of the pre-electrodeposition, the accumu lated Mg in the electrolyte al most reached saturation level. At the same time the concentration of Mg in th e deposits shoots up. This rapid increase in the concentration of Mg in the deposits during stage II is due to the presence of HCP Mg-rich phase

PAGE 76

76 in the deposits. As the deposition progresses, th e deposit is populated mo re with HCP Mg-rich phase and hence the concentration of Mg in it goes up. Due to the saturation reached in the electrolyte during the stage II, a steady state is reached in the electrolyte in terms of Mg concentration. This means that the rate of dissolution of Mg in the electrolyte is almost same as that for deposition. The critical Mg/Al ratio which differentiates the two stages of preelectrodeposition depends on the ons et of nucleation of HCP Mg-ri ch phase in the deposits. The valence of the dissolving Mg was estimat ed to be about 2 which is common for Mg. Therefore, the assumptions of dissolution of Mg as the anodic reaction and deposition of Al and Mg as cathodic reaction are reasonable The valence of 2 for dissolving Mg suggests that there are no electron acceptor molecu les in the electrolyte. The failure of electrolyte (I) in deposition of either Al or Mg suggests that AlEt3 does not have good conductivity and needs addition of a better conducting compound. The Na[AlEt4] must have imparted conductivity to electrolyte ( II) as its addition to electrolyte (I) rendered deposition of both Al and Mg. The breakage of th e Al-H-Al is the reason for the formation of a black film on the Mg anode. The composition (in terms of Mg) of deposit depends on the concentration of accumulated Mg in the electrolyte. The higher concentration of Mg in the deposit obtained from electrolyte (II) than in the case of deposit from elect rolyte (III) may be due to higher concentration of accumula ted Mg in the former electrolyte. A scheme for deposition of Al-Mg alloys is proposed which is cons istent with the above observations.

PAGE 77

77 30405060708020 min 80 min 120 min * Al (+Mg) Al12Mg17Mg (+Al) Cu 2 (degrees)Intensity(arb. units)(111) (200) (220) (311) (100) (002) (330) (101) (332) (422) (102) (110) (103) (112) (004) 30405060708020 min 80 min 120 min * Al (+Mg) Al12Mg17Mg (+Al) Cu 2 (degrees)Intensity(arb. units) 30405060708020 min 80 min 120 min * Al (+Mg) Al12Mg17Mg (+Al) Cu 2 (degrees)Intensity(arb. units)(111) (200) (220) (311) (100) (002) (330) (101) (332) (422) (102) (110) (103) (112) (004) Figure 4-1. The XRD profiles of Al-Mg alloy po wders produced after pre-electrodeposition for different times with major peak s of all the phases indexed. 0 20 40 60 80 100 020406080100120140Pre-Electrodeposition Time (min)Mg or Al (at.%) in deposit Mg Al Pre-ElectrodepositionTime (min) Mg or Al (at.%) in deposit Stage I Stage II 0 20 40 60 80 100 020406080100120140Pre-Electrodeposition Time (min)Mg or Al (at.%) in deposit Mg Al Pre-ElectrodepositionTime (min) Mg or Al (at.%) in deposit 0 20 40 60 80 100 020406080100120140Pre-Electrodeposition Time (min)Mg or Al (at.%) in deposit Mg Al 0 20 40 60 80 100 020406080100120140Pre-Electrodeposition Time (min)Mg or Al (at.%) in deposit Mg Al Pre-ElectrodepositionTime (min) Mg or Al (at.%) in deposit Stage I Stage II Figure 4-2. Magnesium and Aluminum in th e deposits plotted as function of preelectrodeposition time.

PAGE 78

78 Stage I Stage II Measured Calculated 0 0.01 0.02 0.03 0.04 020406080100120140Pre-ElectrodepositionTime (min)Mg/Al in electrolyteStage I Stage II Measured Calculated A 0 1 2 3 4 020406080100120140Pre-Electrodeposition time (min)Conc. of Mg in electrolyte (mg/ml)103 104 105 Conc. of Al in electrolyte (g/l) Stage IStage II (g/l)B Stage I Stage II Measured Calculated 0 0.01 0.02 0.03 0.04 020406080100120140Pre-ElectrodepositionTime (min)Mg/Al in electrolyteStage I Stage II Measured Calculated Stage I Stage II Measured Calculated 0 0.01 0.02 0.03 0.04 020406080100120140Pre-ElectrodepositionTime (min)Mg/Al in electrolyte Stage I Stage II Measured Calculated 0 0.01 0.02 0.03 0.04 020406080100120140Pre-ElectrodepositionTime (min)Mg/Al in electrolyte 0 0.01 0.02 0.03 0.04 020406080100120140Pre-ElectrodepositionTime (min)Mg/Al in electrolyteStage I Stage II Measured Calculated A 0 1 2 3 4 020406080100120140Pre-Electrodeposition time (min)Conc. of Mg in electrolyte (mg/ml)103 104 105 Conc. of Al in electrolyte (g/l) Stage IStage II (g/l)B Figure 4-3. Concentration of Mg in the electrolyte showing A) measured and calculated Mg/Al ratio values in electrolyte, and B) weights of Mg andAl in the electrolyte plotted as function of pre-electrodeposition time.

PAGE 79

79 0 20 40 60 80 00.010.020.030.04Mg/Al in electrolyteMg (at.%) in deposit Stage IStage II Mg/Al in electrolyte Mg(at.%) in deposit 0 20 40 60 80 00.010.020.030.04Mg/Al in electrolyteMg (at.%) in deposit Stage IStage II 0 20 40 60 80 00.010.020.030.04Mg/Al in electrolyteMg (at.%) in deposit Stage IStage II Mg/Al in electrolyte Mg(at.%) in deposit Figure 4-4. Magnesium (at.%) in the deposit plotted as function of Mg/Al ratio in electrolyte 0 25 50 75 100 00.010.020.030.04Mg/Al ratio in electrolyteEfficiency valence of Mg =2 Valence of Mg from calculated values Figure 4-5. Efficiency of the pr e-electrodeposition process as a function of Mg/Al ratio in the electrolyte.

PAGE 80

80 200 m 10 m 10 m 10 m 24 at.%Mg 9 at.%Mg 4 at.%Mg 200 m 10 m 10 m 10 m 24 at.%Mg 9 at.%Mg 4 at.%Mg Figure 4-6. The SEM images of the depos it obtained using electrolyte (II) 200 m 10 m 6 at.%Mg 200 m 10 m 6 at.%Mg Figure 4-7. The SEM images of the depos it obtained using electrolyte (III)

PAGE 81

81 Black film Mg anode Black film Mg anode A 20304050607080901002theta (deg)Intensity (a.u.) Mg MgH2 2 (degrees) Intensity (arb. Units)(100) (002) (101) (102) (110) (103) (112) (201) (110) (101) (200) (211) (321) (222) (411) 20304050607080901002theta (deg)Intensity (a.u.) Mg MgH2 2 (degrees) Intensity (arb. Units) 20304050607080901002theta (deg)Intensity (a.u.) Mg MgH2 2 (degrees) Intensity (arb. Units)(100) (002) (101) (102) (110) (103) (112) (201) (110) (101) (200) (211) (321) (222) (411)B Figure 4-8. Analysis of black film on anode showi ng A) picture of Mg anode with black film on it, and B) the XRD profile of black film shows the presence of MgH2 in the black film.

PAGE 82

82 2NaH + 4AlEt3+ 3Toluene Cathode Anode (Mg)(p/2) Mg+2+ pe-1 (p/2)Mg Al+3+ 3e-1 Al ( p/2 + 1/2)Mg ( p/2 + 1/2)Mg+2+ (p + 1)e-1(p + 3)e-1(p + 1)e-1+ 2e-12[Et3Al -HAlEt3]-1 4AlEt3+ 2H + 2e-12H(ab)+ Mg MgH24AlEt3 3[AlEt4]-1+ Al+3 2Na+1+2[Et3Al -HAlEt3]-1+ 3Toluene 2Na++ 2[AlEt4]-1 2Na[AlEt4] ( p/2 +1/2)Mg+2+ [AlEt4]-1 1/2Mg[AlEt4]2 + (p/2) Mg+2 Mg+2accumulation in electrolyteElectrolyte (III) 2NaH + 4AlEt3+ 3Toluene Cathode Anode (Mg)(p/2) Mg+2+ pe-1 (p/2)Mg Al+3+ 3e-1 Al ( p/2 + 1/2)Mg ( p/2 + 1/2)Mg+2+ (p + 1)e-1(p + 3)e-1(p + 1)e-1+ 2e-12[Et3Al -HAlEt3]-1 4AlEt3+ 2H + 2e-12H(ab)+ Mg MgH24AlEt3 3[AlEt4]-1+ Al+3 2Na+1+2[Et3Al -HAlEt3]-1+ 3Toluene 2Na++ 2[AlEt4]-1 2Na[AlEt4] ( p/2 +1/2)Mg+2+ [AlEt4]-1 1/2Mg[AlEt4]2 + (p/2) Mg+2 Mg+2accumulation in electrolyteElectrolyte (III) Figure 4-9. Scheme of Al-Mg alloy deposition using electrolyte (III)

PAGE 83

83 Table 4-1. Calculations from Mg and Al mass balance schemes in electrolyte (30 mL) and deposits Mg mass balance scheme (all weights are in milligrams) PreElectrodeposition time (min) Anode weight loss Weight of Mg in the deposit Weight of Mg in electrolyte 20 26 2 24 80 122 32 90 120 185 95 90 Al mass balance scheme (all weights are in milligrams) PreElectrodeposition time (min) Initial weight of Al Weight of Al in the deposit Weight of Al in electrolyte 20 3164 17 3147 80 3164 72 3092 120 3164 73 3091 For the Mg mass balance scheme the anode weight loss is assumed to be distributed in both deposit and electrolyte after each pre-electrodepos ition step. For the Al mass balance scheme the initial weight of Al is assumed to be distribut ed in both deposit and electrolyte after each preelectrodeposition step. Table 4-2. Estimate for the valence of dissolving or depositing Mg PreElectrodeposition time (min) 20 80 120 Valence of Mg (n) 1.9.1 1.7.1 2.08.02

PAGE 84

84 CHAPTER 5 ELECTRODEPOSITION OF PURE AL AND AL-MG ALLOYS As seen in Chapter 4 the concentration of th e Mg in the electrolyte depended on the preelectrodeposition time. The effect the concentration of Mg in the electrolyte and that of various parameters such as current density and te mperature and anode composition on the morphology and composition of the Al-Mg alloys is presented in this chapter. In order to establish the general techniques and feasibility of the process of electrodeposition with organometallic-based solutions, the electrodeposition of pure Al was atte mpted as a preliminary work prior to that of Al-Mg alloys. With the gained experience the elec trodeposition of Al-Mg a lloys was carried out. 5.1 Electrodeposition of Pure Al As explained in Section 3.31 the electrolyte used for the depos ition of pure Al consisted of NaF, AlEt3 and toluene as suggested by Lehmkuhl et al. [50]. Electrodeposition was conducted potentiostatically using an applied potential of E =-1.5 V based on the cyclic voltammetry result presented in Figure 3-5(A). In order to understand the role of the complex Na[Et3Al-F-AlEt3] that was expected to form, two different mole ratios, namely, 2:1 (stoic hiometric) and 4:1 of AlEt3/NaF were chosen. The electr odeposition was conducted at 60 oC. For detailed description of electrolyte preparation and the procedur e of electrodeposition refer to Chapter 3. 5.1.1 Current Density Vs Time Results The current density vs time curves gene rated during the electrodeposition for both compositions are shown in Figure 5-1. When a stoichiometric amount of NaF and AlEt3 were used higher current density values were recorded in comparison to the case when there is an excess of AlEt3. Since, there are no side reactions ta king place when a potential of -1.5 V is applied according to the cyclic voltammetry result pr esented in Figure 3-4(A), all the current that is being recorded is a repres entative of the number of Al+3 ions reaching the cathode. Hence

PAGE 85

85 assuming that the complex Na[Et3Al-F-AlEt3] was formed in the solution, the concentration of this complex would be low in the ca se where there is an excess of AlEt3. Since the current density values reported in th e second case (excess of AlEt3) are lower than those in the first case (stoichiometric ratio of NaF to AlEt3) it can be assumed that the Al that is deposited on the cathode is from the Na[Et3Al-F-AlEt3] complex. Hence the complex Na[Et3Al-F-AlEt3] is the one that is electrochemically activ e entity during the electrodeposition. 5.1.2 Morphological Analysis of Pure Al Figure 5-2 shows the SEM images of the de posits obtained with the two electrolyte compositions. The deposits grew by nucleation and gr owth of many particles, each consisting of multiple crystals. It can be observed that at higher current densities with 2:1 ratio of AlEt3:NaF the deposit exhibits a fine structure (see Figure 5-2(A)). Increasing the AlEt3:NaF ratio, which yielded a lower current density re sulted in coarser structure as shown in Figure 5-2(B). This is because when the current density values are higher larger number of sites can be activated and more number of nucleation sites can be rea lized, consequently a fine structure is anticipated.[108] A detailed analys is of morphology revealed that the aluminum crystals exhibit a 5-fold symmetry (see Figure 5-2(C)). This 5-fold symmetry has been observed by several investigators [109-111] and in the electrodeposition of pure FCC meta ls such as Au its presence has been attributed to multiple twinning [112]. There has been a significant amount of work on the mechanism of formation of these twins. In the case of FCC metals it was suggested that the twinning takes place in the early stages of the deposition. A twin was s uggested to be forming over a two dimensional nucleation step or a twin pl ane, (111) in the case of FCC metals [113]. It was also shown that twin nuclei can form only a bove a critical overvoltage or applied potential [114]. Since, the characterization of electrodeposited pure Al is not the focus of the present work, further analysis on the twinning of pure Al was not pursued.

PAGE 86

86 With a preliminary understanding of the role of the complexes in the electrolyte and with some gained experience on handling highly mo isture and oxygen sensitive organometallic chemicals, electrodeposition of Al-Mg all oys was pursued for the present work. 5.2 Electrodeposition of Al-Mg Alloys Electrodeposition of Al-Mg alloys was conduc ted using an organometallic electrolyte based on the work by Lehmkuhl et al. [40, 41]. Th e electrolyte has the following composition 1 mol Na[AlEt4], 2 mol Na[Et3Al-H-AlEt3], 2.5 mol AlEt3 and 6 mol toluene. Detailed description of the method of preparing the electrolyte for Al-Mg alloys deposition is given in Section 3.3.2 of Chapter 3. A rotation speed of 200 rpm was sel ected for our study. This rotation speed was the maximum value at which no wobbling was observed. The maximum value of the rotation speed was chosen to drive as much Mg in the bulk of the electrolyte as possible towards the diffusion layer. The effect of various deposition parameters such as concentration of Mg in the electrolyte, deposition time, current density, temperature and anodic composition on the characteristics of the deposits was studied. The experimental conditions are listed in Table 51. Temperatures for our study were selected based on the melting point of the electrolyte system. Since this electrolyte starts to become more viscous at about 45 oC in comparison to 60 oC two temperatures namely, 90 and 60 oC were selected. The current density values were selected to be in the mass transfer regime so that dendritic growth could be obtai ned. Different pre-electr odeposition times were selected, some of which were same th e values studied in the Chapter 4. In general, the values of the potential during the deposition varied as shown in Figure 5-3. The effect of Mg concentration in the elec trolyte on the morphology and composition of the deposits was studied by making depo sits at a temperature of 90 oC with different preelectrodeposition times from 20 min to 180 min (de posit #1-4). The effect of deposition time was

PAGE 87

87 studied by fabricating deposits with two different electrodeposition times (deposit #4 and 5). For the studying of the effect of current density on the characteristics of the deposits, they were fabricated at 60 oC with different current density values of 60, 100 and 150 mA/cm2 (deposit #68). In order to control the composition of the deposits electrodeposition was done at 60 oC using two anode compositions. The anode compositions were pure Mg and 80 area%Al + 20 area%Mg (deposit #8 and 9). 5.3 Growth and Morphology of the Deposits In general, three stages of growth were id entified in the deposits made in this study. Initially a dense and fine struct ure developed on the substrate wh ich, hereforth will be termed ground. The morphology of the ground in general consisted of hemisphe rical units. Upon this ground two types of growth were re alized. In some deposits an in termediate growth pattern was found, which was coarser than the ground and hen ce, its morphology can be best defined as nodular. The third type of growth was dendritic in nature. Figure 5-4 show s the three stages of growth for two different deposits. In deposit #1 the intermediate growth was limited and the smooth ground surface can easily be recognized (a in Figure 5-4). However, deposit #3 exhibited extensive nodule formation (b in Figure 5-4) with limited dendritic growth (b in Figure 5-4). Several dendritic growth morphologies were observed in the deposits studied here. Depending on the deposition condition more than one dendrite morphology was found in each deposit. However, usually one type of dendrite was dominant for a gi ven deposition condition. Figure 5-5 shows higher magnification images of nodule and dendrite from deposit #4. As suggested by Figure 5-5 the morphology of the de ndrites did not always match that of nodules. Hence, the growth of dendrites can not be consid ered only as an extension of the intermediate growth. Figure 5-6 shows two di fferent types of dendrites form ing on the ground. The first type of dendrites is anistropic in their morphology exhibiting feather-l ike structure (see Figure 5-

PAGE 88

88 6(A)). These feather-like structures are beginni ng to grow from fine anisotropic precursors. Figure 5-6(B) shows the globular units forming over the finer globular units. In all deposits the dendrites could be scraped easily indicating that they were loosely held on the ground. However, the ground was strongly a dherent to the copper substrate. Similar observation was made in the studies of Jovic et al. who prepared Ni-Co alloy powders via electrodeposition [76]. 5.4 Effect of Electrodeposition Parameters on Morphology and Composition Phases of the Deposits Electrodeposition parameters are found to affect the characteri stics of the deposits such as morphology, composition etc. This section is devo ted to investigate the effect of different parameters such as concentration of Mg in the electrolyte, current densit y and electrodeposition time on various characteristics of the deposits. 5.4.1 Effect of Pre-Elec trodeposition Time The effect of Mg ion concentration on the ch aracteristics of the de posits was evaluated by investigating deposits prepared with four different pre-electr odeposition times (deposit #1#4) keeping all other deposi tion parameters constant. The initial electrolyte temperature was at 90 C. The dominant dendrite morphology in these depos its was globular (because of its shape) as shown in Figure 5-7. These dendrites were found to have two distinct surface roughness levels as shown in Figure 5-7(A). Raising the pre-electrodeposit ion time increased the volume fraction of the rough globular morphology. Deposit #1 show ed mostly smooth globules while the dendrites in deposits with 2 and 3 hours pre-elect rodeposition time consisted mainly of the rough globules as shown in Figure 5-7(B). EDS co mpositional analysis, which renders semiquantitative values, revealed that there exists a large differen ce in the composition of rough and

PAGE 89

89 smooth globules. The smooth ones showed 20 to 25at% Mg, however the rough globules had 55 to 85 at% Mg. In addition to a change in the coverage of the rough globules the tendency to dendrite formation was decreased with increasing concentra tion of the Mg ions in electrolyte. Figure 5-8 shows low magnification SEM pictures from de posits #1 and #4. As suggested by Figure 5-8, while dendritic growth is very prevalent in the deposit with 20 min pre-electrodeposition time (see Figure 5-8(A)), deposit #4 with 180 min pre-electrodeposition time exhibited mainly intermediate growth with limited dend rite formation (see Figure 5-8(B)). The XRD profiles for powders fabricated at various pre-electrodepos ition times are shown in Figure 5-9. The powder made of the electrolyt e prepared with 20 min pre-electrodeposition time exhibited only FCC aluminum phase. With an increase in the concentration of Mg in the electrolyte, peaks corresponding to the HCP magn esium phase were found in the XRD profiles. The majority phase at 180 min pre-electrode position time, deposit#4, was the HCP-Mg phase. From Chapter 4 it is known that at 20 min of pre-electrodeposition the Mg in the electrolyte is still in the accumulation stage. Wh en this electrolyte was used for the deposition only FCC Al-rich phase was obser ved (deposit #1) indicating that the amount of Mg in the deposit is low. When an electrolyte, which wa s used for pre-electrode position time of 80 min was used for deposition the HCP-Mg rich phase started to appear. From Chapter 4, 80 min is where the Mg in the electrolyte began to reach stea dy state in the electrolyte (see, Figure 4-2(B)). When this electrolyte was used for deposition the HCP Mg-rich phase started to appear (deposit #2) indicating that whatever Mg dissolved in the electrolyte got eventually deposited on the cathode giving rise to appearance of HCP Mg-ri ch phase. When the electrolyte on which preelectrodeposition was done for 180 min was used for deposition HCP-Mg rich phase was in

PAGE 90

90 majority. This indicates that while a steady state wa s maintained in the electrolyte in terms of Mg The amount of Mg in the deposits was built up with time giving rise to HCP Mg-rich phase as a major one. 5.4.2 Effect of Deposition Time The electrodeposition time for deposits#1 to #4 was 10 min. In order to investigate the effect of deposition thickness on morphology and composition of the dendrites, deposit #5 was prepared under similar conditions as deposit #4, except that the de position time was increased to 30 min. Figure 5-10 shows that the dendrites be came larger but their morphology and roughness remained the same. The magnesium concentratio n range measured by EDS was found to be 5884 at% Mg and 75-80 at% Mg in deposits #4 and #5, respectively. This observation suggests that, due to the anodic dissolution of magnesium during the deposition process there may be an increase in the Mg content of the deposits. 5.4.3 Effect of Current Density The effect of current density was evaluated by making deposits at 60 C and at 60, 100 and 150 mA/cm2, respectively (see Table 5-1). The applied cu rrent density as well as test temperature affected the morphology of these deposits signi ficantly. While the deposits made at 90 C showed predominantly globular dendrite morphology, several othe r varieties were observed in the deposits made at 60 C. At 60 mA/cm2 (deposit #6) faceted large aluminum crystals with low concentrations of Mg (1-4 at. %) were found (see Figure 5-11(A)). Usually these f aceted crystals were octahedral in shape as shown in Figure 511(A). The XRD analysis confirmed the presence of only FCC Al-rich solid solution in the deposit containing this morphology as can be seen in Figure 5-15 (60 mA/cm2) [115]. The octahedral nature of thes e facets suggests that the crystals grew with {111} type habit planes. This observation is consistent with the reported results for the other FCC metals electrodeposited at low deposi tion rates, such as Ag and Au [116, 117]. The

PAGE 91

91 shape of these faceted crystals suggests that ther e is a marked anisotropy in the growth rates of different crystallographic planes. The growth of th ese crystals occurred in such a way that the planes with highest interfacial en ergies disappeared giving rise to the ones with low interfacial energy. These growth rates depend on the conditions in which these crystals (planes) are growing such as current density (over potential), concentrati on of the electrolyte temperature [36, 74, 118]. Here the conditions were probably favourab le for the appearance of <111> planes. With increasing the current density, dendritic growth became more prominent, partially because of the increase in the deposition rate. Both rod-like (see Figure 5-11(B)) and globular (see Figure 5-11(C)) morphologies were observed in deposits #7 and #8. Rod-like dendrites with nominal composition range of 13 to 20 at% Mg were in majority in the sample produced at 100 mA/cm2 (see Figure 5-11(B)). At 150 mA/cm2 the dendrite morphology was predominantly globular. As demonstrated in Figur e 5-12(A), the dendrites that a ppeared globular from top view, in some cases started with a rod-like shap e that evolved into globular morphology. This morphological evolution will be dealt in detail in Chapter 6. The nominal composition range of the globular dendrites varied from 32 at% Mg to 82 at% Mg. Both rough and smooth globules were found as shown in Figure 5-12(B). Similar to the deposits made at 90 C, the high Mg content correlated with the rough morphology. The cross-sectional images suggested that the globular morphology made at 60 oC and a current dens ity of 150 mA/cm2 was porous as shown in Figure 5-13. In addition to a change in the dominant dendrite morphology, the deposit structure was refined significantly with in creasing the current density from 100 mA/cm2 to 150 mA/cm2 as depicted in Figure 5-14. Both the de ndrites and the intermediate growths were decreased in size and increased in density at the higher current density. A de tailed analysis of the morphology and microstructure of the dendrites will be discussed in Chapter 6.

PAGE 92

92 The XRD profiles of deposits #6 to #8 ar e shown in Figure 5-15. When the current densities were 60 and 100 mA/cm2 only FCC Al-rich phase was observed. As the current density was increased to 150 mA/cm2 in addition to the FCC phase, HC P Mg-rich phase was recognized. 5.4.4 Effect of Anode Composition In order to investigate the effect of anode composition on the deposits a deposit was made (deposit #9) at similar parameters as that of deposit #8 but with a different pre-electrodeposition time and anode composition during the deposition. The pre-electrodeposition time in this context was 75 min (longer than that in deposit #8) However, both these durations of preelectrodeposition times fall in stage I of the process during which accumulation of Mg ions in the electrolyte occurs (see, Figur e 4-2 of Chapter 4). The anode used for the deposit #9 had a composition of 80 area% Al + 20 area% Mg. Th e low magnification SEM image of the deposit #9 with that of deposit #10 is shown in Figure 5-16. Figure 5-16 shows that morphologies similar to those of feather-like and gl obules were observed in deposit #9. The nominal composition of feather-like morphology is around 7 at.% Mg. XRD profile of the deposit #9 indicates no HCP Mg-rich phase in contrast to that of depos it #8. The XRD profiles of both these deposits are shown in Figure 5-17. The absence of HCP Mg-ri ch phase with decreased relative composition of anode suggests that, during the deposition proce ss, only lesser amount of Mg is dissolving into the electrolyte in the case of deposit #9 in comparison to de posit #8. In spite of having higher concentration of Mg in the el ectrolyte at the beginning of th e deposition, the Mg that was dissolving during the deposition was not sufficient to reach a saturation level in the electrolyte. Hence, this lesser concentration of Mg in the electrolyte, in the case of deposit #9, is not able to form HCP Mg-rich phase.

PAGE 93

93 5.5 Consistency of the Electrodeposition Process In order to check the consistency within a same deposit and among different deposits conducted under same conditions a set of experime ntal parameters same as that of deposit #8 was chosen. Three depositions were conducted with these parameters. Although the starting temperature for the depositions was 60 oC, the actual temperature of the electrolyte seemed to rise during the electr odeposition at 150 mA/cm2. The temperature vari ation during the three depositions conducted here is s hown in Figure 5-18. The electroche mical changes taking place in the electrolytic bath must be re sponsible for the rise in the temperature during the deposition. Figure 5-19 shows the SEM images of three depos its at different levels of cathode namely top, middle and bottom. Figure 5-19 shows that the deposits are uniform throughout the length of each deposit and also among different deposit s. The globular morphology was the one in majority in all these deposits The XRD profiles of these depos its are shown in Figure 5-20. All the deposits contained peaks corresponding to both FCC Al-ri ch and HCP Mg-rich solid solutions and are almost similar. It can be said from the above results that the depositions conducted here and the electrodeposition process used for the present work in general is consistent nature. 5.6 Discussion The observation of the general ground of the deposits suggests the presence of several nuclei (see Figure 5-6). A schematic of the growth mechanism of the deposits in galvanostatic technique is shown in Figure 5-21. Generally, the process of depositi on starts with formation of a hemispherical nucleus on an atomically rough substrate (see Figure 521(A)) [119]. Since the substrate is atomically rough, t hose nuclei which are at a higher level than the others start growing at faster rates due to la rger availability of ions and hi gher local overpotential values (see Figure 5-21(B)). As these nuclei grow with the deposition time, hemispherical shaped diffusion

PAGE 94

94 fields develop around them [120]. With the further growth of these nuclei the diffusion fields increase in size and start to overlap with di ffusion fields of neighbouring nuclei. When the overlapping occurs, the hemispherical diffusion fields start vanishi ng giving rise to planar fronts of overlapped nuclei. Although, the planar fronts of nuclei are present, the original roughness levels among the individual nuclei persist (see Figure 5-21(C)). Due to the presence of roughness on the growing front, the local overpotentials ar e going be different. These differences in the local overpotentials lead to de velopment of different morphologi es as shown in Figure 5-21(D). More detailed analysis of formation of differe nt morphologies as a func tion deposition rate is discussed in Chapter 6. Aluminum-magnesium phase diagram indicates that under equilibrium conditions the solubility of Al in Mg and vice versa is very limited and there exist two intermetallic compounds, namely Al3Mg2 and Al12Mg17. Electrodeposition is a non-equilibrium processing technique and the XRD results revealed that no intermetalli cs formed under the deposition conditions investigated here. The dominant morphology of the dendrites was found to depend on composition. When the concentration of Mg in the deposit was very low (approximately <5 at%) large crystals developed (Figure 5-11(A)). An in crease in the Mg content resulted in the formation of the rodlike morphology, which was observed in deposits #7 and #8 (Figure 5-11(B)). When the Mg content was above approximately 15 at% the rod-like morphology transitioned to a globular morphology. This evolution can be seen in Figure 511(B), Such a transition along a large dendrite can be seen in Figure 5-12(A). The dendrites that contained higher Mg showed globular morphology. Among these dendrites, tw o types of globules smooth and rough were found. The rough globules always formed over the smooth ones. EDS analysis revealed that

PAGE 95

95 the rough globules usually had Mg content over 50at%. Since the deposits with strongest HCPMg XRD peaks also showed higher per cents of the rough globular morphology it can be concluded that they are HCP magnesium supers aturated solid solution phase and the smooth globules are FCC aluminum super saturated solid solution phase. The Mg content of the deposits was found to depend on the pre-electrodeposition time, electrolyte temperature and current density. Th e raise in the deposit Mg content with preelectrodeposition time, which manifested as an increase in the amount of rough globules (see Figure 5-7), can be attributed to the increase in the amount of Mg ions in the electrolyte. The enrichment of the electrolyte wi th Mg ions also reduced dendrit ic growth tendencies (see Figure 5-8). Results of Lehmkuhl et al.[40, 41] demonstrat e that in electrolytes si milar to the one used here, Mg deposits much faster than Al does and c onsequently a small concentration of Mg in the electrolyte renders a high Mg content deposit. This information with the observed reduction of the dendritic growth suggests that Mg depos ition was mass-transfer limited. A comparison of deposits #2 and #6, which were deposited at the same current density, reveals that a decrease in the electrolyte temperature signif icantly reduces the amount of Mg in the deposit. An increase in the electrolyte temperature is anti cipated to increase the diffusivity of the Mg containing ions and hence its co-deposition rate. Increasing the current density resulted in de posits with higher Mg c ontent. Assuming that the deposition of Mg under the conditions used in this study is mass-transfer limited, one would have expected the Mg content of the deposit to be lowered at higher deposition rates. For example,in Ni-Fe system, which also shows anomalous co-deposition behavior, increasing agitation or increasing the curren t density causes less co-depos ition of the faster depositing element, i.e. iron.[121, 122] The contradictory re sponse observed in the Al-Mg system implies

PAGE 96

96 that the increase in the overpoten tial at higher current densities en courages the deposition of Mg more significantly than does the redu ction of Al ions. This behavior may be partially attributed to the charge differences in Ni-Fe versus Al-Mg systems. In the former system, the faster depositing ion has the same charge (Fe+2) or possibly higher charge (Fe+3) than the slower depositing ion (Ni+2). However, in the latter system the slower depositing ion (i.e. Al+3) has a higher charge in comparison to the Mg+2 ion. It should be noted that the efficiency values in the present deposition process are not 100% (see Fi gure 4-5), Hence, it is possible that the oxidation/reduction rates of other sp ecies as a function of current density indirectly affect the relative deposition rate of Al and Mg. As the curr ent density is increased the nucleation rate is expected to improve and hence result in a finer structure. This is probably the reason for the observed refinement in the size of nodules as well as dendrites with increasing the current density from 100 mA/cm2 to 150 mA/cm2 as shown in Figure 5-14. 5.7 Summary The electrodeposition parameters were chos en such that the deposition took place under mass-transfer limited condition which resulted in dendritic growth in case of Al-Mg alloys. The morphology of the dendrites was dependent on the Mg content of the deposit. Increasing current density, Mg/Al ratio in the electrolyte or temperat ure resulted in a higher concentration of Mg in the deposit. All deposits contained FCC supersat urated Al solid solu tion. The amount of HCP supersaturated Mg solid solution increased with increasing the Mg conten t of the deposit. No intermetallic phase was found in the electrodeposited Al-Mg powders.

PAGE 97

97 0 4 8 12 16 02004006008001000Time (s)Current Density (mA/cm2)NaF+2AlEt3+3Toluene NaF+4AlEt3+6Toluene 0 4 8 12 16 02004006008001000Time (s)Current Density (mA/cm2)NaF+2AlEt3+3Toluene NaF+4AlEt3+6Toluene Figure 5-1. Current vs time curves for electr odeposition of pure Al using two different compositions of electrolyte. 50 m 200 m 20 m A B C 50 m 200 m 20 m A B C Figure 5-2. The SEM images of pure Al deposit from electrolyte with composition A) NaF+ 2 AlEt3+3 Toluene, B) NaF+ 4 AlEt3+6 Toluene, and C) same as B at higher magnification showing five-fold symm etry suggesting twin formation.

PAGE 98

98 -12 -10 -8 -6 -4 -2 0 0500100015002000Time (s)E (V)T: 60oC, i:150 mA/cm2 -12 -10 -8 -6 -4 -2 0 0500100015002000Time (s)E (V)T: 60oC, i:150 mA/cm2 Figure 5-3. Values of potential during a galvanostatic deposition conducted at 60 oC. Figure 5-4. The SEM images showing three stag es of deposit growth: A) ground, B) nodules and dendrites in deposits #1 and #3, and C) schematic showing ground, nodules and dendrites in a deposit.

PAGE 99

99 100 m 100 m100 m A B C 100 m 100 m100 m A B C Figure 5-5. The SEM images showing nodules and dendrites from deposit #4 shown in A) low magnification SEM image B) nodule, and C) dendrite. 20 m 20 m 5 m 10 m A B 20 m 20 m 5 m 10 m A B Figure 5-6. The SEM images of the ground obtained at A) 60 mA/cm2 and 60 oC (deposit #6), B) 60 mA/cm2 and 90 oC (deposit #1).

PAGE 100

100 B A Brough smooth 20 m 20 m B A Brough smooth B A Brough smooth 20 m 20 m Figure 5-7. The SEM images showing A) smooth and rough globules in deposit #1, and B) very rough globules in deposit #3. A B 500 m 500 m A B 500 m 500 m Figure 5-8. Low magnification SEM images showi ng A) extensive dendritic growth in deposit #1, and B) fine nodular growth in deposit #4. 304050607080 2 (degrees)Intensity(arb. units) ** Al (+Mg) Mg (+Al) Cu20 min 80 min 120 min 180 min(111) (220) (200) (311) (100) (002) (101) (102)(110)(103)(112) 304050607080 2 (degrees)Intensity(arb. units) ** Al (+Mg) Mg (+Al) Cu20 min 80 min 120 min 180 min(111) (220) (200) (311) (100) (002) (101) (102)(110)(103)(112) Figure 5-9. The XRD profiles of deposits #1-4 (see Tale 5-1) showi ng the effect of preelectrodeposition time on the phases present.

PAGE 101

101 B A 20 m 20 m B A 20 m 20 m Figure 5-10. The SEM images showing the gr owth of dendrites when deposition time is increased from A) 10 min (deposit #4 ) and B) 30 min (deposit #5). A B C 20 m 20 m 20 m Faceted crystals Rod-like Globular A B C 20 m 20 m 20 m A B C 20 m 20 m 20 m Faceted crystals Rod-like Globular Figure 5-11. The SEM images showing different morphologies of the deposits made at A) 60 mA/cm2, B) 100 mA/cm2, and C) 150 mA/cm2.

PAGE 102

102 10 m 20 m A B 10 m 20 m A B Figure 5-12. The SEM images showing A) rodlike morphology at the root of a globular dendrites, and B) the presence of rough globules over the smooth ones in deposit #8. 5 m 2 m 5 m 2 m Figure 5-13.The SEM images of the cross sec tion of a dendrite rev ealing the porosity.

PAGE 103

103 B 20 m D 20 m 100 m A C 100 m B 20 m D 20 m 100 m A C 100 m Figure 5-14. Low magnification SEM images demons trating the refinement of the dendrites A) and C) as well as the intermediate growth B) and D) with increase in the current density A) and B) 100 mA/cm2 (deposit #7) C) and D) 150 mA/cm2 (deposit #8) 304050607080 2 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) Cu60 mA/cm2100 mA/cm2150 mA/cm2 (111) (220) (200) (311) (100) (002) (101) (102) (110) 304050607080 2 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) Cu60 mA/cm2100 mA/cm2150 mA/cm2 (111) (220) (200) (311) (100) (002) (101) (102) (110) Figure 5-15. The XRD profiles of deposits #6-8 (see Table 5-1) showing the effect of current density on the phases present.

PAGE 104

104 100 m 100 m A B 100 m 100 m A B Figure 5-16. The SEM images showing a general re presentative morphologies in deposits made using anodes having composition of A) pur e Mg (deposit #8), B) 80 area%Al + 20 area% Mg (d,eposit#9) (the rest of the pa rameters were the same in both the cases) (see Table 5-1). 304050607080 *2 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al)Anode: Mg *Anode: 80area% Al + 20area% Mg(111) (220) (200) (311) (100) (002) (101) (102) (110) 304050607080 *2 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al)Anode: Mg *Anode: 80area% Al + 20area% Mg(111) (220) (200) (311) (100) (002) (101) (102) (110) Figure 5-17. The XRD profiles of deposits ma de using anode composition of A) pure Mg (deposit #8), and B) 80 area% Al + 20 ar ea% Mg (deposit #9) (see Table 5-1).

PAGE 105

105 50 60 70 80 90 100 05101520Deposition time (min)Temperature (oC) Deposit(I) Deposit(II) Deposit(III) Figure 5-18. Temperature variati on during deposition at 150 mA/cm2 and starting temperature of 60 oC. (other conditions for deposition are si milar to that of deposit #8, see Table 51). 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 mCathode with deposit Deposit (I) for consistency check Deposit (II) for consistency check Deposit (III) for consistency check 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 mCathode with deposit Deposit (I) for consistency check Deposit (II) for consistency check Deposit (III) for consistency check Figure 5-19. The SEM images at different level of cathode for three depositions done for checking consistency (the deposition conditi ons are same as those of deposit #8, see Table 5-1).

PAGE 106

106 3040506070802 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) *Deposit (I) for consistency check Deposit (II) for consistency check Deposit (III) for consistency check (111) (220) (200) (311) (100) (002) (101) (102) (110) 3040506070802 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) *Deposit (I) for consistency check Deposit (II) for consistency check Deposit (III) for consistency check 3040506070802 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) * 3040506070802 (degrees)Intensity(arb. units)* Al (+Mg) Mg (+Al) *Deposit (I) for consistency check Deposit (II) for consistency check Deposit (III) for consistency check (111) (220) (200) (311) (100) (002) (101) (102) (110) Figure 5-20. The XRD profiles of three depos itions done for checking consistency (the deposition conditions are same as t hose of deposit #8, see Table 5-1). Figure 5-21. General growth mechanism of deposits under galvanostatic technique.

PAGE 107

107 Table 5-1. Experimental parameters in the electrodeposition of Al-Mg alloys Electrodeposition conditions Deposit # Pre-Ed time (min) i (mA/cm2) T (oC) t (min)Substrate Anode composition 1 20 60 90 10 Cu Mg 2 80 60 90 10 Cu Mg 3 120 60 90 10 Cu Mg 4 180 60 90 10 Cu Mg 5 180 60 90 30 Cu Mg 6 60 60 60 15 Cu Mg 7 60 100 60 15 Cu Mg 8 60 150 60 15 Cu Mg 9 75 150 60 30 Cu 80 area%Al +20area%Mg Pre-Ed: Pre-electrodeposition, i current densit y, T: temperature, t:electrodeposition time

PAGE 108

108 CHAPTER 6 MORPHOLOGICAL AND MICROSTRUCTURAL EVOLUTION OF AL-MG DENDRITES Chapter 5 introduced several morphologies of the dendrites that were observed in the deposits produced in our work. It was shown that the morphology of the dendrites in the electrodeposited Al-Mg alloys depends on their Mg content which in turn depends on the different deposition parameters used; such as current density (or applied potential) and temperature. It has been reported in the literature that at low overpotentials (or current densities) faceted crystals [74, 78, 123] as well as epitaxial growth [124] takes place. When the deposition rates are low charge transfer mechanisms domina te the kinetics of deposit growth. Because of the availability of ample time for surface diffusi on, the atoms find the lowest energy state and specific crystallographic facets evol ve owing to the anisotropy in interfacial energy. In the case of the electrodeposition of Zn at low overpotentials it was found that deposits formed with faceted large crystals under activation-controll ed mechanisms [79].The development of such structure was attributed to the differences in the growth rates of different crystallographic planes. On the other hand, at high deposition rates, the kinetics becomes diffusi on controlled [10], which usually results in dendritic growth [76, 83, 125, 126]. For example, during the galvanostatic deposition of Au-Ag alloys, dendritic precursors [85] were observed in the initial stages of deposition under interfac ial energy controlled conditions (s ee Figure 2-13(A)). However, at higher deposition rates, where gr owth is diffusion controlled, cau liflower-like structures were found in Ni-Co system [127]. The globular shap e of these cauliflower-like structures was attributed to the spherical shape of the diffusi on layer near the interface. Recently, Jovic et al. [76] investigated the morphology of dendrites in the multicomponent polycrystalline Ni-Co system. Different morphologies, such as cauliflow er-like and fern-like, were observed in their

PAGE 109

109 studies. They have reported that the morphologies of the deposits in their studies depend on the composition of the powders. Although the morphology of the electrodeposited materials has been studied extensively, the microstructural studies have received much less attention. Fu rthermore, there were no studies on the microstructure and morphologies of the elect rodeposited Al-Mg alloys have been reported in the literature. In this chapte r the microstructure of the various dendrite morphologies in Al-Mg alloys, produced under the conditions of this stud y, and their evolution ar e reported as functions of their composition, and current density. 6.1 Morphological Analysis of Dendrites The various morphologies reported in Chapter 5 were shown to depend mainly on the Mg concentration of the electrolyte, current density and temperature. These parameters were shown to directly affect the amount of Mg in th e deposits. It was found th at disregard of preelectrodeposition time and current dens ity all dendrites formed at 90 oC had a globular morphology. However, deposits made at 60 oC, exhibited a variety of dendritic growth morphologies such as rod-like, feather-like and globular. In this section the conditions for the appearance of of these morphologies will be discussed. 6.1.1 Rod-Like Morphology The SEM image of the rod-like morphology is s hown in Figure 6-1. It can be observed that a bunch of the rod-like entities are growing t ogether.The rod-like morphology was observed in deposits fabricated at 60 oC as a minority morphology, except in the deposit made at the current density of 100 mA/cm2 and temperature of 60 oC with the pre-electrodeposition time of 60 min. This morphology possesses a nominal composition of about 13-20 at.% M g, which suggest that the rod-like morphology is composed of only FCC Al-rich solid solution. As will be shown later this conclusion is confirmed by TEM results.

PAGE 110

110 6.1.3 Feather-Like Morphology The feather-like morphology appeared in almost all deposits made at 60 oC and at current densities higher than 60 mA/cm2. Figure 6-2(A) shows a dendr ite containing feather-like morphology. A closer view of the feather-like morphology is presented in Figure 6-2(B), which shows that this morphology has th e appearance of a feather. When a dendrite consisted only of this morphology, the tip of the dendrite arms were found to be sharp. Detailed observation revealed that this morphology can be viewed as an assembly of small arrow-shaped structural units as shown in Figure 6-2(C). These structural units are assembled in such a way that the higher order arms are attached to the lower order ones in a hierarchical manner. The shape of the feather-like morphology dendrite and its primary arms are clearly defined in Figures 6-2(B) and 6-2(C), respectively.. However, the secondary arms, which are attach ed to the primary arms (see Figure 6-2(C)), are very small in size and di fficult to be recognized. Figure 6-3 shows SEM images of the feather-like dendrites taken in h ead-on position to the dend rite axis and primary arms, respectively. Three sets of primary arms in which each set contains two arms growing away from each other were identified. Each arm has a shape close to a cone. Based on the geometry of these arms, three planes, each comp rising of the dendrite and the arm axes, can be defined. The average angular relationships among these planes are shown in Figure 6-3. In the case of the image taken head-on to the dendrite ax is these angular relations hips suggest that the arrangement of these planes exhi bits a two-fold symmetry about the dendrite axis (see Figure 63(A)). However, it should be noted that while taking SEM images head-on to the dendrites, a perfect head-on position could not be achieved due to the instrument limitations involved in the rotation of the sample stage. Hence, the angles measured on the stems are not the true angles but are only approximate values.

PAGE 111

111 Figure 6-4 presents the SEM images of the feather-like morphology with the beam direction normal to planes defined by dendrite axis and the direction along which the primary arms are oriented. The dendrite axis is marked by an arrow. The average angles that the primary arms make with the dendrite axis are also deno ted. It can be observed that one set of primary arms is making an angle of about 47o and other set of primary arms is making an angle of about 58o with the dendrite axis. Figure 6-5 summarizes these angular relationships among different primary arms and the dendrite axis. These angular relationships suggest that the growth of dendrites and the primary arms are taking place in specific crystallographic directions. Hence Figure 6-5 suggests the crystallo graphic directions along which the dendrites and the primary arms are growing. 6.1.4 Globular Morphology The globular morphology was in major ity in the deposits made at 60 oC and at current densities of 100 and 150 mA/cm2. As mentioned previously, the deposits made at 90 oC contained only this kind of morphology. In order to investig ate the evolution of dendrites with the globular morphology, deposits were made for 1s, 1 min, 15 min and 30 min at the current density of 150 mA/cm2. Figure 6-6 presents SEM pictures of these dendrites. Afte r 1 s of deposition the dendrites appeared on the ground of the deposit (Figure 6-6(A)). Each de ndrite was found to be a cluster of several globules formed upon each other as shown in th e inset of Figure 6-6(A). This observation demonstrates that the globules tend to nucleate and grow on the existing globules rather than on the ground that is formed in the initial stage of the deposition. Also the gl obules that are forming in the latter stages tend to nucleate over the cu sps that are formed by the adjacent globules which were formed in the early stages of dendritic growth.

PAGE 112

112 At slightly longer deposition time, i.e. 1 min, the globules stack over each other as shown in Figure 6-6(B). When the deposition was conduc ted for 15 minutes, the clusters exhibited branching as can be seen in Figure 6-6(C). Each of these branches exhibited higher order arms, formed by clusters of globular morphology. These higher order arms can be seen to be grown only to a limited extent in comparison to the branch es to which they are at tached. The tip of the branches with coarser globules assumed an appr oximate paraboloid shape as shown in Figure 66(D). When the amount of the Mg in the electrolyte increases, nu cleation of the HCP Mg-rich phase occurs over these globular dendrites. This nucleation of the HCP Mg-rich phase manifested itself as globules with rougher texture. Figure 6-7( A) shows the dendrites containing globular morphology at two diffe rent roughness levels namely smooth and rough. The approximately paraboloid shaped tips on the dendrites containing globular morphology can be observed in Figure 6-7(B). As the growth continue d further, the tips of these dendrites coarsen eventually giving rise to the large globules. Similar to the feather-like morphology, the gl obular morphology also consisted of fine structural units. Figure 6-8 shows that both smooth (Figure 6-8(A), (B)) and rough (Figure 68(C), (D)) globules exhibit hierarch ical growth with nanometer sized globules as structural units. The smallest detectable unit had a size of about 100 nm. The fine size of these structural units suggests that the grains in the smooth and rough globules can be in nanometer range. 6.1.5 Morphological Evolution in Dendrites With the increase in the deposition time th e rod-like and feather-like morphologies eventually evolved into globular morphology as shown in Figure 6-9. With the progress of the deposition rough globules can be observed to be forming over the large smooth globules, as presented in previous section, in all these types of dendrites. Th e dendrites which started from

PAGE 113

113 feather-like and globular morphology were found in al l deposits except in t hose deposits with the lowest Mg content. The dendrites which star ted from rod-like morphology were found only in the deposits made at a current density of 100 mA/cm2 and temperature of 60 oC and they were in minority in those deposits. 6.2 Growth Mechanisms of Morphologies In the preceding sections it was demonstrated that feather-like a nd globular dendrite morphologies are constructed by assembly of build ing units, which are very different in nature. The faceted arrow-shaped units of the feathe r-like morphology and its geometrical structure suggest that interfacial energy ma y have a role in its formation. On the other hand the spherical shape of the globules implies diffu sion-controlled growth mechanism. In order to investigate the growth mechanisms of these dendrites, short tim e potentiostatic experiments were conducted at potentials within the range encountered in the galvanostatic depositions (see Figure 53 ). The current density versus time curves recorded during the experiments performed at applied potentials E =-4 V,-6 V, and -8 V for 1 min ar e shown in Figure 6-10(A). As expected, these current density versus time curves indicated that the deposition rate (cu rrent density) increases with the applied potential. Figures 6-10(B)-(D) pr esent the morphologies of the deposits made at E =-4 V, -6 V, and -8 V respectively. At lowe r applied potential, E =-4 V, only feather-like morphology appeared as can be seen in Figure 6-10(B). As the appl ied potential was increased to higher values, the globular morphology was found to be in majority as is demonstrated in Figures 6-10(C) and (D). Since the duration of the depositi on was the same in this set of experiments the amount of the deposits observed increased with in creasing the potential in proportion to the total charge In order to compare the deposits at equal total charge another set of potentiostatic experiments was conducted at two different applied potential valu es namely, E =-4 V and -8 V.

PAGE 114

114 In these experiments the deposition time was varied such that the total charge deposited was kept constant. Deposition at applied po tential of E =-4 V was done for 3 minutes and that at E =-8 V was done for 39 s. The recorded current density va lues matched with those in Figure 6-10(A) and for the present set of experiments are shown in Figure 6-11(A). The examination of deposits revealed that at applied potential of -4 V majority of the dendrit es were feather-like as shown in Figures 6-11(B). Increasing a pplied potential to -8 V rende red only globular morphology of dendrites as shown in Figure 6-11(C). Here, although the durations we re chosen with the expectation to obtain equal amount of deposit, in terestingly the deposit ma de at E =-4 V yielded lesser amount of the dendrites in comparison to the one at E =-8 V. This observation suggests different thickness to the grounds of these two deposits, the ground thickness being more in the case of deposit at E = -4 V. However, the gr ound was not characterized in the our study. The observation of less amount of dendritic growth su ggests that the formatio n of dendrites is more difficult at lower applied potential s. The enhanced coverage with the dendrites observed in the case of deposit made at E =-8 V (see Figure 6-11(C)) in comparison to that in deposit made at E =-4 V (see Figure 6-11(B)) also indi cates that at higher current dens ities recorded in the former case, the number of sites triggering the dendritic growth is more as opposed to that in the latter case. A comparison of Figure 6-10(B) and Figure 6-11(B) demonstrate that an increase in the deposition time from 1 min to 3 min at E =-4V did not change the morphology of the dendrites. This observation suggests that the applied potential of E =-4 V is below that needed for nucleation and growth of the globular morphol ogy. Similarly an increase in the time of deposition from 39 s (see Figure 6-11(C)) to 1 min (see Figure 6-11(C)) did not change the

PAGE 115

115 morphology of the dendrites. In other words the ra te of deposition plays considerable role in deciding the morphologies of the deposits. 6.3 Microstructural Analysis of Dendrites In this section the microstructural analysis of the dendrites with f eather-like and globular morphologies, which were the most observed mo rphologies, using TEM te chniques is discussed. The microstructure of the dendrites with r od-like morphology, which is in minority in the deposits, is also presented here but not discusse d in detail. 6.3.1 Rod-Like Morphology Figure 6-12 shows the microstr ucture of the rod-like mor phology. Figure 6-12(D) is the dark field (DF) image of the rod-like morphol ogy shown in Figure 6-12(B) (Bright Field, BF). The SAED pattern shown in Figure 6-12(C) was taken from the area represented in the DF image. It can be observed from Figures 6-12( B), (C) and (D) that the rod-like morphology consists of nanocrystalline grains. The SAED pattern confirms that the rod like morphology has a FCC crystal structure. This result is consiste nt with the nominal comp osition of the rod-like morphology measured using EDS (13-20 at.% Al) (s ee Section 6.1.2) suggesti ng that the rod-like morphology is made of Al-rich solid solution. Th e ring SAED pattern with consistent spot intensities indicates that there is no preferenti al orientation of the grains in the rod-like morphology. 6.3.2 Feather-Like Morphology Figure 6-13 shows the microstr ucture of the feather-like morphology. The SAED shows a ring pattern indicating that the feather-like morphology cons ists of nanocrystalline grains. Similar to the rod-like morphology the SAED pattern of the feathe r-like morphology suggests that it consists of FCC-Al phase As mentioned in Chapter 5, this morphology contains about 7 at.% Mg confirming the TEM observations. Howe ver, unlike the rod-like dendrites, this

PAGE 116

116 morphology exhibited some crystallo graphic texture as suggested by the SAED pattern presented in Figure 6-13(C), where clustering of the spot s on the {111} ring are distinguishable. Rotation calibration was carried out on the SAED pattern s corresponding to the magnification of the corresponding image. Upon rotation calibration the approximate growth direction of the dendrite and one set of the primary arms are seen to be along the direction of thes e cluster of diffraction spots as seen in Figures 6-12(B) and (C). This observation sugge sts that the {111} planes of the nanocrystals are parallel to each other along these orientations, whic h signify the growth axes in this morphology. Further work is needed to unders tand the crystallographic nature of the dendrite growth in the feather-like morphologies. The grain size of the feather-like morphology wa s measured from dark field (DF) images using about 150 grains. The gr ain size was evaluated by taking measurements in two perpendicular directions on a gr ain and averaging both the values Figure 6-14 shows the grain size distribution in the featherlike morphology. The average grai n size of this morphology is around 42 nm. 6.3.3 Globular Morphology Figure 6-15 presents the TEM analysis of th e microstructure of the fine globules. As shown in Figure 6-15(B) each spherical unit of th is morphology consists of many nanocrystalline grains. The SAED pattern presented in Figure 615(C) shows that the fi ne globules have FCC crystal structure suggesting that they are made of FCC Al solid solution. The diffraction pattern exhibited no preferential orientat ion of the grains in these fine globules. Figure 6-16 presents the grain size distribution in this morphology. The average grain size of the fine globules was measured to be around 32 nm. The microstructural features of the larg e globular morphology are shown in Figure 6-17. The TEM sample shown in Figure 6-17 was pr epared from a deposit made with following

PAGE 117

117 conditions: pre-electrodeposit ion time 60 min, current density during electrodeposition of 150 mA/cm2, temperature of 60 oC and duration of 15 min. A close obs ervation indicates that within the smooth globular morphology there are two grain size regions namely, a coarse-grained core (a in Figure 6-17(B) and (C)) e nveloped by a fine-grained shell (b in Figure 6-17(B) and (D)). The SAED patterns indicate that both the fine-gra ined shell and the coarse-grained core also possess FCC crystal structure sugges ting that they are made of Al solid solutions (see Figures 617(C) and (D)). The rough globules were found to have HCP crystal stru cture indicating that they consist of Mg solid solution (inset in Figure 6-17(E)). The HCP phase also exhibited nanocrystalline micstructure. As seen from the SAED patterns, similar to the fine globular morphology, the large globular morphology does not show any preferred crystallographic orientation of the grains. The microstructures of the coarse-grained regi on and the fine-grained shell are shown in Figure 6-18.and Figure 6-19, respectively. The coar se-grained region exhi bited SAED pattern close to that of a single crysta l indicating that the area in Figur e 6-18(A) consists of only a few grains. The grain size distribution in the fine-grained shell is pr esented in Figure 6-19(D). The average grain size of the fine-grained sh ell is around 16 nm. Figure 6-20 shows the microstructure of the HCP phase. The grain size distribution is shown in Figure 6-20(D). The average grain size of the HCP phase is around 36 nm. Figure 6-21 shows the boundary region betw een FCC and HCP regions (see Figure 621(B)) and that between fine and coarse-grained regions (see Figure 6-21(C )). It can be observed from Figure 6-21 that there is a sharp variation in the grain size between fine-grained FCC and HCP regions, as well as across fine-grain ed FCC and coarse-grained FCC boundary.

PAGE 118

118 6.4 Compositional Analysis The compositional analysis of the dendrites was carried out using Electron Probe Micro Analyzer (EPMA). For compositional analysis th e dendrites were embedded in a polymer along their length and polished as explained in Section 3.5.1. 6.4.1 Feather-like Morphology The composition of the feather-like morphology was observed to be al most constant at about 7 at.% Mg throughout its length as shown in Figure 6-22. The high amount of Al in this morphology is consistent with the presence of F CC phase (see Figure 6-13). This indicates that the FCC phase is a Al-rich solid solution with about 7 at.% Mg in it. 6.4.2 Globular Morphology The compositional analysis was done on th e dendrites containing globular morphology with the electron probe at random locations as shown in Figure 6-23. For the compositional analysis a deposit, made at a pre-electrodeposition tim e of 180 min, was chosen, the conditions of which are shown in Table 6-1. The analysis showed that the composition stayed almost constant from the root of the dendrite to some where in the middle of the dendrite and is around 20 at.% Mg. Beyond this region, the Mg content was much higher and near 70 at.% Mg. These compositions are shown in Figure 6-23(A) and (B ). In order to understand the compositional variation along the dendrites from deposit made at 180 min pre-electrodeposition time, scanning for the composition along a line throughout thei r length was performed. The composition of these dendrites remained constant at about 20 at.% Mg for most of their length. It can be observed from Figure 6-24 that ther e are contours on these dendrites giving rise to variations in the contrast. In the vicinity of these contour s the composition was observed to be between 40-60 at.% Mg. Proceeding further along the line, there was an increase in the composition to about 70 at.% Mg where it almost remained constant as seen in Figure 6-24. This increase in the amount

PAGE 119

119 of Mg was consistent with the nucleation of HCP solid solution (rough globules) over the FCC solid solution (smooth globules) as observed fr om the TEM analysis (see Figure 6-17). Hence the combined results of SAED and compositi on analyses indicate that FCC and HCP phases correspond to Al-rich and Mg-rich solid solu tions respectively. The observation of the intermediate values of composition of about 40-60 at. % Mg may be due to the overlap of these two phases in the interaction volume of the electron beam. The compositional variation in another deposit made at a current density of 150 mA/cm2 and temperature of 60oC and a pre-electrodeposition time of 60 min is shown in Figure 6-25(A) and (B). It can be observed from these figures that the composition of these dendrites was initially around 20 at.% Mg. However with th e duration of the deposition the composition dropped to about 5 at.% Mg unlike in the dendrites from the deposit made at preelectrodeposition time of 180 min. The compositi onal analysis was also done across the boundary of coarse-grained core and fine-grained shell as show n in Figure 6-25(C). Figure 6-25 shows that the composition of the coarse-grained core is about 5 at.% Mg which matches the results in Figure 6-25(A), (B). However the com position increased from coarse-grained core to the fine-grained-shell which is FCC Al-rich phase. However, this region could not be observed in the EPMA results of Figure 6-25 (A) and (B). This may be due to the reason that when the composition was analyzed in the vicinity of the FCC Al-rich and HCP Mg-rich phase region, while still being in the FCC Al-rich region, so me of the HCP Mg-rich region might have got sampled in the interaction volume of the beam. Hence, it is difficult to observe such variations in the EPMA. Although using of 150 mA/cm2 should increase the amount of Mg which gets deposited along the dendrites, a reverse trend is observed here. This may be due to the some irregular convections that might have occured in the electrolyte system.

PAGE 120

120 The compositional analysis across the fine-grained Al-rich shell and Mg-rich phase boundary suggested an increase in the amount of Mg (Figure 626(C)) which is in accordance with the observations from micr oprobe results (see Figure 6-24). 6.5 Discussion Three different morphologies were observed in the dendritic growth of the deposits namely rod-like, feather-like, and globul ar. The rod-like and featherlike morphologies consisted FCC Al-rich solid solution with nanocrystalline structure. The latter showed preferential orientation of the grains in specific crystallographic directio ns. The globular morphology consisted of FCC Alrich phase over which HCP Mg-rich phase nuc leated. Unlike the feather-like morphology this one did not show any pref erential orientation. In our study it was observed that in majority of deposits fabricate d, the dendrites form initially as feather-like or small globular mo rphologies. The investigation of the competition between these two morphologies re vealed that the feat her-like morphology appeared only at low values of applied potential a nd the globular morphology was observe d at higher applied potential levels (see Figures 6-10 and 6-11). The cha nge in the morphology w ith applied pontential indicates that to some extent morphology depe nds on the deposition rates also apart from the composition. For example, at low applied potentia ls the deposition rates are low as indicated by current density values (see Figures 6-10 and 6-11). This suggests that th e depositing ions have more time to diffuse and arrange themselves in a lower energy configuration. Increase in the applied potential from -4 V to -6 V or -8 V corresponded to parallel increase in the deposition rate (see Figure 6-10). The increased de position rate implies that the depositing ions do not have enough time to arrange themselves so as to lower their interfacial energy. Hence, in the case when E =-6 V or -8 V the deposits assumed a globular morphology showing no anisotropy. During galvanostatic el ectrodeposition, because of the substrate

PAGE 121

121 condition and local hydrodynamic state, the requ ired overpotential for nucleation may vary on the substrate surface. This variation will resu lt in the observation of both feather-like and globular morphologies in deposits. The feather-like and the globular morphologies exhibited a hierarchic al growth, i.e. the dendrites developed by formati on of structural units. TEM re sults (see Figure 6-13 and 15) revealed that the structural units for both mor phologies consist grain size in the nanoregime. The main difference between these tw o morphologies was the presence of strong texture in featherlike dendrites. The formation of nanocrystalline dendrites have b een reported in the literature [82, 86]. The overall growth of th e dendrite in relation with the microstructure was explained using a concept of oscillating local overpotential values in relation with the space charges around the growing morphology [87, 128, 129] ..A critical potential is need ed for any nucleation event. Nucleation of a new grain occurs when the local ove rpotential is above a certain critical value. In the present case, because of the high current dens ity values, large number of nuclei form due to which, there is a momentary depletion of cations in the vicinity of cathode. As a result of this depletion, the electric field near the cathode collapses momentar ily. During this time, diffusion field is set up around these nuclei making the pro cess diffusion controlled [130]. Eventually the local concentration of cations build up, during which time only growth of the existing nuclei takes place as the local overpotential is less th an the critical value for nucleation. When the concentration of the cations is bui lt up to such an extent, that the values of the lo cal overpotential are greater than the critical value for nucleati on, new nucleation events trigger. When the new nucleation events trigger depletion of cations takes place and this process continues. The shape of the structural units to form f eather-like dendrites differs from that for globular morphology. In the latter ca se the spherical shape is cont rolled by the shape of local

PAGE 122

122 diffusion layers while in the former it is a result of crystallographic anisotropy. The formation of the arms and their further branching in the feathe r-like dendrites can be at tributed to the local perturbations similar to dendritic growth during solidif ication [131]. In the case of the globular dendrites, the assembly of the spherical units requires the formation of new nucleation sites outside the semi-spherical domain established by a growing globule. It is well known that the process of nucleation is biased towards the areas possessing higher values of local overpotential which can contribute to at least part of the critical overpotential barrier. The cusps formed by the adjacent globules are one such site for nuclea tion. Upon further nucleation and growth processes new globules can be evolved at these locations. One such globule formed from the cusp formed by the existing globules can be observed in the inset of Figure 6-6(A). With the progress of deposition, the globules stack and gr ow away from the cathode as shown in Figure 6-6(B) due to the higher overpotentials existing in that dire ction. Upon further deposition these dendrites containing only globular morphol ogy evolve into a shape somewhat similar to dense branch morphology (DBM) [82, 132-135] as can be seen in Figure 6-6(C). However, each of the branches has arms grown only to a li mited extent (see Figure 6-6(D)). It was demonstrated that a ll dendrites eventua lly evolve into large globular morphology. As the dendrites containing both the globular and feather-like mo rphology grow their tips are sharpened which can be seen in Figures 6-2 and 66. The general paraboloid shape of the tips of dendrites containing globular morphology (see Figur e 6-6) and the sharp tips of the dendrites containing feather-like morphol ogy (Figure 6-2) suggest that these tips are experiencing increased flux of available metal ions in the lateral positions apart from their head-on positions, thus resulting in spherical di ffusion condition (radius of the tip of dendrites << boundary layer thickness) [136]. The onset of spherical diffusion leads to blunting followe d by coarsening of the

PAGE 123

123 tips of dendrites and eventual formation of sm ooth large globular morphology over these tips (see Figure 6-7). It can be obser ved from Figures 6-17 and 25 that the core of the large globular morphology contained about 5 at.% Mg. This low composition of re sulted in the large grain size in the core region of the Al-rich phase. Covering this coarse-grained re gion, there is a fine-grained sh ell with increased amount of Mg. From the Chapter 4 it is known that the c oncentration Mg in the electrolyte gradually increases and reaches a saturation at which point HCP Mg-rich phase starts to appear in the deposits. Since, prior to the formation of the HCP Mg-rich phase the composition of the electrolyte increases in Mg, the amount of Mg in the deposit also should increase. However, this is not the case here. With the applic ation of a current density of 150 mA/cm2 the amount of Mg observed in the FCC Al-rich phase is decreasing. Thus some factors must be hindering the Mg ions from reaching the cathode and their eventual deposition. The reason for this was attributed to the irregularities in convection resulting in low Mg and there by large grain size in large FCC Al-rich globules. 6.6 Summary and Conclusions Dendrites of the Al-Mg alloys exhibited vari ous morphologies namel y, rod-like, featherlike and globular. The rod-like morphology, whic h was observed at a current density of 100 mA/cm2 and temperature of 60 oC contained a nominal compos ition of about 13-20 at.% Mg. The feather-like and globular mor phologies appeared in almost all the deposits which were made at a current density of 150 mA/cm2 and a temperature of 60 oC. This morpholoy contained about 7 at.% Mg. The composition of the globular mor phologywas found to vary along its main growth axis. While the rod-like and feather-like mo rphologies consisted of only FCC Al-rich solid solution, the HCP Mg-rich solid solution was found in the globular morphology. The maximum solubilities of Mg in Al and Al in Mg were found to be approximately 20 at.% and 40 at.%,

PAGE 124

124 respectively. The transition from the FCC phase to the HCP phase was found to be abrupt as evidenced by sharp changes in composition and crys tal structure. All dendri tes studied here were composed of nanocrystalline grains. An investigation of the competition betw een feather-like and globular morphologies revealed that that the former develops at lower applied potential than the latter. Combined morphological and microstructural analysis sugge sts that the feather-l ike morphology grows in specific crystallographic directions. This is co nsistent with the appearance of feather-like morphology at lower deposition rates, suggesting that the depositing ions have enough time to rearrange themselves in orderto reduce the interfacial energy.

PAGE 125

125 20 m 20 m 20 m 20 m Figure 6-1. The SEM images of the rod-like morphology formed at current density of 100 mA/cm2 and temperature of 60 oC. 10 m B 2 m C 10 m A 20 m 10 m B 2 m C 10 m A 20 m Figure 6-2. The SEM images showing A) general feather-like morphology, B), and C) hierarchical behaviour exhi bited by feather-like morphology.

PAGE 126

126 ~51o66o57o2 m ~66o~63o57o1 m A B ~51o66o57o2 m ~66o~63o57o1 m A B Figure 6-3. The SEM images of feather-like mo rphology showing A) primary dendrite and B) secondary dendrite from head-on position. Figure 6-4. The SEM images of feather-like mo rphology showing the angles made by primary arms with the dendrite axis.

PAGE 127

127 2 m (100) (111) (111) 54.7o54.7o70.6o [011] [001][010] (100)45o45o [011] [101][110] (111) 60o60o [011] [110][101] (111) 60o60o 47o 5 m 58o 5 m 58o 5 m 6 6o5 7o2 m 5 7o 2 m 2 m 2 m (100) (111) (111) 54.7o54.7o70.6o (100) (111) (111) (111) (111) 54.7o54.7o70.6o [011] [001][010] (100)45o45o [011] [001][010] (100)45o45o [011] [101][110] (111) 60o60o [011] [101][110] (111) (111) 60o60o [011] [110][101] (111) 60o60o [011] [110][101] (111) (111) 60o60o 47o 5 m 47o 5 m 58o 5 m 58o 5 m 58o 5 m 58o 5 m 6 6o5 7o2 m 5 7o 6 6o5 7o2 m 5 7o 2 m 2 m Figure -6-5.Schematic showing the angular rela tionships and crystallographic directions along which the dendrite and primary arms of the feather-like morphology are growing.

PAGE 128

128 Figure 6-6. The SEM images showing the forma tion of dendrite cont aining globular morphology (images taken from different deposits) after A) 1 s (inset shows the globular morphology forming from the cusp formed by two adjacent globules), B) 1 min, C) 15min, and D) 30 min.(sharp tips of thes e dendrites can be observed here). 20 m rough smooth100 mA/cm2,60oC, 15min (A) 20 m rough smooth100 mA/cm2,60oC, 15min 20 m rough smooth100 mA/cm2,60oC, 15min (A) 4 m 100 m 25 m (B) 4 m 100 m 25 m 4 m 100 m 25 m (B) A B 20 m rough smooth100 mA/cm2,60oC, 15min (A) 20 m rough smooth100 mA/cm2,60oC, 15min 20 m rough smooth100 mA/cm2,60oC, 15min (A) 4 m 100 m 25 m (B) 4 m 100 m 25 m 4 m 100 m 25 m (B) A B Figure 6-7. The SEM images showing A) the HCP Mg-rich rough globules forming over the FCC Al-rich smooth ones, and B) coarseni ng of the smooth globules and eventual formation of HCP Mg-rich rough globules

PAGE 129

129 1 m 5 m A B 500 nm D C globular structural units 20 m 1 m 5 m A B 500 nm D C globular structural units 20 m Figure 6-8. The SEM images showing the hierar chy exhibited by A) smooth globules, and B) rough globules. Figure 6-9. The SEM images showing morphological evolution from A) rodlike and B) featherlike to large globules. Insets show the ma gnified versions of these evolutions.

PAGE 130

130 0 20 40 60 80 1 00 1 20 1 40 020406080Time (s)E = -8V T = 60oC, t = 60sCurrent density (mA/cm2)E = -6V E = -4V Time (s) 3 m 100 mE=-4V A 100 m 100 m 10 m 20 mE=-6VE=-8V B C D 0 20 40 60 80 1 00 1 20 1 40 020406080Time (s)E = -8V T = 60oC, t = 60sCurrent density (mA/cm2)E = -6V E = -4V Time (s) 3 m 100 mE=-4V A 100 m 100 m 10 m 20 mE=-6VE=-8V B C D Figure 6-10. Morphologies as a function of current density A) Current vs time values from the potentiostatic experiments. SEM images of deposits showing B) feather-like morphology at E =-4 V, C) fine globules at E =-6 V, and D) fine globules at E =-8 V (a) 100 m 10 m 100 m 5 m 0 20 40 60 80 100 120 140 060120180240 Time (s)Current Density (mA/cm2)E=-4V E=-8V E=-8V E=-4VT = 60oC, t = 60s B C A (a) 100 m 10 m 100 m 5 m 0 20 40 60 80 100 120 140 060120180240 Time (s)Current Density (mA/cm2)E=-4V E=-8V E=-8V E=-4VT = 60oC, t = 60s B C A Figure 6-11. Morphologies as a function of currnt density A) Current vs time values from the potentiostatic experiments conducted to obt ain same amount of charge. SEM images of deposits showing B) f eather-like morphology at E =4V, 3min, and C) fine globules at E =-8V, 39s

PAGE 131

131 Figure 6-12. Description of rod-like mor phology showing A) SEM image of rod-like morphology, B) the BF image of rod-like mo rphology, C) the SAED pattern showing FCC crystalline structure, and D) th e DF image of rod-like morphology.

PAGE 132

132 ` 10 m 1 m A B Approx. direction of primary and secondary arm growth {200} {111} {220} {311} 1 m D C 10 m 1 m A B Approx. direction of primary and secondary arm growth {200} {111} {220} {311} 1 m D C Figure 6-13. Feather-like morphol ogy showing A) the SEM image of feather-like morphology, B) the BF image of rod-like morphology, C) the SAED pattern (calibrated for rotation) showing FCC crystalline structur e and approximate growth direction of primary and one set of secondary arms of feather-like morphology and D) the DF image of rod-like morphology.

PAGE 133

133 0 2 4 6 8 10 12 14 16 182 .5 1 2 .5 2 2 .5 3 2.5 4 2 .5 5 2 .5 6 2 .5 7 2.5 8 2 .5 9 2 .5 102.5Grain Size (nm)Rel. Frequency (%) 200 nm 200 nm A B D C Avg. Grain size: 42 nm {200} {111} {220} {311} 0 2 4 6 8 10 12 14 16 182 .5 1 2 .5 2 2 .5 3 2.5 4 2 .5 5 2 .5 6 2 .5 7 2.5 8 2 .5 9 2 .5 102.5Grain Size (nm)Rel. Frequency (%) 200 nm 200 nm A B D C Avg. Grain size: 42 nm {200} {111} {220} {311} Figure 6-14. Grain size analysis of the feathe r-like morphology showing A) the BF, B) the DF image of feather-lik e morphology, C) the SAED pattern showing presence of FCC phase, and D) grain size distri bution in feather-like morphology.

PAGE 134

134 2 m 0.5 m A D B 0.5 m C {200} {111} {220} {311} 2 m 0.5 m A D B 0.5 m C {200} {111} {220} {311} Figure 6-15. Fine globular morphology: with A) the SEM image of fine globules, B) the BF image, C) the SAED pattern showing the presence of FCC phase, and D) the DF image.

PAGE 135

135 0 2 4 6 8 10 12 14 16 182.5 12.5 22 .5 32 .5 42.5 52.5 62 .5 72 .5 82.5 92.5 10 2 .5Grain Size (nm)Rel. Frequency (%) 500 nm 500 nm A B D C {200} {111} Avg. Grain size: 32 nm 0 2 4 6 8 10 12 14 16 182.5 12.5 22 .5 32 .5 42.5 52.5 62 .5 72 .5 82.5 92.5 10 2 .5Grain Size (nm)Rel. Frequency (%) 500 nm 500 nm A B D C {200} {111} Avg. Grain size: 32 nm Figure 6-16. Grain size of the fine globular morphology:showing A) the BF, B) the DF image of fine globule, C) the SAED pattern showi ng the presence of FCC phase, and D) grain size distribution in fi ne globular morphology

PAGE 136

136 Figure 6-17. Details of larg e globules.showing A) the SEM image of globular morphology showing rough globules (c) over smooth ones (b); B) the TEM image of a section of globular dendrite showing r ough and smooth globules; C) co arse grained FCC Al-rich solid solution with diffraction pattern. D) fine-grained FCC Al-rich solid solution with diffraction pattern, and E) the HCP Mg-rich solid solution with diffraction pattern. Figure 6-18. Large-grained region of large globules shown by A) the BF, B) the DF image of coarse-grained Al region, and C) the SAED pattern s howing the presence of FCC phase.

PAGE 137

137 0 5 10 15 20 25 30 35 407.512.517.522.527.532.537.542.522.5Grain Size (nm)Rel. Frequency (%) 200 nm 200 nm A B D C Avg. Grain size: 16 nm {200} {111} {220} {311} 0 5 10 15 20 25 30 35 407.512.517.522.527.532.537.542.522.5Grain Size (nm)Rel. Frequency (%) 200 nm 200 nm A B D C Avg. Grain size: 16 nm {200} {111} {220} {311} Figure 6-19. Grain size of fine-grained region of large globular morphology:A) the BF, B) the DF image of fine-grained Al region, and C) SAED pattern showing the presence of FCC phase, and D) grain size distribution.

PAGE 138

138 Figure 6-20.The HCP region of globules.: A) the BF, B) the DF image of HC P Mg-rich phase, C) the SAED pattern showing the pres ence of HCP phase, and D) grain size distribution in Mg rich phase.

PAGE 139

139 5 m 500 nm hcpregion fine grained fccregion 500 nm fine grained fccregion coarse grained fccregion A B C 5 m 500 nm hcpregion fine grained fccregion 500 nm fine grained fccregion coarse grained fccregion A B C Figure 6-21. Large globular morphology showing boundaries be tween different regions: A) the TEM image of dendrite formed by globular morphology, B) boundary between fine-grained FCC and HCP regions, and C) boundary between fine and coarsegrained regions 20 m6.47 77.44 77.3 4.8 9.1 7.4 20 m6.47 77.44 77.3 4.8 9.1 7.4 Figure 6-22. Compositional analysis along th e length of the feat her-like morphology

PAGE 140

140 Figure 6-23. Composition along dendrites contai ning globular morphology from deposit made under conditions: A), and B) pre-electrode position time 180 min, current density of 60 mA/cm2, and temperature of 90 oC

PAGE 141

141 Figure 6-24. Composition along a line in the de ndrites containing globular morphology from the deposit made under conditions: pre-electrode position time 180 min, current density of 60 mA/cm2, and temperature of 90 oC.

PAGE 142

142 A 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 (B) (A) 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 B A 0 5 1 0 1 5 0 1 0 0 2 0 0 3 0 0 D i s t a n c e ( n m ) M g ( a t % ) 0.5 m C 1 0 0 nm. . 0 5 1 0 1 5 0 1 0 0 2 0 0 3 0 0 D i s t a n c e ( n m ) M g ( a t % ) a b b a E D A 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 (B) (A) 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m23.7 19.4 15.2 13.5 6.0 1.2 3.1 67.5 69.5 16.5 3.2 5.0 62.6 65.3 70.3 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 20 m19.6 20.9 19.6 8.8 11.4 8.2 71.4 75.5 20.0 19.8 18.6 7.9 4.2 49.9 B A 0 5 1 0 1 5 0 1 0 0 2 0 0 3 0 0 D i s t a n c e ( n m ) M g ( a t % ) 0.5 m C 1 0 0 nm. . 0 5 1 0 1 5 0 1 0 0 2 0 0 3 0 0 D i s t a n c e ( n m ) M g ( a t % ) a b b a E D Figure 6-25. Compositional analysis in larg e globular morphology: A), composition along dendrites B) compositions along a dendrit e containing globular morphology made at pre-electrodeposition time of 60 min, current density of 150 mA/cm2, and temperature of 60 oC, and C) the TEM image of boundary between coarse-grained FCC Al-rich phase (a) and fine-grain ed FCC Al-rich phase (b). D) High magnification image of C. and E) Co mposition variation across a and b.

PAGE 143

143 Figure 6-26 Compositional analysis between FCC and HCP regions A) boundary between finegrained FCC Al-rich phase (b) and HCP Mg -rich phase,(c), B) high magnification image of A, and C) compositi on (Mg at.%) across b and c.

PAGE 144

144 Table 6-1. Experimental conditions of the samples used for compositional analysis Reference Preelectrodeposition time (min) Current density during electrodeposition (mA/cm2) Temperature (oC) Time (min) Figures 622(A),(B) 180 60 90 30 Figures 622(A),(B) 60 150 60 15

PAGE 145

145 CHAPTER 7 EFFECT OF SUBSTRATE ON NU CLEATION OF HCP MG PHASE It was shown in Chapters 5 and 6 that th e Al-Mg alloys always nucleate as the FCC Al-rich phase either in ground or dendrites on the copper substrate. Eventually, HCP Mgrich phase is formed over this FCC Al-ri ch phase. The amount of the FCC-phase preceding the HCP-phase depends on the concentration of Mg in the electrolyte. However, even at high concentrations of Mg in the electrolyte the nucleating phase was observed to be only FCC-phase. These obser vations suggest that perhaps the energy barrier for nucleation of the HCP-phase on the FCC-copper is quite high. Therefore, experiments were designed to investigate whether an HCP substrate such as pure Mg would reduce this energy barri er and allow the nucleation of the HCP-Mg rich phase. Understanding the effect of substrate can have significant impact where it is desirable to coat a surface with the HCP-Mg phase. Al so, for hydrogen storage applications, where the aim is to fabricate the HCP-Mg rich phase for producing magnesium hydride, the presence of the FCC-Al rich may not be de sirable. Therefore, the present chapter is dedicated to the understanding th e effect of substrate on th e nucleation of the HCP-Mg phase. 7.1 Effect of Electrolyte Composition on the Nucleation of the HCP Phase As shown in Chapter 4, during pre-electrode position, the electrolyte becomes richer in Mg in stage I before reaching a steady st ate in stage II (see Figure 4-3). At the same time the concentration of Mg in the electrodeposit during the stage I was maintained low during stage I before increasing rapidly in stage II (see Figure 42). The rapid increase in the Mg composition during stage II was acco mpanied with an increase in the HCP Mgrich phase in the deposits. These observations suggest that the HCP Mg-rich phase can

PAGE 146

146 nucleate on the FCC-Al rich phase when the c oncentration of Mg in the electrolyte is high enough. Figure 7-1 shows a dendrite from a deposit made for 30 minutes on a fresh Cu cathode after pre-electr odeposition for 3 hours. The EPMA compositional map revealed the presence of an initial Al-rich region (<25% Mg) at the root of the dendrite followed by the Mg-rich dendrite (>70%Mg). This behavior was typical of all dendrites investigated. This observation reveals that, when a new c opper cathode is used the FCC Al-rich phase nucleates, suggesti ng that the concentration of th e Mg in the electrolyte is not the only requirement for the nucle ation of the HCP-Mg rich phase. 7.2 Effect of Substrate on Nucleation of HCP Phase In order to investigat e the effect of substrate on the nucleation of the HCP-Mg rich phase short term deposits were made on thr ee different substrates namely, Cu, Mg, and amorphous graphite, using an electrolyte wi th a composition in stage II, where it was shown that the HCP-Mg rich phase can nucle ate and grow on the F CC-Al rich phase. As far as the effect of substrate is concerned, ma ny factors including crys tal structure, lattice parameter, surface and interfacial energies elastic modulus, presence of oxides and surface preparation are expected to infl uence the nucleation energy barrier. The Cu and Mg substrates were elec tropolished and graphite substrate was machined before using them for deposition, as mentioned in Section 3.1. For all the deposition experiments reported in this chap ter the pre-electrode position was done for 3 hrs using Cu substrate at a temperature of 90 oC and current density of 60 mA/cm2. Under these conditions, the HCP-Mg phase nucleated and grew on the FCC-Al rich phase in the steady-state regime (stage II). After the pre-electrodeposition, the Cu cathode was removed and a fresh cathode of either Cu, Mg or graphite was used for electrodeposition. The electrodeposition was performed at a current density of 150 mA/cm2 and temperature

PAGE 147

147 of 90 oC for 1 min. After deposition the deposits were cleaned using the same procedure as explained in Section 3.5.1. Figure 7-2 presents the SEM images of the deposits produced on different substrates with their nominal composition af ter pre-electrodeposit ion as evaluated by EDS. The nominal compositions suggest that th e ground on all three substrates is rich in Al, which indicates the formation of FCC Al-ri ch solid solution (see Figures 7-2(A)-(C)). It can be observed from Figure 7-2(B) that there are some large portions of the Mg substrate where no deposit was formed. In the case of graphite substrate (see Figure 72(C)) the grinding marks left by the SiC paper can be detected. The preferential formation of the dendrites along these ma rks is apparent from Figure 7-2(C). It was anticipated that using a Mg substr ate with an HCP crystal structure would aid in nucleation of HCP Mg-ri ch phase. On the contrary to the expectation, the nominal composition of the ground from the deposit made using Mg substrate showed Al to be in majority. One possible reason for this could be that since Mg is very active, as soon as it is polished and rinsed with toluene an oxi de film might have formed on the surface before its transfer to the glove box. Hence, in the case of the Mg substrate the substrate preparation had to be improved to eliminate or avoid the formation of the oxide film. In order to remove the oxide film on th e Mg substrate, it was ground with 800 grit paper in the glove box before using it for elec trodeposition. The grinding of the substrate with the grit paper was believed to remove the oxide layer and reveal the oxide-free surface. To compare the results of deposit on the oxide-free Mg substrate, another deposit was made on a Cu substrate which was al so ground as mentioned above. The SEM images of these two deposits with their nom inal composition are presented in Figure 7-3.

PAGE 148

148 As can be seen from Figure 7-3 the nomi nal composition of the deposit made using oxide-free Cu substrate is less than 20 at.% Mg. These results s uggest that using the oxide-free Cu substrate still resulted in th e formation of only FCC Al-rich solid solution (see Figure 7-3(A)). The no minal composition of the depos it made on the oxide-free Mg substrate however was much higher (>50 at.% Mg) as shown in Figure 7-3(B). However, in the case of the Mg substrate it was diffi cult to conclude whether HCP Mg-rich phase is formed based on the nominal compositions reported here. The reason for this uncertainty is that when the electron beam in EDS s canned an area of interest for composition, it might have sampled the substrate also. In other words the interaction volume of the electron beam used in EDS might have had c ontributions from both the Mg substrate and the Mg in the deposit. This double source of Mg might have resulted in the higher values for compositions of Mg than th e actual values in the deposit made on the Mg substrate. To avoid this ambiguity in measuring the composition of the deposit made on the Mg substrate, the deposit was scraped caref ully from the substrate and mounted on a graphite stub. Composition was measured on the deposit that was mounted on the graphite stub, the results of which are show n in Figure 7-4. As shown in Figure 7-4 the ground from the deposit made on the Mg subs trate showed a nominal composition of more than 50 at. % Mg suggesting the nuclea tion of HCP Mg-rich phase. The presence of HCP Mg-rich phase was confirmed by XRD. The XRD profile of this deposit is shown in Figure 7-5. Since the deposition was done for 1 min, only a small amount of powder was produced. This is the reason for the relative p eak heights in the XRD profile to be only slightly higher than that of the background. The combined results of nominal composition

PAGE 149

149 and XRD analysis confirm the formation of the HCP Mg-rich phase on the oxide free Mg substrate under present conditions. 7.3 Discussion The results indicate that the nucleation of the HCP-Mg rich phase depends on the substrate. FCC Al-rich solid solution nucleat ed on Cu, graphite and on oxide-containing Mg substrate. The nucleation of Mg was po ssible only on the oxide free Mg substrate. The selectivity of nucleation of different phase s towards different subs trates indicates that the nucleation barrier has a role to play. Nucleation barrier can be related to the classical capillarity effect of the nucle ating phase on the substrate [137] According to this theory, the nucleating phase is assumed to have the shape of a spherical cap as shown in Figure 7-6, where is the interfacial energy, M is the nu cleating phase, sub is the substrate. The contact angle depends on the mechanical balance of the interfacial energies of substrate-electrolyte, substrate-nucleating phase and electrolyte-nucleating phases as shown in equation (7-1). cos s ubelecMelecMsub (7-1) The energy barrier for nucleation is given by, 3 216 3Melec VGf G (7-2) where 323coscos 4 f (7-3) here GV is the chemical free energy change per unit volume of nuclei formation and f( ) is the wetting factor which varies from 0-1 for = 0-180o. = 0 indicates that there is no barrier for nucleation. The GV depends on the adsorption of depositing

PAGE 150

150 material on the substrate. Different substrates offer different number of adsorption sites. For example, metallic substrates offer more number of adsorption sites than covalent substrates due to the absence of the directi onality and covalent na ture of the bonds [138]. This is the reason for finding large areas on the Mg substrate covered with oxide where no deposit was formed (see Figure 7-2(B)). Also in the case of the graphite substrate (see Figure 7-2(C)) the grinding marks caused defect s such as kinks, steps etc. on the surface of the substrate. Since these defects are high energy regions, th ey provided for part of the energy needed for nucleation. Hence, larger fraction of dendrites formed along side of these defects in the case of depos it on the graphite substrate. According to equation (7-1) the wetting factor depends on different interfacial energies. The contact angle and thus the nuc leation barrier would be lowered when the interfacial energy between s ubstrate and electrolyte ( sub-elec) is increased or that between depositing metal and substrate ( M-sub) is lowered. One of the factors that affects the interf acial energy between the substrate and the depositing metal is the interatomic distan ce mismatch. The lower this interatomic distance mismatch, the lower is the inte rfacial energy between substrate and the depositing metal and higher is its probability of nucleation on the substrate. Since the close packed planes are low energy planes in crystals, they offer maximum nucleation barriers for the depositing phases. Hence, fo r our study the mismatch between the close packed atomic positions is considered as an estimate for the nucleation barrier. For the FCC and HCP systems mismatch between {111}<110> and {0001}<11 20> directions were considered.

PAGE 151

151 Lattice parameter for the case of FCC Al-rich phase was calculated using Maximum Likelihood Estimate (MLE) method. assuming the peaks are Pearson VII functions [139]. The value was found to be 4.075.002 In the case of HCP Mg-rich phase the lattice parameter could not be cal culated from MLE method as this method is not applicable to HCP systems. There have been many instances in the literature where the molar volume [140] and lattice parameters for the Mg-Al system have been calculated [141-143]. For our study the data by Hardie and Perkins [141] is used. Hardie and Perkins prepared the Mg-Al alloys in different co mpositions ranging from 0.96 to 6.89 at.% Al. Since the solubility of Al in Mg is very less the alloys were quenched to retain these compositions thereby avoiding formation of any stable phases. The variation of lattice parameter ( a ) as a function of concentration of Al (taken from Hardie and Perkins [141]) is plotted in Figure 7-7. As can be seen from Figure 7-7 the lattice parameter follows a linear trend with the concentration of Al. Hence, this lattice parameter was extrapolated to find its value at 30 at.% Al which is about the typi cal concentration found in our study. The extrapolated value of the lattice parameter is 3.092 This value is used for the HCP Mg-rich phase to calcul ate the lattice mismatch with different substrates. For Cu and Mg substrates the lattice parameter ( a in the case of Mg) values were taken from the JCPDS files. Th e lattice parameter of Cu is 3.615 [144] and that of Mg is 3.209 [145] respectively. In case of oxi de-containing Mg substrate the oxide was assumed to be MgO. The lattice parameter of MgO was taken from Smith et al. [146] and it is 4.212 Using the above mentioned lattice parameter values mismatch was calculated between depositing phases and di fferent substrates considered. Figure 7-8 shows these mismatch values. Figure 7-8 s hows that the FCC Al-rich phase has lower

PAGE 152

152 mismatch in the case of Cu and oxide-containing Mg substrates. However, in the case of the Mg substrate the HCP Mg-rich phase has the lower mismatch. These mismatches indicate that there is higher probability of nucleation of HCP Mg-rich phase only on the Mg substrate. In all other cases there is hi gher probability of nucleation of FCC Al-rich phase. As can be seen from the present analys is the nucleation of different phases is in accordance with the observed trends in the lattice mismatch. 7.4 Summary The interfacial energies play role in deciding the nucleating phase on any given substrate. The lower the interfacial ener gy between a given substrate and nucleating phase, the higher is the probability of that phase nucleating on that given substrate. One of the main factors that deci de the nucleation of phases is crystal structure. HCP Mg-rich phase nucleated on preferably on the oxide-free mg substrate as their crystal structures are same. In all other cases nucleation of FCC Al-rich phase was preferred due to matching crystal structures. One other factor that can decide the nucleation barriers is the interatomic distance mismatch. An estimate was given in our study for the interatomic distance mismatch between nucleating phases an d various substrates. According to this estimate the nucleation of HCP Mg-rich pha se is favoured only on the oxide free Mg.

PAGE 153

153 2 1 372.6 85.4 82.0 75.98 8 077.4 20 m 2 1 372.6 85.4 82.0 75.98 8 077.4 20 m Figure 7-1. Composition along a dendrite from a deposit made at 180 min of preelectrodeposition containing FCC Al-rich phase at the root and HCP Mg-rich phase on Al-rich phase. Figure 7-2. The SEM images of deposits on different substrates: A) deposit on Cu substrate showing Al in majority, B) deposit on oxide-containing Mg substrate showing Al in majority. Large areas where no deposit formed can be seen, and C) deposit on amorphous graphite substrate showing Al in majority large number of dendrites formed along the grind marks can be seen (all compositions are measured as at. % Mg)

PAGE 154

154 Figure 7-3. The SEM images of A) deposit on Cu substrate (ground with 800 SiC paper) showing Al in majority, and B) depos it on Mg substrate (ground with 800 SiC paper) showing Mg in majority. Figure 7-4. The SEM images of deposit made on oxide-free Mg substrate mounted on graphite stub. The composition values suggest that HCP Mg-rich phase is nucleated on this substrate.

PAGE 155

155 304050607080 2 (degrees)Intensity(arb. units) * Al (+Mg) Mg (+Al) 304050607080 2 (degrees)Intensity(arb. units) * Al (+Mg) Mg (+Al) Figure 7-5. The XRD profile from deposit ma de on oxide-free subs trate confirming the formation of HCP Mg-rich phase. Figure 7-6. Schematic showing different interfaci al energies at a hemispherical nucleus.

PAGE 156

156 3.040 3.060 3.080 3.100 3.120 3.140 3.160 3.180 3.200 3.220 01020304050 Hardie&Perkins Linear ExtrapolationAl (at.%)Lattice Parameter (Ao) 3.040 3.060 3.080 3.100 3.120 3.140 3.160 3.180 3.200 3.220 01020304050 Hardie&Perkins Linear ExtrapolationAl (at.%)Lattice Parameter (Ao) Figure 7-7. Lattice parameter of HCP Mg-rich phase as a function of at.% Al. From 0.96 at.% Al to 6.89 at.% Al data taken from Hardie and Perkins [142]. From 6.89 at.% Al to 40 at.%, the lattice para meter values are extrapolated ones. Oxide-free Mg Cu Mg with oxide layer Lattice Mismatch (%)Substrate Oxide-free Mg Cu Mg with oxide layer Lattice Mismatch (%)Substrate Figure 7-8.Interatomic distance mismatch be tween nucleating materials (FCC Al-rich and HCP Mg-rich phases) and different subs trates. Only oxide-free Mg substrate is favourable for nucleation of HCP Mg-ri ch phase. On all other substrates the feasibility of nucleation of FCC Al-rich phase is higher due to lower lattice mismatch values.

PAGE 157

157 CHAPTER 8 CONCLUSIONS Nanocrystalline Al-Mg alloys were fabr icated in the form of powders using electrodeposition technique in a ga lvanostatic mode. The electrolyt e used for our study had the following composition Na[AlEt4] + 2 Na[Et3Al-H-AlEt3] + 2.5 AlEt3 + 3.3 Toluene (where Et=C2H5). Pure Mg (99.99%) was used as anode. Ma gnesium was introduced into the electrolyte using a process which was termed pre-elect rodeposition. Pre-electrodeposition was conducted at a current density of 60 mA/cm2 and at a temperature of 90 oC for different times. The effect of pre-electrodeposition time was st udied using 60 mA/cm2 at the electrolyte temperature of 90 oC. The Al-Mg alloy powders were fa bricated at 60, 100 and 150 mA/cm2 current densities at constant pre-electrodeposition time and electrolyte temperature of 60 oC. Nucleation of different phases was investigated by conducting electrodepos ition on various substrates, namely, Cu, Mg and graphite. The results of this st udy led to the following conclusions. A small concentration of Mg in the electrolyte is sufficien t for obtaining large amount of Mg in the deposits. The amount of Mg in the depo sit increases rapidly wh en the Mg/Al ratio in the electrolyte reaches a critical value. The formation of HCP Mg-rich solid solution in the deposits was found to be associated with the rapi d increase in the amount of Mg in the deposits at this critical value of Mg/Al ratio. The presence of Na+ ions in the electrolyte is necessa ry for the dissociation of the AlEt3 and consequent deposition of Al. When the electrolyte contains Na[Et3Al-H-AlEt3] complex, the breakage of the Al-H-Al bonds resu lts in the formation of MgH2 on the anode surface. Increasing the current density, concentration of Mg in the electrolyte or the temperature resulted in a higher concentration of Mg in the deposits. Since electrodeposition was conduc ted at conditions far

PAGE 158

158 from equilibrium, only supersaturated FCC Al-ri ch and HCP Mg-rich phases were found in the deposits studied here. The deposits made under the conditions used in this study consist of a ground, upon which dendritic growth took place. The dendritic mor phologies observed can be devided into three categories namely rod-like, feather-like, and globular. Short time poten tiostatic depositions indicated that the feather-like morphology formed at lower deposition rates whereas, the globular morphology was observed at hi gher deposition rates. All dendrite morphologies observed initially c onsist of nanocrystalline FCC Al-rich solid solution. The feather-like morphology, which exhi bited heavily testured grains showed pronounced crystallographic orient ation of arms. These arms form along the <110> and <100> directions because of higher growth velocities along these directions. In a galvanostatic deposition th ere is a morphological evolutio n from feather-like or fine globules to large globular morphology. This mor phological evolution is attributed to the enhanced lateral flux of the depositing ions near the sharp tips of the de ndrites. This enhanced lateral flux of the depositing ions is believed to have triggered a spherical diffusion condition. The large globular morphology exhibited two roughness levels, where the rough globules always grew over the smooth ones. The combined TEM and compositional analysis revealed that the smooth globules consisted of FCC Al-rich so lid solution and the rough ones consisted of HCP Mg-rich solution. The nucleation of the phases was found to be substrate dependent. The HCP Mg-rich phase was found to have a high nucleation barrier on coppe r, graphite of passivated Mg surfaces and it nucleates only on the oxide-free Mg substrate.

PAGE 159

159 LIST OF REFERENCES [1] Gross KJ, Chartouni D, Leroy E, Zuttel A, Schlapbach L, J. Alloys Compd., 1998;269:259. [2] Dornheim M, Klassen T, Bormann R, Jpn. Soc. Promotion of Sci., 2004;18:1. [3] McNicol BD, Rand DAJ, Williams KR, J. Power Sources, 2001;100:47. [4] Schlapbach L, Mater. Res. Soc. Proc., 2002:675. [5] Irani RS, Mater. Res. Soc. Proc., 2002;27:680. [6] Wolf J, Mater. Res. Soc. Proc., 2002;27:684. [7] Bowmann RC, Fultz B, Mater. Res. Soc. Proc., 2002;27:688. [8] Zhou L, Renewable and Su stainable Energy Rev., 2005;9:395. [9] Song Y, Guo ZX, Yang R, Phys. Rev. B., 2004;69:094205. [10] Zaluska A, Zaluski L, Strom-Olsen JO, Appl. Phys. A., 2001;72:157. [11] Shang X, Bououdina M, G uo ZX, J. Alloys Compd., 2003;349:217. [12] Fichtner M, Fuhr O, Kirche r O, J. Alloys Compd., 2003;356-357:418. [13] Bouaricha S, Dodelet JP, Guay D, Huot J, Boily S, Schultz R, J. Alloys Compd, 2000;297:282. [14] Fossdal A, Brinks HW, Fichtner M, Hauback BC, J. A lloys Compd., 2005;387:47. [15] Arroyo y de Dompablo ME, Ce der G, J. Alloys Compd., 2004;364:6. [16] Fichtner M, Fuhr O, J. Alloys Compd., 2002;345:286. [17] Ashby EC, Kobetz P, Inorg. Chem., 1966;5:1615. [18] Luo W, Gross KJ, J. Alloys Compd., 2004;385:224. [19] Suryanarayana C, Int. Mater. Rev., 1995;40:41. [20] Gleiter H, Prog. Mater. Sci., 1989;33:223. [21] Granqvist CG, Buhrman RA, J. Appl. Phys., 1976;47:2200. [22] Zaluska A, Zaluski L, Strom-Ol sen JO, J. Alloys Compd., 1999;288:217. [23] Zaluski L, Zaluska A, Tessier P, SrromOlsen JO, Schulz R, J. Mat. Sci., 1996;31:695. [24] Suryanarayana C, Prog. Mater. Sci., 2001;46:1.

PAGE 160

160 [25] Nieman GW, Weertman JR, Si egel RW, Scripta Mater., 1990;24:145. [26] Ebrahimi F, Bourne GR, Kelly MS E. MT, Nanostruct. Mater., 1999;11:343. [27] Ebrahimi F, Ahmed Z, Morgan KL, Mater. Res. Soc. Proc., 2001;634:B2.7.1. [28] Lu L, Wang LB, Ding BZ, Lu K, J. Mater. Res., 2000;15:270. [29] Karimpoor AA, Erb U, Aust KT, Palumbo G, Scripta Mater., 2003;49:651. [30] Natter H, Schmelzer M, Hempelmann R, J. Mater. Res., 1998;13:1186. [31] Czerwinski F, J. Electrochem. Soc., 1996;143:3327. [32] Cheung C, Djuanda F, Erb U, Nanostruct. Mater., 1995;5:513. [33] Sehayek T, Bendikov T, Vaskevich A, R ubinstein I, Adv. Funct. Mater., 2006;16:693. [34] Fratesi R, Roventi G, Giuliani G, To machuk CR, J. Appl. Electrochem., 1997;27:1088. [35] Li H, Ebrahimi F, Mater. Sci. Eng. A, 2003;347:93. [36] Bdikin IK, Strukova GK, Zver kov SA, Acta Crystallogr., 2002;58:C25. [37] Tochitskii TA, Jones GA, Blythe HJ, Fedosyuk VM, Castro J, J. Magn. Magn. Mater., 2001;224:221. [38] Mayer A. US Patent, 4,778,575. 1988. [39] Mayer A, J. Electrochem. Soc., 1990;137:2806. [40] Lehmkuhl H, Mehler K, Reinhold B, Bonga rd H, Tesche B, Materwiss. Werkstofftech, 2000;31:889. [41] Lehmkuhl H, Mehler K, Reinhold B, Bonga rd H, Tesche B, Adv. Eng. Mater., 2001;3:412. [42] Mayer A, J. Electrochem. Soc., 1990;137:2806. [43] Imamura H, Masanari K, Kusuhara M, Kats umoto H, Sumi T, Sakata Y, J. Alloys Compd., 2005;386:211. [44] Grochala W, Edwards PP, Chem. Rev., 2004;104:1283. [45] Leng HY, Ichikawa T, Isobe S, Hino S, Hanada N, Fuji H, J. Alloys Compd., 2005;404406:443. [46] Gremaud R, Borgschulte A, Chacon C, Van Mechelen JIM, Schreuders H, Zuttel A, Hjorvasson B, Dam B, Greisse n R, Appl. Phys. A., 2006;84:77.

PAGE 161

161 [47] Gremaud R, Borgschulte A, Lohstroh W, Schreuders H, Zu ttel A, Dam B, Griessen R, J. Alloys Compd., 2005;404:775. [48] Dong HW, Ouyang LZ, Sun T, Z hu M, J. Rare Earths, 2008;26:303. [49] Brenner A. Electrodepositi on of alloys: Principles and practice. New York: Academic Press, 1963:42. [50] Lehmkuhl H, Mehler K, Landau U., The principles and techniques of electrolytic aluminum deposition and disso lution in organoaluminum elect rolytes. In: Gerischer H, Tobias C, editors. Advances in electroche mical science and engineering, 3. VCH, 2001. p.163. [51] Zhao Y, VanderNoot TJ, Electrochim. Acta, 1997;42:3. [52] Great MWM, J. Electrochem. Soc., 1985;132:1308. [53] Couch DE, Brenner A, J. Electrochem. Soc., 1952;99:234. [54] Brenner A, Couch DE. U.S. Patent 2,651,608, 1953. [55] Ishibashi N, Yoshio M, Electrochim. Acta, 1972;17:1343. [56] Yoshio M, Ishibashi N, J. Appl. Electrochem., 1973;3:321. [57] Ziegler K, Lehmkuhl H, Angew. Chem., 1955;67:424. [58] Ziegler K, Lehmkuhl H, Z.Anorg.Allg.Chem., 1956;283:414. [59] Liebenow C, Yang Z, Lobitz P, Electrochem. Commun., 2000;2:641. [60] Viestfrid Y, Levi MD, Gofer Y, Aurbach D, J. Electroanal. Chem., 2005;576:183. [61] Garst JF, Soriaga MP, Coord. Chem. Rev., 2004;248:623. [62] Schlenk W, Schl enk W, Ber., 1929;62:920. [63] Schlenk W, Ber., 1931;64:735. [64] Dessy R, Handler GS, Wotiz JH, Hol lingsworth CA, J. Am. Chem. Soc., 1957;79:3476. [65] Dessy R, Handler GS, J. Am. Chem. Soc., 1958;80:5824. [66] Cope A, J. Am. Chem. Soc., 1935;57:2238. [67] Noller CR, White WR, J. Am. Chem. Soc., 1937;59:1354. [68] Ziegler K, Koster R, Lehmkuhl H, Reinert K, Liebigs.Ann.Chem., 1960;629:33.

PAGE 162

162 [69] Rundle RE, J. Am. Chem. Soc., 1947;69:1327. [70] Astruc D. Organometallic chemistr y and catalysis. Berlin: Springer, 2007:324. [71] Lewis P, Rundle R, J. Chem. Phys., 1953;21:986. [72] Atwood JL, Hrncir DC, Rogers RD, Ho ward JAK, J. Am. Chem. Soc., 1981;103:6787. [73] Medley JH, Fronczek FR, Ahmad N, Day MC, Rogers RD, Kerr CR, Atwood JL, J.Crystallogr. Spectrosc. Res., 1985;15:99. [74] Siegfried MJ, Choi KS, Angew. Chem., 2005;44:3218. [75] Wang XF, Xu JJ, Chen HY, J. Phys. Chem. C, 2008;112:7151. [76] Jovic VD, Jovic BM, Pavlovic MG, Ma ksimovic V, J. Solid State Electrochem., 2006;10:959. [77] Wang ZL, J. Phys. Chem., 2000;104B:1153. [78] Siegfried MJ, Choi KS, Angew. Chem., 2008;47:368. [79] Lopez CM, Choi KS, Langmuir, 2006;22:10625. [80] Fang J, You H, Zhu C, Kong P, Shi M, Song X, Ding B, Chem. Phys. Lett., 2007;439:204. [81] Pal BN, Chakravorty D, J.Phys. Chem. B, 2006;110:20917. [82] Liu J, Fu Y, Guo A, Wang C, Hua ng R, Zhang X, J. P hys. Chem. C, 2008;112:4242. [83] Jovic VD, Maksimovic V, Pavlovic MG Popov KI, J. Solid State Electrochem., 2006;10:373. [84] Nakanishi S, Fukami K, Sakai S, Yoshihiro N, Chem. Lett., 2002;31:636. [85] Bozzini B, De Gaudenzi GP J. Electroanal. Chem., 2004;563:133. [86] Fleury V, Nature, 1997;390:145. [87] Rosso M, Electrochim. Acta, 2007;53:250. [88] Chazalviel JN, Phys. Rev. A, 1990;42:7355. [89] Murray JL, Bull. Alloy Phase Diagrams, 1982;3:60. [90] Walsh FC, Herron MJ, J. Phys. D, 1991;24:217. [91] Iwakura C, Nohara S, Inoue H, Fukumoto Y, Chem. Commun., 1996:1831.

PAGE 163

163 [92] Imamura H, Takesue Y, Akimoto T, Ta bata S, Journal of Alloys and Compounds, 1999;295:564. [93] Imamura H, Takesue Y, Tabata S, Shiget omi N, Sakata Y, Tsuchiya S, Chem. Commun., 1999:2277. [94] Li H. Materials Sc. Eng., PhD. Gain esville: University of Florida, 2004. p.42. [95] Ahmed Z. Materials Sc. Eng. Gain esville: University of Florida, 2001. [96] Lehmkuhl H. Aachen: Rheinische-Wes tfalische Technische Hochschule, 1954. [97] Zakharkin LI, Gavrilenko VV, J. Gen.Chem. USSR., 1962;32:688. [98] Ziegler K, Lehmkuhl H, Lindner E, Chem. Ber., 1959;92:2320. [99] Connor JH, Reid WE, Wood GB J. Electrochem. Soc., 1957;104:38. [100] Brenner A, Sligh JL, Trans.Inst.Met.Finish., 1971;49:71. [101] Liebenow C, Yang Z, Lobitz P, Bull. Electrochem., 1999;15:424. [102] Petty RL, Davidson AW, Klienbe rg J, J. Am. Chem. Soc., 1954;76:363. [103] Rausch M, McEwen W, Kleinbe rg J, J. Am. Chem. Soc., 1955;77:2093. [104] Greenblatt JH, J. El ectrochem. Soc., 1956;103:539. [105] Bonitz E, Chem. Ber., 1955;88:742. [106] Bonitz E, Angew. Chem.., 1955;67:525. [107] Lehmkuhl H, Kobs HD, Liebigs. Ann. Chem., 1968;705:1. [108] Gunawardena GA, Hills GJ, Montenegro I, Electrochim. Acta, 1977;23:693. [109] Faust JW, Ogburn F, Kahan D, Ruff AW, J. Electrochem. Soc., 1967;114:1311. [110] Faust JW, John HF, J. Phys. Chem. Solids, 1964;25:1407. [111] Kozlov VM, Bicelli LP, J. Cryst. Growth, 1997;177:289. [112] Bozzini B, Fanigliulo A, Se rra M, J. Cryst. Growth, 2001;231:589. [113] Pangarov N, Phys. Status Solidi, 1967;20:371. [114] Kozlov VM, J. Cryst. Growth, 1996;165:421. [115] Sankara Sarma VT, Ebrahimi F, J. Electrochem. Soc., 2008;155:D363.

PAGE 164

164 [116] Radmilovic V, Popov KI, Pavlovic MG, Di mitrov A, Jordanovic SH, J. Solid State Electrochem. 1998;2:162. [117] Lu DL, Tanaka K, J. Solid State Electrochem., 1997;1:187. [118] Lu DL, Tanaka K, J Sol.State Electrochem., 1997;1:187. [119] Hill G, Pour AK, Scharifker B, Electrochim. Acta, 1983;28:891. [120] Scharifker B, Hill G, Electrochim. Acta, 1983;28:879. [121] Leith SD, Ramli S, Schwartz T, J. Electrochem. Soc., 1999;146:1431. [122] Grimmet DL, Schwartz M, Nobe K, Plat. Surf. Finish, 1988;75:94. [123] Siegfried MJ, Choi KS, Adv. Mater., 2004;16:1743. [124] Allongue P, Maroun F, Current Opin ions Solid State Ma ter. Sci., 2006;10:173. [125] Jovic VD, Maksimovic V, Pavlovic MG, Jovic BM, Recent Developments Adv. Mater. Processes, 2006;518:307. [126] Jovic VD, Jovic BM, Pavlovi c MG, Electrochim. Acta, 2006;51:5468. [127] Popov KI, Grgur BN, Pavlovic MG, Radm ilovic V, J. Serb. Chem. Soc., 1993;58:1055. [128] Fleury V, Kaufman J, Hibbert B, Phys. Rev. E, 1993;48:3831. [129] Rosso M, Chassaing E, Fleury V, Ch azalviel JN, J. Electroanal. Chem., 2003;559:165. [130] Popov KI, Djokic SS, Grgur BN. Fundamenta l Aspects of Electrometallurgy. New York: Kluwer Academic Publishers, 2002:137. [131] Doherty RD, Feest EA, Holm K, Metall. Mater. Trans. B., 1973;4:115. [132] Barkey D, Garik P, Ben-Jacob E, Miller B, Orr B, J. Electrochem. Soc., 1992;139:1044. [133] Barkey D, J. Electrochem. Soc., 1991;138:2912. [134] Ben-Jacob E, Garik P, Grier D, Superlattices Microstruct., 1987;3:599. [135] Ben-Jacob E, Garik P, Nature, 1990;343:523. [136] Despic AR, Popov HI.. Modern Aspects of Electrochemistry. New York: Plenum Press, 1972:127. [137] Ohring M. The Materials Science of Thin Films. San Diego: Academic Press, 1992:46. [138] Liu Y, Overzet L, Goeckner M, Thin Solid Films, 2007;515:6730.

PAGE 165

165 [139] Tsur Y, Randall CA, J. Am. Ceram. Soc., 2000;83:2062. [140] Hallstedt B, Comput er Coupling Phase Diagrams Thermochemistry, 2007;31:292. [141] Hardie D, Perkins RN, Phil. Mag. A, 1959;4:815. [142] Luo HL, Chao CC, Duwez P, Trans. Metall. Soc. AIME, 1964;230:1488. [143] Frebel M, Behler K, Pr edel B, Z. Metallkd., 1976;67:228. [144] JCPDS, Card 04-0836, Joint Committee on Po wder Diffraction Standa rds. International Center for Diffraction Data, 1996. [145] JCPDS, Card 35-0821, Joint Committee on Po wder Diffraction Standa rds. International Center for Diffraction Data, 1996. [146] Smith JR, Hong T, Srolowitz DJ, Phys. Rev. Lett., 1994;72:4021.

PAGE 166

166 BIOGRAPHICAL SKETCH Sankara Sarma V Tatiparti was born in 1979, March 14, in Khammam, Andhra Pradesh, India. He secured his bachelors degree in metallu rgical engineering from the Indian Institute of Technology Madras in 2001. He obtai ned his masters degree in metallurgical engineering and materials science from the Indian Institute of Technology Bombay in 2004. Sankara joined the Materials Science and Engineering Department at University of Florida in Fall 2004 for his PhD under the guidance of Dr. Fereshteh Ebrahimi He is expecting his PhD in Fall 2008.