Interplay Between Structure and Properties in Dicyanoaurate-Based Coordination Polymers

INTERPLAY BETWEEN STRUCTURE AND PROPERTIES IN DICYANOAURATE-BASED COORDINATION POLYMERS

Julie Lefebvre

B.Sc, McGill University, 2002

A THESIS SUBMITTED IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY in the Department of Chemistry

All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author. Library and Bibliotheque et 1*1 Archives Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition

395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A0N4 Canada Canada

Your file Votre reference ISBN: 978-0-494-46809-8 Our file Notre reference ISBN: 978-0-494-46809-8

NOTICE: AVIS: The author has granted a non L'auteur a accorde une licence non exclusive exclusive license allowing Library permettant a la Bibliotheque et Archives and Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par Plntemet, prefer, telecommunication or on the Internet, distribuer et vendre des theses partout dans loan, distribute and sell theses le monde, a des fins commerciales ou autres, worldwide, for commercial or non sur support microforme, papier, electronique commercial purposes, in microform, et/ou autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in et des droits moraux qui protege cette these. this thesis. Neither the thesis Ni la these ni des extraits substantiels de nor substantial extracts from it celle-ci ne doivent etre imprimes ou autrement may be printed or otherwise reproduits sans son autorisation. reproduced without the author's permission.

In compliance with the Canadian Conformement a la loi canadienne Privacy Act some supporting sur la protection de la vie privee, forms may have been removed quelques formulaires secondaires from this thesis. ont ete enleves de cette these.

While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. Canada APPROVAL Name: Julie Lefebvre

Degree: Doctor of Philosophy

Title of Thesis: Interplay Between Structure and Properties in Dicyanoaurate-based Coordination Polymers

Examining Committee: Chair Dr. Paul C.H. Li Associate Professor, Department of Chemistry

Dr. Daniel B. Leznoff Senior Supervisor Associate Professor, Department of Chemistry

Dr. Ross H. Hill Supervisor Professor, Department of Chemistry

Dr. Gary W. Leach Supervisor Associate Professor, Department of Chemistry

Dr. Paul W. Percival Internal Examiner Professor, Department of Chemistry

Dr. Jeffrey R. Long External Examiner Associate Professor, Department of Chemistry University of California - Berkeley

Date Defended/Approved: March 27, 2008

11 SIMON FRASER UNIVERSITY SFU LIBRARY

Declaration of Partial Copyright Licence

The author, whose copyright is declared on the title page of this work, has granted to Simon Fraser University the right to lend this thesis, project or extended essay to users of the Simon Fraser University Library, and to make partial or single copies only for such users or in response to a request from the library of any other university, or other educational institution, on its own behalf or for one of its users.

The author has further granted permission to Simon Fraser University to keep or make a digital copy for use in its circulating collection (currently available to the public at the "Institutional Repository" link of the SFU Library website at: ) and, without changing the content, to translate the thesis/project or extended essays, if technically possible, to any medium or format for the purpose of preservation of the digital work.

The author has further agreed that permission for multiple copying of this work for scholarly purposes may be granted by either the author or the Dean of Graduate Studies.

It is understood that copying or publication of this work for financial gain shall not be allowed without the author's written permission.

Permission for public performance, or limited permission for private scholarly use, of any multimedia materials forming part of this work, may have been granted by the author. This information may be found on the separately catalogued multimedia material and in the signed Partial Copyright Licence.

While licensing SFU to permit the above uses, the author retains copyright in the thesis, project or extended essays, including the right to change the work for subsequent purposes, including editing and publishing the work in whole or in part, and licensing other parties, as the author may desire.

The original Partial Copyright Licence attesting to these terms, and signed by this author, may be found in the original bound copy of this work, retained in the Simon Fraser University Archive.

Simon Fraser University Library Burnaby, BC, Canada

Revised: Fall 2007 Abstract

This research project has focused on the design and synthesis of capping ligand-free dicyanoaurate-based coordination polymers. Dicyanoaurate is a building block that readily bridges two transition metals and increases the connectivity of the system through gold-gold interactions.

A series of isostructural M(/i-OH2)2[Au(CN)2]2 coordination polymers (Cu, Ni, Co, Fe, Mn) was prepared. Their structure contains an unprecedented motif in which metal centers are doubly bridged by water molecules, generating chains. The mag netic interactions present in these water-bridged polymers were investigated; both ferromagnetic (Cu, Ni, Co) and antiferromagnetic interactions (Fe, Mn) were ob served. Muon spin relaxation studies showed different magnetic ground states for these polymers, including long-range ordered (Cu, Tc = 0.2 K) and spin-glass states

(Ni, Tf = 3.5 K). Several CujA^CN^^analyte)^ polymers (analyte = dimethylsulfoxide, N,N- dimethylformamide, water, pyridine, ammonia) were prepared and found to be vapochromic, i.e. change colour upon exposure to volatile analytes. The analyte incorpo rated can easily be identified as each colour is specific to one analyte, with maximum reflection ranging from 433 to 560 nm. The infrared cyanide vibration signature of each polymer is also unique and allows another way of identifying the analyte present. The analyte can be exchanged at room temperature, without any thermal treatment. A survey of analyte sensing capabilities was performed by replacing the Cu ions with

Ni and Co ions in M[Au(CN)2]2(analyte):E. + The preparation of [cation]{M[Au(CN)2]3} polymers (cation = K+, [PPN] , [nBu4N]+; M = Ni, Co) was also investigated. The incorporation of different cations

m affected the structural arrangement of the anionic framework, but did not prevent the 3-D connectivity between the metal centers. Despite the high connectivity, very weak antiferromagnetic interactions were observed which suggests that dicyanoaurate is a poor mediator of magnetic exchange. Studies by transmission electron microscopy showed the presence of trace amounts of NiO and Au nanoparticles in hydrothermally recrystallized samples of dicyanoaurate-containing coordination polymers. These products were hardly de tectable by standard analysis techniques, but the NiO nanoparticles dominated the magnetic response at low temperatures. This finding illustrates the care that must be taken when investigating samples prepared by hydrothermal methods.

Keywords: coordination polymers; cyanometallate; vapochromism; molecular mag netism; gold-gold interactions.

IV To those whom I love, To those who cared and understood, To those who offered their support, To the one without whom this would have been impossible, I dedicate this thesis.

v The most important thing in science is not so much to obtain new facts as to discover new ways of thinking about them." — Sir William Lawrence Bragg, 1965

"On fait la science avec des faits, comme on fait une maison avec des pierres: mais une accumulation de faits n 'est pas plus une science qu'un tas de pierres n'est une maison."

— Henri Poincare, LA SCIENCE ET L'HYPOTHESE , 1905

vi Acknowledgments

First and foremost, I would like to thank my senior supervisor, Daniel Leznoff, for his guidance, support, trust and encouragement throughout all these years. His love of chemistry was so great that it was contagious! Learning from him was a pleasure, especially if you could understand his handwriting. I would like to thank my supervisory committee members, Ross Hill and Gary Leach, for their help and support during my degree. Seeing the "big picture" was not always easy, but I'm glad they insisted! A special thank also to Ross for his collaboration and thorough scientific thought process in investigating the formation of nanoparticles impurities. A special thank to my external examiner, Jeffrey Long, for taking the time to read my thesis, being present at my defense and providing useful comments. I would also like to thank Paul Percival, my internal examiner, for carefully reading my whole thesis, giving me useful feedback and teaching me a few grammatical rules! I wish to deeply thank Jeff Sonier not only for performing the muon spin relaxation experiments with us, but also for all his time spent analyzing the data and teaching me about this unusual technique. I'd like to thank the members of his group, especially Fergal Callaghan, Christina Kaiser, Vighen Pacradouni and Pooja Tyagi, for spending numerous hours collecting and analyzing the data over the last five years. I'd like to thank the TRIUMF staff of the Center for Molecular and Materials Science (CMMS) for their 24-hour support during our experiments. I'd like to thank Jim O'Brien, application engineer at Quantum Design, for help ing me in my struggles with the SQUID magnetometer, Miki Yang for performing the elemental analysis and Brian Patrick (UBC) for collecting several powder X-ray

vn diffractograms over the last two years. I thank Michael Katz and Raymond Batchelor for watching over my shoulders and making sure that no crystallographic mistakes were made. I'd like to thank Xin Zhang for teaching me how to use our powder X- ray diffractometer and for all his time and devotion at making the instrument run. I'd like to thank Dev Sharma for his wisdom and help to perform reactions under hydrothermal conditions safely. For financial support over the last five years, I thank NSERC of Canada, FQRNT (Quebec), Simon Fraser University and the World Gold Council. Science does not evolve by itself, and without interactions with the "outside world", no progress could be made. I want to thank all those with whom I bounced back ideas and had useful discussion. Especially, I'd like to express my gratitude to Michael Katz for all these discussions we had about science (and gossip), for teaching me tons of new words, and being such a good friend. I'd like to thank also all the Leznoff group members, past and present, who made all these years pass by so fast. I also thank Daniel Chartrand and the few other undergraduate students who worked with me on different projects. Many many thanks to Simon Trudel who was there since the beginning to discuss science at any time of the day, to teach me about Mossbauer spectroscopy, magnetism in a physicist-way and the wonderful world of nanoparticles, to spend time at TRIUMF by my side and several hours imaging my samples by TEM. Enfin, un merci particulier a ma famille et a tous mes amis. Votre support moral a toujours ete tres apprecie et, grace a vous, j'ai persevere jusqu'au bout!

Merci a tous!

viii Contents

Approval ii

Abstract iii

Dedication v

Quotation vi

Acknowledgments vii

Contents ix

List of Tables xvi

List of Figures xx

List of Schemes xxiv

List of Abbreviations xxvi

1 Introduction 1 1.1 An introduction to coordination polymers 1 1.1.1 Cyanometallate building blocks 3

1.1.2 The use of [Au(CN)2]~ units as linear cyanometallate building blocks 6 1.2 Objectives of this research project 8

ix 1.3 An introduction to molecular magnetism 9 1.3.1 Principles, definitions and fundamental equations 9 1.3.2 Single-ion effects 16 1.4 Characterization Tools 19 1.4.1 Determination of chemical composition 19 1.4.2 Information obtained by spectroscopy 19 1.4.3 Investigation of structural arrangement using crystallography . 24 1.4.4 Investigation of magnetic properties by SQUID magnetometry 26 1.4.5 Investigation of magnetic properties using muon spin relaxation 28

2 M(//-OH2)2[Au(CN)2]2: Magnetic exchange through unusual double aqua-bridges 33 2.1 Introduction and research objectives 33 2.2 An uncommon structural motif: double aqua-bridges 34 2.2.1 Preparation and identification of products 34 2.2.2 Structure determination 37

2.2.3 Fe(^OH2)(jU-OH)[Au(CN)2]2 46 2.2.4 Thermal stability 49 2.2.5 Removal of the water molecules in the Cu-based system .... 50 2.2.6 Discussion about the synthesis and structural arrangements . 52 2.3 Magnetic exchange through double aqua-bridges 54

2.3.1 Cu(/x-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 55

2.3.2 Ni(At-OH2)2[Au(CN)2]2 61

2.3.3 Co(/>OH2)2[Au(CN)2]2 67

2.3.4 Fe(//-OH2)2[Au(CN)2]2 and Mn(/x-OH2)2[Au(CN)2]2 69

2.3.5 Magnetic properties of Fe(/i-OH2)(M-OH)[Au(CN)2]2 71 2.3.6 Preliminary /iSR results 72 2.4 Discussion 73 2.4.1 Magneto-structural correlations 73 2.4.2 Magnetic exchange through double aqua-bridges 74

x 2.4.3 Importance of weak interchain interactions 79 2.4.4 Summary of magnetic behaviours 84 2.5 Conclusion 85 2.6 Future work 86 2.7 Experimental Section 87 2.7.1 Reagents and general procedures for characterization 87 n 2.7.2 Synthesis of [ Bu4N][Au(CN)2]-0.5H2O 89

2.7.3 Synthesis of Cu(/z-OH2)2[Au(CN)2]2 90

2.7.4 Synthesis of Cu[Au(CN)2]2 90

2.7.5 Synthesis of Ni(/u-OH2)2[Au(CN)2]2 90

2.7.6 Synthesis of Co(Ai-OH2)2[Au(CN)2]2 91

2.7.7 Synthesis of Fe(//-OH2)2[Au(CN)2]2 91

2.7.8 Synthesis of Fe(/x-OH2)(//-OH)[Au(CN)2]2 92

2.7.9 Synthesis of Mn(/x-OH2)2[Au(CN)2]2 92

2.7.10 Synthesis of Ni[Ag2(CN)3][Ag(CN)2] 93 2.7.11 X-ray crystallographic analysis 93 2.7.12 Details on SQUID magnetometry experiments 94 2.7.13 Experimental details on Mossbauer spectroscopy 95 2.7.14 //SR measurements 96

3 M[Au(CN)2]2(analyte)2: coordination polymers and their vapochromic properties 98 3.1 Introduction 98 3.1.1 Coordination polymers as vapochromic materials 98 3.1.2 Research objectives 100

3.2 Cu[Au(CN)2]2(DMSO)2: Preparation and characterization 100 3.2.1 Synthesis of polymorphs 100

3.2.2 Cu[Au(CN)2]2(DMSO)2 (green) 101

3.2.3 Cu[Au(CN)2]2(DMSO)2 (blue) 102

3.2.4 Physical properties of Cu[Au(CN)2]2(DMSO)2 107

3.3 Vapochromic behaviour of the Cu[Au(CN)2]2(DMSO)2 polymorphs . 109

XI 3.3.1 Identification of adsorbed guest by infrared and UV-vis spec troscopies 109 3.4 Structure-properties relationships of the

Cu[Au(CN)2]2(analyte)x polymers 113

3.4.1 Cu[Au(CN)2]2(DMF) 113

3.4.2 Cu[Au(CN)2]2(pyridine)2 116

3.4.3 Cu[Au(CN)2]2(CH3CN)2 118

3.4.4 Cu[Au(CN)2]2(dioxane)(H20) 120

3.4.5 Physical Properties of Cu[Au(CN)2]2(analyte)cc 120 3.5 Substituting the Cu(II) ions for Ni(ll) and Co(Il) 124

3.5.1 M[Au(CN)2]2(DMSO)2 124

3.5.2 Ni[Au(CN)2]2(DMF)2 128

3.5.3 M[Au(CN)2]2(pyridine)2 131 3.5.4 Physical properties of Ni- and Co-based polymers 134 3.5.5 Comparing the vapochromic behaviour of Ni- and Co-based polymers 140 3.6 Discussion 143 3.6.1 Polymorphism in coordination polymers 143 3.6.2 Solution synthesis vs solvent exchange in the solid state .... 144 3.6.3 Comparison of structural motifs adopted upon modification of metal center and analyte molecules 146

3.6.4 Responses of the M[Au(CN)2]2(analyte)x polymers upon expo sure to analyte vapours 148 3.7 Conclusions 151 3.8 Future work 152 3.8.1 Modification of the analytes 152

3.8.2 M[Au(CN)2]2 as sensors 153 3.8.3 Modification of the building blocks 154 3.9 Experimental Section 156

3.9.1 Synthesis of Cu[Au(CN)2]2(DMSO)2 (green) 156

3.9.2 Synthesis of Cu[Au(CN)2]2(DMSO)2 (blue) 156

xii 3.9.3 Synthesis of Cu[Au(CN)2]2(DMF) 157

3.9.4 Synthesis of Cu[Au(CN)2]2(pyridine)2 157

3.9.5 Synthesis of Cu[Au(CN)2]2(CH3CN)2 158

3.9.6 Synthesis of Cu[Au(CN)2]2(dioxane)(H20) 158

3.9.7 Synthesis of Cu[Au(CN)2]2(NH3)4 159

3.9.8 Synthesis of Ni[Au(CN)2]2(DMSO)2 159

3.9.9 Synthesis of Ni[Au(CN)2]2(DMF)2 159

3.9.10 Synthesis of Ni[Au(CN)2]2(pyridine)4 160

3.9.11 Synthesis of Co[Au(CN)2]2(DMSO)2 160

3.9.12 Synthesis of Co[Au(CN)2]2(pyridine)2 160

3.9.13 Synthesis of Co[Au(CN)4]2(DMSO)4 161 3.9.14 Details on structure determination of

Cu[Au(CN)2]2(analyte):c 161 3.9.15 Details on structure determination of

M[Au(CN)2]2(analyte)x (M = Ni and Co) 162

[cation]{M[Au(CN)2]3} polymers: Templating effects of the cation 166 4.1 Incorporation of non-coordinating building blocks 166 4.1.1 Research objectives 168 4.2 Modification of the 3-D superstructure through incorporation of K+, n + [PPN]+ and [ Bu4N] cations 169

4.2.1 Synthesis and structural characterization of K{Ni[Au(CN)2]3} 169 4.2.2 Synthesis and structural characterization of

[PPN]{M[Au(CN)2]3} polymers 172 4.2.3 Synthesis and structural characterization of n [ Bu4N]{M[Au(CN)2]3} polymers 174 4.2.4 Synthesis and structural characterization of n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] polymer 178 4.2.5 Solid-state UV-vis-NIR absorption spectroscopy of

[cation]{M[Au(CN)2]3} 181 4.2.6 Thermal stability and structural rearrangement 184

xm 4.3 Effect of structural changes on the magnetic behaviour 186

4.3.1 Magnetic properties of [cation]{Ni[Au(CN)2]3} 186

4.3.2 Magnetic properties of [cation]{Co[Au(CN)2]3} 189

4.3.3 Magnetic properties of K{Fe[Au(CN)2]3} 190 4.4 Discussion 192

4.4.1 Synthesis of M(^-OH2)2[Au(CN)2]2 vs K{M[Au(CN)2]3} ... 193 4.4.2 Impacts of the cations on the structural arrangement 194 4.4.3 Impacts of the cations on the physical properties 195 4.5 Conclusions and future work 198 4.6 Experimental Section 200

4.6.1 Synthesis of K{Ni[Au(CN)2]3} 200

4.6.2 Synthesis of K{Fe[Au(CN)2]3} 201

4.6.3 Synthesis of [PPN][Au(CN)2] 201

4.6.4 Synthesis of [PPN]{Ni[Au(CN)2]3} 202

4.6.5 Synthesis of [PPN]{Co[Au(CN)2]3} 203 n 4.6.6 Synthesis of [ Bu4N]{Ni[Au(CN)2]3} 203 n 4.6.7 Synthesis of [ Bu4N]{Co[Au(CN)2]3} 204 n 4.6.8 Synthesis of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] .... 204 4.6.9 Attempts involving [DAMS]+ 205 4.6.10 Details of X-ray structural determinations 206

5 Hydrothermal synthesis of [Au(CN)2]-based polymers and related nanoparticle impurities 210 5.1 Introduction 210 5.1.1 Crystallization of coordination polymers 210 5.1.2 Research objectives 211

5.2 Characterization of K{Ni[Au(CN)2]3} samples prepared at elevated temperatures 212 5.2.1 Magnetic behaviour of samples prepared at 125 °C 212 5.2.2 Transmission electron microscopy studies 215 5.2.3 Modification of the reaction conditions 219

xiv 5.3 Discussion 224 5.3.1 Chemical identity of the nanoparticles 224 5.3.2 Nanoparticle formation route 226 5.3.3 Magnetic behaviours 227 5.3.4 Shortcomings of conventional analytical methods 232 5.3.5 Presence of nanoparticles in other systems 233 5.4 Conclusions 234 5.5 Experimental Section 235

5.5.1 Hydrothermal synthesis of K{Ni[Au(CN)2]3} at 125 °C .... 235

5.5.2 Hydrothermal synthesis of K{Ni[Au(CN)2]3} at 135 and 165 °C 235

5.5.3 Control experiment with K[Au(CN)2] 236 5.5.4 SQUID Magnetometry 236 5.5.5 Transmission electron microscopy 236

6 Global conclusions and perspectives 237

A Summary of Crystallographic Data 241

A.l Fractional atomic coordinates for the M(/i-OH2)2[Au(CN)2]2 coordina tion polymers 241

A.2 Fractional atomic coordinates for the M[Au(CN)2]2(analyte)s coordi nation polymers 247

A.3 Fractional atomic coordinates for the [cation]{M[Au(CN)2]3} coordina tion polymers 257

References 264

Index 291

xv List of Tables

1.1 Typical isomer shift and quadrupole splitting observed for low spin and high spin six coordinate Fe(n) and Fe(m) complexes 24

2.1 UCN vibration frequency for the M(/x-OH2)2[Au(CN)2]2 complexes and

Fe(/i-OH2)(//-OH)[Au(CN)2]2 36

2.2 Unit cell parameters for the M(/x-OH2)2[Au(CN)2]2 and Fe(/A-OH2)(M-

OH)[Au(CN)2]2 complexes 39

2.3 Selected bond lengths and angles for Ni(>-OH2)2[Au(CN)2]2 42

2.4 Decomposition temperatures of the M(/x-OH2)2[Au(CN)2]2 complexes

and Fe(/x-OH2)(/u-OH)[Au(CN)2]2 49 2.5 Comparison between the magnetic behaviours of the different Ni(/x-

OH2)2Ni containing complexes 76 2.6 Comparison between the magnetic behaviours of the different M(/x-

OH2)2[Au(CN)2]2 polymers 85 2.7 Crystallographic data and structural refinement details for

Ni(^-OH2)2[Au(CN)2]2 97

3.1 Selected bond lengths and angles for Cu[Au(CN)2]2(DMSO)2 (green) 104

3.2 Selected bond lengths and angles for Cu[Au(CN)2]2(DMSO)2 (blue) . 106 3.3 Maximum solid-state visible reflection and UQN absorptions for different

Cu[Au(CN)2]2(analyte)x complexes 112

3.4 Selected bond lengths and angles for Cu[Au(CN)2]2(DMF) 114

3.5 Selected bond lengths and angles for Cu[Au(CN)2]2(pyridine)2 .... 118

xvi 3.6 Thermal stability of the CufA^CN^j^analyte)^ polymers: decompo sition temperature and products obtained 121

3.7 Effective magnetic moment determined for the Cu[Au(CN)2]2(analyte)a; coordination polymers at different temperatures 122

3.8 Unit cell parameters determined for M[Au(CN)2]2(DMSO)2 (M = Zn, Cu (blue polymorph) and Ni) 126

3.9 Unit cell parameters determined for M[Au(CN)2]2(DMF)2 (M = Co,

Ni) and Co[Au(CN)2]2(pyridine)2 130 3.10 Absorbance maxima observed in the solid-state UV-Vis-NIR absorption

spectra of the M[Au(CN)2]2(analyte)x (M = Co and Ni) coordination polymers and their respective assignment 136

3.11 Effective magnetic moment determined for the M[Au(CN)2]2(analyte)x coordination polymers at different temperatures 139

3.12 VCN absorptions for different M[Au(CN)2]2(analyte)x complexes . . . 141 3.13 Crystallographic data and structural refinement details for the two

Cu[Au(CN)2]2(DMSO)2 polymorphs 163 3.14 Crystallographic data and structural refinement details for

Cu[Au(CN)2]2(DMF) and Cu[Au(CN)2]2(pyridine)2 164 3.15 Crystallographic data and structural refinement details for

Co[Au(CN)4]2(DMSO)4 165

4.1 Selected bond lengths and angles for K{Ni[Au(CN)2]3} 170

4.2 Selected bond lengths and angles for [PPN]{Ni[Au(CN)2]3} 174 n 4.3 Selected bond lengths and angles for [ Bu4N]{Ni[Au(CN)2]3} 178

4.4 Selected bond lengths and angles for {Co[Au(CN)2]2(H20)2}- n [ Bu4N][Au(CN)2] 181 4.5 Absorption maxima observed in the solid-state UV-Vis-NIR absorp

tion spectra of the [cation]{M[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}- n [ Bu4N][Au(CN)2] coordination polymers and their respective assign ment 183

4.6 Thermal stability of the [cation]{M[Au(CN)2]3} polymers 185

xvn 4.7 Effective magnetic moment determined for the [cation]{M[Au(CN)2]3} coordination polymers at different temperatures 188 4.8 Crystallographic data and structural refinement details for

K{Ni[Au(CN)2]3} and [PPN]{Ni[Au(CN)2]3} 207 4.9 Crystallographic data and structural refinement details for n n [ Bu4N]{Ni[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] 208

4.10 Unit cell parameters determined for [PPN]{Co[Au(CN)2]3} and n [ Bu4N]{Co[Au(CN)2]3} 209

A.l Fractional atomic coordinates and equivalent isotropic thermal para-

metersfor Ni(Ai-OH2)2[Au(CN)2]2 241

A.2 Fractional atomic coordinates for Cu(/x-OH2)2[Au(CN)2]2 244

A.3 Fractional atomic coordinates for Co(/i-OH2)2[Au(CN)2]2 244

A.4 Fractional atomic coordinates for Fe(//-OH2)2[Au(CN)2]2 245

A.5 Fractional atomic coordinates for Mn(/^-OH2)2[Au(CN)2]2 245

A.6 Fractional atomic coordinates for Fe(//-OH2)(//-OH)[Au(CN)2]2 .... 246 A.7 Fractional atomic coordinates and equivalent isotropic thermal para

meters for Cu[Au(CN)2]2(DMSO)2 (blue) 247 A.8 Fractional atomic coordinates and equivalent isotropic thermal para

meters for Cu[Au(CN)2]2(DMSO)2 (green) 250 A.9 Fractional atomic coordinates and equivalent isotropic thermal para

meters for Cu[Au(CN)2]2(DMF) 251 A. 10 Fractional atomic coordinates and equivalent isotropic thermal para

meters for Cu[Au(CN)2]2(pyridine)2 252

A.ll Fractional atomic coordinates for Ni[Au(CN)2]2(DMSO)2 253

A.12 Fractional atomic coordinates for Ni[Au(CN)2]2(DMF)2 254

A.13 Fractional atomic coordinates for Co[Au(CN)2]2(pyridine)2 255

A.14 Fractional atomic coordinates for Co[Au(CN)4]2(DMSO)4 256 A. 15 Fractional atomic coordinates and equivalent isotropic thermal para

meters for K{Ni[Au(CN)2]3} 257

xvm A. 16 Fractional atomic coordinates and equivalent isotropic thermal para

meters for [PPN]{Ni[Au(CN)2]3} 258 A. 17 Fractional atomic coordinates and equivalent isotropic thermal para n meters for [ Bu4N]{Ni[Au(CN)2]3} 259 A. 18 Fractional atomic coordinates and equivalent isotropic thermal para n meters for {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] 262

xix List of Figures

1.1 Solid-state structures of [C4H8NH2][Au(CN)2] and Cu(en)2[Au(CN)2]2,

which both contain [Au(CN)2]~ units held by aurophilic interactions. 7 1.2 Temperature dependence of the magnetic susceptibility and the ef fective magnetic moment for a paramagnetic system, as well as for systems in which ferromagnetic or antiferromagnetic interactions are present (A.). Field dependency of the magnetization predicted for a paramagnetic system and compared to typical curves observed for fer romagnetic and antiferromagnetic systems (B.) 12 1.3 Number of positrons detected in the detectors as a function of time for a sample with a local magnetic field transverse to the initial spin polarization (A.) and the corresponding asymmetry spectrum (B.). . 32

2.1 FT-IR spectra of the M(>-OH2)2[Au(CN)2]2 complexes 35

2.2 Mossbauer spectrum of Fe(/i-OH2)2[Au(CN)2]2 37

2.3 Powder X-ray diffractogram of M(//-OH2)2[Au(CN)2]2 38

2.4 Structure of Ni(//-OH2)2[Au(CN)2]2 41 2.5 Powder X-ray diffractogram predicted by the proposed models for Ni(/i-

OH2)2[Au(CN)2]2 43 2.6 Powder X-ray diffractogram predicted by the proposed model for Cu(/z-

OH2)2[Au(CN)2]2 45

2.7 Mossbauer spectrum of Fe(/x-OH2)(/x-OH)[Au(CN)2]2 47

2.8 Effective magnetic moment of Cu(/i-OH2)2[Au(CN)2]2 and

Cu[Au(CN)2]2 56

XX 2.9 Asymmetry spectra and fitted parameters for Cu(/i-OH2)2[Au(CN)2]2 58

2.10 Asymmetry spectra and fitted parameters for Cu[Au(CN)2]2 60 2.11 Magnetic susceptibility and effective magnetic moment of

Ni(/i-OH2)2[Au(CN)2]2 62

2.12 Field dependence of the magnetization of Ni(//-OH2)2[Au(CN)2]2 • • • 63

2.13 Temperature dependence of the critical field for Ni(/i-OH2)2[Au(CN)2]2 64 2.14 Temperature and frequency dependence of the two components of the

ac susceptibility for Ni(//-OH2)2[Au(CN)2]2 65

2.15 Asymmetry spectra for Ni0u-OH2)2[Au(CN)2]2 66 2.16 Effective magnetic moment and magnetization as function of tempera

ture and field for CoQu-OH2)2[Au(CN)2]2 68 2.17 Effective magnetic moment and magnetization of

Fe(^-OH2)2[Au(CN)2]2 and Mn(//-OH2)2[Au(CN)2]2 70

2.18 Effective magnetic moment of Fe(/i-OH2)(//-OH)[Au(CN)2]2 71

3.1 Extended solid-state structure of Cu[Au(CN)2]2(DMSO)2 (green) ., 103

3.2 Extended solid-state structure of Cu[Au(CN)2]2(DMSO)2 (blue) ... 105

3.3 Thermogravimetric analysis of the two Cu[Au(CN)2]2(DMSO)2 poly morphs 107

3.4 Cu[Au(CN)2]2(DMSO)2 exposed to various solvent vapours 110 3.5 Solid-state UV-vis-NIR transmission spectra of

Cu[Au(CN)2]2(analyte):E Ill

3.6 Extended solid-state structure of Cu[Au(CN)2]2(DMF) 115

3.7 Extended solid-state structure of Cu[Au(CN)2]2(pyridine)2 117

3.8 Powder X-ray diffractogram of Ni[Au(CN)2]2(DMSO)2 compared with the diffractogram of the proposed structural model 125 3.9 Comparison between powder X-ray diffractograms of

Co[Au(CN)2]2(DMSO)2 and Mn[Au(CN)2]2(H20)2 127 3.10 Comparison between powder X-ray diffractograms of

Ni[Au(CN)2]2(DMF)2 and Co[Au(CN)2]2(DMF)2 129

3.11 Structural model proposed for Ni[Au(CN)2]2(DMF)2 130

xxi 3.12 Powder X-ray diffractogram of M[Au(CN)2]2(pyridine)2 133

3.13 Structural model proposed for Co[Au(CN)2]2(pyridine)2 134

3.14 Solid-state UV-vis-NIR absorbance spectra of M[Au(CN)2]2(analyte);r 135

3.15 Solid-state structure of Co[Au(CN)4]2(DMSO)4, showing the molecular complexes interacting through N-Au interactions 154

4.1 Solid-state structure of K{Ni[Au(CN)2]3} 171

4.2 Extended solid-state structure of [PPN]{Ni[Au(CN)2]3} 173 4.3 Comparison between the powder X-ray diffractogram of the

[PPN]{M[Au(CN)2]3} complexes 175 n 4.4 Solid-state structure of [ Bu4N]{Ni[Au(CN)2]3} 176 4.5 Comparison between the powder X-ray diffractograms of the n [ Bu4N]{M[Au(CN)2]3} complexes 179 n 4.6 Solid-state structure of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] . . 180

4.7 UV-vis-NIR spectra of [cation]{M[Au(CN)2]3} 182

4.8 Effective magnetic moment for K{Ni[Au(CN)2]3} 187

4.9 Effective magnetic moment of [PPN]{M[Au(CN)2]3} 189 4.10 Effective magnetic moment and Mossbauer spectra of

K{Fe[Au(CN)2]3} 191

5.1 Effective magnetic moments of the K{Ni[Au(CN)2]3} RT and H-125 samples 213 5.2 ZFC measurements under different external fields and field dependence

of TB for sample H-125 214 5.3 In-phase and out-of-phase components of the ac susceptibility for the

RT and H-125 K{Ni[Au(CN)2]3} samples 215 5.4 Bright field (A.) and high resolution (B.) transmission electron mi croscopy images of nanoparticles present in sample H-125, selected area electron diffraction (C.) and size distribution of the nanoparticles (D.) 217 5.5 HAADF scanning transmission electron microscopy image of the H- 125 sample and energy dispersive X-ray spectra for areas highlighted. 218

xxii 5.6 Comparison of the ZFC magnetic behaviours of samples RT, H-125, H-135 and H-165 220 5.7 HAADF scanning transmission electron microscopy image showing the 200-300 nm particles and the filament-like particles present in the H- 165 sample and selected area electron diffraction of an area containing only filament-like particles 221 5.8 HAADF scanning transmission electron microscopy image of the H- 165 sample and energy dispersive X-ray spectra for areas highlighted. 221 5.9 Comparison of the measured X-ray diffractogram of H-165 with the

diffractograms predicted for K{Ni[Au(CN)2]3}, Au and NiO 222 5.10 HAADF scanning transmission electron microscopy images of the H- 135 sample and energy dispersive X-ray spectra for different high lighted areas 223 5.11 Field dependence of the blocking temperature TB in H-125, following the de Almeida-Thouless law 230

A.l Powder X-ray diffractogram predicted by the proposed model for Co(/z-

OH2)2[Au(CN)2]2 242 A. 2 Powder X-ray diffractogram predicted by the proposed model for Fe(/x-

OH2)2[Au(CN)2]2 242 A.3 Powder X-ray diffractogram predicted by the proposed model for Mn(/x-

OH2)2[Au(CN)2]2 243 A.4 Powder X-ray diffractogram predicted by the proposed model for Fe(/i-

OH2)(/i-OH)[Au(CN)2]2 243

XXlll List of Schemes

1.1 Different superstructures resulting from the assembly of simple building blocks 2 1.2 Structural motifs observed with cyanometallate building blocks: Prus sian Blue array; square-grid array; diamond-like array. 5 1.3 Orientation of the magnetic moments in paramagnetic, ferromagnetic and antiferromagnetic systems 10 1.4 Energy diagram for an S = 1 system, showing the effect of Zeeman splitting in the presence of an applied magnetic field (A.). Energy diagram for a dimer of S — | centers in the presence of a magnetic field (B.) 14 1.5 Zero-field splitting energy diagram of an S = 1 ion with the relative energy of each state 17 1.6 Experimental set-up used to determine the solid-state UV-vis-NIR absorbance and reflection spectra 20 1.7 Nuclear transitions probed by 57Fe Mossbauer spectroscopy and asso ciated Mossbauer spectrum 23 1.8 Schematic of the experimental geometry used to carry out the /iSR measurements 31

2.1 Asymmetric diamond motif in Cu(/i-OH2)2[Au(CN)2]2 resulting from the bridging mode of the water molecules 46

2.2 Proposed 2-D square grid structural motif for Cu[Au(CN)2]2 51 2.3 Possible magnetic pathways in M(/i-OH2)2[Au(CN)2]2 • 73

xxiv 4 Schematic representation of the interchain magnetic interactions lead ing to long-range order and spin-glass like magnetic states in the M(/x-

OH2)2[Au(CN)2]2 and Fe(/x-OH2)(A*-OH)[Au(CN)2]2 polymers 80

1 Analyte molecules incorporated into the M[Au(CN)2]2(analyte)x poly mers presented in this chapter 100 2 Geometrical parameters to determine the r-value of a five-coordinate metal center 102

3 Different structural models observed for the M[Au(CN)2]2(analyte)a;

polymers, all resulting from the distortion of the basic M[Au(CN)2]2 square-grid array. 147 _ 1 Shape and coordination modes of [N(CN)2] and [Au(CN)2]~ units. . 167 n 2 "Shape" of the different cations: K+, [PPN]+and [ Bu4N]+ 168 3 Magnetic exchange occurring in Prussian Blue between the Fe(lll) cen ters through the Fe(n) centers due to partial electron derealization. Only one electron spin is favoured during electron transfer (shown in red circle) which leads to overall ferromagnetic orientation of the for mally Fe(lll) magnetic moment 197 4 "Shape" of different ligands and cations: bipyridine, terpyridine, [DAMS]+, and [TTF]n+ 200 1 Magnetic nanoparticles in the superparamagnetic and blocked states when cooled in zero field 229

xxv List of Abbreviations

l-D one dimensional 2-D two dimensional 3-D three dimensional °C degree Celsius a,/3,7 unit cell parameters (angles) 7 muon gyromagnetic ratio S isomer shift X wavelength

^max abs. wavelength at maximum absorbance

''max refl. wavelength at maximum reflection 6 Weiss constant or diffraction angle V linear absorption coefficient »+ positively charged muon VB Bohr magneton Veff effective magnetic moment //SR muon Spin Relaxation

VCN cyanide vibration frequency

VH muon precessing frequency P crystal density

Tc correlation time X magnetic susceptibility

XXVI A Angstrom a, b, c unit cell parameters (length) ac alternating current Anal. analysis B, local magnetic field at the muon site

Bext external magnetic field

Bs(y) Brillouin function BT transverse magnetic field BF bright field bipy 2,2'-bipyridine n Bu butyl group (CH3-(CH2)2-CH2-) C Curie constant Calcd. calculated D zero-field splitting parameter DAMS 4-[4-(dimethylamino)-Q!-styryl]-N-methylpyridinium dc direct current dca dicyanamide anion, [N(CN)2]~ deg degree(s) DMF N, iV-dimethylforrnarnide DMSO dimethyl sulfoxide emu electromagnetic units EDXS energy dispersive X-ray spectroscopy en 1,2-ethylenediamine FC field cooled fee face centered cubic FFT fast Fourier transform FT-IR Fourier transform infrared g gram 9 g Lande factor H magnetic field strength

XXVll H magnetic field vector HAADF high angle annular dark field HRTEM high resolution transmission electron microscopy HS high-spin J magnetic coupling parameter K Kelvin kB Boltzmann constant kJ kilo joule L ligand LF longitudinal field LS low-spin M magnetization or molarity mL milliliter mmol millimol

NA Avogadro's number NMR nuclear magnetic resonance Oe Oersted Ph phenyl ppm parts-per-million PPN+ bis(triphenylphosphoranylidene)ammonium cation S total spin quantum number s strong (intensity) SEM scanning electron microscope, microscopy SQUID Superconducting Quantum Interference Device STEM scanning transmission electron microscope, microscopy T temperature or Tesla

TB blocking temperature

Tc critical or Curie temperature

TF freezing temperature TEM transmission electron microscope, microscopy tren tris(2-aminoethyl)amine

xxvm TTF tetrathiafulvalene U(iso) isotropic thermal parameter v. br. very broad VOC volatile organic compound w weak Z number of asymmetric units per unit cell z number of neighboring metal centers ZF zero-field ZFC zero-field cooled zfs zero-field splitting

xxix Chapter 1

Introduction

1.1 An introduction to coordination polymers

Research into new functional materials with useful and tunable conductive, (1_3) mag netic (4'5) or optical (6~8) properties or the formation of porous materials (9~12) has re cently attracted a lot of attention. The generation of many such properties is enhanced by, or depends on, the synthesis of materials with high structural connectivity. In re cent years, supramolecular coordination polymers have been investigated due to their potential of generating such materials. (12>13) Polymers result from the assembly of molecular building blocks. In a coordination polymer, these building blocks are composed of inorganic metal ions and bridging ligands. Scheme 1.1 illustrates some examples of coordination polymers resulting from the assembly of different building blocks. As shown in Scheme 1.1, the metal ions can be imagined as assembling points for the bridging ligands, which form connections through metal-ligand coordinate bonds. This particular method of assembly is often referred to as the directional bonding approach as it uses metal-ligand bonding to direct the structures formed. (14^

1 Chapter 1. Introduction

A. x cH*-a +

B. x o#-a + 2x

C. 9 9 *V-»-a + x O {% transition metal ion

Scheme 1.1: Different superstructures resulting from the assembly of simple building blocks, showing how the structure can be affected by the number of available coordi nation sites around the metal ions (A. and B.) as well as the effects of introducing an additional capping ligand (C). The areas highlighted in gray represent the repeating unit in each coordination polymer. Chapter 1. Introduction 3

The characteristics of the building blocks ideally control the extended superstruc ture/13'15) The metal ions are chosen according to their number of available coordi nation sites and their coordination geometry preferences. For example, Schemes 1.1A and 1.1B compare the superstructure obtained as the number of available coordination sites is increased from two to four. Rigid or flexible ligands, possessing at least two functional groups capable of bind ing to a metal center, can be used to bridge two metal centers. Most ligands contain carboxylate, cyanide or pyridine-type groups. (13>16) Weak interactions between the ligands, such as hydrogen bonding or 7T-7T stacking, can also influence the overall superstructure obtained. (17) Additional non-bridging ligands can also be incorporated to cap the metal ions and restrain the growth of the coordination polymer in a particular manner. (13) Scheme 1.1C illustrate such an example where a chelating ligand occupies two co ordination sites on a metal ion and prevents the formation of a layered superstructure such as that shown in Scheme LIB. By modifying the building blocks, enormous potential for functionality and com plexity of these modular materials is created. (16'18)

1.1.1 Cyanometallate building blocks

Systems that employ cyanometallate building blocks, [M(CN)a;]n~, as bridging groups have been widely investigated. (19_24) In concert with metal ions or their coordinately unsaturated complexes, they can create multidimensional M-CN-M' networks. n 21 Cyanometallate groups, [M(CN)I] ~, are very stable( ) and very few ligands are capable of displacing C-bound cyanide under mild conditions. (25) The cyanide ligand has a negative charge and can act both as a a donor and n acceptor. (26' The carbon and the nitrogen ends of the cyanide ion act as Lewis bases that can coordinate two metal cations acting as the corresponding Lewis acids. The carbon end is considered a strong field ligand whereas the nitrogen end is a weak field ligand that can stabilize high-spin states. (21^ The ability of the cyanide ligand to bridge two different (M-CN- M') or similar (M-CN-M) metal centers makes the cyanometallate ion an excellent Chapter 1. Introduction 4 design element for coordination polymers. (21'2?)

The cyanometallate-based Fe4[Fe(CN)6]3-14H20 material, better known as Prus sian Blue, is perhaps the oldest known coordination polymer and occupies a spe cial place in the field of coordination polymers. It was initially prepared in 1704 ^28^ and used as a pigment for paints, inks and dye for several decades due to its dark blue color. The structure of Prussian Blue remained a mystery for several cen turies. More than two hundred years after the first report on the synthesis, powder (29) and single-crystal (30) X-ray diffraction experiments revealed its basic structural motif (Scheme 1.2A): a three-dimensional cubic array of Fe(ll)-CN-Fe(lll), with Fe centers occupying the corners and the cyanide groups lying on the edges. 4 Due to the ratio of Fe(n) to Fe(lll), 25 % of the [Fe(CN)6] ~ lattices sites are vacant, and water molecules are attached to the Fe(lll) centers instead (forming Fe(lll)-OH2 units). The remaining water molecules occupy the cavities in the cubic framework. Prussian Blue was found to ferromagnetically order at 5.6 K^31) and, in a drive to generate materials that magnetically order at higher temperatures, extensive stud ies on Prussian Blue analogues were conducted. (22>23>32~34) By exploiting the modular nature of coordination polymer synthesis, the two iron sites (Fe(n) and Fe(m)) of Prussian Blue have been systematically substituted by other transition metals. While the basic three-dimensional structural motif remained essentially unchanged for these metal-substituted polymers, the magnetic ordering temperature varies widely depend ing on the metal centers present. To date, KV(ll)[Cr(lll)(CN)6]-2H20 has the highest observed critical temperature at 376 K.(34) Theoretical studies have been performed to explain the observed magnetic behaviours^23,35) and predictions have been made that ordering up to 550 K could be achieved with specific pairs of metal centers. (33) Other modifications of the Prussian Blue system have yielded porous materials that show significant H^-uptake^36'37) or vapochromic behaviour. (38) A more substantial modular substitution is the alteration of the key Prussian 3 4- Blue building block, namely the octahedral cyanometallate [Fe(CN)6] / , by non- octahedral cyanometallates [M(CN)a;]n~. Cyanometallates can assume a number of different geometries depending on the nature of the metal cation (M), ranging from two-coordinate linear complexes such as [Au(CN)2]~ to eight-coordinate complexes Chapter 1. Introduction 5

Scheme 1.2: Different structural motifs observed with cyanometallate building blocks: A. Prussian Blue array; B. square-grid array; C. diamond-like array. The cyanide groups in B. and C. have been removed for clarity.

3 27 such as [Mo(CN)8] -.( ) 2 For example, four-coordinate square-planar [M(CN)4] ~ (M = Ni, Pd, Pt) building blocks were found to generate square-grid arrays (Scheme 1.2B) when reacted with other transition-metal cations (M').(19'39'40) On the other hand, the reaction of M' ions with tetrahedral [M(CN)4]2~ units (M = Zn, Cd) favours the formation of diamond like networks (Scheme 1.2C).(41)

General synthesis of coordination polymers

The synthesis of cyanometallate-based coordination polymers is usually straight forward. It often involves the simple mixing, at room temperature, of a solution of ra_ cyanometallate [M(CN);c] and a solution of coordinately unsaturated metal cation M'. The second metal M' must have at least two vacant coordination sites available to form N-cyanide linkages with the cyanometallate in order to permit an extended network to be formed. The second metal center M' can be pre-coordinated with a capping ligand, thereby exercising some control over the type of structure formed and the extent of M'-M' connectivity. Ligand-based functionality can also be introduced into the polymer. Chapter 1. Introduction 6

Thus, from the general synthesis shown in Equation 1.1 (which excludes the pos sibility of charged frameworks), it should be clear that alteration of (i) the metal centers, M and M', (ii) the preferred geometries of both metal centers and (hi) the capping ligands (L), provides the flexibility desired to assemble solids with tunable properties. (27>42~~46)

+ [M(CN)J"- + [LyMT - [LyM'UMiCNUrn (1.1)

Indeed, an understanding of the factors controlling the arrangement of the building blocks is a vital prerequisite for the rational design of systems with useful materials properties. As a result, much research has focused on probing this vast structural diversity, and some of the resulting physical (especially magnetic) properties. (21>27)

1.1.2 The use of [Au(CN)2]~ units as linear cyanometallate building blocks

Recently, linear cyanometallate building blocks such as [Au(CN)2]~^47^ and [Ag(CN)2]~(48^ have been used to synthesize a variety of coordination polymers. The networks that can be obtained using two-coordinate, linear [M(CN)2Jn~ depend greatly on the preferred geometry of the second metal center M' as well as on the presence of capping ligands in the coordination sphere of this cation. For exam ple, the structure of Zn[Au(CN)2J2 consists of a quartz-like network in which all the 49 [Au(CN)2]~ units bridge the tetrahedral Zn(ll) centers.( ^ When an octahedral metal 50 51 center is involved, cubic-type arrays of K{M[Au(CN)2]3} can also be obtained. ( > ) If capping ligands are utilized, a wide range of architectures can be obtained. For example, with a tetradentate capping ligand, only two sites are available on octahe dral M' and chains of alternating M'(ll) and [Au(CN)2]~ are formed, with additional unbound anions present, as in [Ni(tren)Au(CN)2][Au(CN)2].(52-1 The dimensionality of the polymers can be increased by decreasing the number of sites occupied by the capping ligands. Cu(pyrazine)[Au(CN)2]2 is an example of a 3-D coordination poly mer made of grid-shaped sheets of Cu[Au(CN)2]2 linked together by bridging pyrazine ligands. ^53^ Chapter 1. Introduction 7

Figure 1.1: A. Solid-state structure of [C4H8NH2][Au(CN)2], containing stacks 56 ) of [Au(CN)2]~ units held by aurophilic interactions. ^ - B. Structure of + Cu(en)2[Au(CN)2]2: chains of repeating [(en)2Cu-NC-Au-CN-] units (one is high lighted in gray) that interacts with the free [Au(CN)2]~ units via aurophilic interac tions. (57)

Aurophilic interactions as a tool to increase structural connectivity

Aggregations in the solid-state between closed-shell (5d10) Au(i) ions, leading to the formation of dimers and oligomers, have been frequently observed by X-ray 54 55 crystallography. ( - ) For example, the structure of [C4H8NH2][Au(CN)2], which is shown in Figure 1.1A, contains stacks of [Au(CN)2]~ units with Au-Au distances of 3.0795(4) A.(56) In 1988, Hubert Schmidbaur introduced the term "aurophilicity" to describe this phenomenon of Au-Au bonding. (58) Viable aurophilic interactions are considered to exist between Au(i) ions within a distance smaller than the sum of the van der Waals radii for gold, which corresponds to 3.6 A.^55'59^ Distances as short as 2.78 A have been observed in some complexes. (60) In addition to an increase in structural connectivity, compounds containing au rophilic interactions often show intense photoluminescence in the UV-vis region. (54) The strength of such interactions between two Au(l) ions was investigated by vari able temperature NMR spectroscopy. (59^ These studies have shown that the bonding strength lies between approximatively 20 and 50 kJ mol-1/55'59'61^ which is com parable to that observed for hydrogen bonds. ^62^ These interactions tend to govern Chapter 1. Introduction 8 the solid-state structure adopted, but dissociation into smaller oligomeric or distinct monomeric species seems to dominate in solution over the formation of Au(i)-Au(i) bound units. ^ Studies on the oligomerization through aurophilic interactions in solution, as a function of concentration, temperature and solvent, have been performed by Howard H. Patterson (63>64) and the critical aqueous concentration determined for

KAu(CN)2 was found to be 0.02 M at 300 K. Previous work in the Leznoff group has surveyed the potential of using Au(l)- containing building blocks to generate coordination polymers and using aurophilic 47 53 57 65 interactions as a tool to increase the network connectivity. ( > > > ) For example, as shown in Figure LIB, the structure of Cu(en)2[Au(CN)2]2 contains coordinate bonds + 57 that generate chains with repeating [(en)2Cu-NC-Au-CN-] units. ( ) The presence of aurophilic interactions between the chains and the free [Au(CN)2]~ units increases the connectivity in the system from 1-D to 2-D. Several other cyanometallate poly mers where Au(i)-Au(l) interactions play an important role in increasing the network connectivity can be found in the literature. (51>57>66)

1.2 Objectives of this research project

The general purpose of this thesis is to examine the preparation of capping ligand- free bimetallic [Au(CN)2]-based coordination polymers and investigate their solid- state structures as well as their physical properties. As physical properties are often a result of structural arrangement, an emphasis will be put on establishing relationships between the two. Amongst the previously reported capping ligand-free coordination polymers con taining [Au(CN)2]~ units are the luminescent Ln[Au(CN)2]3 • xH20 (Ln = La, Gd, 67 69 (49) Sm, Eu, Tb, Dy; x = 2 or 3)( " ) and Zn[Au(CN)2]2 systems and the piezo 5o:i electric KCo[Au(CN)2]3^ coordination polymer. However, no work was done on the magnetic properties of capping ligand-free coordination polymers containing only

[Au(CN)2]~ bridging units. The type and strength of magnetic interactions mediated via [Au(CN)2]~ units will be especially addressed throughout this thesis. Chapter 1. Introduction 9

The construction of open frameworks accessible to vapours and gases will be tar geted. Instead of using capping ligands, reactions will be carried out in donor solvents, that may be incorporated into the final product. The synthesis of frameworks containing metal centers bridged in three dimensions is also of interest as capping ligands usually prevent such a framework type from be ing obtained. High metal connectivity is generally required for cooperative magnetic behaviour to be present (13) and this will be investigated. In addition to metal connec tivity, the impacts of the framework structure on the observed magnetic properties will also be examined.

1.3 An introduction to molecular magnetism

1.3.1 Principles, definitions and fundamental equations

The magnetic behaviour observed in transition metal ions arises from the magnetic moments of unpaired electrons and the motion of electrons in orbitals. In the presence of an external magnetic field H, these magnetic moments can align with the field and generate a magnetization, M, along the direction of H.

McxH (1.2)

The total magnetization of a sample can be measured experimentally with a Super conducting Quantum Interference Device (SQUID) magnetometer, which is explained in more detail in section 1.4.4. The relationship in Equation 1.2 can be written in terms of the magnetic suscep tibility, X) such that: M M = XH or X = -jj • (1-3) The magnetic susceptibility, %, is thus an indication of the response of the magnetic moments of a sample to an external magnetic field. The total molar magnetic susceptibility8, for a sample can be expressed as a sum

M v molecular weight The total molar magnetic susceptibility is defined as XMtotai H sample mass Chapter 1. Introduction 10

Scheme 1.3: Orientation of the magnetic moments, in zero applied magnetic field, for (A.) a paramagnetic system, (B.) a ferromagnetic system and (C.) an antiferromagnetic system. Note that in a paramagnetic system, the magnetization vectors fluctuate freely in all orientations in space.

of a diamagnetic and a paramagnetic component:

XM total = XM + XM- I1-4)

The diamagnetic susceptibility, XM> *S a property of all matter and is due to the pres ence of paired electrons interacting with the magnetic field. It is temperature inde pendent and it has a negative value, usually on the order of -10~6 to -10~4 cm3 mol-1, depending on the molecular weight of the complex. On the other hand, the para magnetic susceptibility, XM> *S positive, usually temperature dependent and is due to the presence of unpaired electrons in the system. (70'71) In systems containing metal ions with unpaired electrons, the paramagnetic contribution is usually at least two orders of magnitude larger than XM- In coordination polymers, the paramagnetic susceptibility is the variable of interest as it reports on the magnetic moments present in a sample and how they interact with each other and with the applied magnetic field. It can be isolated by correcting for the diamagnetic contribution, x^, which is estimated from tabulated values. (71) Prom now on, XM wm be simply referred to as XM-

Paramagnetic systems

A system containing magnetic moments that are independent from each other and free to align with the magnetic field is defined as a paramagnetic system. Scheme 1.3A illustrates an example of a paramagnetic system in zero applied magnetic field. Chapter 1. Introduction 11

For a paramagnetic system, the magnetic susceptibility is inversely proportional to the temperature and follows the Curie expression:

XM = j, (1.5)

N g2n%S(S + l) C A 3kB in which NA is Avogadro's number, g is the Lande factor, \iB is the Bohr magneton -5 -1 -1 -1 _1 constant (4.6686 x 10 cm Oe ), kB is the Boltzmann constant (0.6950 cm K ) and S is the total spin value. (71) The parameter C is usually referred to as the Curie constant. A typical plot of the magnetic susceptibility as a function of temperature for a paramagnetic system is shown in Figure 1.2A, and is labelled as P. From Equation 1.5, it can be seen that for a paramagnetic system (P), a plot of XMT as a function of temperature should yield a constant value (C).

It is also very common to report the effective magnetic moment, fieff, which is related to the magnetic susceptibility by:

= 2.828^/XMT

where the unit of \xe^ is the Bohr magneton, \xB (in the cgsemu units system). A plot of the temperature dependence of fieff also gives a constant value for a paramagnetic system (Figure 1.2A, bottom). The effective magnetic moment of a spin-only para magnetic system can be estimated using the following equation:

n.eff = gy/S(S + l). (1.7)

To calculate the spin-only value, g is usually assumed to be equal to two. Figure 1.2B shows a typical isothermal plot of the magnetization of a sample as a function of applied magnetic field. The field dependency of the magnetization, at a given temperature, for a system of non-interacting magnetic moments, can be Chapter 1. Introduction 12

I AF U 'I p ^—— /AF i t i i i , ,i , 1 t i , 25 50 75 100 0 10 20 30 40 50 60 70 Field [ kOe ] Temperature [ K ]

Figure 1.2: Temperature dependence of the magnetic susceptibility (XM) and the ef fective magnetic moment (/J>eff) for a paramagnetic system (P, black), as well as for systems in which ferromagnetic (F, blue) or antiferromagnetic (AF, red) interactions are present. B. Field dependence of the magnetization (M) predicted for a para magnetic 5 = 1 system (P, black) at 1.8 K (g = 2.2) and compared to typical curves observed for ferromagnetic (F, blue) and antiferromagnetic systems (AF, red). described by the following equation: (70>71)

M = NAgnBS-Bs{y) (1.8) where Bs(y) is the Brillouin function, which is defined as:

25 + 1 (2S+1 B (y) = coth y (1.9) s 25 coth I ——25— y h (is gi^BS H y

The magnetization is predicted to reach a saturation value, Msaturation, which should correspond to an alignment of all the magnetic moments in the sample with the external magnetic field. Chapter 1. Introduction 13

Magnetic interactions

Deviations from the predicted values, either as a function of temperature (Equa tion 1.5) or field (Equation 1.8), could indicate the presence of magnetic interactions between the magnetic moments in the system.b As is illustrated in Scheme 1.3, the magnetic moments can spontaneously orient either parallel (Scheme 1.3B) or antiparallel (Scheme 1.3C) to each other. When the magnetic moments orient in a parallel fashion (Scheme 1.3B), these interactions are defined as ferromagnetic, whereas an antiparallel alignment is called antiferromagnetic (Scheme 1.3C). Antiferromagnetic interactions cause a smaller magnetization to be observed, com pared to the value for a paramagnetic system, whereas a larger magnetization is ob served when ferromagnetic interactions are present. Figure 1.2 shows the effects of ferromagnetic (F) and antiferromagnetic (AF) interactions on the magnetic suscepti bility (XM), effective magnetic moment (neg) and magnetization (M) determined for a sample as a function of temperature and field. To account for the interactions between the magnetic moments, the Weiss constant 9 can be added to the Curie expression (Equation 1.5) to obtain what is known as the Curie-Weiss equation: which contains the same parameters as Equation 1.5 with the exception of 9.{71) The Weiss constant is usually used to describe very weak interactions and does not relate in any way to the structural arrangement of the metal centers with respect to each other.

Modelling magnetic interactions

In order to predict the magnetization (and susceptibility) of a sample at a finite temperature, the different energy levels and their respective populations have to be

bDeviations can also result from single-ion effects and some of them are discussed in more detail in section 1.3.2. Chapter 1. Introduction 14

m„ B.

^Total * o -/ -l -J-g/^H

0 H- Ku = ° H

Scheme 1.4: A. Energy diagram for an S = 1 system, showing the effect of Zeeman splitting in the presence of an applied magnetic field. B. Energy diagram for a dimer of S = \ centers, with Sxotai = 0 or Sxotai = 1, in the presence of a magnetic field.

known. The magnetic moment of an ion in a state n is given by

dEn Vn = dH ;i.n) and the overall magnetization of a sample is equal to

M = NA^2iinPn (1.12)

where Pn is the Boltzmann factor, which gives the probability that a state with energy 0 En is occupied at a given temperature. In the simplest case, in the absence of a magnetic field and any type of interactions, the different ms states of a metal ion are degenerate in energy. As illustrated in Scheme 1.4A, upon application of a magnetic field H, the degeneracy of the different energy levels is removed. The energy of each state is defined by:

^Zeeman — ~g^B S • H (1.13)

Ezeeman = msgn,BH (1.14)

This phenomenon is known as Zeeman splitting (70) and describes the behaviour observed for paramagnetic systems.

c 1 The Boltzmann factor is denned as: Pn = exp[—^n/(fcsjf )]/J]TOexp[—£m/(fcsT)]. Chapter 1. Introduction 15

A general equation to evaluate the magnetization of a sample by taking into ac count each differently populated level was developed by Van Vleck. ^ Most magnetic models use this equation to predict the magnetization and magnetic susceptibility of a sample once the energy of each level is known. When the connectivity between the metal centers is known, it is possible to model the magnetic behaviour and make correlations between the observed magnetic prop erties and the type and strength of the interactions present between the metal centers. In the case of a dimer containing two metal centers (SA and SB) interacting with each other, there are two possible configurations: ferromagnetic alignment (Sxotai — SA + SB) or antiferromagnetic alignment {Sxotai — SA~ SB)- Scheme 1.4B illustrates the energy diagram for such a system. The Hamiltonian representing the interactions between the two metal centers is called the Heisenberg Hamiltonian and can be written as:(7°)

li-Heisenberg = ~~ J ^A " ^B (•'•••'•") and has the following eigenvalues

^Heisenberg = —^~ST(ST + 1) (1-16)

This splits the two energy levels in the absence of a magnetic field by — J. The parameter J has a negative value if antiferromagnetic interactions are present (ground state is ST = 0) and positive if the ground state is ferromagnetic (ST = 1)- When a magnetic field is present, Zeeman splitting further removes the degeneracy of the ms states. Several models have been developed to evaluate the strength of the magnetic interactions, J, between metal centers in different structural arrangements. Models are known for chains, layered compounds and even 3-D connected systems. In most cases, the magnetic exchange is considered to be isotropic between each pair of metal centers and one J value is obtained when fitting the data to the appropriate model. The magnetic models used in this thesis will be described in the appropriate sections. Chapter 1. Introduction 16

Time scale effects

The type of magnetic behaviour observed depends greatly on the time scale of the measurements and of the spin dynamics. For example, if the time scale for a mea surement is very short, the spins may appear to be static whereas on a longer time scale the same spins could appear to be dynamic. This will be addressed in more detail later in this thesis when comparing the magnetic properties determined using techniques with different measurement time scales, such as SQUID magnetometry, muon spin relaxation and Mossbauer spectroscopy.

1.3.2 Single-ion effects

The magnetic properties observed for certain metal ions differ from Curie behaviour, even though no magnetic interactions are involved between the metal centers. This is due to single-ion effects which are intrinsic to some metal ions. The principal single- ion effects observed with the metal ions used in this thesis are zero-field splitting and spin-orbit coupling.

Zero-field splitting

As shown in Scheme 1.5, axial crystal field effects in octahedral Ni(n) ions are known to remove the degeneracy of the 3A2 ground state, in the absence of a magnetic field, by splitting the ms = 0 from the m^ = ±1 states. (70'71) The splitting energy is defined as D and is usually on the order of a few wavenumbers for octahedral Ni(li) ions. This phenomenon is known as zero-field splitting and can generally be observed in any system with a total spin quantum number larger than |, such as high-spin octahedral Fe(ii) and Mn(ii) ions. This uniaxial distortion results in an anisotropy of the magnetic properties. The field dependence of the energy levels is not the same whether the distortion axis is parallel or perpendicular to the orientation of the magnetic field (note that Scheme 1.5 only shows the dependence when the field is parallel). Parallel (x\\) and perpendicular (x±) components of the magnetic susceptibility must then be defined. Chapter 1. Introduction 17

D + gzMBHz D -gz^H

H = 0 increasing H • z c? z Scheme 1.5: Zero-field splitting energy diagram of an S = 1 ion with the relative energy of each state (assuming no other type of interactions).

When the energy of each level is incorporated into the van Vleck equation, the following equation for the zero-field splitting of isolated S = 1 ions is obtained:

X|| +2x± Xzfs = ^^ (1-17) 2 2NAg fi% XII kBT \ 1 + 2ev 2 2NAg n% / l-e» x± D I 1 + 2e» -D

kBT where D is the zero-field splitting parameter. (7°) In this equation, only one g-Lande factor is taken into account (g = g\\ ~ g±). Compared to the expected value for a paramagnetic system, zero-field splitting causes a decrease in susceptibility and effective magnetic moment at low tempera tures (depending on the magnitude of D/ksT). The temperature dependence of the magnetic moment for a zero-field split system would be similar to that observed for a system with very weak antiferromagnetic interactions. The value of D can be es timated using Equations 1.17 if no interactions are assumed to be present between the metal centers. For highly symmetrical Ni(il) complexes, D has a very small value (~ 0-1 cm-1) whereas larger values (> 4 cm-1) are observed for asymmetric Ni(ll) complexes. (71^ Chapter 1. Introduction 18

Spin-orbit coupling in octahedral Co (II) ions

Spin-orbit coupling arises from the interactions of the electron spin magnetic moment s with the orbital momentum I, which is generated by the orbital motion of the electrons. When spin-orbit interactions are weaker than Coulombic interactions,d the electron spins interact among themselves to form the total spin angular moment S, while the orbital momenta combine into a total orbital angular momentum L. These two momenta couple together to form the total angular momentum J. This phenomenon is known as LS-coupling or Russell-Saunders coupling. (73)

J = L + S (1.18) where L = YJ ^ and S = Vj Sj i % The Hamiltonian describing the spin-orbit coupling can be written as:

fts.o.c. = AL-S (1.19) with the eigenvalues:

EL,s,j = ^ [J(J + 1) - L(L + 1) - S(S + 1)] (1.20) where J can take the values from L — S to L + S. In the above equations, A is the spin-orbit coupling parameter, which is negative if the electronic shell is more than half full.6 For octahedral Co(ll) ions, A is on the order of 170 cm"M70' For octahedral Co(ll) ions, spin-orbit coupling results in a much larger value for the magnetic susceptibility and the effective magnetic moment. Values of 4.7 to 5.2 [is are usually observed at room temperature for compounds containing octahedral Co(n) ions compared to the predicted spin-only value of 3.89 [XB- AS the temperature is lowered, depopulation of the higher energy levels occurs and a continuous drop in effective moment over a large temperature range is observed.

dOn the other hand, when spin-orbit coupling dominates (such as in heavier atoms), the jj-coupling scheme must be used. eThe parameter ( is also used to define spin-orbit coupling and is related to A by: A = ±(£/2S), the positive sign being for a less than half full electronic shell and the negative sign for a more than half full shell. Chapter 1. Introduction 19

1.4 Characterization Tools

1.4.1 Determination of chemical composition

The chemical composition of coordination polymers can be determined using standard carbon, hydrogen and nitrogen (CHN) elemental analysis. The results are compared to the predicted percentages for each element based on the proposed chemical formula. Elemental analysis for specific metal ions could also be performed, but this technique was not used in this thesis. Thermogravimetric analysis is usually performed to determine the thermal stabil ity of a coordination polymer and to confirm the chemical composition of the sample. From the relative weight losses observed at different temperatures, it is possible to identify the lost fragments based on the chemical composition.

1.4.2 Information obtained by spectroscopy

Infrared spectroscopy

FT-IR spectroscopy is an invaluable tool to study cyanometallate coordination poly mers. The frequency corresponding to the vibration of a cyanide group (VCN), which usually ranges between 2000 and 2200 cm-1, is very sensitive to the environment of this cyanide group/21) Bonding of the cyanide ligand (through the N atom) to a metal ion can either be primarily a CN—>M' a interaction or a M'—>NC n interaction. In the case of CN—>M' a bonding, the cyanide molecular orbital involved is antibonding with respect to the cyanide triple bond. Removal of electron density from that orbital stabilizes the cyanide group. As a consequence, upon coordination to M', the force constant increases and the cyanide vibration frequency shift to higher energy. (21>74) When considerable 7r-back bonding occurs, from the d orbitals of the metal ion to the antibonding 2p7r* orbital of the cyanide group, the C=N bond is weakened and the force constant of the vibration decreases. Such situation is usually observed in coordination polymers containing electron rich metal centers in a low oxidation state, (75) such as Pb(n) in Pb(H20)[Au(CN)2]2. Chapter 1. Introduction 20

\. detector B. light source y to the detector

Mrom the light source N\ sample

Scheme 1.6: A. Experimental set-up used to determine the solid-state UV-vis-NIR absorbance and reflection spectra; B. Cross-section showing the relative position of the optical fibers near the sample.

Infrared spectroscopy can also be used to confirm the presence of the building blocks in the coordination polymer, such as capping ligands and solvent molecules, and also to shed light on the structural arrangement of the cyanide ligands (terminal M-CN vs bridging M-CN-M' groups). Small chemical and structural changes in the M' coordination sphere can cause an additional shift in vibration frequency which can be used to discriminate between different compounds.

Solid-state UV-vis-NIR spectroscopy

Solid-state UV-vis-NIR spectroscopy can be performed to assess the color of the coor dination polymer and investigate the coordination number, geometry and ligand field of the transition metal ions. The experimental set-up used to determine the absorbance and reflection spectra of a solid-state sample is shown in Scheme 1.6. The sample is irradiated with a broad spectrum light source (combination of tungsten and deuterium lamps) using an optical fiber and the light reflected/emitted from the sample surface is collected by a second set of optical fibers positioned around the first optical fiber (Scheme 1.6B.). The incident angle is not chosen as 90° to avoid direct reflection of the incident light beam and saturation of the detector. In this type of measurement, the light collected at the detector is a combination of light emitted by the sample and light reflected by the surface. Reflection can also be Chapter 1. Introduction 21 specular (the angle of reflection is the same as the angle of incidence) and/or diffuse (the angle of reflection is not equal to the angle of incidence). As the surface is not perfectly flat, it is possible to detect both type of reflections with our experimental set-up. The reflection of the solid-state sample at a specific wavelength, %R\, is calculated as:(76) S X x 100 %Rx = ( R XZ^) % (1-21) where S\ is the sample intensity, D\ is the dark intensity and R\ is the reference intensity, all determined at the same wavelength A. A non-luminescent white powder (e.g. magnesium oxide) is usually used as the reference sample for these measurements. The dark intensity corresponds to the intensity measured by the detector when the shutter at the light source is closed. The absorbance of the solid-state sample at a specific wavelength, A\, is then calculated as: (76) ^-"""•(frt;) (122) Several assumptions are made when making such measurements. It is assumed that the surface of the sample is similar to that of the reference sample and light is scattered back to the detector in the same way. For the same experimental geometry, a rougher sample surface would however scatter a different amount of light back to the detector and the intensity recorded would be altered compared to the reference sample. Also, as the sample is irradiated with a broad spectrum light source and the detector collects all wavelengths simultaneously, a luminescent sample could show %RA values larger than 100 % in the region where the sample emits. As a consequence, the relative absorbance or reflection values as a function of wavelength can be obtained but the exact value cannot be determined.

Mossbauer spectroscopy

Coordination polymers containing Fe ions can also be investigated using 57Fe Mossbauer spectroscopy. Mossbauer spectroscopy can give information on the Chapter 1. Introduction 22 oxidation state and the spin configuration of 57Fe ions, as well as magnetic ordering occurring in the sample. ^ The information obtained with this technique comple ments the information obtained by crystallography and SQUID magnetometry. In contrast to other spectroscopic methods used to characterize coordination poly mers, 57Fe Mossbauer spectroscopy probes a nuclear transition from the I — \ state to the / = | state of the Fe ions. This transition can be induced by using 7-rays of the appropriate energy. In order to obtain the proper excitation energy for 57Fe, a radioactive 57Co source is used. The 57Co atoms decay to an excited state of 57Fe (J = |), which then relaxes and emits the 7-rays used to probe the sample. This yields a source of photons (14.4 keV) with a very narrow energy distribution. If the sample and the source contain 57Fe nuclei in an identical electronic environ ment, the emitted 7-rays will have exactly the right energy to excite the sample. The energy of the different nuclear states depends on the distribution of electron density around the nucleus. Modification of the chemical environment or oxidation state af fects the electron distribution and, as a consequence, the nuclear transition energy probed by Mossbauer spectroscopy. If there is a difference in nuclei environment be tween the source and the sample, the emitted 7-rays will not have the appropriate energy to be absorbed by the sample. The Doppler effect is used to tune the excitation energy. The source is accelerated linearly toward and away from the sample and the relative motion causes a shift in energy received by the sample. To probe all possible transitions in the sample, the velocity of the source is oscillated between set values (which usually ranges between -12 and +12 mm s_1, or smaller) and the relative transmission is plotted as a function of velocity to create a Mdssbauer spectrum. Scheme 1.7 illustrates the different types of Mossbauer spectra for a non- magnetically ordered sample. In the case where the nuclei environment is the same for the source and the sample, no motion is required between the two and absorption occurs at a velocity of 0 mm s"1 (Scheme 1.7A). However, when the environment differs between the nuclei in the source and the sample, the maximum absorption (minimum transmission) occurs when the source is moving (Scheme 1.7B). The shift in velocity is known as the isomer shift or chemical shift, 5, which is directly related to the difference in energy (AE). Chapter 1. Introduction 23

-4-20246 -4-20246 -4-20246 Velocity [ mm s ] Velocity' [ mm s ] Velocity [ nun s ]

Scheme 1.7: Nuclear transitions probed by 57Fe Mossbauer spectroscopy and associ ated Mossbauer spectrum. A. 57Fe atoms in the sample have the same nuclear state energies as those in the source (S = 0). B. The transition energy between the two states in the sample differs from that of the source and a non-zero isomer shift, S, is observed in the Mossbauer spectrum. C. Quadrupole splitting, AEQ, resulting from the splitting of the mj = ±| and mi = ±| energy levels.

From the isomer shift, it is possible to determine the oxidation state of the Fe ion in the sample. For example, the isomer shift observed for high spin Fe complexes goes from ~ 2 mm s_1 for ions in the +1 oxidation state to ~ -0.8 mm s_1 for ions in the +6 oxidation state. (78) Another feature that can be observed in a Mossbauer spectrum results from the fact that a nucleus with I > 1 can be considered as an electric quadrupole. This quadrupolar nucleus interacts with the electric field gradient created by the surround ing asymmetric electron distribution. For 57Fe atoms, this interaction results in the splitting of the m/ = ±| and m/ = ±| levels of the excited state (Scheme 1.7C). The difference in energy causes two absorption bands to be observed in the Mossbauer spectrum, with a velocity difference, referred to as the quadrupole splitting, of AEQ. Chapter 1. Introduction 24

Table 1.1: Typical isomer shift (5) and quadrupole splitting (A£7Q) observed for low spin and high spin six coordinate Fe(n) and Fe(m) complexes. (77>78)

Fe(ll) Fe(lll)

Spin state 5 AEQ 5 AEQ mm s_1 mm s_1 mm s_1 mm s_1 High spin 0.6-1.7 2-4 0.2 - 0.5 0.1 - 0.8 Low spin -0.3 - 0.5 0-1 -0.2-0.2 1-2

For a perfectly symmetrical electron cloud surrounding the nucleus, no quadrupole splitting is observed whereas for a very asymmetric electron distribution, a very large quadrupole splitting (up to ~ 4 mm s_1) is observed. When quadrupole splitting is present, the isomer shift is denned as the position of the middle point between the two absorption bands. The spin configuration (low vs high spin) is obtained by looking at the quadrupole splitting. Typical values observed for Fe(ll) and Fe(ni) complexes in either the low or high spin state are reported in Table 1.1. From the quadrupole splitting, it is also possible to learn about the symmetry around the metal center. Additional information can be obtained by Mossbauer spectroscopy about the magnetic order in a sample due to further splitting of the mi energy levels. This aspect is however neglected here as none of the Fe-containing samples studied in this thesis magnetically order over the temperature range studied.

1.4.3 Investigation of structural arrangement using crystallo graphy

The basic principles of X-ray diffraction by a crystalline sample rely on Bragg's law, which states that:

nA = 24Hsin0 (1.23) Chapter 1. Introduction 25 where n is the diffraction order, A is the radiation wavelength, 9 is the angle between the crystal planes and the incident beam and dhu is the distance between the crystallographic planes with hkl indices. In an X-ray diffraction experiment, the intensity of diffracted light is measured as a function of 6, keeping A at a fixed value which depends on the source (n is equal to 1). The intensity of a diffraction peak depends in turn on the scattering factor of the atoms present in the specific planes as well as on the occurrence of constructive or destructive interference. For more detailed information on X-ray crystallography, the reader is directed to the book by Warren. ^79^ Two different types of experiments can be performed with coordination polymers, depending if the sample is a microcrystalline powder or if crystals with sub-millimeter dimensions can be grown.

Single crystal X-ray diffraction

Single crystal X-ray diffraction can be used to determine the structural arrangement of the building blocks. One of the major problems to solving the crystal structure of a coordination polymer is the obtention of single crystals of sufficiently high quality. To favour the formation of suitable crystals, different crystallization techniques can be used: change in concentration, modification of crystallization solvent or use of solvent mixtures, modification of counterions, etc. Diffusion of the reagents through gels or H-shaped tubes can also be used to slow down the crystal growth. Solvothermal recrystallization is another technique employed to crystallize coordination polymers. This technique relies on the use of heat to solubilize an otherwise insoluble material and then very slow cooling allows control of the polymer growth. This method will be discussed in more detail in Chapter 5.

Powder X-ray diffraction

As opposed to single-crystal diffraction, powder diffraction relies on the assumption of a randomly oriented sample, such that all angles of incidence are present at once. X-ray diffraction with powdered samples is commonly used to verify phase purity. Chapter 1. Introduction 26

Diffractograms of powdered sample are usually plotted as a function of 29 angle for a specific wavelength, or as a function of dhki distances/ A crystalline impurity mixed with the desired coordination polymer can be detected by the presence of additional peaks in the diffractogram. It is also possible to distinguish between polymorphic coordination polymers, which are polymers with the same chemical composition but with different structural arrangement (see section 3.2). It is also possible, but more difficult, to determine the crystal structure of a co ordination polymer from its powder X-ray diffractogram by using the Rietveld anal ysis method/80^ Such a technique is becoming more accessible with the increase in computing power of personal computers. The technique requires very high quality data, especially for structures containing a large number of atoms per unit cell, and synchrotron data is usually used. In the case of coordination polymers, structure determination is a little bit easier as building blocks can only adopt a limited number of orientations relative to each other. The problem is simplified by only considering the position and orientation of these large units instead of the individual position of every constituent atom.

1.4.4 Investigation of magnetic properties by SQUID magnetometry

Principles

SQUID magnetometry measurements are based on the principle that, when a magne tized sample is moved through a coil, an electric current is generated in this coil. To perform the experiment, the sample is mounted at the end of a rod, and moved through a pick up coil. Prom the electric current generated as a function of the position of the sample with respect to the coil, the magnetization of the sample can be determined and the magnetic susceptibility calculated. The axis along which the

fPlotting as a function of dhki distance is advantageous as this value does not depend on the radiation source, which make comparison easier between diffractograms collected on different instru ments. Also, this value represents real distances between atomic planes in the crystal structure, which is more meaningful to the understanding of the structural arrangement. Chapter 1. Introduction 27 sample is moved and the external magnetic field is applied is defined as the z-axis.

Different measurements

For a coordination polymer containing metal centers with unpaired electrons, the magnetic susceptibility is usually determined as a function of temperature in a fixed applied magnetic field. Using a standard SQUID magnetometer, it is generally pos sible to vary the temperature between 1.8 and 400 K. The temperature dependent magnetic behaviour observed for a specific coordination polymer can then be related to the type of magnetic phenomena present in the sample (as explained in section 1.3; see Figure 1.2A for some examples). Measurements can also be carried out as a function of applied magnetic field at a given temperature. Most measurements are carried out in a constant magnetic field (as shown in Figure 1.2B), referred to clS £L dc field. The magnetic field along the z-axis, during one measurement, can also oscillate between two set values (±x Oe), at a fixed frequency. An alternating magnetic field is referred to as an ac field. When a measurement is performed with an ac field, the sample is kept fixed in the center of the pickup coil, while the orientation of the field changes. The maximum amplitude of the applied field is much smaller than in dc measurements and is usually smaller than 5 Oe. This type of experiment looks at the dynamics of the magnetic moments. In a paramagnetic system, the spins are dynamic and can follow the alternating orientation of the applied magnetic field. However, magnetic interactions in a sample can prevent fast reorientation of the magnetic moments and a lag in magnetization will be observed. When using an ac magnetic field, the ac-susceptibility of the sample, %', is ob tained. In a dc experiment, the magnetic susceptibility is defined as the slope of the magnetization curve in the low to moderate field range (Figure 1.2B, Equation 1.3). Similarly, in an ac experiment, the ac-susceptibility (x') is defined as the instantaneous slope:

The ac-susceptibility (x')> also called the in-phase component, is related to the dc Chapter 1. Introduction 28

susceptibility through the following equation:

x' = XCOS(f (1.25) where

(1.26)

In a paramagnetic system, no phase shift is present and the out-of-phase component (x") is equal to zero. Ac susceptibility measurements are usually performed as a function of temperature upon cooling. Phase transitions from a paramagnetic state are usually characterized by the onset of a non-zero out-of-phase component near the transition. The frequency of the oscillating field can also be varied (usually between 0 and 1500 Hz) and the frequency dependence of the two components can be determined.

1.4.5 Investigation of magnetic properties using muon spin relaxation

The muon spin relaxation or /iSR technique can be used to investigate magnetic ordering in a sample.^81'82) The fundamental principles of /xSR are similar to those of more traditional magnetic resonance techniques such as NMR and ESR. In contrast to bulk magnetic susceptibility determined by SQUID magnetometry, /xSR is a local magnetic probe that is highly sensitive to weak internal magnetic fields, short-range magnetic order and disordered magnetism. Furthermore, /xSR is sensitive to spin fluctuation rates in the range 104-1012 Hz, which is beyond the range accessible with ac-susceptibility magnetometry. Recently, /LtSR has been applied to the study of molecule-based low dimensional magnets, (83~85) including transition metal- dicyanamide polymers(86) and layered organic and inorganic hybrid systems.<87) Chapter 1. Introduction 29

Principles of the /xSR technique

The /iSR method involves the implantation of muons into the sample. Muons are short-lived subatomic particles, with a mean lifetime (rM) of 2.197 fis, which possess a spin S = \. Muons can be negatively or positively charged, but for the purpose of /^SR experiments, positively charged muons are chosen. Due to their charge, muons will stop preferentially in certain interstitial sites in the crystal lattice to minimize their potential energy. More specifically, positively charged muons, fi+, are repelled by nuclei but attracted by electron clouds. Positively charged muons are thus best suited to relay information about the electronic properties. The number of different preferred muon stopping sites in the crystal lattice is usually small, especially in crystal systems of high symmetry. Once implanted into a sample, each muon can be used as a very sensitive probe of magnetism as it couples to the local magnetic field (Biocai) via its spin. In the presence of a local magnetic field, the muon's spin will start precessing (at a frequency Pfj) around the direction of the field, changing its overall polarization along its initial axis. The precessing frequency of the muon's spin is given by:

Vfi = -^-B local (1-27) where 7^ is the muon's gyromagnetic ratio (0.0852 /us-1 G_1). The time evolution of the muon spin polarization reflects the local magnetic field at the muon stopping site(s).

Positively charged muons spontaneously decay into two neutrinos (ue, V^) and a positron (e+). + + /x —>• e + ue + v^ As will be explained below, the //SR technique relies on the fact that the positron is primarily emitted in the direction of the muon's spin at the time of the decay event.

Experimental setup of the /xSR technique

Scheme 1.8A shows a schematic of a zero-field /iSR (ZF-//SR) set up. A polycrystalline sample is thermally anchored to the cold finger of a dilution refrigerator through a Chapter 1. Introduction 30 silver backing plate. Silver is used since it is diamagnetic and does not modify the direction of the muon spin polarization. Each muon is produced from the decay of a pion. Due to parity violation and conservation of momentum and spin, the muon spin is oriented antiparallel to the muon momentum. (81'82) In a ZF-//SR experiment, the travelling direction of the muon beam is defined as the z-axis and the muon beam is considered nearly 100 % spin- polarized (Pz(0)) in the opposite direction (—z-axis). A counter placed upstream of the sample is used to detect an incoming positive muon (/i+) and establish time t = 0 at the electronic clock. The muon then pene trates the sample and stops at one energetically favoured site in the crystal lattice (Scheme 1.8B). To prevent muon-muon interactions, only one muon is implanted at a time in the sample. If a second incoming muon is detected by the counter before the decay of the previous muon, the event is discarded. After interacting with the local environment (Scheme 1.8C), the muon decays and emits a positron (e+) in the direc tion of its spin (Scheme 1.8D). The forward (F) or backward (B) detector senses the decay positron at time t, which is determined by the electronic clock. Approximately 107 decay events are averaged to obtain a time-differential histogram, N(t). For each positron detector (B,F) this is described by:

F T NB,F{t) = N*' e-V » [1 ± aBiFPz(t)] + BB,F (1.28)

r Here, N^ is a normalization constant, rM is the mean muon lifetime, aB,F is the intrinsic asymmetry of each positron detector, and BBF is a time-independent back ground. After subtracting the experimentally determined random backgrounds BB,F from Equation 1.28, the /xSR "asymmetry" spectrum is formed as follows:

N A{t) = aoPi(t) = j(t) - NF(t) w zK v ; ° ' NB(t) + NF(t) where aB — aF = ao- Note that this isolates the desired information, which is the time evolution of the muon spin polarization, Pz(t), and eliminates the exponential decay due to the muon's lifetime (e~t/T>'). If the sample is diamagnetic, the polarization of the muon spin will not be af fected and the asymmetry function will be time independent in zero field. On the Chapter 1. Introduction 31

Sample holder Ag backing plate

+ Backward e + detector Forward e detector jU+ detector

Electronic Clock STOP time = t

local local local magnetic *v environment in sample *

Scheme 1.8: A. Schematic of the experimental setup used to carry out the ZF-/^SR measurements. The incoming muon beam has a spin polarization antiparallel to the beam direction (2-axis). See text for more details. B-D. Schematic of a decay event after interaction with the sample: B. a muon is inserted into one crystalline domain of the sample having a local magnetic field (Biocai) along a specific direction; C. the muon's spin starts to precess around B;oca/; D. decay event emitting two neutrinos {ve,ni blue) and a positron (e+, red) in the muon's spin direction. Chapter 1. Introduction 32

A. i , i ii | , i -r B. Wo I JrxNF(t) ^ I | 3

10 4 6 time [ JJS ] time [ /«]

Figure 1.3: A. Number of positrons (e+) detected in the backward (-) and forward (•-) detectors as a function of time for a sample with a local magnetic field transverse to the initial spin polarization. B. Asymmetry spectrum determined using equation 1.29.

other hand, if the magnetic field associated with an unpaired electron is present, the polarization will change as a function of time according to the strength and orienta tion of the magnetic field. For example, Figure 1.3 shows what would be observed for a theoretical sample containing only one magnetically ordered domain and pro ducing a magnetic field transverse to the initial muon spin polarization. The number of positrons detected by the two detectors at different times, N(t), is illustrated in Figure 1.3A and the corresponding asymmetry spectrum is shown in Figure 1.3B. However, the samples investigated are often composed of randomly oriented polycrys- tals and more complicated spectra, resulting from the different domain orientations, are usually observed. Other types of magnetic states can also be investigated and the results observed for these will be explained in more detail in Chapter 2. Similar experiments can also be done while applying a magnetic field along the ini tial muon spin polarization axis (—z-axis). This technique is referred to as longitudinal field //SR or LF-y^SR. The information obtained from an LF experiment differs from that obtained by ZF-//SR experiments. Using LF-/xSR on coordination polymers, information on the spin dynamics can be obtained. Chapter 2

M(/i-OH2)2[Au(CN)2]2: Magnetic exchange through unusual double aqua-bridgesa

2.1 Introduction and research objectives

A small number of capping ligand-free dicyanoaurate-based coordination polymers has been reported in the past. For example, pseudocubic Prussian Blue type (51) 50 structures of KM[Au(CN)2]3 (M = Fe, Co( )), interpenetrated 3-D quartz 88 ( 49 like arrays of Co[Au(CN)2]2( ) and Zn[Au(CN)2]2, - ^ and square-planar sheets 51 of Mn[Au(CN)2]2-2H20,( ) have been synthesized. A series of 3-D arrays of

aPart of the work presented in this chapter is also reported in: J. Lefebvre, F. Callaghan, M. J. Katz, J. E. Sonier, D. B. Leznoff, "A new basic motif in cyanometallate coordination polymers: structure and magnetic behaviour of M(/i-OH2)2[Au(CN)2]2 (M = Cu, Ni)", Chem. Eur. J., vol. 12, pp. 6748-6761, 2006.

33 Chapter 2. Magnetic exchange through double aqua-bridges 34

67 69 Ln[Au(CN)2]3-2,3H20 (Ln = La, Gd, Sm, Eu, Tb, Dy) have also been prepared, t " ) Similar to the situation for Prussian Blue and its many analogues, despite an aqueous synthesis, these [Au(CN)2]-based products either do not retain water at all, or the water molecules bind to the metal cation without significantly impacting the basic structural motif. In this chapter, the synthesis of different capping ligand-free [Au(CN)2]-based coordination polymers with Cu(ll), Ni(li), Co(n), Fe(ll), Fe(lll) and Mn(n) tran sition metal ions was attempted. A series of coordination polymers of the form M(//-OH2)2[Au(CN)2]2 will be presented along with the related Fe(/i-OH2)(/i-

OH)[Au(CN)2]2 polymer. As opposed to the previously reported polymers, these compounds have a very unusual basic structural motif in which the water molecules play a key structural role; the structure type is unique with respect to cyanometallate-based polymers and, more broadly, is rare in aqueous coordination chemistry. Because the structural motif often defines the physical properties (e.g., magnetism) of a particular compound, the magnetic behaviour of the M(/z-OH2)2[Au(CN)2]2 poly mers has been investigated by direct current (dc) and alternating current (ac) SQUID magnetometry. To obtain further information on the magnetic properties of these systems, particularly in zero-applied field, the muon spin-relaxation (/xSR) technique has also been applied (an introduction of this technique is given in section 1.4.5). '81)

2.2 An uncommon structural motif: double aqua- bridges

2.2.1 Preparation and identification of products

The addition of an aqueous solution of K[Au(CN)2] to an aqueous solution of

Cu(C104)2-6H20 afforded an immediate green precipitate in high yield. The rela tive amounts of carbon, hydrogen and nitrogen determined by elemental analysis for this product are consistent with the empirical formula being Cu(H20)2[Au(CN)2]2. To Chapter 2. Magnetic exchange through double aqua-bridges 35

\ Cu(n) k *>— A Ni(n) kL Co(n) k Fe(n) ,L Mn(n) f V Fe(m) . 1 . . . . 1 . i . . . . 2500 2250 2000 1750 Wavenumber [ cm" ]

Figure 2.1: Comparison between the FT-IR spectra of each M(//-OH2)2[Au(CN)2]2 complex (M(n) = Cu, Ni, Co, Fe, Mn) and Fe(/x-OH2)(/z-OH)[Au(CN)2]2 (Fe(m)). reflect the structural arrangement discussed below, this complex will later be referred to as Cu(/x-OH2)2[Au(CN)2]2 and so forth for the other analogous complexes.

The similar reaction between K[Au(CN)2] and Ni(N03)2-6H20 also yielded an im mediate green precipitate. Results from elemental analysis suggested that the com position of this product was consistent with Ni(/x-OH2)2[Au(CN)2]2. The FT-IR spectra of both products show similar features (Figure 2.1 and Ta ble 2.1). Bands corresponding to cyanide vibrations (VCN) are observed at 2217 (s), 2194 (vw) and 2171 (s) cm"1 for the Cu-containing product and 2214 (s), 2204 (sh), 2170 (s) cm-1 for the Ni-containing product.

The similar reaction between K[Au(CN)2] and Co(ll), Fe(ll) and Mn(ll) salts did not yield analogous products. However, when the reaction of n [ Bu4N][Au(CN)2]-0.5H2O with a series of M(C104)2-6H20 salts in either acetonitrile (M = Co, Mn) or an acetonitrile/water mixture (M = Fe) was performed, imme diate precipitates (pink for M = Co, white for M = Mn and pale yellow for M = Fe) were obtained. The general formula of each product was also determined to be Chapter 2. Magnetic exchange through double aqua-bridges 36

Table 2.1: VCN vibration frequency for each M(/x-OH2)2[Au(CN)2]2 complex (M = Cu,

Ni, Co, Fe, Mn) and Fe(//-OH2)(^-OH)[Au(CN)2]2 (Fe(m)).

~M VCN vibration frequency* cm-1 Cu(n) 2217(s) 2194(vw) 2171(s) Ni(n) 2214(s) 2204(sh) 2170(B) Co(n) 2204(s) - 2168(B) Fe(ll) 2196(B) - 2166(B) Mn(ll) 2196(m) 2161(sh) 2158(B) Fe(m)* 2184(B) - 2168(sh) t Band intensity: (s): strong, (m): medium, (sh): shoulder, (vw): very weak.

t Fe(/x-OH2)(/x-OH)[Au(CN)2]2

M(/i-OH2)2[Au(CN)2]2 from the results obtained by elemental analysis.

Figure 2.1 shows the FT-IR spectra of all M(//-OH2)2[Au(CN)2]2 complexes in the cyanide vibration region. Each product was found to have two main bands at tributable to cyanide vibration at ~ 2200 and ~ 2170 cm-1 (Table 2.1).

The VQN vibration frequencies of all M(/>OH2)2[Au(CN)2]2 complexes are shifted -1 toward higher energy relative to the 2141 cm stretching frequency of K[Au(CN)2]. This suggests that the cyanide groups are iV-bound to the transition metal centers or hydrogen bound to the water molecules. (21)

Mossbauer spectroscopy

To confirm the oxidation state and assess the spin configuration of the Fe centers in the Fe-containing complex, Mossbauer spectroscopy was performed. Figure 2.2 shows the

Mossbauer spectrum of Fe(/x-OH2)2[Au(CN)2]2 collected at 4.5 K. A large quadrupole pair is observed, with an isomer shift (5) of 1.25(2) mm s_1 and a quadrupole splitting (AEQ) of 2.98(2) mm s-1. These values indicate the presence of high spin Fe(ll) 78 centers in Fe(Ai-OH2)2[Au(CN)2]2.( ) Chapter 2. Magnetic exchange through double aqua-bridges 37

0.0

I -0.5 I

% -l.o

-1.5

-4-3-2-101234 Velocity [mm s" ]

Figure 2.2: Mossbauer spectrum of Fe(/i-OH2)2[Au(CN)2]2 at 4.5 K. The solid line corresponds to the best fit using a quadrupole split doublet.

2.2.2 Structure determination

Powder X-ray diffraction

The powder X-ray diffractogram of each M(/i-OH2)2[Au(CN)2]2 complex was mea sured and is shown in Figure 2.3. They were found to show the same basic features, with similar peak positions and relative intensities. A slight shift toward larger dhki values can be observed for the Mn(ll) complex compared to the other complexes. A few additional peaks in the 1.5-3.5 A d^ki range can also be observed for some compounds. The powder X-ray diffractograms of the M(/x-OH2)2[Au(CN)2]2 complexes (M = Cu, Co, Fe, Mn) can be indexed to similar or related unit cells. The unit cell para meters determined for each complex are reported in Table 2.2. The comparable FT-IR spectra and powder X-ray diffractograms of the M(//- OH2)2[Au(CN)2]2 complexes altogether suggest that the building blocks arrangement is similar in all complexes. Chapter 2. Magnetic exchange through double aqua-bridges 38

r—I—|—I I I—I | 1 ' I • • • • I ' •

\cu(n) LMJVA_KJ

Ni(ll) b L*JUUUJUAMA1_ 1

1i LA^JAALL Co(il)

Fe(ll) ^^.JWAAAJL

Mn(ll) VXA_MJUJVIAJ_

^j^jOkx 1 Fe(m) _J 1 L_J 1 L, t.,.. 6 7 10 11 d,,,[A]

Figure 2.3: Comparison of the powder X-ray diffractograms of M(/x-OH2)2[Au(CN)2]2 (M(n) = Cu, Ni, Co, Fe, Mn) and Fe(^-OH2)(/i-OH)[Au(CN)2]2 (Fe(m)). Chapter 2. Magnetic exchange through double aqua-bridges 39

Table 2.2: Unit cell parameters for the M(,u-OH2)2[Au(CN)2]2 (M(n) Ni, Cu, Co,

Fe and Mn) and FeOLi-OH2)Gu-OH)[Au(CN)2]2 (Fe(in)) complexes.

Ni Cu Co Sample initial single powder and single powder powder crystal crystal crystal isyste m orthorhombic orthorhombic monoclinic monoclinic space group Cmmm Immm P2 or P2/m P2/m a, A 6.375(3) 6.374(3) 6.335 6.4023 b, A 3.3186(11) 3.3183(11) 20.509 20.6094 c, A 10.252(2) 20.512(5) 3.482 3.3876 a, deg 90.0 90.0 90.0 90.0 P, deg 90.0 90.0 90.93 90.478 7, deg 90.0 90.0 90.0 90.0 Volume.,A 3 216.89(13) 433.9(3) 452.324 446.99

Fe Mna Fe(m) Sample powder powder powder crystal isyste m monoclinic monoclinic orthorhombic space group P2/m P2/m Cmmm a, A 6.428 6.50 6.26 b, A 20.670 20.78 3.59 c, A 3.465 3.49 10.34 a, deg 90.0 90.0 90.0 /3, deg 90.68 90.4(3) 90.0 7, deg 90.0 90.0 90.0 Volume.,A 3 460.32 471.38 231.98 aThere is a large uncertainty associated with the (3 angle. Chapter 2. Magnetic exchange through double aqua-bridges 40

Crystal growth

The growth of crystals of M(/>OH2)2[Au(CN)2]2 suitable for single crystal X-ray stud ies was found to be very challenging. Attempted crystal growth using H-shaped tubes, gel diffusion techniques, layered solvents and more standard variations of concentra tion and temperature in water or water/solvent mixtures generated only microcrystalline powder. Only small poor-quality crystals of Ni(//-OH2)2[Au(CN)2]2 (less than 0.05 mm per side) were obtained by hydrothermal recrystallization; these crystals and the bulk Ni(/>OH2)2[Au(CN)2]2 powder had comparable FT-IR spectra and were used to collect single crystal X-ray data.

Ni(/i-OH2)2[Au(CN)2]2

The cell parameters initially determined for Ni(/Li-OH2)2[Au(CN)2]2 are presented in Table 2.2. From the initial single crystal X-ray analysis, the basic structural motif of

Ni(/U-OH2)2[Au(CN)2]2 could be determined. Each Ni(ll) center is coordinated by four bridging water molecules, generating Ni(/^-OH2)2Ni diamond chains in the 6-direction (Figure 2.4, Table 2.3). The two other coordination sites are occupied in a trans fashion by the N-cyanide groups of the linear [Au(CN)2]~ units, thereby yielding a slightly distorted octahedral environment around the Ni(ll) center. The closest Au- Au distance, 3.3183(11) A, indicates the presence of aurophilic interactions. The structural model obtained from the initial single crystal X-ray data gener ates a random half-occupancy of the nickel and oxygen atoms; Figure 2.4B shows the contents of one unit cell from the initial single crystal data analysis. However, some discrepancies can be noticed when comparing the powder X-ray diffractogram predicted by this model with the one obtained experimentally (Figure 2.5). A few key observed reflections (dhki = 4.65 and 6.15 A) are not predicted by this model. The relative intensities of these reflections are very weak (less than 5 %) and, under our experimental conditions for the single crystal data collection, fall below the 3a signal-to-noise ratio cutoff used to discriminate between background noise and real signal originating from the sample. This half-occupancy indicated by the initial single crystal data analysis could be Chapter 2. Magnetic exchange through double aqua-bridges 41

A. f 4 I B. Aul' t C1 * t . i Aul' f NlfNil fNil't 'Aul' ^ ^ Aul- I i 01 t4 i 1

Figure 2.4: A. Basic structural motif of Ni(//-OH2)2[Au(CN)2]2. B. Unit cell con tents obtained from initial single-crystal X-ray data analysis, for which all Ni(ll) and 0 atoms have a half-occupancy, hydrogen atoms are omitted for clarity. Extended structure after doubling the unit cell along the c-axis, in the sheet model (C.) and in the ribbon model (D.) (Ni, green; Au, yellow; C, grey; N, blue.). E. Hydrogen-bonding interactions between non-coordinated N atoms and water molecules. Chapter 2. Magnetic exchange through double aqua-bridges 42

Table 2.3: Selected bond lengths (A) and angles (deg) for Ni(^-OH2)2[Au(CN)2]2

Bond Lengths Ni(l)-N(l) 2.0133(5) Au(l)-Au(l*) 3.5932(12) Ni(l)-0(1) 2.156(11) 0(1)-H(11) 0.93 Ni(l)-Ni(l') 3.3183(11) H(ll)---N(ll#) 1.788 Au(l)-Au(l') 3.3183(11) 0(1)-N(11#) ' 2.709 Angles Ni(l)-0(1)-Ni(l') 100.6(7) 0(1)-Ni(l)-0(1") 79.4(7) 0(1)-Ni(l)-N(l) 90.00(2) 0(1)-H(11)-N(11#) 169.8 Symmetry operations: (') x, y + 1, z; (") —x, —y+l,z; (#) x + \, y — \, z + \; (*) -x + \, -y + \,-z + \

due either to a truly disordered structure or to the existence of a superstructure. The former was discounted on the basis of the additional weak peaks observed in the powder diffractogram. In the latter case, doubling the unit cell along the c-axis would remove the half occupancy. In order to maintain the formula unit and charge balance, two of the four Ni(ll) sites must then be eliminated, as well as their bound water molecules. To ensure physical feasibility, one site must be removed from each a6-plane and the choice of which sites are eliminated determines the supramolecular arrangement of the basic structural motif: either 2-D sheets (Figure 2.4C) or 1-D ribbons (Figure 2.4D) can be obtained. The powder X-ray diffractograms predicted using the unit cell with a doubled c- axis and placement of atoms in the 2-D sheet and 1-D ribbon model respectively were generated and were compared to the experimental diffractogram obtained for Ni(/x-

OH2)2[Au(CN)2]2 (Figure 2.5). The sheet model was rejected due to poor agreement between the predicted and experimental results in the 4.5-6.5 A data range (inset Figure 2.5). On the other hand, the diffractogram simulated for the ribbon model was found to be very similar to the experimental data. All the reflections, including the key weak peaks at 4.65 and 6.15 A, are present in both diffractograms and the same relatives intensities are observed. Using the cell information obtained from the powder diffractogram analysis, the Chapter 2. Magnetic exchange through double aqua-bridges 43

1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 9.0 9.5 10.0 10.5 11.0

Figure 2.5: Comparison of the powder diffractogram obtained experimentally for Ni(/x- OH2)2[Au(CN)2]2 (i-, blue) with the one predicted by the single crystal solution (ii., purple), the ribbon model (iii., green) and the 2-D sheet model (iv., orange). single crystal diffraction data was recollected on the same crystal with the doubled cell. The single-crystal solution that is obtained corresponds to the ribbon model as determined with the powder diffractogram, with no disorder or occupancy issues remaining. The basic structural motif shown in Figure 2.4A is still preserved in the rib bon model: Ni(/z-OH2)2Ni diamond chains are found along the 6-axis, with pendant

[Au(CN)2]~ units above and below the chains parallel to the c-direction. The hydro gen atoms likely lie symmetrically above and below the Ni202 diamond plane. These ribbons are offset with each other along the a- and 6-axes and interact through hydro gen bonds that involve water molecules in one chain and terminal N-cyanide atoms of the four neighbouring chains (Figure 2.4E, Ni-OH- • • NC, dQ-N = 2.709 A). Chapter 2. Magnetic exchange through double aqua-bridges AA

In light of the structure determined for the Ni(/^-OH2)2[Au(CN)2]2 complex, struc tural models are proposed for the analogous M(/x-OH2)2[Au(CN)2]2 complexes.

Cu(//-OH2)2[Au(CN)2]2

The powder X-ray diffractogram of Cu(/i-OH2)2[Au(CN)2]2 could be best indexed to a monoclinic unit cell (Table 2.2). Attempts to index the powder diffractogram of

Cu(/i-OH2)2[Au(CN)2]2 to an orthorhombic unit cell failed, as several peaks below 3.1 A were not predicted. The dimensions of the monoclinic cell obtained by indexing are similar to those of Ni(/i-OH2)2[Au(CN)2]2. However, the b- and c-axes in Cu(/>OH2)2[Au(CN)2]2 are switched versus the Ni analogue. The c-lattice parameter of Cu(/i-OH2)2[Au(CN)2]2 is consistent with the one obtained in the final unit cell determined for the Ni-containing analoguous polymer (6Cu = cNi.final). The lower symmetry and small differences between the unit cells can be attributed to the presence of Jahn-Teller distorted Cu(ll) centers in Cu(/i-OH2)2[Au(CN)2]2 com pared to the more symmetric Ni(ii) centers in Ni(/i-OH2)2[Au(CN)2]2.

Cu(/Li-OH2)2[Au(CN)2]2 likely adopts the same structural motif and packing as de scribed for the Ni analogue (Figure 2.4). Figure 2.6 compares the experimentally ob tained diffractogram with the diffractogram predicted using atomic positions similar to those in Ni(//-OH2)2[Au(CN)2]2. The atomic coordinates for Cu(/x-OH2)2[Au(CN)2]2 are reported in Table A.2. This model predicts the same peak positions and relative intensities as the ones observed experimentally. The axial and equatorial sites of the Jahn-Teller distorted Cu(ll) centers in Cu(/x-

OH2)2[Au(CN)2]2 were assigned based on the frequencies of the cyanide stretches observed in the FT-IR spectrum. The [Au(CN)2]~ units are believed to be equatorially iV-bound to the Cu(ll) centers, as indicated by the large shift of the cyanide vibration -1 -1 band, from 2141 cm in K[Au(CN)2] to 2217 cm . The frequency corresponding to vibration of axially bound Cu(ll)-[Au(CN)2] units usually shows a smaller shift from the frequency observed in free [Au(CN)2]~ units compared to an equatorially bound 57 -1 unit.^ ) The band at 2171 cm was assigned to the pendant side of the [Au(CN)2]~ Chapter 2. Magnetic exchange through double aqua-bridges 45

Ky \J,

ii.

J—i.-i 11 I i i i i L ,> t—i i L-1...I—J i—J i i i__t L-i..i—i i_J i i i u-J—i i t.,.t.J..iii i t i—L,••>.,.i, I, '/> 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 10.0 10.5 11.0

Figure 2.6: Comparison of the powder diffractogram obtained experimentally for

Cu(//-OH2)2[Au(CN)2]2 (i., orange) and the diffractogram predicted by the proposed structural model for Cu(//-OH2)2[Au(CN)2]2, based on the structure of the analogous Ni-containing polymer (ii., black).

unit, which hydrogen bonds to the water molecules, shifting its frequency of vibration. The four water molecules in the Cu(ll) coordination sphere occupy the two axial sites and the remaining two equatorial sites. As a consequence, the oxygen atoms bridge the Cu(n) centers in an equatorial-axial fashion. This bridging combina tion generates an asymmetric diamond chain motif (Scheme 2.1), although accurate Cu-0 distances and Cu-O-Cu angles could not be determined. The asymmetric

Cu(/x-OH2)2Cu diamond chains are likely responsible for the reduction in unit cell symmetry (from orthorhombic to monoclinic), as the chains lie in the distorted ac plane, along the c-direction.

M(//-OH2)2[Au(CN)2]2 (M = Co, Fe and Mn)

Similar reasoning can be used to explain the unit cells obtained for the Co, Fe and Mn- containing M(/x-OH2)2[Au(CN)2]2 complexes. The monoclinic unit cells differ slightly from that of Ni(//-OH2)2[Au(CN)2]2, after switching the b- and c-axes. The small differences suggest a distortion of the coordination sphere around the metal centers or a packing distortion in the ac plane, similarly to the Cu-containing analogous Chapter 2. Magnetic exchange through double aqua-bridges 46

/\ /

Scheme 2.1: Asymmetric diamond suggested to be present in Cu(/i-OH2)2[Au(CN)2]2 from the equatorial-axial bridging mode of the water molecules polymer. The diffractograms predicted when using similar atomic coordinates to the Cu analogue match well the experimentally obtained diffractograms (see Appendix A,

Figures A.l, A.2 and A.3). The atomic coordinates for Co(//-OH2)2[Au(CN)2]2, Fe(y^-

OH2)2[Au(CN)2]2 and Mn(^-OH2)2[Au(CN)2]2 are reported in Tables A.3. A.4 and A.5 respectively. The small increase in the a (6.335-6.50 A) and b (20.509-20.78 A) parameters when going from the Cu(ll)- to Fe(ll)-containing unit cells could be attributed to the increasing size of the metal centers. The differences between the c and j3 parameters could be due to the differences in the geometry of the M(//-OH2)2M bridges: symmet ric vs asymmetric diamond motif, and larger vs shorter M-O-M angles. However, as for the Cu analogue, the position of the oxygen atoms could not be determined accurately and, as a consequence, conclusions cannot be drawn about the exact ge ometries.

2.2.3 Fe(ju-OH2)(ju-OH)[Au(CN)2]2

Synthesis

When an aqueous solution of Fe(C104)3-6H20 was mixed with an aqueous solution containing three equivalents of K[Au(CN)2], an immediate deep orange precipitate was obtained. The FT-IR spectrum of this product showed two overlapping bands, with maximum at 2184 cm-1 and 2168 cm-1, attributable to cyanide stretching vibrations. Chapter 2. Magnetic exchange through double aqua-bridges 47

-4-3-2-10 1 2 3 4 Velocity [ mm s" ]

Figure 2.7: Mossbauer spectrum of Fe(//-OH2)(/>OH)[Au(CN)2]2 at 4.5 K. The solid line corresponds to the best fit using a quadrupole split doublet.

Determination of chemical composition

To assess the oxidation state of the Fe centers in the product, Mossbauer spectroscopy was performed. Figure 2.7 shows the Mossbauer spectrum acquired at 4.5 K. A narrow quadrupole pair with an isomer shift (S) of 0.47(3) mm s-1 and a quadrupole splitting (AEQ) of 0.88(3) mm s_1 can be observed. These values are typical for paramagnetic high-spin Fe(lll) centers. (78) No additional peaks are present in the spectrum, which suggests the presence of only one type of Fe center. From the relative amounts of carbon, hydrogen and nitrogen determined by ele mental analysis and taking into account the Mossbauer spectrum, the chemical com position was determined to be Fe(/x-OH2)(//-OH)[Au(CN)2]2. Chapter 2. Magnetic exchange through double aqua-bridges 48

Structural Characterization

The powder X-ray diffractogram of Fe(/i-OH2)(/x-OH)[Au(CN)2]2 was found to have similar features to that of the M(//-OH2)2[Au(CN)2]2 coordination polymers (Fig ure 2.3). The low intensity peaks at around 4.65 and 6.15 A are however ab sent in Fe(/u-OH2)(Ai-OH)[Au(CN)2]2. The powder diffractogram of Fe(/u-OH2)(/i-

OH)[Au(CN)2]2 can be indexed to a few unit cells with similar (or related) dimen sions. Amongst the best possibilities, a unit cell with dimensions similar to those of the initial, smaller unit cell determined for Ni(/i-OH2)2[Au(CN)2]2 by single crystal diffraction was obtained. These unit cell parameters are reported in Table 2.2.

A general structural model is proposed for Fe(/x-OH2)(/i-OH)[Au(CN)2]2 using atomic coordinates similar to those of the Ni-containing unit cell shown in Figure 2.4B. Note that in this model, the transition metal (Ni or Fe(lll)) and oxygen atoms have a half occupancy. As was explained above and shown in Figure 2.5, the diffractogram predicted for this small unit cell with half occupancies does not include the peaks at 4.65 and 6.15 A, which are characteristic of the ribbon model, or the peaks at 5.4 and 6.4 A, which are characteristic of the sheet model.

The half occupancies imply that the structure of Fe(^-OH2)(/x-OH)[Au(CN)2]2 consists of a random mixture of the sheets and chains/ribbons structures shown in Figures 2.4C and 2.4D.b In other words, defects are present along the chains and chain fragments are connected by [Au(CN)2]~ units to form partial sheets. In this model, both the hydroxide and water groups act as bridging ligands between the Fe(lll) centers. These two groups could be randomly distributed along the struc ture or could alternate in a regular pattern (either aqua-hydroxo bridges or double aqua-bridges alternating with double hydroxo-bridges), but there is no experimental evidence to distinguish between these different situations. The diffractogram predicted by this structure matches well the experimentally obtained diffractogram (see Appendix A, Figure A.4). The atomic coordinates for

Fe(/i-OH2)(yu-OH)[Au(CN)2]2 are reported in Table A.6. The position of the oxygen

bIt is assumed that when an Fe(lll) center is absent the associated oxygen atoms are also absent and vice versa. Chapter 2. Magnetic exchange through double aqua-bridges 49

Table 2.4: Decomposition temperature of each M(/i-OH2)2[Au(CN)2]2 complex (M(n) = Cu, Ni, Co, Fe, Mn) and Fe(/x-OH2)Gu-OH)[Au(CN)2]2 (Fe(m)).

M Loss of water Loss of CN gr

atoms does not affect the diffractogram to a great extent and, hence, could not be accurately determined. It is believed that slightly different positions could be adopted by the oxygen atoms of the hydroxide and water groups.

2.2.4 Thermal stability

Thermogravimetric analysis was performed for all M(//-OH2)2[Au(CN)2]2 complexes. Table 2.4 compares the temperatures at which weight losses are observed for each

M(/x-OH2)2[Au(CN)2]2 complex. For the Cu(ll), Ni(il), Co(ll) and Mn(ll)-containing complexes, two distinct weight losses occurred for each complex between 25 and 500 °C. The relative weights were found to correspond to the loss of two water molecules followed by the loss of two cyanogen, (CN)2, molecules and gain of oxygen. The final weight observed for each complex is consistent with the formation of MOa; (where x = 1 or 1.5) and Au (2 equivalents). As the thermogravimetric analysis experiments were performed in air, uptake of oxygen by the transition metal to form Chapter 2. Magnetic exchange through double aqua-bridges 50 the more stable metal oxides is expected.

The Fe(^-OH2)2[Au(CN)2]2 polymer decomposes differently, with two losses oc curring below 300 °C. The two losses could be attributed to the sequential loss of half an equivalent of water followed by the remaining 1.5 equivalent or to the loss of ad sorbed surface water followed by the loss of two equivalents of water. However, within experimental errors, the two possibilities cannot be distinguished. Above 340 °C, a third weight loss is observed which is attributed to the loss of the cyanogen groups and the gain of oxygen. The Fe(yLt-OH2)(/i-OH)[Au(CN)2J2 polymer decomposes initially over a very large temperature range, losing what corresponds approximately to a hydroxide ion and a water molecule. The second weight loss occurs over a much narrower temperature range, which is similar to the temperature at which the other M[Au(CN)2J2 frame works decompose.

2.2.5 Removal of the water molecules in the Cu-based system

Anhydrous green-brown Cu[Au(CN)2J2 was prepared in bulk by thermally remov ing the water molecules from Cu(/Li-OH2)2[Au(CN)2]2 at 180 °C for several hours. The dehydration temperature was chosen following the thermogravimetric analysis experiments. Elemental analysis showed no hydrogen in the product and no peaks characteristic of Cu(//-OH2)2[Au(CN)2]2 were observed in its FT-IR spectrum. These results confirm the complete removal of the water molecules.

The diffractogram of Cu[Au(CN)2]2 only shows a small number of broad peaks that are shifted from that observed for Cu(/i-OH2)2[Au(CN)2]2- When the water molecules are thermally removed, the remaining framework, Cu[Au(CN)2]2, most likely under goes a significant structural rearrangement to re-occupy some of the vacated metal coordination sites. -1 The FT-IR spectrum of Cu[Au(CN)2]2 only shows one band, at 2191 cm , at tributable to a cyanide stretching vibration. The presence of only one band suggests a structure in which all the cyanide groups are equivalent to each other, while the Chapter 2. Magnetic exchange through double aqua-bridges 51 •l^t^lT Cu2+ i^T~* -M-ft Au(CN); 1 I I

Scheme 2.2: Proposed 2-D square grid structural motif for Cu[Au(CN)2]2.

frequency shift of the cyanide vibration indicates that the [Au(CN)2]~ units are co ordinated to the Cu(ll) centers through the N-atom. A square-grid structure as shown in Scheme 2.2, where all the Cu(n) centers - adopt a square planar geometry and are surrounded by four bridging [Au(CN)2] units, could be formed. Such a motif has been observed for related complexes such as 51 66 Mn[Au(CN)2]2(H20)2( ) and Co[Au(CN)2]2(DMF)2( ) (see also Chapter 3 for more examples). However, the metal centers in these other compounds have an octahedral geometry and analyte (H20 or DMF) molecules occupy the two axial sites on each side of the square-grid.

Cu[Au(CN)2]2 was found to be relatively stable in the solid state under atmo spheric conditions, as no colour change from green-brown to a brighter pale green could be observed over a period of several hours. However, when FT-IR spectra were acquired as a function of time on the same pressed KBr pellet containing Cu[Au(CN)2]2, changes could be observed and two -1 new bands in the VCN region (with frequencies of 2217 and 2172 cm ) started to grow. After two hours, the FT-IR spectrum contained features associated with the

Cu[Au(CN)2]2 and Cu(^-OH2)2[Au(CN)2]2 complexes. These changes in the FT-IR spectrum indicate that re-hydration of Cu[Au(CN)2]2 back to Cu(//-OH2)2[Au(CN)2]2 can occur after several hours under certain conditions. A full conversion was not how ever observed during these experiments. The KBr salt used as matrix for the FT-IR Chapter 2. Magnetic exchange through double aqua-bridges 52

measurements is hygroscopic and could accelerate the rehydration of Cu[Au(CN)2]2.

2.2.6 Discussion about the synthesis and structural arrange ments

Different synthetic routes

The aqueous reactions involving Co(ll), Fe(ll) and Mn(ll) metal centers with

K[Au(CN)2] did not yield the analogous M(/Lt-OH2)2[Au(CN)2]2 polymers as ex pected, but different products were collected. For example, the aqueous reac 51 tion of Fe(C104)2-6H20 with K[Au(CN)2], which was previously reported^ ), af fords the formation of a coordination polymer with a different chemical composi tion, K{Fe[Au(CN)2]3}, irrespective of the Fe(n):[Au(CN)2]~ ratio in solution (see Chapter 4 for more details on this polymer). The similar aqueous reaction between 51 Mn(C104)2-6H20 and K[Au(CN)2] was also reported. ( ) It allows the formation of

Mn[Au(CN)2]2(H20)2, a polymorph of Mn(//-OH2)2[Au(CN)2]2. The chemical compo sition is the same for both polymers, but the arrangement of the building blocks differs.

The structure of Mn[Au(CN)2]2(H20)2 consists of square grid arrays of Mn[Au(CN)2]2 with water molecules bound to the Mn(ll) centers in a trans fashion on each side of the grids. Note that, in this coordination polymer, the water molecules are not bridging.

The reaction of Co(C104)2-6H20 with K[Au(CN)2], on the other hand, yielded a mixture of [Au(CN)2]-containing products that could not be separated. The aqueous 88 5 synthesis of Co[Au(CN)2]2^ ^ and K{Co[Au(CN)2]3}( °) was previously reported and, despite the different reaction conditions, these products could be present in the mix ture obtained. A change in solvent, from water to acetonitrile, allowed the obtention of a pure product.

For the reactions carried out in acetonitrile, the starting materials, M(C104)2-6H20 ra and [ Bu4N][Au(CN)2]-0.5H2O are believed to be the source of water molecules for the formation of M(/x-OH2)2[Au(CN)2]2. The hydrated salts of M(C104)2-6H20 (M = Co, 2+ (89 Fe, Mn) all contain [M(H20)6] ions, ) from which the formation of M(/^OH2)2M bridges is possible when reacted with [Au(CN)2]~ anions. Chapter 2. Magnetic exchange through double aqua-bridges 53

Formation of Fe(/u-OH2)(Ai-OH)[Au(CN)2]2

The chemical identity of the product formed by the aqueous reaction of Fe(lll) with

K[Au(CN)2] was assessed using a combination of characterization methods. Results from elemental analysis suggested that only two [Au(CN)2]~ units per Fe center are present in the product, rather than the expected Fe[Au(CN)2]3(H20)x product. In fact, the results of elemental analysis suggested that the composition, within exper imental errors, could be identical to that of Fe(^-OH2)2[Au(CN)2]2 (which contains Fe(ll) centers). However, Mossbauer spectroscopy confirmed the presence of only Fe(lll) centers in the product. From these results and to charge balance the com plex, it was suggested that a hydroxide group replaces a water molecule in Fe(^-

OH2)2[Au(CN)2]2, yielding the Fe(//-OH2)(//-OH)[Au(CN)2]2 complex. 3+ 90 Hydrolysis of [Fe(H20)6] is known to occur in aqueous solution: ( )

3+ 2+ 305 [Fe(H20)6] ^ [Fe(H20)5(OH)] + H+ K = 1(T (2.1)

Hydrolysis of the remaining water molecules around the Fe(lll) center is also possible. 2+ From the formation constant, it can be seen that [Fe(H20)5(OH)] will form unless the solution is very acidic. The presence of [Au(CN)2]~ in solution can modify this 2 equilibrium if [Fe(H20)5(OH)] + is consumed, as the Fe(/x-OH2)(/j-OH)[Au(CN)2]2 polymer forms and precipitates immediately. This would in turn favour the hydrolysis 3+ of more [Fe(H20)6] .

An uncommon motif

The chain motif adopted by the M(/i-OH2)2[Au(CN)2]2 coordination polymers is un precedented in cyanometallate chemistry (19>21>27) and, more generally, very rare in co ordination chemistry. Compared to the large number of oxo- and hydroxo-bridged metal centers, only a small number of mono or doubly aqua-bridged dimers are known for each type of metal center: Cu(ll),(91-94) Ni(ll),(95-100) Co(ll),(95'100-102) Fe(n) (102>103) and Mn(ll)(102'104). In most cases, a second ligand, often a carboxylate group, is also bridging the metal centers in addition to the water molecules. 2+ The structure of a chain system with a [Cu(/u-OH2)2(/x-L-alanine)] repeating unit Chapter 2. Magnetic exchange through double aqua-bridges 54 was also reported/105^ but no double aqua-bridges have been observed to form chains with any other metal center. M(//-OH2)2M units have also been observed 106 in the 3-D framework of Mn2[Ru(CN)6]-8H20/ ) as well as the cluster-containing

[Co2(H2O)4][Re6S8(CN)6]-10H2O and [Co(H20)3]4[Co2(H20)4][Re6S8(CN)6]3-44H20 coordination polymers. (38) A few examples of aqua-hydroxide bridges have been reported with different tran sition metal centers. To our knowledge however, only one example of an Fe(ll) dimer with an aqua-hydroxo bridging motif can be found in the literature (107) and no Fe(m)- containing dimer or polymer with this bridge is known. The aurophilic interactions could be playing a role in the formation of these rare double aqua-bridges, by bringing the metal centers in close proximity. The metal centers along the chains in M(/z-OH2)2[Au(CN)2]2 could be considered as quadruply bridged through two aqua-bridges and two NCAu-AuCN bridges.

The M(^-OH2)2[Au(CN)2]2 and Fe(//-OH2)(//-OH)[Au(CN)2]2 polymers were found to be stable up to 75 - 215 °C depending on the metal center (Table 2.4).

The bridging M(/i-OH2)M coordination motif as well as the hydrogen bonding inter actions between the water molecules and [Au(CN)2]~ units most likely stabilize the polymers and prevent the loss of the water molecules at low temperature.

The thermal stability of the Fe(/i-OH2)2[Au(CN)2]2 polymer differs from that of the other analogues. Upon heating in air, oxidation of the Fe(li) centers could be occurring, which would enable a different decomposition pathway. No investigations were performed to test this hypothesis.

2.3 Magnetic exchange through double aqua- bridges

The magnetic properties of the M(/i-OH2)2[Au(CN)2]2 polymers, between 300 and 1.8 K and in different applied fields, were investigated by SQUID magnetometry. To obtain further information, particularly in zero applied field, the muon spin relaxation (/zSR) technique was also applied. (81) (See section 1.4.5 for more explanations on the Chapter 2. Magnetic exchange through double aqua-bridges 55 technique of /xSR.) The /^SR experiments and related calculations were performed in collaboration with Prof. Jeff Sonier (Department of Physics, Simon Fraser University).

2.3.1 Cu(ju-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2

SQUID magnetometry results

The magnetization of Cu(/x-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 was measured and their respective magnetic susceptibility and effective magnetic moments (faff) were determined as a function of temperature in a field of 10 kOe (Figure 2.8). The effective moments observed between 300 K and 100 K are 1.81 /IB and 1.89 [iB for

Cu(//-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 respectively. Below 50 K, the effective magnetic moment increases with decreasing temperature for Cu(//-OH2)2[Au(CN)2]2 while it decreases for the anhydrous Cu[Au(CN)2]2.

_1 Above 50 K, the inverse of the susceptibility (%M ) for Cu(/i-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 have a linear dependence with temperature. They can be fit to the Curie-Weiss expression (Equation 1.10). For Cu(/i-OH2)2[Au(CN)2]2, a 9 value of + 5.1(3) K was obtained (g = 2.084(2)), while the fit to the Cu[Au(CN)2]2 data yielded a 9 value of- 8.2(6) K (g = 2.211(4)). The value of the effective magnetic moment observed at room temperature for

Cu(/i-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 is typical for isolated Cu(n) centers. The increase in the effective moment at low temperature suggests the presence of ferromagnetic interactions in Cu(/Lt-OH2)2[Au(CN)2]2, while the drop observed for

Cu[Au(CN)2J2 is indicative of antiferromagnetic interactions.

The magnetic susceptibility of Cu(^-OH2)2[Au(CN)2]2 was fit to the high- temperature series expansion derived by Baker (Equation 2.3) for the regular infinite Chapter 2. Magnetic exchange through double aqua-bridges 56

B. —I 1 1 1 ir i 1 p—| 1 1 r-

300

i 200 k

^ 100

0 50 100 150 200 250 300 50 100 150 Temperature [ K ] Temperature [ K ]

Figure 2.8: Temperature dependence of the effective magnetic moment (fieff) of Cu(/x- OH2)2[Au(CN)2]2 (o) and Cu[Au(CN)2]2 (A) under a 10 kOe applied field. The solid line represents the best fit to the Baker expression, including a mean field correction. B. Inverse of the susceptibility for Cu(//-OH2)2[Au(CN)2]2 (o) and Cu[Au(CN)2]2 (A) as a function of temperature and theoretical fits (solid lines) to the Curie Weiss ex pression.

Heisenberg chain model for S = \ with the following Hamiltonian: (108)

H = -jY,St-Si+l (2.2)

2 NA9 ^B fA\i Xchain - ^^ -^BJ W A = 1.0 + 5.7979916?/ + 16.902653 y2 + 29.376885 y3 + 29.832959 yA + 14.036918 y5 B = 1.0 + 2.7979916 y + 7.0086780 y2 + 8.653644 y3 + 4.5743114 y4

y = J/2kBT where J is the exchange coupling constant between the metal centers along the chain. This fit yielded an exchange coupling constant, J, of 0.37(2) cm-1 and a g value Chapter 2. Magnetic exchange through double aqua-bridges 57

of 2.134(3), consistent with weak ferromagnetic interactions along a Cu(jii-OH2)2Cu chain. To account for interactions present between the chains, a mean field approxi mation was incorporated into the Baker expression (Equation 2.4): (71)

Xchain /~ ,\ XMF l ~ l-XcKain-{2zJ'lNAg^l) ' ' where z is the number of interacting neighbouring chains and J' is the coupling constant between each chain (xchain being defined in Equation 2.3). The best fit to Equation 2.4 yielded a J value of 0.96(5) cm-1, a zJ' value of - 0.58(4) cm-1 and a g value of 2.122(2). However, this fit should be treated with caution since the similar magnitudes of J and zJ' (0.96(5) and -0.58(4) cm-1 respectively) limits the validity of the mean field approximation. (71) These two fits suggest ferromagnetic interactions along the chain and weak antiferromagnetic interactions between the chains.

//SR results for Cu(/i-OH2)2[Au(CN)2]2

In order to investigate further the magnetic state of Cu(//-OH2)2[Au(CN)2J2, zero- field muon spin relaxation (ZF-/^SR) measurements were performed down to 0.015 K. Figure 2.9 shows the ZF-//SR asymmetry spectra acquired at different temperatures for Cu(/Lt-OH2)2[Au(CN)2]2 upon cooling from 1.0 K to 0.020 K. The asymmetry spectra above 0.20 K were fit to a power exponential relaxation function for the sample signal, such that:

K A(t) = a0Pz(t) = ase-^ + aAg (2.5) where as and aAg are amplitudes reflecting the fraction of muons that stop in the sample and the silver backing plate, respectively. The background signal originates primarily from the silver backing plate, and is both non-relaxing and temperature independent. (81) Below 0.20 K, a coherent precession signal is observed. The asymmetry spectra be low 0.20 K were fit to the following equation which includes an oscillating component Chapter 2. Magnetic exchange through double aqua-bridges 58

0.40 * i i • • | i 1 5 i •

B. ..i? 4 r v2 | 3 j 2 a 7 vi i* 1 •i 0 • , . i . . . i . . r-

0.0 0.2 0.4 0.6 Temperature [ K ]

Figure 2.9: A. ZF asymmetry spectra acquired for Cu(^-OH2)2[Au(CN)2]2 at different temperatures ranging from 1.0 K to 0.02 K. The solid curves are the fits to Equa tions 2.5 and 2.6. The spectra have been vertically offset from each other for clarity. Temperature dependences of the fitted parameters with their respective error bars: B. ui (o) and v

(each variable will be explained below).

A(t) = CL0Pz(t)

Alt A2t -At = CLS 11 [/cos(27ri/it + ^i)e" (l-/)cos(27n/2t + ^2)e- ]+^

+ aAg (2.6)

The presence of a precession signal in the asymmetry spectrum indicates the pres ence of a local magnetic field due to the onset of long-range magnetic order. For a polycrystalline sample randomly oriented, there is a | probability that the local magnetic field at the muon site is parallel to the muon spin direction. Similarly, there Chapter 2. Magnetic exchange through double aqua-bridges 59 is a | probability for the local magnetic field to be transverse to the muon spin. Con sequently, only | of the muon spins precess about the local magnetic field B^ with a frequency equals to z/M (Equation 1.27).

In Cu(^-OH2)2[Au(CN)2]2, the fit is greatly improved by introducing two well- defined muon precession frequencies, v\ and v2. This indicates that two magnetically non-equivalent muon stopping sites exist in Cu(yU-OH2)2[Au(CN)2J2- The temperature dependences of V\ and i/2 are shown in Figure 2.9B. Above 0.20 K, both v\ and v2 are equal to zero, but below that temperature, V\ and v2 increase rapidly to reach a constant non-zero value at 0.15 K. The fraction of muons that sense an average local field B\ = fi/j/j, is /, and the fraction of muons that sense the average local field B2 = ^2/7^ is (1-/)- The fits to the asymmetry spectra below 0.20 K yielded an / value of |. In Equation 2.6, ip\ and

A = ^>i • fol TV, (2-7) where v^ = 7M-Bext is the Larmor frequency of the muon in an external magnetic field

Bext. Figure 2.9C shows the temperature dependence of A, which exhibits a cusp-like behaviour at 0.2 K. This is indicative of a spin-freezing phase transition (Tf = 0.2 K). The critical slowing down of magnetic fluctuations on approach of the freezing temperature from above results in a growth of TC, which in turn causes an increase in A. Below Tf, the freezing out of the magnetic excitations reduces A.

The maximum in A(T) and the simultaneous increase of V\ and v2 confirm the existence of a transition to a magnetically ordered state at Tf m 0.20 K.

/iSR results for Cu[Au(CN)2]2

ZF-/iSR measurements were also performed on the dehydrated Cu[Au(CN)2]2 com plex. The asymmetry spectra obtained are shown in Figure 2.10A. As opposed to Chapter 2. Magnetic exchange through double aqua-bridges 60

A. 0.20

0.16

0.12

0.08

0.04

0.00 0 2 4 6 8 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Time [ [is ] Temperature [ K ]

Figure 2.10: A. ZF asymmetry spectra acquired as a function of temperature for Cu[Au(CN)2]2 at 3.2 (•), 1.0 (o) and 0.05 K (A). The solid curves are fits to Equa tion 2.5. B. Temperature dependence of the relaxation rate A (o) and the power K (•) from the fits. The solid curves are guides for the eye.

Cu(//-OH2)2[Au(CN)2]2, there is no oscillatory component in the asymmetry spec trum at low temperatures which suggests that there is no long-range magnetic order in Cu[Au(CN)2]2. The asymmetry spectra were fit to Equation 2.5 (Figure 2.10). The values obtained for A and K are plotted in Figure 2.10B as a function of temperature. The relaxation rate A increases slowly with decreasing temperature, which is consistent with the slowing down of fluctuating Cu(ll) spins. The slowing down of the fluctuating spins causes the mean square transverse field (f?f,) felt by the muon's spin during its lifetime to increase. The power K (Figure 2.10B) has a value close to unity at 3.2 K, as expected for rapidly fluctuating Cu(ll) spins. (109) We note that for a single muon stopping site, K is equal to 2 in a dense system of static moments. At low temperatures, K ranges between 1 and 2, suggesting a superposition of field distributions from magnetically Chapter 2. Magnetic exchange through double aqua-bridges 61 inequivalent muon stopping sites. In order to determine whether the magnetic moments sensed by the muons are static or fluctuating at low temperatures, longitudinal field (LF) //SR measurements were performed on Cu[Au(CN)2]2. We found that at 0.40 K, a LF of only 100 G was sufficient to completely decouple the muon spin from the local magnetic field and cause the relaxation of the asymmetry spectrum, A(t), to vanish. This indicates that the relaxation observed in the ZF asymmetry spectrum at this temperature is due primarily to randomly oriented static Cu(ll) spins. (109)

2.3.2 Ni(^-OH2)2[Au(CN)2]2

SQUID magnetometry results

The magnetic susceptibility (XM) for Ni(/x-OH2)2[Au(CN)2]2 was determined upon cooling with a range of applied external dc magnetic fields (5 Oe to 10 kOe). Selected plots of XM as a function of temperature are shown in Figure 2.11 A. Between 300 and 25 K, the magnetic susceptibility of Ni(/ii-OH2)2[Au(CN)2]2 is field independent and a slow increase can be observed as the temperature decreases. The inverse of the susceptibility is linear with temperature above 50 K. Fitting the data obtained above 50 K with an external field of 1 kOe to the Curie-Weiss expression (Equation 1.10) yielded a 6 value of + 5.1(2) K (g = 2.256(1)). A field-dependent magnetic behaviour is observed below ca. 15 K. When exposed to a magnetic field of 1 kOe, the magnetic susceptibility increases rapidly below 10 K with decreasing temperature, reaching 2.18 emu mol-1 at 1.8 K. When the external magnetic field is smaller than 1 kOe, the magnetic susceptibility still increases below 10 K as the temperature decreases, but levels off between 2.0 and 1.8 K. The maximum value in susceptibility may correspond to a peak or a plateau, however the data does not allow discrimination of the two possibilities. The presence of a maximum in susceptibility could indicate antiferromagnetic interactions between the Ni(ll) centers.

Figure 2.11B shows the effective magnetic moment (faff) of Ni(^-OH2)2[Au(CN)2]2 as function of temperature at different magnetic fields (50 Oe to 10 kOe). The ef fective moment observed at 300 K is 3.2 /J,B and remains relatively constant as the Chapter 2. Magnetic exchange through double aqua-hridges

B. 7PT—i—i—i—i—|—i—i—i—i—|—i—i—i—i - 7.0 k

booooooooooooooooooo^

1 • • ' • ' • • • ' f • ' • 100 200 300J

4 a a -4f7 * ^> *46 A &A66J A • A . A

5 10 15 20 25 5 10 15 20 25 Temperature [ K ] Temperature [ K ]

Figure 2.11: A. Temperature-dependent magnetic susceptibility (XM) of Ni(//-

OH2)2[Au(CN)2]2 measured in a 1 kOe (o), 500 Oe (v), 50 Oe (D), 10 Oe (A) dc field. B. Effective magnetic moment (/J.eff) of Ni(/u-OH2)2[Au(CN)2]2 in a 1 kOe (o), 5 kOe (D), 10 kOe (A) dc field. Inset shows the behaviour in a field of 1 kOe between 300 and 1.8 K.

temperature is lowered down to 100 K. A maximum in effective moment is then ob served at temperatures ranging from 3.4 to 6.4 K, depending on the strength of the external magnetic field. The observed room temperature moment is as expected for magnetically dilute

5 = 1 Ni(ii) centers (with a g value of 2.2). The maximum in effective moment (fieff) occurs at a higher temperature than the maximum in susceptibility (XM), which suggests that ferromagnetic interactions are also present between the metal centers. Zero field cooled (ZFC) and field cooled (FC) magnetization measurements were also conducted on Ni(/Lt-OH2)2[Au(CN)2]2- The applied magnetic field upon warming ranged from 10 to 800 Oe. No obvious difference was observed between the ZFC and the FC magnetization measurements between 1.8 and 100 K. Isothermal magnetization measurements were performed as the applied magnetic Chapter 2. Magnetic exchange through double aqua-bridges 63

A. 2.25 2.00 1.75 <—< 1.50

s 1()0 0.75 0.50 0.25 0.00 0 10 20 30 40 50 60 70 ""'o.O 0.5 1.0 1.5 2.6" Field [kOe] Field [kOe]

Figure 2.12: A. Isothermal magnetization curve of Ni(//-OH2)2[Au(CN)2]2 at 1.8 K (•) compared with the Brillouin function for an S = 1 paramagnetic system (with g = 2.2). B. Field dependence of the dc magnetization (M, •) and of the in-phase component of the ac magnetization (M, o) at 1.8 K, in the low-field region. Line is a guide for the eye. field was increased from 0 to 70 kOe at 1.8 K (Figure 2.12A). At 1.8 K, the mag netization curve has a sigmoidal shape: at low fields, the magnetization increases slowly with increasing applied field, but between 700 and 800 Oe, the magnetization reaches an inflection point, after which it increases more rapidly with increasing field (Figure 2.12B). The slope finally decreases above 1 kOe. The magnetization at 1.8 K does not reach saturation over the field range studied (a saturation value of 2.2 NA^B is expected for a Ni(ll) complex with g = 2.2). The sigmoidal shape was not observed when the measurements were performed at higher temperature.

The ac susceptibility of Ni(//-OH2)2[Au(CN)2]2 as a function of the applied dc field at 1.8 K was measured (Figure 2.12B). The in-phase component M' increases to a maximum value at 700 Oe, and then decreases with increasing field, reaching 6.05 x 10~5 NA/^B at 25 kOe. The position of the maximum of the in-phase component M' Chapter 2. Magnetic exchange through double aqua-bridges 64

1 • • • i • • • • i ' ' ' ' i ' • ' ' i • ' • • -

• • . O 600 U_l -o "oB 400 MH "^ o /III)

'-H

u 0 ^-<—•—i—•—i—•—•—•—•—i i i i i i i i i i • " • i i_^ 1.5 2.0 2.5 3.0 3.5 4.0 Temperature [ K ]

Figure 2.13: Temperature dependence of the critical field for Ni(/z-OH2)2[Au(CN)2]2 determined from field-dependent ac susceptibility measurements (•). Transition tem perature (Tf) at zero-field obtained from //SR experiments (o).

corresponds to the position of the inflection point of the isothermal dc magnetization curve. The ac results are consistent with the dc magnetization measurements performed at the same temperature. This is expected since the in-phase component M' is a measure of the instantaneous slope dM/dH (M being the magnetization) observed in dc experiments. The critical field at which a maximum is observed in the in-phase component M' was determined as a function of temperature up to 3.2 K (Figure 2.13). As the temperature increases, the maximum shifts to lower field and above 3.2 K, no distinct maximum in the in-phase component could be observed. Temperature dependent ac susceptibility measurements were performed upon cool ing for Ni(//-OH2)2[Au(CN)2]2 with an applied ac field of 5 Oe and a driving frequency ranging from 1.00 to 1488.10 Hz (no dc field was applied). The results are shown in Figure 2.14. As the temperature decreases, the in-phase component of the ac sus ceptibility {XM') increases and reaches a maximum at low temperature. As shown in Figure 2.14B, a frequency dependence is observed below 4 K. The position of the maximum shifts to higher temperature (2.16 to 2.42 K) as the frequency is increased Chapter 2. Magnetic exchange through double aqua-bridges 65

A. t I | I I II | H • • | I I I I | I I • • | • I I • | • II I | • I • I | I I I I | 1I 0 9 8 1.5 7 ; § 6 53 1.0 5 4 3 0.5 2 1

0.0 "it • t t t t • • H 0 • • • .TWWM i i 2345 6 789 10 2.0 2.5 3.0 Temperature [ K ] Temperature [K]

Figure 2.14: A. Temperature and frequency dependence of the in-phase (XM, empty symbols) and out-of-phase (XM" , filled symbols) components of the ac susceptibility of Ni(/x-OH2)2[Au(CN)2]2 measured under zero dc field with an ac field of 5 Oe (1.00 (o, •); 1488.10 Hz (•, •)). B. Frequency dependence of the in-phase component at low temperature: 1.00 (o), 10.00 (v), 38.58 (A), 997.34 (0), 1488.10 Hz (•). Lines are guides for the eye.

from 1.00 to 1488.10 Hz. An increase in the out-of-phase component of the ac sus ceptibility, XM\ is observed below 2.5 K and a frequency dependence can also be observed.

/LtSR results

In order to clarify the nature of the phase transition suggested by the maxima in ef fective moment (faff) and susceptibility (XM) observed in the SQUID measurements,

ZF-//SR experiments were performed on Ni(/w-OH2)2[Au(CN)2]2. The ZF-//SR asym metry spectra for Ni(/i-OH2)2[Au(CN)2]2 (Figure 2.15) were best fit with the following function: At A(t) = a0Pz(t) = as C^e- + ^e~A + aAg (2.8) Chapter 2. Magnetic exchange through double aqua-bridges 66

0.30 i i i i—i | i i i i > ' ' ' i

0.25 U

0.20

'0.15

0.10

0.05

0.00 i i i i • i 2 4 6 5 10 15 Time [ \xs ] Temperature [ K ]

Figure 2.15: A. ZF asymmetry spectra for Ni(jU-OH2)2[Au(CN)2]2 at 14.5 (A), 5.0 (•) and 0.015 K (o). The solid curves are fits to Equation 2.8. The temperature dependencies of (B.) the relaxation rates A (•) and (C.) A (o). Lines are guides for the eye.

where the term containing as describes the signal originating from muons stopping in the sample and, as before, a,Ag is the temperature-independent background signal.

This is a similar function to that used in fitting the Cu(/i-OH2)2[Au(CN)2]2 data (Equation 2.6), although in this case an oscillatory component is not required. There is no indication of long-range magnetic order in Ni(^-OH2)2[Au(CN)2]2 down to 0.015 K. Figures 2.15B and 2.15C show the temperature dependences of the relaxation rates, A and A, of the | and | components respectively. A maximum can be observed in A(T) followed by a rapid decrease of A to zero as the temperature falls below 3.6 K. This indicates that a spin freezing transition occurs at Tf ~ 3.6 K. Below Tf, the initial relaxation is so fast that the early part of the signal is not observed, and only the slow-relaxing "| tail" is fully visible. At these temperatures, Chapter 2. Magnetic exchange through double aqua-bridges 67 the relaxation rate A is very large, indicating that the distribution of local static fields at the muon sites is extremely broad. All of these observations are consistent with the Ni moments freezing into a state below Tf that resembles a spin-glass. (n°)

2.3.3 Co(/i-OH2)2[Au(CN)2]2

The magnetization of Co(/i-OH2)2[Au(CN)2J2 as a function of temperature was mea sured upon cooling and the corresponding magnetic susceptibility and effective mag netic moment were determined. Figure 2.16A shows the temperature dependence of the effective magnetic moment determined in an external field ranging from 1 kOe to 30 kOe. At 300 K, the effective magnetic moment has a value of ~5.0 /ig, irrespective of the magnetic field. Between 300 and 25 K, the effective magnetic moment is nearly field independent and a slow decrease is observed (inset Figure 2.16A). At 25 K, val ues of 4.5-4.6 HB are observed under the different fields. Below 25 K, the effective moment determined under a field of 1 kOe increases steadily to reach a value of 5.67 jj,B at 1.8 K. When the sample is cooled in a field of 5 kOe or more, the effective moment also increases below 25 K, but reaches a maximum before decreasing until the temperature reaches 1.8 K. The temperature at which a maximum occurs is field dependent: the maximum occurs at 3.0 K, 5.5 K, and 9.0 K respectively in a magnetic field of 5, 10 and 20 kOe. When a field of 30 kOe is applied, the effective moment remains stable between 25 and 15 K and then drops to reach a value of 2.48 HB at 1.8 K. The values observed at room temperature (in every applied field) are larger than the value expected for a spin-only S — § system (3.87 /jg). They however fall in the range of values usually observed for high-spin Co(n) centers in an octahedral environment (4.7-5.2 HB)^111^ for which spin-orbit coupling is important (see sec tion 1.3.2). The initial decrease in effective moment is attributed to the depopulation of the spin-orbit coupled excited states. The different behaviours below ~ 20 K suggest a field-dependent magnetic state. The increase in effective moment and plateau, observed below 25 K, indicate the presence of ferromagnetic interactions between the Co(ll) centers. Chapter 2. Magnetic exchange through double aqua-bridges 68

A. I i i i • • i i i i • i i • • • i • • • • i • • • • i j 6.0

5.0 591819$ $ $ 8 $ 8 j =C 4.0

'% 3.0 =3.

2.0

1.0 0 100 200 300 0.0 J ' • ' ' f • f * • I » i > i I i i * * I i i i i I QQ n.... i i 0 10 20 30 40 50 0 10 20 30 40 50 60 70 Temperature [ K ] Field [ kOe ]

Figure 2.16: A. Temperature dependence of the effective magnetic moment (/ieff) °f Co(/i-OH2)2[Au(CN)2]2 measured in a 1 (o), 5 (v), 10 «>), 20 (•) and 30 kOe (A) dc field. B. Field dependence of the magnetization of Co(/i-OH2)2[Au(CN)2]2 (•) at 1.8 K compared with the Brillouin function (solid line) for an S = § paramagnetic system.

The isothermal magnetization determined at 1.8 K for Co(/i-OH2)2[Au(CN)2]2 as the field is increased from 0 to 70 kOe is shown in Figure 2.16B. A rapid increase between 0 and 10 kOe can be seen followed by a slower increase to reach a value of

2.39 fiB at 70 kOe. A comparison is made with the behaviour expected for a paramagnetic S = | system following the Brillouin function (Equation 1.9). Between 0 and 12 kOe, the magnetization of Co(/Li-OH2)2[Au(CN)2]2 is larger than the predicted values, but falls below the expected values above 13 kOe and never reaches the predicted saturation value. The faster increase of the magnetization in the small field region is also con sistent with the presence of ferromagnetic interactions in Co(/x-OH2)2[Au(CN)2]2. Chapter 2. Magnetic exchange through double aqua-bridges 69

2.3.4 Fe(^-OH2)2[Au(CN)2]2 and Mn(^-OH2)2[Au(CN)2]2

The magnetic susceptibility of the Fe(/u-OH2)2[Au(CN)2]2 and Mn(/it-

OH2)2[Au(CN)2]2 coordination polymers was determined as a function of tem perature upon cooling in a 1 kOe external field (Figure 2.17A). The effective magnetic moment determined for Fe(/W-OH2)2[Au(CN)2]2 was found to be relatively stable between 300 and 75 K at value of 5.3 \XB- Below 75 K, the effective magnetic moment drops to reach 2.0 JIB at 1.8 K. Similarly, the effective magnetic moment of

Mn(/it-OH2)2[Au(CN)2]2 was found to have a value of 5.6 ^B between 300 and 50 K, before dropping to reach 2.1 JIB at 1.8 K. For both polymers, no maximum could be observed in the temperature dependence of the magnetic susceptibility. The temperature dependence of the effective magnetic moment for both polymers was found to be independent of the external field in the 100-5000 Oe range. The room temperature effective magnetic moment determined for the Fe(il)- containing polymer is slightly larger than the expected value for a spin-only S = 2 paramagnetic system (4.89 /J,B). On the other hand, for the Mn(n)-containing poly mer, the value is slightly smaller than the spin-only value for a S = § paramagnetic system (5.91 HB)- The drop in effective magnetic moment in both polymers sug gests the presence of antiferromagnetic interactions occurring at low temperatures, but rather weak as no maximum in the magnetic susceptibility was observed.

The magnetic susceptibility of Mn(/i-OH2)2[Au(CN)2]2 was fit to the analytical expression derived by Fisher (Equation 2.9) for an infinite chain of equally spaced classical spins. (112) Due to the classical approximation, this expression is best when used for large values of S, usually | and above, and was thus not used to fit the

Fe(/z-OH2)2[Au(CN)2]2 data.

NA9^%S(S+1) fl+u\ (9Q) K Xchain - 3ksT \l-uj '

JS(S + 1) k T u = coth B kRT yjs{s + i) In this equation, J represents the exchange coupling constant along the chain. The fit to this equation, shown in Figure 2.17A, yielded a J value of -0.33(2) cm-1 (g = Chapter 2. Magnetic exchange through double aqua-bridges 70

1 I I I I | I I I I J I I I » | I 'T I I '|"l I II | II I I 'I T '—i— i > i ' i • • i > i > i A. B. 4.0

6.0 r—i . nnnnoaoooouuuaol ^ 3.0 . • " 5.0 2.0 ^.. 1.0 - ^ s •*. 4.0 0.0 i i i i i i i i i i i • i n— 1 i i • i •'" i-' v" • i" ' i - C. 5.0 , 1 . 3.0 =e 4.0 D " D ^ 3.0 D I__J „nD° 2.0 . ?0 o"°D _ • ° • ?, 1.0 - / " 1.0 1 * * • ' ' • * * * ' » * » * * • • • • ' ' • ' ' ' • • • • ' ft ft 1 . 1 50 100 150 200 250 300 0 10 20 30 40 50 60 70 Temperature [ K ] Field [kOe]

Figure 2.17: A. Temperature dependence of the effective magnetic moment (fieff) of

Fe(/i-OH2)2[Au(CN)2]2 (•) and Mn(//-OH2)2[Au(CN)2]2 (o) measured in 1 kOe dc field; the solid line corresponds to the fit to Equation 2.9 for the Mn-containing poly mer. Field dependence of the magnetization of (B.) Fe(/x-OH2)2[Au(CN)2]2 (•) and

(C.) Mn(/x-OH2)2[Au(CN)2]2 (•) at 1.8 K compared with the corresponding Brillouin function (solid lines) for an S = 2 and S = | paramagnetic system respectively.

1.92(2)) for Mn(/x-OH2)2[Au(CN)2]2. This small negative coupling constant suggests weak antiferromagnetic interactions along the chains. The isothermal magnetization of the two polymers was determined as a function of field, between 0 and 70 kOe, at 1.8 K. Figure 2.17B and 2.17C show a comparison between the results obtained experimentally for Fe(/x-OH2)2[Au(CN)2]2 and Mn(/x-

OH2)2[Au(CN)2]2 and the Brillouin functions (Equation 1.9) for S = 2 and S = § paramagnetic systems respectively. In both polymers, the magnetization has a zero value under zero applied magnetic field. As the field is increased, the magnetization increases steadily to reach 2.69 and

3.8 NAUB under a field of 70 kOe for the Fe(ll)- and Mn(ll)-containing polymers respec tively. Differences can be observed between the expected behaviour for a paramagnetic system and the experimental data, which suggest that the Fe(/^-OH2)2[Au(CN)2]2 and Chapter 2. Magnetic exchange through double aqua-bridges 71

[ I I I I I I I I I | I I I I | I I I I | I T I l-T'T-r1"!-! |~

4.0

f^ -.°5 : -< -id ^ l_l =tt =c 2.0 r • •

i n r • • • ' I . ... I ... • I • • • • I . • . • I . . . . I 0 50 100 150 200 250 300 Temperature [ K ]

Figure 2.18: Temperature dependence of the effective magnetic moment (faff) of Fe(//- OH2)(jU-OH)[Au(CN)2]2 measured in 10 kOe dc field.

Mn(/i-OH2)2[Au(CN)2]2 polymers are not in a purely paramagnetic state at 1.8 K. The smaller magnetization over the field range studied is consistent with the presence of antiferromagnetic interactions between the M(ll) centers at 1.8 K.

2.3.5 Magnetic properties of Fe(/z-OH2)(/x-OH)[Au(CN)2]2

The magnetic behaviour of Fe(/x-OH2)(/x-OH)[Au(CN)2]2 was determined as a func tion of temperature and is shown in Figure 2.18. At 300 K, in a magnetic field of 10 kOe, the effective moment has a value of 3.59 /J,B, which drops steadily upon cooling until 20 K (2.76 /«#). Below that temperature, a steeper drop in effective moment occurs and a final value of 1.65 HB was determined at 2.0 K. No maximum could be observed in the temperature dependence of the susceptibility. The value observed at room temperature is much smaller than the expected value for an S = | system, 5.91 \XB- The room temperature value along with the continuous decrease in effective magnetic moment could indicate the presence of antiferromagnetic interactions between the Fe(lll) centers in Fe(yu-OH2)(//-OH)[Au(CN)2]2. Chapter 2. Magnetic exchange through double aqua-bridges 72

2.3.6 Preliminary /iSK results

ZF-jiSR experiments were performed for the M(^-OH2)2[Au(CN)2]2 (M = Co(n),

Fe(ll) and Mn(ll)) and Fe(^-OH2)(/i-OH)[Au(CN)2]2 coordination polymers and the asymmetry spectra were determined at different temperatures. Over the temperature range studied, no precessing signal was observed in the asymmetry spectra of any polymer. This suggests that no internal magnetic field resulting from long-range ordering, is present in these systems. However, from the temperature dependence of the relaxation rate of the muon polarization function A(T) determined for each polymer, it was determined that all of them undergo a spin-freezing transition, similar to that of the Ni(n) analogous polymer. This is indicated by a saturation of the muon spin relaxation rate below ~2 K (similar to the behaviour shown in Figure 2.15B for the Ni(ll) analogue). For the Co(/x-OH2)2[Au(CN)2]2 and Fe(/>OH2)(/i-OH)[Au(CN)2]2 polymers, below the transition, slower relaxation rates were observed compared to the Ni(ll)-containing polymer, whereas very fast relaxation rates were observed for the Fe(ll) and Mn(ll)- containing polymers. The magnitude of the relaxation rate is an indication of the width of the local internal field distribution sensed by the muon: a slow relaxation implies a narrower distribution (Co(ll) and Fe(lll)) while a fast relaxation implies a broader field distri bution (Fe(n) and Mn(li)). The width of the local static field distribution is affected by both the size of the local magnetic moments and the degree of the correlation between the magnetic moments.0 Further data analysis is in progress in the Sonier group to fully characterize the spin-frozen zero-field magnetic ground states of these polymers.

CA higher degree of correlation gives rise to a narrower field distribution and, hence, a slower relaxation rate. Chapter 2. Magnetic exchange through double aqua-bridges 73

< J, •

Scheme 2.3: Possible magnetic pathways in M(/i-OH2)2[Au(CN)2]2: A. interactions along the M(//-OH2)2M chains (Ji), viewed down the c-axis; B. interchain interactions (J2, J3), viewed down the chains. Dotted lines indicate hydrogen bonding.

2.4 Discussion

2.4.1 Magneto-structural correlations

Magnetic exchange between the metal centers in the M(/>OH2)2[Au(CN)2]2 and Fe(//-

OH2)(/i-OH)[Au(CN)2]2 coordination polymers can occur through several magnetic pathways, which are illustrated in Scheme 2.3. The interactions mediated through the short M^-OH^M bridges, Ji, are probably the strongest possible interactions in every polymer. The type and strength of interactions mediated by the oxygen atom in such a bridge depends on the extent of overlap between the magnetic orbitals of the metal centers and the orbitals of the oxygen atoms; (71) this, in turn, depends on the structural arrangement, i.e. the bond lengths and angles in the M(//-OH2)2M core. Since the exact positions of the oxygen atoms could not be determined for most of the M(/x-OH2)2[Au(CN)2]2 polymers, a quantitative comparison cannot however be made. Chapter 2. Magnetic exchange through double aqua-bridges 74

Other possible magnetic interaction pathways, albeit weaker than J\ due to the longer distances involved, likely utilize hydrogen-bonding between the [Au(CN)2]~ units and the water molecules. This can occur either through the M-OH- • • N- • • H0-

M (J2) or M-OH- • • NCAuCN-M (J3) pathways. Each of these building blocks has previously been reported as a mediator of mag netic exchange. Hydrogen-bonding, involving different donor and acceptor atoms, was shown to favour magnetic coupling in different systems.^113' More specifically, water molecules were reported to mediate both ferro- and antiferromagnetic interac tions through hydrogen-bonding between different spin carriers, including Cu(ll)/114^ Cr(m),(115) nitronylnitroxide radicals/116^ and mixed Ni(n)-radical systems.^117) On the other hand, the [Au(CN)2]~ unit has been observed to mediate magnetic interac tions between several first-row transition metals at low temperature. (57>65) The differences between the magnetic behaviour observed for the M(/i-

OH2)2[Au(CN)2]2 and Fe(/x-OH2)(£t-OH)[Au(CN)2]2 polymers can arise mainly from (1) the slight structural differences in the double aqua-bridges (distances, angles, asymmetric vs symmetric motif), (2) the differences in the metal centers' geometry (highly symmetrical vs distorted) and (3) the number of unpaired electrons (1-5 unpaired electrons) and their associated magnetic orbitals. A deeper understanding of the magnetic behaviour of each M(/i-OH2)2[Au(CN)2]2 polymer can be gained by combining the SQUID magnetometry and ZF-//SR results, and some magneto-structural correlations can be drawn.

2.4.2 Magnetic exchange through double aqua-bridges

The magnetic behaviour observed for the Cu-, Ni- and Co-containing M(/x- OH2)2[Au(CN)2J2 polymers indicates the presence of weak ferromagnetic interactions between the M(ll) centers. As mentioned above, these interactions are most likely present along the chain axis, and mediated through the double aqua-bridges (Ji > 0). The sign of the interactions implies that the magnetic orbitals involved are orthogonal to each other in this direction. (71) Chapter 2. Magnetic exchange through double aqua-bridges 75

Cu(/z-OH2)2[Au(CN)2]2

According to the structure proposed for Cu(/z-OH2)2[Au(CN)2]2, the Cu(n) centers are in an axial-equatorial arrangement along the Cu(/U-OH2)2Cu chain axis. Despite the exhaustive magnetic studies carried out for a range of equatorial-equatorial oxygen- bridged Cu(n) dimers,'71'118' which include hydroxo, aqua, acetate, carboxylate, and alkoxo-bridged systems, the magnetic behaviours of axial-equatorial oxygen-bridged Cu(ll) dimers have not been investigated. Given the lack of data, no clear magneto- structural trends have been reported. '94' In general, very weak interactions are mediated through axial-equatorial asym metric Cu(n) bridges, due to poor orbital overlap between the metal center and the bridging atom occupying the axial positions. This interaction weakens as the axial distance increases, which causes a decrease in orbital overlap. For example, complexes with a Cvi-Oaxiai distance between 2.3 and 2.4 A were found to be ferromagnetic with a coupling constant J of up to 10 cm"1.'119,120' In contrast, a larger distance (~2.5 A) generated a weakly antiferromagnetic complex (J ~ -0.5 cm"1).'121' Extended Hiickel calculations performed on similar axial-equatorial dichloro- bridged Cu(li) systems'122'123' showed that, for an ideal geometry with 90° angles, no magnetic coupling should exist between the Cu(ll) centers. Any small ferro- or antiferromagnetic coupling observed for such systems was attributed to structural deviations from ideality. Reliable bond lengths and angles involving the oxygen atoms could not be de termined for Cu(/i-OH2)2[Au(CN)2]2 and, therefore, it is difficult to put the struc tural parameters into context. Nevertheless, the small exchange coupling constant obtained by fitting the data with the Baker expression (Equation 2.3, Jsaker = J\ = 0.37(2) cm-1) agrees with the few previously reported values for axial-equatorial oxygen-bridged Cu(ll) systems. (119>120)

Ni(/x-OH2)2[Au(CN)2]2

Very few aqua-bridged Ni(il) dimers or chains have been reported,'95"99' and of these, the magnetic properties were rarely investigated. Antiferromagnetic interactions were Chapter 2. Magnetic exchange through double aqua-bridges 76

Table 2.5: Comparison between the magnetic behaviours of the different Ni(/Lt- OH2)2Ni containing complexes.

Complex Ni-0 Ni-O-Ni O-Ni-0 J (A) (deg) (deg) (cm-1) Ni(/i-OH2)2[Au(CN)2]2 2.156(11) 100.6(7) 79.4(7) >0 96 Ni2(Hpdc)2(/x-H20)2(H20)2( ) 2.075(2) 100.94(9) 79.05 -1.6(2) 2.145(2) 10 Ni2(/i-OH2)2(O2CFcCO2)2(bipy)2( °) 2.108(3) 99.83 80.17 8.15 2.138(3)

observed in the double-stranded chain [Ni2(Hpdc)2(/Lt-H20)2(H20)2] (H3pdc = 3,5- pyrazoledicarboxylic acid), in which the water molecules are linking two [Ni(Hdcp)] chains together. ^96^ The double aqua-bridge was determined to be the dominant mag netic pathway in this system, with a coupling constant of -1.6(2) cm-1.

On the other hand, ferromagnetic interactions were found to be present in [Ni2(//-

OH2)2(02CFcC02)2(bipy)2] (Fc = Ferrocene; bipy = 2,2'-bipyridine), which contains aqua-bridged Ni(ll) dimers linked through ferrocenedicarboxylic ligands. (10°) Magnetic interactions through the aqua-bridges were determined to be stronger, with a coupling constant of + 8.15 cm-1. The similar structural parameters (Table 2.5) but different magnetic behaviour obtained for these aqua-bridged Ni-containing compounds limit the discussion of any general magneto-structural correlations regarding Ni(/x-OH2)2Ni units. More vari ables, such as the complete coordination sphere of the Ni(n) ions (orientation of the magnetic orbitals) and the presence of other bridging ligands, would also need to be taken into account to explain the different type of magnetic exchange.

Fe(^-OH2)2[Au(CN)2]2 and Mn(p-OH2)2[Au(CN)2]2

Despite their similar structure, the magnetic behaviour of the Fe(ll) and Mn(li)- containing M(,ii-OH2)2[Au(CN)2]2 polymers differs greatly from that of the Cu(ll), Ni(ll), and Co(ll)-containing polymers. The magnetic data determined for these two Chapter 2. Magnetic exchange through double aqua-bridges 77 polymers indicate that weak antiferromagnetic interactions are present between the metal centers at low temperature. As for the other polymers, it is proposed that these interactions, despite the different sign, are mediated through the short M(/i-OH2)2M pathways (J\). The different type of interactions could be due to the larger number of magnetic orbitals (4 and 5) involved in the Fe(ll) and Mn(n)-containing polymers. Despite the orthogonality of some magnetic orbitals, the presence of any orbital overlap would yield overall antiferromagnetic interactions between the metal centers. It is also possible that the M-O-M angle varies amongst the different M(/J,- OH2)2[Au(CN)2]2 polymers. Larger M-O-M angles are usually associated with the mediation of antiferromagnetic interactions in several systems, whereas smaller an gles mediate ferromagnetic interactions. An increase in M-M distance was observed when going from the Ni(ll) structure (3.3183(11) A) to the Fe(ll) (3.465 A) and the Mn(ll) structure (3.49 A), which can be attributed in part to an increase in atomic radii. Longer M-M distances can impose larger M-O-M angles, unless there is a large increase in M-0 distances. The change in angle could allow more orbital overlap and cause stronger antiferromagnetic interactions between the metal centers. However, as mentioned earlier, the positions of the oxygen atoms could not be determined with certainty for most M(/i-OH2)2[Au(CN)2]2 polymers and no structural comparison can be made to explain the difference in interactions. The magnetic data reported for the few doubly aqua-bridged Fe(ll) and Mn(ll) dimers known also indicated that very weak antiferromagnetic interactions were me diated through these bridges/102"104) The reported coupling constant for these com plexes ranges between -0.02 and -1.38 cm-1 (with M-CHV1 angle varying between 100 and 113°). These coupling constants are on the same order of magnitude as the value obtained for the Mn(/i-OH2)2[Au(CN)2]2 polymer using the Fisher expression (JFisher = J\ = -0.33(2) CHI"1). Chapter 2. Magnetic exchange through double aqua-bridges 78

Fe(/x-OH2)(/i-OH)[Au(CN)2]2

The magnetic properties observed for Fe(/i-OH2)(/i-OH)[Au(CN)2]2 also suggest that antiferromagnetic interactions are mediated along the shortest J\ magnetic pathway.

The nature of this pathway however differs in Fe(/i-OH2)(/i-OH)[Au(CN)2]2 and could consist of either alternating double aqua-bridges/double hydroxide-bridges or mixed aqua-hydroxide bridges.

The presence of stronger magnetic interactions in Fe(//-OH2)(/i-OH)[Au(CN)2]2, compared to the M(/i-OH2)2[Au(CN)2]2 polymers, likely result from the hydroxide groups present between the Fe(in) centers. The properties of hydroxide bridges have been widely investigated, especially in Cu(ll) dimers, and were shown to be good mediators of both ferro- and antiferromagnetic interactions with coupling constant ranging from -509 to 178 cm"M124' For the doubly hydroxide-bridged dimers, the M-O-M angles as well as the position of the hydrogen atoms with respect to the

M02M plane (in or out of the plane) were found to be determinant. Small M-O-M angles and large out of plane M-O-H angles were found to favour strong ferromagnetic interactions in Cu(il) dimers. (125) The reason why aqua-bridges are weak mediators of magnetic interactions com pared to the hydroxide-bridges is unclear at this point. Slight structural differences between the two bridges are sometimes observed, with M-0 distances being slightly shorter in a hydroxide bridge (most likely due its charge). (107) These structural differ ences could explain in part the different magnitudes in coupling constants, but other variables such as the different molecular orbitals involved and the extent of overlap with those of the metal centers should also be taken into consideration. Theoretical calculations should be performed to compare the magnetic exchange capability of wa ter vs hydroxide bridges and explain the different mediator strength observed for the two bridges. Very few examples of aqua-hydroxide bridged complexes are known and the mag netic behaviour of most of them have not been investigated. As mentioned earlier, an aqua-hydroxide bridged high-spin Fe(ll) dimer has been reported. The magnetic 3+ behaviour of [Fe2(^-OH2)(/i-OH)(TPA)2] (TPA = tris(picolylamine)) shows signs of Chapter 2. Magnetic exchange through double aqua-bridges 79 significant antiferromagnetic interactions and a coupling constant of-9.6 cm-1, medi ated through this mixed bridge, was determined. (lor) Despite the differences between this dimer and Fe(/^-OH2)(/i-OH)[Au(CN)2]2 (including the different S value), it is likely that the aqua-hydroxide bridge mediates similar type of interactions in both systems.

2.4.3 Importance of weak interchain interactions

Interchain magnetic interactions, although weak, must be present to explain the type of magnetic ground state obtained in every M(/U-OH2)2[Au(CN)2]2 and Fe(/x-OH2)(/t/- OH)[Au(CN)2J2 polymer. These non-negligible interactions are believed to be medi ated through hydrogen-bonding and [Au(CN)2]~ units, along the J2 and J3 pathways.

Long-range order in Cu(/x-OH2)2[Au(CN)2]2

In Cu(/i-OH2)2[Au(CN)2]2, interactions between the ferromagnetically coupled chains yield a long-range magnetically ordered state below 0.20 K (TN), as detected by the ZF-/^SR experiments. Observation of a three-dimensionally ordered system im plies that magnetic interactions through the long J2 and J3 pathways must be non- negligible. It also suggests that magnetic exchange along J2 must be ferromagnetic, as antiferromagnetic interactions along J2 would lead to a frustrated system, irrele vant of the sign of J3 (see below for more explanations). Depending on the type of magnetic exchange along J3, the overall long-range order in Cu(/i-OH2)2[Au(CN)2]2 could be either ferromagnetic (J3 > 0) or antiferromagnetic (J3 < 0). Scheme 2.4A summarizes the magnetic model proposed for Cu(^-OH2)2[Au(CN)2]2- When the experimental data was fit with a mean field approximation (Equa tion 2.4), which does not differentiate between the different possible interchain mag netic pathways, a negative zJ' coupling constant was obtained. This would be consis tent with overall antiferromagnetic ordering occurring in Cu(/x-OH2)2[Au(CN)2]2 and a negative J3 parameter. Recent theoretical studies on quasi 1-D systems linked the ratio of the ordering temperature and the coupling constant along the chain (T/v/J) to the interchain J' Chapter 2. Magnetic exchange through double aqua-bridges 80

Scheme 2.4: Schematic representation of the magnetic interchain interactions leading to different magnetic ground states: A. long-range order in Cu(/>OH2)2[Au(CN)2]2 (J2 > 0), B. a spin-glass like state in the intrachain ferromagnetically coupled (J\ > 0) Ni and Co-containing polymers (J2 < 0), and C. a spin-glass like state in the intrachain antiferromagnetically coupled (Jx < 0) Fe(/x-OH2)(/x-OH)[Au(CN)2]2 and Fe(ll) and Mn(n)-containing M(/>OH2)2[Au(CN)2]2 polymers. value through the empirical formula: (126^ T, J' = N (2.10) _4cv/ln(M) + Iln(ln(M))

where c = 0.233 and A = 2.6 for an S = \ system. If the TN and J values determined -1 for Cu(^-OH2)2[Au(CN)2]2 are used in Equation 2.10, a J' value of -0.08 cm is predicted by this formula. This value is similar to the zJ' value obtained using Equation 2.4: if z is assumed to be four or six, J' would equal —0.14 or —0.09 cm-1 respectively. This is consistent with the structure of Cu(/i-OH2)2[Au(CN)2]2 where either four or six chains can be considered as neighbouring chains (Scheme 2.3). Chapter 2. Magnetic exchange through double aqua-bridges 81

Spin-glass like magnetic states containing ferromagnetically coupled chains

Despite the similar type of coupling along the chains, the magnetic behaviours of the

Ni and Co-containing M(/u-OH2)2[Au(CN)2]2 polymers were found to differ from that of the Cu(li) analogue at lower temperature.

In Ni(//-OH2)2[Au(CN)2]2, the maxima/plateaus observed in susceptibility (XM) (Figure 2.11), when a small field is applied, suggest that the ferromagnetically coupled chains interact in an antiferromagnetic manner with each other through the J2 or J3 pathways. If the external field is increased (larger than 1 kOe), these weak interchain interactions can be overcome and, as the chains align with the external field, an increase in susceptibility is observed. This field-dependent behaviour at low temperature would be consistent with either antiferromagnetic ordering occurring between the chains or the formation of an overall spin-glass like magnetic state. A spin-glass state is defined as a frozend magnetic state in which spins are oriented in a random fashion. This term is also used more broadly to describe randomly oriented magnetic clusters or domains. The combination of ferromagnetic interactions along the chain and antiferromagnetism/spin-glass behaviour between the chains would also explain the sigmoidal shape of the isothermal magnetization curve at low temperature. Below the critical field, the magnetization increases very slowly due to interchain interactions, but these are overcome above a critical field causing the magnetization to increase faster. At 1.8 K, the critical magnetic field was determined to be 700 Oe. As the temperature increases, the critical field decreases, as it becomes easier to align the chains with a smaller external field. This type of magnetic behaviour is similar to that of a metamagnetic system. (70>127) Such systems are commonly composed of ferro- or ferrimagnetic sheets or chains that interact or order antiferromagnetically at low temperature and field, but undergo a transition to a ferromagnetic state at a critical field. Examples of metamagnetic coordination polymers include monometallic azido-Ni(n) (128>129) and Fe(ll)^130^ com plexes, malonato-Cu(ll) (131) systems as well as a range of bimetallic or metal-radical

dThe spins appear frozen on the time scale of the measurements. Chapter 2. Magnetic exchange through double aqua-bridges 82 systems. (132_135) However, metamagnet type transitions have not, to our knowledge, been reported in the case of the zero-field state being a spin-glass.

The field-dependent behaviour of the Co(/x-OH2)2[Au(CN)2]2 polymer (below 25 K) is shifted toward lower temperatures compared to that of the Ni-analogue.

While Ni(/x-OH2)2[Au(CN)2]2 shows a maximum in effective moment at 6.4 K under

a field of 10 kOe, the effective moment of Co(/i-OH2)2[Au(CN)2]2 reaches a maximum at 5.5 K under the same field. No maximum in effective moment can be seen in a field of 1 kOe, but the steady increase suggests that a maximum might occur below the range of temperature reachable with the SQUID magnetometer. This suggests the presence of weaker interchain interactions in the Co-based system.

The Co(/u-OH2)2[Au(CN)2]2 polymer also differs from the Ni(//-OH2)2[Au(CN)2]2 polymer as no maximum in susceptibility (or plateau) and no inflection point in the isothermal magnetization curve (at 1.8 K) can be observed. These differences discard the possibility of a transition to a metamagnetic or spin-glass state, in the magnetic field range studied, above 1.8 K. A transition could however be happening at lower temperature or lower field, which would be consistent with weaker interchain interactions. To clarify the nature of the magnetic state of the Ni and Co-containing polymers under zero-field, that is, whether antiferromagnetic ordering or a spin-glass like be haviour is operative, the results of the ZF-/uSR experiments were invaluable. Indeed, the asymmetry spectra of both polymers, down to 0.015 K, indicate that no long- range order, whether ferromagnetic or antiferromagnetic, is present in zero external field. However, the results are characteristic of a spin-glass like state, below 3.6 K for

Ni(^-OH2)2[Au(CN)2]2 and ~ 2.0 K for Co(/u-OH2)2[Au(CN)2]2. This means that, on the time scale of the /iSR measurements, which is much shorter than that of the SQUID measurements, the magnetic moments arc frozen in a random orientation.

For Ni(|U-OH2)2[Au(CN)2]2, the presence of an out-of-phase signal and the fre quency dependence of both phase components in the ac susceptibility (Figure 2.14) also support the formation of a spin-glass system. Although a difference is usually observed between the ZFC and FC magnetization Chapter 2. Magnetic exchange through double aqua-bridges 83 measurements performed on a spin-glass system, this was not observed for Ni(/^-

OH2)2[Au(CN)2]2. This was attributed to the fact that the glass transition temper ature approaches the temperature limit of the SQUID magnetometer and that the system cannot completely freeze under the experimental conditions. The formation of a spin-glass generally requires an element of spin frustration from competing magnetic interactions. As shown in Scheme 2.4B, a negative coupling con stant J2 in the Ni and Co-containing polymers would lead to an overall spin frustration if J3 has a non-negligible value, whether positive or negative. As a result, a frozen disordered magnetic state is obtained below a critical temperature which depends on the magnitude of the competing coupling constants. Weaker coupling along the J3 pathway in Co(/i-OH2)2[Au(CN)2]2, compared to the Ni analogue, could explain the smaller level of frustration (higher spin correlation) observed by //SR and the lower critical temperature/field observed by SQUID magnetometry.

Disordered spin-frozen magnetic states containing antiferromagnetically coupled chains

As shown in Schemes 2.3A and 2.4C, when antiferromagnetic coupling is present along the chains (J\ < 0), any interactions between the chains along the J3 pathway would lead to strong magnetic frustration, irrespective of J2. The frozen disordered state observed by //SR for the Fe(ll) and Mn(n)-containing polymers confirms that, similarly to the other M(/i-OH2)2[Au(CN)2]2 polymers, mag netic interchain interactions exist along the J3 pathways. The higher level of magnetic disorder (lower degree of spin correlation) observed for these two polymers could result from the combination of competing Ji and J3 interactions.

On the other hand, the structure of the Fe(/Lt-OH2)(/x-OH)[Au(CN)2]2 polymer differs from that of the M(ll)-containing polymers, not only due to the presence of hydroxide bridges but also due to the crystallographic random half-occupancy of the Fe(lll) centers. Due to this random occupancy, the J3 pathway is not present in all the unit cells, but only in half of them, which should decrease the overall amount Chapter 2. Magnetic exchange through double aqua-bridges 84

of frustration in the system. Also, the structural model proposed for Fe(/i-OH2)(/i-

OH)[Au(CN)2]2 (section 2.2.3, Figure 2.4B) suggests the presence of Fe-NCAuCN-Fe bridges which generates an additional pathway between the metal centers. Mediation of magnetic interactions through this [Au(CN)2]~ unit could be either ferro- or antiferromagnetic. In both cases, this additional path would not introduce an element of magnetic frustration in the system. As a result of these structural differences, despite the presence of antiferromagnetic interactions along the chain (Jx < 0), a higher degree of spin correlation is present in the Fe(^-OH2)(/x-OH)[Au(CN)2]2 polymer compared to the Fe(ll) and Mn(ll)-containing M(^-OH2)2[Au(CN)2]2 polymers. More calculations on the /xSR results are in progress in the Sonier group to analyze the differences between the spin-frozen states observed for each M(/x-OH2)2[Au(CN)2]2 polymer.

2.4.4 Summary of magnetic behaviours

Table 2.6 summarizes the magnetic behaviours observed for the different M(/U-

OH2)2[Au(C.N)2]2 and Fe(/^-OH2)(//-OH)[Au(CN)2]2 polymers.

For the ferromagnetically coupled chains, the key J2 pathway is predicted to be positive in Cu(/i-OH2)2[Au(CN)2]2 and negative in Ni(/x-OH2)2[Au(CN)2]2 and Co(/x-

OH2)2[Au(CN)2]2. The extra magnetic orbitals in Ni(ll) and Co(ll) compared to Cu(ll) appears to provide an overlap that yields interchain antiferromagnetic interac tions. On the other hand, the Jahn-Teller bond lengthening in Cu(/U-OH2)2[Au(CN)2]2 effectively weakens all three magnetic interactions pathways, leading to a magnetic phase-transition one order-of-magnitude lower than the Ni or Co analogues. For the antiferromagnetically coupled chains, the non-negligible coupling along the Js pathway is the key to the overall interchain frustrations. The larger the J3 coupling constant is, the greater the magnetic frustrations (lower degree of correlation) will be.

As for the Ni and Co polymers, it is likely that the J2 pathway is also antiferromagnetic due to the orbital overlap, but the data did not allow to distinguish between the two types of interactions. Chapter 2. Magnetic exchange through double aqua-bridges 85

Table 2.6: Comparison between the magnetic behaviours of the different M(/i- OH2)2[Au(CN)2]2 polymers.

Metal S Intrachain Interchain Magnetic phase center interactions interactions State Transition h Jz temperature Cu(n) l >0 >0 ^0 long-range 0.20 K 2 0.37(2) cm"1 order Ni(n) 2 >0 <0 7^0 spin-glass 3.5 K 2 (disordered) 3 a Co(n) 2 >0 <0 ^0 spin-frozen -2.0 K weaker coupling highly correlated

Fe(n) 4 <0 7^0 spin-frozen" -2.0 K 2 considerable lower correlation

a Mn(n) 5 <0 7^0 spin-frozen -2.0 K 2 -0.33(2) cm"1 considerable lower correlation

Fe(ni) 5 n.a.b spin-frozena -2.0 K 2 <0 highly correlated

a Magnetic state currently being analyzed from //SR data. 6 This pathway differs in the Fe(/i-OH2)(M-OH)[Au(CN)2]2 structure.

2.5 Conclusion

A new basic structural motif among cyanometallate-based coordination polymers was identified in the five nearly isostructural M(//-OH2)2[Au(CN)2]2 (M = Cu, Ni, Co,

Fe, Mn) coordination polymers. The central M(/z-OH2)2M aqua-bridged chain motif is a surprisingly rare structural feature, and should serve as a new "point of refer ence" for further modification and research. Along with this family of isostructural polymers, the preparation of the related Fe(yU-OH2)(/i-OH)[Au(CN)2]2, in which one water molecule was replaced by a hydroxide group, was also presented. Investigations on the magnetic properties of these aqua-bridged polymers revealed Chapter 2. Magnetic exchange through double aqua-bridges 86 that the metal centers were ferromagnetically coupled along the chains in the Cu, Ni and Co-containing polymers, whereas antiferromagnetic coupling was present in the Fe and Mn-containing polymers. This study illustrates the importance of weak interactions and their power to im pact magnetic properties. Through interchain interactions involving hydrogen bond ing and [Au(CN)2]~ units, the chains were found to order magnetically in Cu(/x-

OH2)2[Au(CN)2]2 and form a spin-glass in Ni(/i-OH2)2[Au(CN)2]2 at low temperature. A series of frozen disordered states, with different degree of spin correlation, were ob tained for the Co(ll), Fe(ll) and Mn(ll)-containing M(^-OH2)2[Au(CN)2]2 polymers as well as for Fe(/x-OH2)(jU-OH)[Au(CN)2]2- Further analysis of the /iSR data will be required to confirm the type of interchain interactions present in these polymers.

2.6 Future work

As the interchain magnetic interactions are mediated through hydrogen bonding, it would be interesting to replace the hydrogen atoms by deuterium atoms and investi gate the changes in magnetic properties. The preparation of Ni(/i-OD2)2[Au(CN)2]2 was attempted using deuterated water as the solvent, but the presence of H20 from the starting materials could not be excluded. No significant differences were observed in the magnetic properties of the resulting product. This could indicate that either the deuteration process was not complete or that the replacement of hydrogen for deuterium atoms does not affect the magnetic pathways. The synthesis of a pure

Ni(/i-OD2)2[Au(CN)2]2 product should be attempted to answer that question. In addition to playing a structural role, the [Au(CN)2]~ units are also intimately involved in the mediation of magnetic interactions between the chains (J3). Since it is known that the isostructural [Ag(CN)2]~ units are generally able to mediate stronger (48) magnetic interactions than [Au(CN)2]~ units, it would be interesting to replace the [Au(CN)2]~ units by [Ag(CN)2]~ units and study the effects on the magnetic properties. The presence of a stronger magnetic pathway could yield higher ordering temperatures or could yield stronger magnetic frustration as two magnetic pathways compete to align the spins. Chapter 2. Magnetic exchange through double aqua-bridges 87

There is, however, some challenge in using [Ag(CN)2]~ units. This building block is less stable than [Au(CN)2]~ in solution and tends to rearrange to form - (48) the [NCAgCNAgCN]~ (i.e. [Ag2(CN)3] ) building block. Preliminary attempts at preparing the Ni(//-OH2)2[Ag(CN)2]2 analogous polymer from aqueous solutions yielded a Ni[Ag2(CN)3][Ag(CN)2] product. Different synthetic routes should be at tempted to make the desired [Ag(CN)2]-containing products.

2.7 Experimental Section

2.7.1 Reagents and general procedures for characterization

The experimental details described here apply to work reported in this chapter as well as all following chapters. All manipulations were performed in air unless otherwise noted. All reagents were purchased from commercial sources such as Sigma-Aldrich and Strem Inc. and used as received. CAUTION. Although we have experienced no difficulties, perchlorate salts are potentially explosive and should be used in small quantities and handled with care.

General procedures for hydrothernial synthesis.

Hydrothermal reactions were carried out in sealed 5 mL glass ampoules inserted into a 125 mL stainless steel reaction vessel, with 30-40 mL of water placed outside the ampoule to equalize the pressure. The vessel was heated in a Lindberg Heavy-Duty furnace equipped with a programmable temperature controller. The thermal profile will be specified for each reaction.

FT-IR spectroscopy, elemental analysis and thermogravimetric analysis

The infrared spectrum of each compound reported in this thesis was recorded on a Thermo Nicolet Nexus 670 FT-IR spectrometer with samples prepared as KBr pressed pellets. The resolution of the instrument was set at 1 cm-1 and the spectra were collected between 400 and 4000 cm-1. Chapter 2. Magnetic exchange through double aqua-bridges 88

All microanalyses (carbon, hydrogen and nitrogen) were performed at Simon Fraser University by Mr. M. K. Yang using a computer-controlled Carlo Erba (Model 1106) CHN analyzer. Thermogravimetric analyses (TGA) of all complexes were performed in an air atmosphere on a Shimadzu TGA-50 instrument at a rate of 5 °C per minute. Due to balance drift, the observed weight losses were approximated to the closest half integer.

UV-Vis-NIR absorption spectroscopy

Solid-state UV-Vis-NIR absorption spectra were measured by reflectance using an Ocean Optics SD2000 spectrophotometer equipped with tungsten halogen and deu terium lamps. Prior to analysis, the samples were ground into fine powders and deposited on a glass plate. Magnesium oxide powder, prepared in the same fashion, was used as a reference.

Powder X-ray diffraction

The powder X-ray diffractograms of compounds reported in this thesis were collected on several instruments. Generally, to verify sample purity, diffractograms were mea sured on a Rigaku RAXIS-Rapid Auto diffractometer. The Rigaku RAXIS-Rapid

Auto is equipped with a Cu Ka source (A = 1.54056 A). Samples were mounted on a glass fiber using grease and were exposed, as the phi axis was spinning (10 deg sec-1), for a period of 40 to 60 minutes (with the chi axis fixed at 0 ° and omega fixed at 90°). Other diffractometers used for data acquisition, especially to obtain high resolution diffractograms, will be described as necessary.

Single crystal X-ray diffraction

The following experimental details apply to all crystal structures determined and reported throughout this thesis (Chapter 2-5), unless otherwise indicated. Single crystals were mounted on glass fibers using epoxy adhesive. Single crystal X-ray diffraction data were recorded at room temperature on an Enraf Nonius CAD4F Chapter 2. Magnetic exchange through double aqua-bridges 89

diffractometer equipped with a Mo Ka source and controlled by the DIFRAC pro gram. (136) The NRCVAX Crystal Structure System was used to perform the psi-scan absorption correction and data reduction, including Lorentz and polarization correc tions. (137) The crystal structure of each compound was solved using the Sir 92 routine and refined through Fourier techniques using the CRYSTALS software package. ^138^ Unless otherwise stated, all non-hydrogen atoms are refined anisotropically. Full matrix least-squares refinement on F was performed for each compound and the re sulting R% and wR2 can be found in the corresponding crystallographic data tables (along with the number of reflections included and the number of parameters in volved). Hydrogen atoms were placed in geometric positions in all compounds and their coordinates allowed to ride with their associated atoms. Diagrams were prepared using Ortep-3 (version 1.076)(-139^ and POV-Ray (version 3.6.0). (14°)

SQUID magnetometry

Magnetization measurements for each compound reported in this thesis were per formed as follows unless otherwise noted. The instrument used was a Quantum Design MPMS-XL-7S SQUID magnetometer with an Evercool-equipped liquid helium dewar. Microcrystalline samples were packed in gelatin capsules and mounted in diamagnetic plastic straws. Direct current (dc) magnetization was measured for all samples upon cooling from 300 to 1.8 K under an applied dc field of 1 kOe or as indicated. The mag netic susceptibility of each compound was corrected for the diamagnetic contribution of the constituent atoms using Pascal's constants. (71) Other magnetic experiments will be described accordingly in the following chap ters.

n 2.7.2 Synthesis of [ Bu4N][Au(CN)2]-0.5H2O

n A 30 mL aqueous solution of [ Bu4N]Br (2.325 g, 7.212 mmol) was added to a

30 mL aqueous solution of K[Au(CN)2] (2.002 g, 6.949 mmol) while stirring. A white solid immediately precipitated and was stirred for 30 minutes. The powder Chapter 2. Magnetic exchange through double aqua-bridges 90

n of [ Bu4N][Au(CN)2]-0.5H2O was isolated by filtration and dried overnight. Yield:

3.409 g, 98.0 %. Anal. Calcd for C18H37N3AuOo.5: C, 43.20; H, 7.45; N, 8.40. Found: C, 43.24; H, 7.48; N, 8.11. IR (KBr): 2962 (s), 2932 (s), 2875 (m), 2865 (sh), 2146 (s), 1630 (w), 1488 (m), 1463 (m), 1384 (w), 1156 (w), 1111 (w), 1054(w), 1030 (w), 882 (m), 805 (w), 740 (m), 490 (s) cm-1.

2.7.3 Synthesis of Cu(/i-OH2)2[Au(CN)2]2

A 10 mL aqueous solution of Cu(C104)2-6H20 (0.259 g, 0.699 mmol) was prepared and added to a 10 mL aqueous solution of K[Au(CN)2] (0.403 g, 1.40 mmol). A pale green powder of Cu(^-OH2)2[Au(CN)2]2 was formed immediately and was then filtered and air-dried. Yield: 0.380 g, 91.0 %. Anal. Calcd for C4H4N4Au2Cu02: C 8.04, H 0.67, N 9.38. Found: C 8.18, H 0.71, N 9.22. IR (KBr): 3246 (m), 2217 (s), 2194 (vw), 2171 (s), 1633 (w) cm"1.

2.7.4 Synthesis of Cu[Au(CN)2]2

Cu(/i-OH2)2[Au(CN)2]2 was heated (150 °C) in vacuo for at least one hour to yield the green-brown Cu[Au(CN)2]2 product. No VCN peaks corresponding to Cu(/i-

OH2)2[Au(CN)2]2 could be observed in the FT-IR spectrum of the product obtained.

Anal. Calcd for C4N4Au2Cu: C 8.56, H 0, N 9.98. Found: C 8.68, H trace, N 9.80. IR (KBr): 2191 (s), 1613 (vw), 530 (m) cm"1.

2.7.5 Synthesis of Ni(^-OH2)2[Au(CN)2]2

Ni(NOs)2-6H20 (0.291 g, 1.00 mmol) was dissolved in water (2 mL) and an aqueous solution (2 mL) of K[Au(CN)2] (0.575 g, 1.99 mmol) was added dropwise. A pale blue- green precipitate was formed within several seconds, collected by filtration, and air- dried. The composition of the precipitate was confirmed to be Ni(/x-OH2)2[Au(CN)2]2 by conducting elemental analysis and TGA. Yield: 0.433 g, 73.4 %. Anal. Calcd for

C4H4N4Au2Ni02: C 8.11, H 0.68, N 9.45; found: C 8.31, H 0.74, N 9.20. IR (KBr): 3416 (br), 3116 (br), 3033 (br), 2214 (s), 2204 (sh), 2170 (s), 1537 (m), 910 (m), Chapter 2. Magnetic exchange through double aqua-bridges 91

-1 756 (w) cm . Care must be taken during the synthesis of Ni(/u-OH2)2[Au(CN)2]2 as blue K{Ni[Au(CN)2]3} (discussed in Chapter 4) can be obtained as a side product if a slight excess of K[Au(CN)2] is used, if the precipitate is left in solution for several days, or if the reaction is very concentrated. The purity of Ni(/>OH2)2[Au(CN)2]2 can be determined by powder X-ray diffraction.

Poor-quality pale blue-green crystals of Ni(/i-OH2)2[Au(CN)2]2 were obtained by recrystallization under hydrothermal conditions. In this reaction, 0.050 g of Ni(/x-

OH2)2[Au(CN)2]2 were added to a 5 mL pyrex ampoule containing 3 mL of water. The ampoule was sealed and heated to 125 °C for 6 h in a hydrothermal vessel and was then cooled down to 100 °C at 0.5 °C h_1, held at that temperature for 12 h, and then cooled to room temperature at 1 °C h-1.

2.7.6 Synthesis of Co(//-OH2)2[Au(CN)2]2

n A 15 mL acetonitrile solution of [ Bu4N][Au(CN)2]-0.5H2O (0.903 g, 1.81 mmol) was added to a 15 mL acetonitrile solution of Co(C104)2-6H20 (0.360 g, 0.984 mmol). An immediate pink precipitate was formed. The solution containing the precipitate was covered and left to sit overnight. The pink powder was collected the following day by filtration using a Buchner funnel and air-dried. The composition was determined to be Co(/z-OH2)2[Au(CN)2]2. Yield: 0.509 g, 94.9 %. Anal. Calcd for C4H4N4Au2Co02: C 8.10, H 0.68, N 9.45. Found: C 8.37, H 0.77, N 9.33. IR (KBr): 2204 (s), 2167 (s), 1530 (w), 1384 (w), 889 (w), 753 (w), 542 (w), 471 (w) cm"1.

2.7.7 Synthesis of Fe(/i-OH2)2[Au(CN)2]2

n [ Bu4N][Au(CN)2]-0.5H2O (0.085 g, 0.17 mmol) was first dissolved in a 98:2 acetonitrile/water mixture (1 mL). Fe(C104)2-6H20 (0.036 g, 0.10 mmol) was then rapidly dissolved in 1 mL of the same solvent mixture. The [Au(CN)2]~ solution was im mediately added to the Fe(ll) solution while stirring manually. A pale yellow peach precipitate was instantly formed. It was left to settle for 5 minutes and then filtered on a Buchner funnel, using all the solution to wash the product. The composition was determined to be Fe(/i-OH2)2[Au(CN)2]2. Yield: 0.037 g, 74 %. Anal. Calcd for Chapter 2. Magnetic exchange through double aqua-bridges 92

C4H4N4Au2Fe02: C 8.14, H 0.68, N 9.50. Found: C 8.34, H 0.85, N 9.41. IR (KBr): 2196 (s), 2165 (s), 1524 (w), 1094 (w), 877 (w), 747 (w), 535 (w) cm-1. Shortly after drying, the powder was stored under nitrogen to prevent decomposition.

2.7.8 Synthesis of Fe(//-OH2)(/u-OH)[Aii(CN)2]2

An aqueous solution (10 mL) of K[Au(CN)2] (0.581 g, 2.02 mmol) was added dropwise to an aqueous solution (10 mL) of Fe(C104)3-6H20 (0.359 g, 0.777 mmol). A very fine dark orange precipitate was formed immediately. The reaction mixture was transferred to a test tube and centrifuged for approximately 2-5 minutes. The over lying solution was then removed. To rinse the product and wash away any salt, water (10 mL) was added to the test tube and mixed with the precipitate. The mixture was centrifuged after which the overlying solution was removed. The rinsing process was repeated two more times. After removing the final overlying solution, the wet powder was left to dry in the test tube. The composition of the precipitate was deter mined to be Fe(/x-OH2)(/x-OH)[Au(CN)2]2. Yield: 0.305 g, 66.4 %. Anal. Calcd for

C4H3N4Au2Fe02: C 8.16, H 0.51, N 9.51. Found: C 8.13, H 0.70, N 9.47. IR (KBr): 2184 (s), 2168 (sh,s), 1612 (w), 669 (w), 518 (w) cm"1.

2.7.9 Synthesis of Mn(/i-OH2)2[Au(CN)2]2

n A 6 mL acetonitrile solution of [ Bu4N][Au(CN)2]-0.5H2O (0.314 g, 0.627 mmol) was added to a 6 mL acetonitrile solution of Mn(C104)2-6H20 (0.116 g, 0.321 mmol). An immediate white precipitate was formed. The powder of Mn(/>OH2)2[Au(CN)2]2 was collected by filtration, air-dried for a short period of time and then kept under N2.

Yield: 0.163 g, 88.3 %. Anal. Calcd for C4H4N4Au2Mn02: C 8.16, H 0.68, N 9.51. Found: C 8.33, H 0.85, N 9.28. IR (KBr): 2196 (w-m), 2161 (sh), 2158 (s), 1625 (w), 1518 (w), 1115 (w), 1090 (w), 859 (w), 741 (w), 527 (w), 460 (w) cm-1. If left uncovered and exposed to air, a small amount of the powder was found to turn yellow or brown. Chapter 2. Magnetic exchange through double aqua-bridges 93

2.7.10 Synthesis of Ni[Ag2(CN)3][Ag(CN)2]

An aqueous solution (2 mL) of Ni(N03)2-6H20 (0.029 g, 0.10 mmol) was added dropwise to an aqueous solution (3 mL) of K[Ag(CN)2] (0.060 g, 0.30 mmol). A pale blue precipitate was formed immediately. It was collected by nitration, and air-dried. The composition of the precipitate was found to be Ni[Ag2(CN)3][Ag(CN)2]. Anal. Calcd for C5HoN5Ag3Ni: C 11.72, H 0, N 13.67; found: C 11.86, H traces, N 13.39. IR (KBr): 3564 (br), 3443 (br), 2167 (w), 2134 (s), 2129 (m), 2126 (s), 2122 (s), 1621 (w), 1599 (w), 422 (m) cm"1.

2.7.11 X-ray crystallographic analysis

Powder diffraction

Powder X-ray diffraction data for Cu(/i-OH2)2[Au(CN)2]2 and Cu[Au(CN)2]2 were col lected by Prof. Ken Sakai (Kyushu University, Japan) using a RINT2000 diffractometer equipped with a Cu rotating-anode source (50 kV and 100 mA) and a scintillation- counter detector. Samples were mounted in a tube and irradiated in a scan step of 0.02° at a scan speed of 10° min-1. Data was measured between 4 and 135° in 26.

Powder X-ray diffraction data for M(yu-OH2)2[Au(CN)2]2 (M = Ni, Co, Mn) and

Fe(//-OH2)(//-OH)[Au(CN)2]2 was collected by Dr. Brian O. Patrick (University of British Columbia, Canada) using a Bruker D8 Advance diffractometer equipped with a Cu sealed-tube source (powered at 40 kV and 40 mA), a graphite monochromator, and a scintillation detector. The diffractogram of the Fe(/^-OH2)2[Au(CN)2]2 complex was collected (by myself) on a similar Bruker D8 Advance instrument at Simon Fraser University.

For Ni(/i-OH2)2[Au(CN)2]2, data was collected from 3 to 70° in 26 by using a step of 0.02° and a total counting time of 1.5 s per step. To obtain a higher-quality diffractogram in the 13 to 21° range, data was also collected with a total counting time of 130 s per step (same step size).

For M(/x-OH2)2[Au(CN)2]2 (M = Co, Fe, Mn) and Fe(/>OH2)(/i-OH)[Au(CN)2]2, data was collected from 5 to 60° in 26 by using a step of 0.02° and a total counting Chapter 2. Magnetic exchange through double aqua-bridges 94 time of 3.9 s per step.

Indexing of the powder diffractograms of M(/i-OH2)2[Au(CN)2]2 and Fe(/>OH2)(/> 141 OH)[Au(CN)2]2 was performed by using WinPLOTR.^ ) The powder X-ray diffrac tograms of the Co-, Fe- and Mn-containing M(/i-OH2)2[Au(CN)2]2 complexes were also indexed using the DASH software. (142^ The simulation of powder diffractograms from atomic coordinates and comparison with experimentally obtained powder diffrac tograms were conducted by using POWDER CELL. (143^

Single-crystal diffraction

Single-crystal X-ray diffraction data for the Ni(//-OH2)2[Au(CN)2]2 compound were collected as described above. Diffraction peaks were very broad, consistent with poor crystal quality. The Au and Ni atoms were refined with anisotropic thermal parameters, whereas the C, N, and O atoms were refined with isotropic thermal para meters. The hydrogen atom was geometrically placed and its position and isotropic thermal parameter were not refined.

Crystallographic data for Ni(//-OH2)2[Au(CN)2]2 are collected in Table 2.7 while selected bond lengths and angles are reported respectively in Table 2.3.

2.7.12 Details on SQUID magnetometry experiments

To prevent possible rehydration, the Cu[Au(CN)2]2 sample was packed into a cylin drical, airtight PVC sample holder. (144) In addition to correcting for the diamagnetic contribution of the constituent atoms, the data were corrected for the signal of the sample holder. Direct current (dc) magnetization of all compounds was measured upon cooling from 300 to 1.8 K under an externally applied dc field. The magnetic field was fixed at the following values for each compound: 10 kOe for the Cu- and Fe(lIl)-containing samples, 10 Oe, 50 Oe, 500 Oe, 1 kOe, 5 kOe and 10 kOe for the Ni-containing sample, 1, 5, 10, 20 and 30 kOe for the Co-containing sample and 100 Oe, 1 kOe and 5 kOe for the Fe(ll)- and Mn-containing samples. The magnetization as a function of dc field strength (from 0 to 70 kOe) was also recorded at 1.8 K; in each case, the sample Chapter 2. Magnetic exchange through double aqua-bridges 95 was first cooled from 100 K under zero-applied field.

Ni(//-OH2)2[Au(CN)2]2

Zero-field cooled (ZFC) and field cooled (FC) magnetization measurements were per formed on Ni(//-0H2)2[Au(CN)2]2. For each ZFC measurement, the sample was cooled from 100 K to 1.8 under zero-applied field. The desired field was then applied and the magnetization was measured as the temperature was increased. In the case of a FC measurement, the sample was cooled from 100 K down to 1.8 K under a set magnetic field. The magnetization was then measured upon warming under the same magnetic field as the cooling sequence. Experiments were carried using fields of 10, 100, and 800 Oe. Isothermal magnetization as a function of dc field strength (from 0 to 70 kOe) was recorded at several other temperatures (2.0, 2.5, 3.0, 3.5, 4, and 10 K). Alternating current (ac) susceptibility measurements in zero-applied dc field were performed for Ni(//-OH2)2[Au(CN)2]2 as a function of temperature (upon cooling be tween 50 and 1.8 K). At each temperature, the amplitude of the ac field was fixed at 5 Oe and the operating frequency was varied between 1.00 and 1488.10 Hz. The isothermal ac susceptibility of Ni(^-OH2)2[Au(CN)2]2 as a function of the applied dc field, from 0 to 70 kOe, was also measured at several temperatures (1.8, 2.0, 2.2, 2.4, 2.5, 2.8, 3.0, 3.2, and 3.5 K), with an ac field of 5 Oe and a driving frequency of 997.34 Hz.

2.7.13 Experimental details on Mossbauer spectroscopy

Mossbauer spectra were acquired on a WEB Mssbauer spectroscopy system connected to a Janis Research variable temperature SHI-850 Cryostat and a closed cycle refrig erator. A 57Co (in rhodium matrix) source was used for the experiment. The 7-rays detector was a Reuters-Stokes Kr-C02 proportional counter. Each powdered sample (60-100 mg) was loaded in a 1 cm x 1 cm parafilm envelope, which was sealed with Kapton tape. The envelope was then fixed to the sample holder rod using Kapton tape. The Mossbauer spectrum was collected at 300 and 4.5 K. Chapter 2. Magnetic exchange through double aqua-bridges 96

The velocity was scanned between -4 and 4 mm s_1, using a constant acceleration triangle waveform. No external magnetic field was applied during measurement. The spectra were collected over several days until a good signal to noise was obtained. The spectra were referenced and scaled with respect to an Fe foil spectrum measured on the same instrument at 295 K in zero field. Fitting of the data was performed using the WMOSS software.

2.7.14 //SR measurements

The /xSR experiments were performed on the M15 surface muon channel at the TRI- University Meson Facility (TRIUMF) in Vancouver, Canada. For each compound, two pressed polycrystalline pellets with masses ranging from 250 to 500 mg and a diameter of 1 cm were prepared and varnished onto a silver backing plate. The silver backing plate was then thermally anchored to the cold finger of a 3He/4He dilution refrigerator. Figure 1.8 shows the arrangement of scintillation detectors used in the experiment. The initial polarization, Pz(0), was directed antiparallel to the beam momentum p^ (i.e. along the z-axis). The //SR measurements were taken both in zero external magnetic field (ZF), and in a longitudinal field (LF) geometry with the magnetic field applied parallel to Pz(0). Chapter 2. Magnetic exchange through double aqua-bridges

Table 2.7: Crystallographic data and structural refinement details for OH2)2[Au(CN)2]2.

Ni(^-OH2)2[Au(CN)2]2 Empirical formula C4H4N4Au2Ni02 Fw 592.75 Colour Green-blue Shape chunk Dimension, mm3 0.02 x 0.03 x 0.05 Crystal system orthorhombic Space group (#) Immm (71) a, A 6.374(3) 6, A 3.3183(11) c, A 20.512(5) a, deg 90 (3, deg 90 7, deg 90 Volume, A3 433.9(3) Z 2 A, A 0.70930 Data range, deg 4-60 Transmission range - Pealed, g cm"3 4.537 /i, mm-1 35.845 Reflections, parameters 217, 14 Ria (/ > 2.5a(I)) 0.0370 a wR2 {I > 2.5a(I)) 0.0376 goodness of fit 1.0503 a w 2 l 2 2 Function minimized J2 (\^o\ \FC\) where w~ = a {F0) + 0.0001Fo , /2 R = Z\\F0\-\FC\\/J:\F0\,RV = [j:w(\F0\-\Fc\r/j:w\F0\Y - Chapter 3

M[Au(CN)2]2(analyte)x coordination polymers and their vapochromic properties81

3.1 Introduction

3.1.1 Coordination polymers as vapochromic materials

Vapochromic materials, which display absorption or luminescence changes upon expo sure to vapors of volatile organic compounds (VOCs), have been a focus of attention due to their potential applications as chemical sensors. (38>145~150) The majority of vapochromic systems contain metal ions. For example, a vapo chromic sensor commonly used in most laboratories to detect the presence of water is

aReproduced in part with permission from: J. Lefebvre, R.J. Batchelor, D. B. Leznoff, Cu[Au(CN)2]2(DMSO)2: Golden polymorphs that exhibit vapochromic be haviour", J. Am. Chem. Soc, vol. 126, pp. 16117-16125, 2004. Copyright 2004 American Chemical Society.

98 Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 99

C0CI2. A small amount of C0CI2 is added to desiccant materials and, upon hydration, this Co(n) complex goes from a deep blue-purple colour to a pale pink colour. This color change is attributed to modifications in the metal ion coordination sphere. Several coordination polymers also display vapochromic properties toward different volatile analytes. For example, when exposed to certain organic solvents, the extended 2+ 4 Prussian Blue Co -[Re6Q8(CN)6] ~ (Q = S, Se) system yields dramatic changes in the visible spectrum that are attributable to the analyte impacting the geometry and hydration around the Co(ll) centers. (38) Modification of metal-metal interactions can also yield colour changes by altering the visible absorption and emission spectra. Several vapochromic compounds con taining Au(l),(145~148) Pd(ll),(149) and Pt(ll)(149'150) have been recently reported. In the linear {Tl[Au(C6Cl5)2]}n polymer, weak interactions between the Tl(i) atoms and the adsorbed VOC molecules modify slightly the absorption, and more significantly the emission spectra, which result mainly from Tl(l)-Au(i) interactions.^148^ Examples of vapochromic molecular complexes include the family of

[Pt(CN-R)4][M(CN)4] (R = iso-C3H7 or C6H4-CnH2n+1; n = 6, 10, 12, 14 and M = Pt, Pd) complexes in which changes in the emission spectra occur when metal- metal distances are modified due to the presence of VOC molecules in lattice voids; small changes in the absorption spectrum can also be observed/149'151) Another ex ample is the trinuclear Au(l) complex with carbeniate bridging ligands, for which the luminescence is quenched in the solid-state when CeF6 vapor is adsorbed due to the disruption of Au-Au interactions/147) Some of these vapochromic materials have recently been incorporated in chemical sensor devices. For example, [Au-(PPh2C(CSSAuC6F5)PPh2Me)2][C104] has been used in the development of an optical fiber volatile organic compound sensor. (152) A vapochromic light-emitting diode (153) and a vapochromic photodiode^154) have also been built using tetrakis(p-dodecylphenylisocyano) platinum tetranitroplatinate and bis(cyanide)-bis(p-dodecylphenylisocyanide)platinum(n), respectively. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 100

o o V II X n l [1 ™3 H3C—^N

H3C^ ^CH, H3C ^CH- DMSO DMF pyridine dioxane ammonia acetonitrile

Scheme 3.1: Analyte molecules incorporated into the M[Au(CN)2]2(analyte)a; poly mers presented in this chapter.

3.1.2 Research objectives

It was demonstrated in Chapter 2 that water molecules could be incorporated in [Au(CN)2]-based coordination polymers, binding to the transition metal ions. To further investigate this family of polymers, the preparation of a series of

M[Au(CN)2]2(analyte)x, incorporating different types of analyte molecules and tran sition metal ions, was attempted. Donor molecules such as N, iV-dimethylformamide (DMF), dimethyl sulfoxide (DMSO), pyridine, acetonitrile and ammonia were chosen as analytes (Scheme 3.1). Similarly to water molecules, these molecules contain a functional group that can bind to the transition metal ions (or not) depending on the nature of the system. The structural arrangements and the magnetic properties of these polymers were studied as a function of analyte incorporated. In addition, the vapochromic behaviours of these M[Au(CN)2]2(analyte):r polymers were investigated and are reported.

3.2 Cu[Au(CN)2]2(DMSO)2: Preparation and characterization

3.2.1 Synthesis of polymorphs

The reaction of Cu(ll) with K[Au(CN)2] in dimethyl sulfoxide (DMSO) produced upon slow evaporation a mixture of two compounds that differ in colour and crystal shape. Modification of the reaction concentration allowed each product to be isolated Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 101 separately. In dilute solution (total concentration of reagents is lower than 0.2 M), green crystals formed slowly, whereas blue crystals were obtained rapidly in a highly concentrated solution (total concentration of reagents is higher than 0.5 M). The FT-IR spectra of the two products show different features: the green com pound shows two bands attributable to cyanide vibration at 2183 and 2151 cm-1, whereas the blue compound has four bands of different intensity at 2206, 2193, 2175 and 2162 cm-1. In the two complexes, the higher-energy bands likely correspond to bridging cyanide groups (Cu-NC-Au), whereas the lower-energy bands are due to cyanide vibration in either free or loosely bound [Au(CN)2]~ units.^ Elemental analysis showed that the two products have the same relative amounts of carbon, hydrogen and nitrogen, which suggests that both products have the same empirical formula. The results for the two products are consistent with the formula being Cu[Au(CN)2]2(DMSO)2. The crystal structure of each complex was determined by single-crystal X-ray diffraction. The results obtained confirmed the chemical composition suggested by elemental analysis for the green and blue products and revealed two different polymeric networks.

3.2.2 Cu[Au(CN)2]2(DMSO)2 (green)

The structure of Cu[Au(CN)2]2(DMSO)2 (green) contains five-coordinate Cu(ll) cen ters (Figure 3.1A). Their geometry is neither purely square pyramidal nor trigonal bipyramidal. To better describe the geometry of a five-coordinate metal center, a r-value can be determined. (155) It is defined as an index of the degree of trigonality for five-coordinate metal centers:

where a and [3 correspond to basal angles or largest angles (with f3 > a by definition) around the metal center (Scheme 3.2). A r value of 0 corresponds to a pure square pyramidal geometry (as f3 = a — 180°) and a value of 1 corresponds to a pure trigonal bipyramidal geometry (as f3 = 180° and a = 120°). The Cu(ll) center in Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 102

L, *__M_—> L3 T Sa^"""»n

Scheme 3.2: Geometrical parameters to determine the r-value of a five-coordinate metal center

Cu[Au(CN)2]2(DMSO)2 (green) has a r-value of 0.44 (using 0 = 167.0° and a = 140.9°, Table 3.1). This suggests that the coordination geometry could be considered nearly equally distorted from either polyhedron.

The Cu(n) center in Cu[Au(CN)2]2(DMSO)2 (green) is bound to two DMSO molecules through the oxygen atoms in a trans fashion (^O-Cu-O = 167.06°) and three N-bound [Au(CN)2]~ groups (Figure 3.1A). Selected bond lengths and angles for Cu[Au(CN)2]2(DMSO)2 (green) are listed in Table 3.1. The asymmetric unit con tains two different [Au(CN)2]~ units: a Cu(ll)-bridging moiety that generates a 1-D zigzag chain, and a Cu(ll)-bound dangling group (Figure 3.1A). The chains stack on top of each other parallel to the (lOl)-plane, forming stacks of chains that are offset to allow interdigitation of the dangling [Au(CN)2]~ units (Figure 3.IB). Each chain is connected to the four neighboring chains through Au-Au interactions of 3.22007(5) A between the Au(l) atoms of each dangling group and the Au(2) atoms of the chain backbone (Figure 3.IB). The DMSO molecules occupy the channels be tween the chains; these channels are delineated by both [Au(CN)2] groups and Au-Au bonds (Figure 3.1C).

3.2.3 Cu[Au(CN)2]2(DMSO)2 (blue)

The structure of the blue Cu[Au(CN)2]2(DMSO)2 polymorph contains Cu(ll) cen ters in a Jahn-Teller distorted octahedral geometry (Figure 3.2A). The two DMSO molecules are bound through the oxygen atom in a ds-equatorial fashion (Z o-Cu-0 — Chapter 3. M[Au(CN)2]2(an&lyte)x and their vapochromic properties 103

Figure 3.1: A. Extended 1-D zigzag chain structure of Cu[Au(CN)2]2(DMSO)2 (green); DMSO-methyl groups were removed for clarity. B. Schematic representa tion showing offset stacks of chains viewed down the (lOl)-plane (slightly tilted). Au-Au bonds connect bridging and dangling [Au(CN)2]~ units of neighboring chains (yellow lines). C. 3-D structure formed via Au-Au bonding, viewed down the a-axis. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 104

Table 3.1: Selected bond lengths (A) and angles (deg) for Cu[Au(CN)2]2(DMSO)2 (green)

Bond Lengths Au(l)-Au(2) 3.22007(5) Cu(l)-N(2) 2.107(18) Cu(l)-0(1) 1.949(7) Cu(l)-N(3) 1.965(11) Angles 0(1)-Cu(l)-0(1*) 167.0(6) N(3)-Cu(l)-N(3*) 140.9(8) 0(1)-Cu(l)-N(2) 96.5(3) Cu(l)-0(1)-S(l) 127.2(6) 0(1)-Cu(l)-N(3) 87.3(4) Au(2)-Au(l)-Au(2) 171.73(3) 0(1*)-Cu(l)-N(3) 88.4(4) C(l)-Au(l)-C(2) 180 N(2)-Cu(l)-N(3) 109.5(4) Symmetry operations: (*) — x 4-1, y, —z + |; (') — x + \, — y — §, z + 1.

95.2°) rather than in the nearly 180°-arrangement in Cu[Au(CN)2]2(DMSO)2 (green).

Selected bond lengths and angles for Cu[Au(CN)2]2(DMSO)2 (blue) are found in Ta ble 3.2. The four remaining sites (two axial and two equatorial) are occupied by

N-bound cyanide groups of bridging [Au(CN)2]~ units, generating corrugated 2-D sheets (Figure 3.2B). These 2-D layers stack (Figure 3.2C) and are held together by weak Au(l)-Au(2) interactions of 3.419(3) A and perhaps weak Au(3)-• • Au(4) contacts of 3.592(4) A.

The structural arrangement adopted in Cu[Au(CN)2]2(DMSO)2 (blue) was also 156 observed in the related Zn[Au(CN)2]2(DMSO)2 complex. ( ) The unit cell determined for Zn[Au(CN)2]2(DMSO)2 has similar dimensions. However, it is of higher symmetry, being monoclinic rather than triclinic. Lacking a Jahn-Teller axis, the Zn(il) centers are not as distorted as the Cu(n) centers and the four Zn-N distances fall in the 2.11-

2.17 A range (compared to the 1.97-2.42 A range observed in Cu[Au(CN)2]2(DMSO)2 (blue)). Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 105

Figure 3.2: A. Coordination sphere around the Cu(n) centers in Cu[Au(CN)2]2(DMSO)2 (blue); B. Extended structure showing the 2-D corru gated layers, viewed down the a-axis. DMSO-methyl groups were removed for clarity. C. Layers stacked via aurophilic interactions to yield a 3-D network. DMSO molecules were removed for clarity. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 106

Table 3.2: Selected bond lengths (A) and angles (deg) for Cu[Au(CN)2]2(DMSO)2 (blue)

Bond Lengths Au(l)-Au(2) 3.419(3) Au(3)---Au(4) 3.592(4 Cu(l)-0(1) 2.02(3) Cu(2)-0(3) 1.97(3 Cu(l)-0(2) 1.95(3) Cu(2)-0(4) 2.29(3 Cu(l)-N(ll) 2.42(4) Cu(2)-N(12) 2.11(4 Cu(l)-N(21) 1.97(4) Cu(2)-N(22) 2.37(5 Cu(l)-N(31) 2.42(4) Cu(2)-N(32) 2.03(5 Cu(l)-N(41) 1.99(4) Cu(2)-N(42) 2.00(5 Angles 0(l)-Cu(l)-0(2) 95.2(12) 0(3)-Cu(2)-0(4) 93.0(12 0(1)-Cu(l)-N(ll) 85.9(12) 0(3)-Cu(2)-N(12) 87.8(15 0(2)-Cu(l)-N(ll) 86.4(12) 0(4)-Cu(2)-N(12) 87.0(14 N(ll)-Cu(2)-N(21) 92.7(14) N(12)-Cu(2)-N(22) 92.3(16 N(ll)-Cu(2)-N(31) 172.7(13) N(12)-Cu(2)-N(32) 172.0(17 N(ll)-Cu(2)-N(41) 92.6(14) N(12)-Cu(2)-N(42) 95.2(17 N(21)-Cu(2)-N(31) 92.6(15) N(22)-Cu(2)-N(32) 91.4(17 N(21)-Cu(2)-N(41) 90.7(15) N(22)-Cu(2)-N(42) 91.2(17 N(31)-Cu(2)-N(41) 92.3(14) N(32)-Cu(2)-N(42) 91.8(18 Cu(l)-0(1)-S(l) 124.9(17) Cu(2)-0(3)-S(3) 125.4(20 Cu(l)-0(2)-S(2) 124.4(19) Cu(2)-0(4)-S(4) 127.9(18 C(ll)-Au(l)-C(12) 172.7(25) C(31)-Au(3)-C(32) 172.6(18 C(21)-Au(2)-C(22*) 175.9(23) C(41*6)-Au(4)-C(42) 177.9(20 Symmetry operations: (*) — x + 1, — y + 1, —z + 1; (*6) —x + 1, —y, —z + 1; (')x + 2,y,z-l; ('fe) x - 2, y, z + 1. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 107

100

^ 85

I 75

60 50 100 150 200 250 300 350 400

Temperature [ °C ]

Figure 3.3: Thermogravimetric analysis of the green (A., green) and blue (B., blue)

Cu[Au(CN)2]2(DMSO)2 polymorphs.

3.2.4 Physical properties of Cu[Au(CN)2]2(DMSO)2

As polymorphs, Cu[Au(CN)2]2(DMSO)2 (green) and Cu[Au(CN)2]2(DMSO)2 (blue) clearly have significantly different solid-state structures and it follows that their phys ical and chemical properties may also vary. This is the case for their solid-state optical reflection spectra, which show Amaxrefl. of 550(7) and 535(15) nm respectively (Table 3.3, Figure 3.5, see section 3.3.1).

Thermal stability

When examining the thermal stabilities of the green and blue Cu[Au(CN)2]2(DMSO)2 polymorphs by thermogravimetric analysis (Figure 3.3), differences and similarities could be observed.

Three different weight losses can be observed for Cu[Au(CN)2]2(DMSO)2 (green) in the 150-190, 210-250 and 310-330 °C temperature ranges. The first two losses Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 108 have roughly the same magnitude (~ 13 %) and they can be attributed to the loss of the DMSO molecules in a stepwise manner (theoretically 11 % per DMSO molecule).

The third loss likely corresponds to the simultaneous loss of two cyanogen (CN)2 units (theoretical: 14 %, experimental: 11 %). For the blue polymorph, a total of four weight losses can be observed. The first two losses (~ 7 % each) occur between 100-135 °C and 150-190 °C. Losses are then observed in the 210-250 °C (~ 13 %) and 310-330 °C (10 %) temperature ranges. The last two weight losses are comparable to those of the green Cu[Au(CN)2]2(DMSO)2 polymorph. The differences in thermal stability and decomposition mechanism may be ex plained by their different structures. Cu[Au(CN)2]2(DMSO)2 (blue) has four crystallographically distinct DMSO molecules, compared to only two for the green polymorph. These crystallographically different molecules can be lost at different temperatures (each of the DMSO molecule accounts for 7 % of the mass if the molecular formula is defined as {Cu[Au(CN)2]2(DMSO)2}2). The first two DMSO molecules appeared to be lost in two separate steps (100-135 and 150-190 °C), followed by the loss of the two others in one step (210-250 °C). The loss of the last two DMSO molecules in

{Cu[Au(CN)2]2(DMSO)2}2 (blue) occurs at the same temperature as the loss of the second molecule in Cu[Au(CN)2]2(DMSO)2 (green). As for Cu[Au(CN)2]2(DMSO)2 (green), the drop above ~ 310 °C can be attributed to the decomposition of the remaining framework and the loss of the all cyanide groups. Hence, the thermal stabilities of the two polymorphs with respect to the loss of the first DMSO molecules are significantly different. However, above ~ 200 °C, both polymorphs decompose similarly. Differential scanning calorimetry experiments were performed to investigate pos sible conversion from the blue form to the green form. However, no peak attributable to a phase transition was observed below the decomposition temperature of the blue form (100 °C). This suggests that the thermal interconversion in the solid state from

Cu[Au(CN)2]2(DMSO)2 (blue) to Cu[Au(CN)2]2(DMSO)2 (green) does not occur. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 109

3.3 Vapochromic behaviour of the

Cu[Au(CN)2]2(DMSO)2 polymorphs

3.3.1 Identification of adsorbed guest by infrared and UV-vis spectroscopies

Even though the two Cu[Au(CN)2]2(DMSO)2 polymorphs are thermally stable up to at least 100 °C, they were found to be unstable under ambient air conditions. When exposed to humidity (at room temperature), a colour change to pale green and a loss of single crystallinity was observed over a period of several hours for both polymorphs (Figure 3.4). Elemental analysis and thermogravimetric analysis on the decomposed products gave results different from that of either Cu[Au(CN)2]2(DMSO)2 polymorph. The relative amounts of carbon, nitrogen and hydrogen suggested the presence of two water molecules in the compounds instead of two DMSO molecules, yielding the general Cu[Au(CN)2]2(H20)2 formula. The FT-IR spectrum and powder

X-ray diffractogram of the two converted Cu[Au(CN)2]2(DMSO)2 polymorphs were indistinguishable to those of the Cu(/x-OH2)2[Au(CN)2]2 coordination polymer pre sented in Chapter 2.

This suggests that the two DMSO molecules of Cu[Au(CN)2]2(DMSO)2 (in either polymorph) are replaced by two water molecules, in the solid state, at room tempera ture. In addition, this solvent exchange seems to be complete/quantitative as no sign of DMSO is present in the FT-IR spectrum of the final product.

Despite the different solid-state structures, both Cu[Au(CN)2]2(DMSO)2 poly morphs convert to the same Cu(yu.-OH2)2[Au(CN)2]2 complex (Table 3.3) as evidenced by the FT-IR spectra and powder X-ray diffractograms of both products.

If Cu(//-OH2)2[Au(CN)2]2 is exposed to an atmosphere containing more DMSO than water, a colour change to a darker green can be observed. The FT-IR spectrum of this later product as well as its diffractogram were very similar to the Cu[Au(CN)2]2(DMSO)2 (green) polymorph. This suggests that the Cu(/x-

OH2)2[Au(CN)2]2 complex converts back to Cu[Au(CN)2]2(DMSO)2, when exposed Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 110

#• v-,:- W W DMSO(blue) Water CH3CN DMF x~2 x = 2 x-2 x-1

Dioxane / Water Pyridine NHL,

Figure 3.4: Powder sample of Cu[Au(CN)2]2(DMSO)2 (blue) exposed to various sol vent vapours. The values of x refers to the number of solvent molecules incorporated in Cu[Au(CN)2]2(analyte)s. to an excess of DMSO. Upon conversion, the compound adopts only the green poly morph structural arrangement, as no sign of the blue polymorph can be seen in the FT-IR spectrum or X-ray diffractogram. This happens despite the fact that the origi nal DMSO-complex to which water was added was the Cu[Au(CN)2]2(DMSO)2 (blue) polymorph. The exchange of DMSO for water or vice versa can be observed visually from the associated color change (Figure 3.4 and 3.5 and Table 3.3). The Cu(/z-

OH2)2[Au(CN)2]2 complex was determined to have a maximum visible reflection

(Amaxrefl.) of 535(5) nm which differs from that of both Cu[Au(CN)2]2(DMSO)2 poly mers. It is important to remember that the colour of a complex is not only determined by the position of Amax refl. but also by the broadness and shape of the peak. This is the case for Cu(At-OH2)2[Au(CN)2]2 (535(5) nm) vs Cu[Au(CN)2]2(DMSO)2 (blue) (535(15) nm) that have different colours due to different peak shape (Figure 3.5).

The two Cu[Au(CN)2]2(DMSO)2 polymorphs also display a colour change when exposed to a variety of other analyte vapors. Figures 3.4 and 3.5 show the changes occurring when the blue polymorph is exposed to different vapours. The chemical Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 111

400 450 500 550 600 650 700 750 800 Wavelength [ nni ]

Figure 3.5: Solid-state UV-vis-NIR transmission spectra of Cu[Au(CN)2]2(analyte):E: blue DMSO polymorph (A., pale blue), green DMSO polymorph (B., green), water (C, pale green), dioxane (D., red), DMF (E., orange), pyridine (F., dark gray) and ammonia (G., blue).

composition of the final products was determined by a combination of FT-IR spec troscopy, elemental analysis and thermogravimetric analysis. Molecules of DMF, pyridine, dioxane, acetonitrile and ammonia were found to react with the two Cu[Au(CN)2]2(DMSO)2 polymorphs. In each case, no DMSO molecule was present in the final product, but molecules of the exposed vapour were incorporated into the solid polymers. The number of analyte molecules incorporated per Cu[Au(CN)2]2 was found to be constant for a given vapour. The values of x are reported in Figure 3.4. When Cu[Au(CN)2]2(DMSO)2 was exposed to dioxane, the final product was found to always contain one equivalent of dioxane and one equivalent of water.

When the Cu[Au(CN)2]2(DMSO)2 polymers were exposed to alcohol vapour (i.e. methanol and ethanol), no color change could be observed and indistinguishable FT- IR spectra were obtained, indicating that the system was not sensitive to that type of analyte. Table 3.3: Maximum solid-state visible reflection (Amaxrefl., nm) and cyanide vibration frequencies (VCN, cm *) for different Cu[Au(CN)2]2(analyte)^ complexes

VCN Complex ^max refl. from solution from absorption 6 Cu[Au(CN)2]2(DMSO)2 (green) 550(7) 2183(s), 2151(B) 2184(B), 2151(s) Cu[Au(CN)2]2(DMSO)2 (blue) 535(15) 2206(m), 2193(B), 2175(m), (broad) 2162(m) a Cu[Au(CN)2]2(DMF) 498(7) 2199(B) 2199(s) Cu[Au(CN)2]2(pyridine)2 480(15) 2179(m), 2167(B), 2152(m), 2179(m), 2167(B), 2152(m), (broad) 2144(m) 2144(m)a c Cu(/z-OH2)2[Au(CN)2]2 535(5) 2217(B), 2194(w), 2172(B) 2217(B), 2194(W), 2171(s) 2216(B), 2196(w), 2171(s)a Cu[Au(CN)2]2 560(20) 2191(B) (v. broad)

Cu[Au(CN)2]2(CH3CN)2 2297(w), 2269(w), 2191(B) a Cu[Au(CN)2]2(dioxane)(H20) 505(15) 2201(B), 2172(W) 2200(s), 2174 (w) (broad) a Cu[Au(CN)2]2(NH3)4 433(7) 2175(m), 2148(s) a b Solvent adducts were made from Cu[Au(CN)2]2(DMSO)2 (blue). Solvent adducts were made from Cu(/x-OH2)2[Au(CN)2]2- c Solvent adducts were made from Cu[Au(CN)2]2(DMSO)2 (green). Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 113

Each Cu[Au(CN)2]2(analyte)x complex can be distinguished by its color (Fig ure 3.4 and 3.5 and Table 3.3). In addition, the UCN region of the FT-IR spectrum for each solvent adduct is characteristic for that solvent (Table 3.3). The number of different vibration bands observed changes as well as the frequency at which they occur and the relative intensities of all of them. Importantly, this solvent exchange is reversible, thus permitting dynamic solvent sensing. Hence, starting with a solid sample of a CufA^CN^^analyte^ complex, addition of a different solvent vapor generates a new complex. The only exceptions occur with pyridine and ammonia, which are not easily displaced by other solvents. These two solvents are very strong donors and are likely to bind more strongly to the Cu(ii) centers.

3.4 Structure-properties relationships of the

Cu[Au(CN)2]2(analyte)x polymers

In order to understand the vapochromic behaviour of the Cu[Au(CN)2]2(analyte)a; family and the structural rearrangements that occur as analyte molecules are ex changed, attempts were made at synthesizing the different solvent adducts in solution from the appropriate solvent. The synthesis and characterization of the water-containing polymer (from solution) was also investigated and was emphasized in Chapter 2 (section 2.2.1, 2.2.2 and 2.3.1).

3.4.1 Cu[Au(CN)2]2(DMF)

The reaction of Cu(ll) with K[Au(CN)2] in DMF afforded a mixture of dark blue- green powder and very small crystals of the same colour. The FT-IR spectrum of the bulk product showed only one band in the cyanide vibration region at 2199 cm"1.

When Cu(/i-OH2)2[Au(CN)2]2 was dissolved in DMF and the solution was left to evaporate slowly, crystals suitable for single crystal X-ray diffraction analysis were obtained. The FT-IR spectrum of these crystals is indistinguishable to that of the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 114

Table 3.4: Selected bond lengths (A) and angles (deg) for Cu[Au(CN)2]2(DMF)

Bond Lengths Au(l)-Au(l*a) 3.3050(12) Cu(l)-N(2) 1.990(11) Au(2)-Au(2*6) 3.1335(13) Cu(l)-N(3) 1.961(10) Cu(l)-0(1) 2.202(12) Cu(l)-N(4'6) 1.982(10) Cu(l)-N(ll'a) 1.958(10) 0(1)-C(5) 1.202(17) Cu- • • Cu 4.018 Au(l)-Au(2) 3.383 Angles N(l'a)-Cu(l)-N(2) 89.8(4) 0(1)-Cu(l)-N(l'a) 95.1(5) N(l'°)-Cu(l)-N(3) 88.7(5) 0(1)-Cu(l)-N(2) 98.3(5) N(4'6)-Cu(l)-N(2) 89.6(5) 0(1)-Cu(l)-N(3) 92.4(5) N(4'6)-Cu(l)-N(3) 89.3(5) 0(1)-Cu(l)-N(4'6) 98.1(5) N(l'a)-Cu(l)-N(4'6) 166.7(5) N(2)-Cu(l)-N(3) 169.2(5) C(3)-Au(2)-C(4) 175.4(6) Cu(l)-0(1)-C(5) 125.4(13) C(l)-Au(l)-C(2) 176.0(6) Symmetry operations: (*a) —x — 1, y, —z + §; (*b) —x — 1, y, —z + \\ (,a) x, 6 c ,d -y, z - |; C ) x,-y-l,z + i; (' ) x, -y, z + ±; ( ) x,-y-l,z-\.

product obtained when using a mixture of Cu(ll) and [Au(CN)2]~ salts as start ing materials. The comparable powder X-ray diffractograms also confirmed that the structural arrangement was the same in both products. The results of CHN elemental analysis for both products were identical and suggested that the chemical composition was Cu[Au(CN)2]2(DMF).

The structure of Cu[Au(CN)2]2(DMF) was determined by X-ray crystallogra phy. Selected bond lengths and angles for Cu[Au(CN)2]2(DMF) are listed in Ta ble 3.4. Figure 3.6A shows the square-pyramidal geometry of the Cu(ll) centers in

Cu[Au(CN)2]2(DMF) (using Equation 3.1, the r value was determined to be 0.04).

The four basal sites are occupied by N-bound cyanide groups of bridging [Au(CN)2]~ units, and the apical site is occupied by an O-bound DMF molecule. The alter nation of Cu(ll) centers and [Au(CN)2]~ units generates a 2-D square grid motif with all the DMF molecules pointing in the same direction, either above or below the plane of the sheet (Figure 3.6B). This grid is similar to that observed in the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 115

Figure 3.6: A. Coordination sphere around a Cu(ii) center in Cu[Au(CN)2]2(DMF);

B. Extended structure of Cu[Au(CN)2]2(DMF) showing the 2-D square-grid motif; C. Side view of three bilayers connected via aurophilic interactions; D. Offset between two square-grid layers forming one bilayer, also connected via aurophilic interactions. Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 116

Cu[Au(CN)2]2(DMSO)2 (blue) complex if one DMSO molecule were removed and the corrugation reduced. The layers stack to form bilayers, with the DMF molecules of each layer pointing toward the second layer of the bilayer (Figure 3.6C). The two layers are offset with respect to each other (Figure 3.6D) and are held together by Au(l)-Au(l*a) and Au(2)-Au(2*6) interactions of 3.3050(12) A and 3.1335(13) A. The bilayers stack on top of each other (Figure 3.6C), with the Cu(ll) centers on the bottom of one bilayer lying directly above the Cu(il) centers on the top the adja cent bilayer, with a Cu-Cu distance of 4.018 A. The bilayers are connected together through Au(l)-Au(2) interactions of 3.383 A. When looking at the space-filling model, no empty channels are present in the structure of Cu[Au(CN)2]2(DMF) due to the offset stacking of the grids.

3.4.2 Cu[Au(CN)2]2(pyridine)2

The reaction of Cu(n) and K[Au(CN)2] in a mixture of water, methanol and pyridine (for solubility purposes) afforded the formation of a mixture of blue powder and dark blue crystals. The FT-IR spectrum of the powder and the crystals were indistinguish able and showed four different bands in the cyanide vibration region, between 2144 and 2179 cm-1 (Table 3.3). The results from CHN elemental analysis for the bulk product were consistent with the chemical formula being Cu[Au(CN)2]2(pyridine)2.

The structure of Cu[Au(CN)2]2(pyridine)2 was determined by single crystal X- ray diffraction and is shown in Figure 3.7. Selected bond lengths and angles for

Cu[Au(CN)2]2(pyridine)2 are listed in Table 3.5.

The Cu(n) centers in Cu[Au(CN)2]2(pyridine)2 adopt a Jahn-Teller distorted octa hedral geometry (Figure 3.7A). The two axial sites and two of the equatorial sites are occupied by N-bound bridging [Au(CN)2]~ units, with Cu-N distances of 2.532(9) A and 2.016(9) A respectively. The N-bound pyridine molecules occupy two trans = equatorial sites (dcu-N(pyridine) 2.007(7), ZN-CU-N = 180.0°).

As observed for Cu[Au(CN)2]2(DMF), infinite 2-D square grid layers are obtained from the alternation of Cu(n) centers and [Au(CN)2]~ units (Figure 3.7B and C). No Chapter 3. M[A u (CN)2]2 (analyte) x and their v&pochromic properties 117

Figure 3.7: A. Coordination sphere around a Cu(ll) center in Cu[Au(CN)2]2(pyridine)2. B. Extended structure showing a 2-D square grid layer. C. Side view of a 2-D layer showing the pyridine ligands situated above and below the plane. D. Offset stacking of the 2-D square grids allowing 7T-7T interactions between every other sheet (one example of 7T-7T interactions is highlighted). Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 118

Table 3.5: Selected bond lengths (A) and angles (deg) for Cu[Au(CN)2]2(pyridine)2

Bond Lengths Cu(l)-N(l) 2.016(9) Cu(l)-N(3) 2.007(7) Cu(l)-N(2*a) 2.532(9) Angles N(l)-Cu(l)-N(2*a) 89.5(4) N(l)-Cu(l)-N(3) 90.0(3) N(l')-Cu(l)-N(2*a) 90.5(4) N(l)-Cu(l)-N(3') 90.0(3) N(2*a)-Cu(l)-N(3) 90.4(3) N(2*Q)-Cu(l)-N(3') 89.6(3) C(2)-Au(l)-C(l) 177.8(4) Symmetry operations: (*a) x-l, -y + \, z-\; (*6) x + 1, -y + \, z + \; (') ~x + 1, -y, -z + 1. aurophilic interactions are present between the Au(i) atoms of neighboring sheets, but 7T-7T interactions (with a distance of ~ 3.3 A between the pyridine planes) are found between pyridine rings of every other sheet (Figure 3.7D). Thus, the square-grid array present in Cu[Au(CN)2]2(DMSO)2 (blue) is maintained but in this case the sheets are completely flat as in Cu[Au(CN)2]2(DMF), rather than being corrugated. The 180° disposition of the pyridine rings (vs the cis orientation of the DMSO molecules in

Cu[Au(CN)2]2(DMSO)2 (blue)) also serves to separate the sheets, disrupting potential intersheet Au-Au interactions.

3.4.3 Cu[Au(CN)2]2(CH3CN)2

The reaction of Cu(C104)2-6H20 with K[Au(CN)2] in acetonitrile produced immedi ately two precipitates, one green and one white. When the mixture was isolated by filtration and exposed to ambient atmosphere, the dark green powder changed colour and became pale green upon drying. The FT-IR spectrum of the final mixture was very similar to that of Cu(yU-OH2)2[Au(CN)2]2 with additional bands attributable to ClOj vibrations. The formation of Cu[Au(CN)2]2(analyte)a; from the two salts should generate two equivalents of KC104 as a side product. In addition, a solubility test confirmed that KCIO4 is insoluble in acetonitrile. Hence, the white powder is as sumed to be KCIO4 and is responsible for the additional band observed in the FT-IR Chapter 3. M[Au(CN)2]2(sinalyte)x and their vapochromic properties 119 spectrum.

To avoid conversion of the dark green product to the Cu(/^-OH2)2[Au(CN)2]2 com plex, a different isolation method was employed. When the dark green precipitate (mixed with the KCIO4 powder) was isolated by decantation followed by solvent eva poration under vacuum, no colour change was observed. The FT-IR spectrum of this later mixture contained bands attributable to ClOj vibrations as well as different bands in the cyanide vibration frequency range (2297, 2269, 2191 cm-1). No band -1 around 2217 or 2171 cm , characteristic for the Cu(//-OH2)2[Au(CN)2]2 complex, could be observed. This indicates that no conversion, even partial, to the Cu(/i-

OH2)2[Au(CN)2]2 complex has occurred. The results obtained by elemental analysis suggested that the mixture was com posed of Cu[Au(CN)2]2(CH3CN)2(KC104)2. This is consistent with the precipitation of two equivalents of KCIO4 per equivalent of polymer formed. Two weak bands, at 2269 and 2297 cm-1, were attributed to the presence of acetonitrile in the complex. The bands corresponding to the vibration of the cyanide groups in acetonitrile (measured in a liquid film) are usually observed at 2254 and 2293 cm-1. The shift to higher energy could suggest that the molecules of acetonitrile are weakly N-bound to the Cu(ll) centers.

The presence of only one band corresponding to a [Au(CN)2]-based cyanide vi -1 bration (2191 cm ) suggests that all [Au(CN)2]~~ groups are in a similar environ ment. In addition, the frequency shift of the cyanide vibration band indicate bind ing of the [Au(CN)2]~ unit to the Cu(ll) centers, likely in a bridging mode like in 1 Cu[Au(CN)2]2(DMF), which has a band at 2199 cm" . No single crystal suitable for X-ray diffraction could be obtained. Also, the dark green Cu[Au(CN)2]2(CH3CN)2 complex was found to convert to the water adduct very rapidly when exposed to ambient moisture, which limited the characterization by powder X-ray diffraction (which requires 30—45 minutes of exposure to ambient atmosphere). However, it could be inferred from the FT-IR spectrum that the structure of

Cu[Au(CN)2]2(CH3CN)2 is similar to that of Cu[Au(CN)2]2(DMF) and composed of square grid arrays of Cu[Au(CN)2]2 with molecules of acetonitrile on each side of the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 120 grid.

3.4.4 Cu[Au(CN)2]2(dioxane)(H20)

The reaction of Cu(ll) with K[Au(CN)2] in a dioxane/water solution (for solubility purposes) yielded an immediate pale blue-green precipitate. No colour change was observed upon drying in ambient air. The results of CHN elemental analysis were consistent with the chemical composition being Cu[Au(CN)2]2(dioxane)(H20).

The FT-IR spectrum of Cu[Au(CN)2]2(dioxane)(H20) showed two bands in the cyanide vibration region at 2201 and 2172 cm-1. The shift of the cyanide stretching frequency from free [Au(CN)2]~~ to higher energy suggests that the [Au(CN)2]~ units are N-bound to the Cu(ii) centers or hydrogen bound to the water molecules, as in

Cu(//-OH2)2[Au(CN)2]2.

However, no crystal of Cu[Au(CN)2]2(dioxane)(H20) suitable for structure de termination could be prepared and the poor powder X-ray diffractogram prevented indexing and structure determination.

3.4.5 Physical Properties of Cu[Au(CN)2]2(analyte)x

Thermal stability

The thermal stability of the Cu[Au(CN)2]2(analyte)x complexes was determined using thermogravimetric analysis. Table 3.6 shows the temperature at which each of them decomposes along with the weight losses (expected and observed) and corresponding molecular fragments. In all cases, the solvent molecules are lost first, followed by the decomposition of the remaining Cu[Au(CN)2]2 framework between ~ 280 and 360 °C, with some decomposing over a larger temperature range. Prom the relative weight, it is proposed that the final decomposition product in each case contains Au and CuO. The formation of CuO at elevated temperature can be expected as the decomposition is carried out in air. Table 3.6: Thermal stability of the Cu[Au(CN)2]2(analyte)a; polymers: decomposition temperature and products obtained

Complex Temperature Fragment lost % Weight (°C) or product obtained calculated observed Cu[Au(CN)2]2(DMF) 195 - 280 -DMF lT5 13.5 310 - 355 - 2 C2N2 + O 13.8 11.5 400 CuQ + 2 Au 74.6 80.0 Cu[Au(CN)2]2(pyridine)2 155 - 190 - 1 pyridine 11.0 11.0 210 - 260 - 1 pyridine 11.0 12.5 310 330 - 2 C2N2 + 0 12.2 9.0 400 CuQ + 2 Au 65.8 66.0 Cu(//-OH2)2[Au(CN)2]2 140 - 180 - 2 water 6.0 5.5 320 - 360 -2 C2N2 + 0 14.7 15.5 400 CuQ + 2 Au 79.2 76.0 Cu[Au(CN)2]2 260-350 - 2 C2N2 + 0 15.2 14.0 400 CuQ + 2 Au 81.7 81.0

Cu[Au(CN)2]2(dioxane)(H20) 150 - 280 - dioxane - H20 15.9 17.5 290 - 330 - 2 C2N2 + 0 13.2 10.0 400 CuQ + 2 Au 70.9 71.0

Cu[Au(CN)2]2(NH3)4 50 - 95 - 1 NH3 2.7 3.0 115 - 220 - 3 NH3 8.1 7.5 280 - 350 - 2 C2N2 + 0 14.0 14.0 400 CuO + 2 Au 75.2 74.5 Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 122

Table 3.7: Effective magnetic moment (faff) determined for the Cu[Au(CN)2]2(analyte)a; coordination polymers in the high temperature range (T) and at 2.0 K.

a Analyte T M Veff,2.0K 9 J -1 (K) M (VB) (cm ) DMSO (green) 300 - 25 2.00(3) 1.74 DMSO (blue) 300 - 25 •1.90(3) 1.67 2.2353(8) 0.395(5) DMF (x =1) 300 - 60 2.10(3) 1.24 2.439(2) 1.53(2) Pyridine 300 - 25 2.20(3) 2.06 2.543(2) 0.174(5) Dioxane/water (x -=1/1) 300 - 25 1.95(3) 1.51 Ammonia (x =4) 300 - 25 1.80(3) 1.46 a For every complex, x = 2 unless otherwise noted.

Magnetic properties

The magnetic behaviour of each Cu[Au(CN)2]2(analyte)x complex was determined as the temperature was decreased from 300 K to 2 K. For each polymer, the effective magnetic moment remains relatively stable at high temperature and drops at low temperature to reach a minimum value at 2.0 K. The effective moments observed in the high temperature range and at 2.0 K are reported in Table 3.7. At 300 K, the effective moment was found to range between 1.80 and 2.15 HB for all complexes, and values of

1.46-2.06 nB are reached at 2.0 K. The magnetic behaviour of Cu[Au(CN)2]2(DMF) differs slightly from the others as the effective moment starts to decrease at 50 K and reaches a lower minimum value at 2.0 K (1.24 /j,B). These behaviours are consistent with mostly isolated S = \ Cu(ll) centers. (71) The range in effective moment at room temperature can be explained by slightly different values of the ^-factor for each complex, due to their different spheres of coordination. The drop in effective moment indicates that very weak antiferromagnetic interactions are present between the metal centers at low temperature. To estimate the magnitude of these interactions, the data obtained for the poly mers containing 2-D arrays of Cu(n) centers were fitted to the Lines model (Equa tion 3.3). The Hamiltonian describing a 2-D Heisenberg antiferromagnet can be Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 123 expressed as:

nn where Yl runs over aU pairs (nn) of nearest neighbor spins i and j and assumes that the same exchange coupling constant J is present within each pair of metal centers.b Using this Hamiltonian, the inverse susceptibility was determined by Lines to be equivalent to the following series expansion: (157) ^f = 3, + f;^ (3.3) *~ n=l 6 - 7TO) P'4> where Cn are numerical coefficients for a given spin value and the other variables have their usually definition.0 The values of g and J obtained with this fit are reported in Table 3.7. The obtained g values, especially for the DMF and pyridine adduct, were found to be unusually large for isolated Cu(n) ions (gcu usually ranges between 2.0 and 2.3). This is an indication that the Lines model is not a good model to obtain an accurate value for the strength of such weak interactions. This model works best when a clear deviation from linearity can be observed in the inverse of the susceptibility (i.e. when a maximum is present in the magnetic susceptibility). Nonetheless, the magnitude of the coupling constants determined when fitting to this equation is similar to what was reported for several other [Au(CN)2]-bridged coordination polymers.^57'158)

A larger decrease in effective moment was observed for the Cu[Au(CN)2]2(DMF) polymer and, as a result, a larger coupling constant (1.53(2) cm-1), although still relatively small, was obtained when fitting the data to the Lines model (Equation 3.3). Stronger interactions in the DMF-containing polymer could be explained by a better overlap of the magnetic orbitals through the equatorially bridged [Au(CN)2]~ units.

bDue to the notation used in this equation, a positive value of J is indicative of antiferromagnetic coupling. c For S = \, Cx = 4, C2 = 2.667, C3 = 1.185, C4 = 0.149, C5 = -0.191, C6 = 0.001. Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 124

The very close proximity of the layers (Cu-Cu 4.018 A, Figure 3.6C) compared to the other complexes could also favor through-space antiferromagnetic interactions.

3.5 Substituting the Cu(u) ions for Ni(n) and Co(n)

3.5.1 M[Au(CN)2]2(DMSO)2

Ni[Au(CN)2]2(DMSO)2

The reaction of Ni(ll) with K[Au(CN)2] in DMSO afforded the formation of a blue turquoise precipitate. The FT-IR spectrum of the isolated product shows two strong bands in the cyanide vibration region (2189 and 2180 cm-1) as well as bands at tributable to sulfoxide (S=0) vibration (1032-956 cm-1). From the elemental analysis results the chemical formula was assessed to be Ni[Au(CN)2]2(DMSO)2. No single crystal of this product, suitable for X-ray crystallography, could be obtained. However, the powder X-ray diffractogram of Ni[Au(CN)2]2(DMSO)2 was found to show similar features to the diffractogram of the Cu[Au(CN)2]2(DMSO)2

(blue) and Zn[Au(CN)2]2(DMSO)2 coordination polymers (Figure 3.8). The diffrac togram of Ni[Au(CN)2]2(DMSO)2 could be indexed to a monoclinic unit cell, with parameters very similar to those of the Zn analogue (Table 3.8). It is suggested that the structure of Ni[Au(CN)2]2(DMSO)2 is similar to the structure of the Cu(ll) analogue shown in Figure 3.2: octahedral Ni(ll) ions are surrounded by two DMSO molecules in a cis fashion and connected to four other Ni(ii) centers via [Au(CN)2]~ units generating, as a result, corrugated 2-D sheets. As in Cu[Au(CN)2]2(DMSO)2 (blue), the layers stack on top of each other in an offset fashion and two different Au-Au distances are observed (3.5 and 3.8 A). Fi gure 3.8 shows the diffractogram predicted by this proposed structural model, using the cell parameters obtained by indexing and atomic coordinates similar (Appendix A) to those of the Zn analogue.

It is suggested that the structure of Ni[Au(CN)2]2(DMSO)2 is closer to that of the Zn analogue as both Ni(n) (d8) and Zn(ll) (d10) octahedral centers are not expected Chapter 3. M[Au(CN)2]2(snialyte)x and their vapochromic properties 125

Figure 3.8: Comparison between the powder X-ray diffractograms predicted for Zn[Au(CN)2]2(DMSO)2 (Zn, green) and the blue Cu[Au(CN)2]2(DMSO)2 polymorph (Cu, blue) with the diffractogram obtained experimentally for Ni[Au(CN)2]2(DMSO)2 (Ni, orange). Ni* (black) represents the diffractogram predicted by the proposed struc tural model for Ni[Au(CN)2]2(DMSO)2. to exhibit Jahn-Teller distortion, while octahedral Cu(ll) (d9) is very often distorted.

The FT-IR spectrum of Ni[Au(CN)2]2(DMSO)2 also shows very similar features to the spectrum of Zn[Au(CN)2]2(DMSO)2, which has two bands in the cyanide vibration region (2186 and 2175 cm-1). The slight shift in vibration frequency between the Ni and Zn analogous polymers can be attributed to the different transition metal bound to the nitrogen atom of each cyanide group.

Co[Au(CN)2]2(DMSO)2

When Co(/x-OH2)2[Au(CN)2]2 was immersed in DMSO, a slight colour change to a different shade of pink occurred slowly. The resulting powder was isolated after three Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 126

Table 3.8: Unit cell parameters determined for the M[Au(CN)2]2(DMSO)2 coordina tion polymers (M = Zn, Cu (blue polymorph) and Ni), by either single crystal X-ray diffraction or powder X-ray diffraction.

Zn Cu(blue) Ni Co Sample crystal0. crystal powder powder crystal system monoclinic triclinic monoclinic monoclini space group P2i/c PI P2i/c P21/n a, A 7.8403(11) 7.874(7) 7.63 12.97 b, A 12.8395(12) 12.761(11) 12.84 14.45 c, A 16.4572(17) 16.207(13) 16.30 8.59 a, deg 90.0 89.61(7) 90.0 90.0 P, deg 98.891(10) 82.29(7) 98.89 94.47 7, deg 90.0 88.57(7) 90.0 90.0 Volume, A3 1636.77 1613.2(24) 1577.65 1605.88 ° From reference (156).

days to ensure complete conversion. The FT-IR spectrum of the isolated product contains only one band in the cyanide vibration frequency range (2176 cm-1) and does not show the characteristic bands for the Co(//-OH2)2[Au(CN)2]2 starting material. Results from elemental analysis suggest that the chemical formula of this product is

Co[Au(CN)2]2(DMSO)2. This reaction did not yield single crystals suitable for X-ray analysis, but only microcrystalline powder. Figure 3.9 shows the powder diffractogram collected for

Co[Au(CN)2]2(DMSO)2. The diffractogram was indexed and the unit cell parameters obtained are reported in Table 3.8. The diffractogram and the unit cell parameters do not match that of the Zn, Cu or Ni-containing M[Au(CN)2]2(DMSO)2 polymers presented above (Figure 3.8 and Table 3.8). This suggests that the Co-containing system adopts a different structural arrangement. Due to the poor quality of the diffractogram, no structure refinement could be performed, but a general model can be proposed based on comparisons with other known complexes.

The frequency of cyanide vibration in Co[Au(CN)2]2(DMSO)2 is shifted with re spect to the frequency observed in systems containing free [Au(CN)2]~ units, which Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 127

-1—i—i—i—|—i—i—i—i—|—i—T-,,,I—i—|—i—i—i—i—|—i—i—i—i—1~

Co JLUXJOLJU. JL JU_ Is' I LI UJI. Ill M Mn

J A• • • • 5 6 7 10 ii

Figure 3.9: Comparison between the powder X-ray diffractogram determined exper imentally for Co[Au(CN)2]2(DMSO)2 (purple) and the diffractogram predicted for 51 Mn[Au(CN)2]2(H20)2 (blue) from the single crystal structure.( )

suggests that the [Au(CN)2]~ units are N-bound to the Co(n) centers. The presence of only one vibration frequency suggests that all the cyanide groups are in a similar envi ronment. This is not the case in the corrugated sheet structure as some cyanide groups are trans to a DMSO molecule and some are trans to another cyanide group. As a consequence, the FT-IR spectrum of the Zn, Cu and Ni-containing systems all showed at least two different bands attributable to cyanide vibrations. All the cyanide groups could however be structurally equivalent if the structure of Co[Au(CN)2]2(DMSO)2 contained square-grid arrays of Co[Au(CN)2]2 with DMSO molecules bound to Co(n) ions in a trans fashion, on each side of the grid. -1 The cyanide vibration frequency observed for Co[Au(CN)2]2(DMSO)2 (2176 cm ) is similar to the frequency observed in the related Co[Au(CN)2]2(DMF)2 coordi nation polymer (2179 cm"1).'66' This polymer contains 2-D square-grid arrays of

Co[Au(CN)2]2 with DMF molecules bound to the Co(ll) ions on both sides of the grids. The powder diffractogram of Co[Au(CN)2]2(DMSO)2 does not however match that of Co[Au(CN)2]2(DMF)2, which can be due to the presence of different ligands (DMF vs DMSO) or could indicate a different packing of the layers.

Figure 3.9 compares the diffractogram of Co[Au(CN)2]2(DMSO)2 with that of Chapter 3. M[Au(CN)2]2(a,nalyte)x and their vapochromic properties 128

Mn[Au(CN)2]2(H20)2, which also adopts a square-grid M[Au(CN)2]2 structural ar rangement. (51) The main features in the two diffractograms are similar, which could suggest that the packing of the M[Au(CN)2]2 arrays in both polymers is comparable. Differences in the molecules bound on the metal center (water vs DMSO) could ac count for the differences observed in the diffractogram.

The main structural differences between the Co[Au(CN)2]2(DMF)2 and

Mn[Au(CN)2]2(H20)2 is that the Co[Au(CN)2]2 layers stack in pairs whereas the

Mn[Au(CN)2]2 layers are equally spaced and do not form pairs of layers. It is hence suggested that the Co[Au(CN)2]2(DMSO)2 polymer contains flat and equally spaced square-grid arrays, with DMSO molecules bound in a trans fashion.

3.5.2 Ni[Au(CN)2]2(DMF)2

When Ni(/x-OH2)2[Au(CN)2]2 was immersed in DMF for one day, a color change from green to blue was observed. The FT-IR spectrum of the isolated product showed only one band attributable to cyanide vibration, at 2189 cm-1. This spectrum differs from that of Ni(/x-OH2)2[Au(CN)2]2, which suggests that conversion occurred. Results from elemental analysis were consistent with the chemical composition being Ni[Au(CN)2]2(DMF)2. This indicates that replacement of the water molecules by two DMF molecules occurred. The composition of the Ni(ll)-containing DMF adduct differs from the composition of the Cu(ll)-containing polymer in which only one DMF molecule was incorporated. The composition is however analogous to that 66 of the previously reported Co[Au(CN)2]2(DMF)2 coordination polymer/ )

The powder X-ray diffractogram of Ni[Au(CN)2]2(DMF)2 was obtained and is shown in Figure 3.10. The diffractogram shows features similar to those present in the diffractogram of the Co[Au(CN)2]2(DMF)2 polymer. ^ The unit cell parameters determined by indexing are reported in Table 3.9. The unit cell dimensions were found to be similar to those of Co[Au(CN)2]2(DMF)2 (Table 3.9), after switching the a and c axis. The FT-IR spectrum of Ni[Au(CN)2]2(DMF)2 is also similar to that of the Co(ll) analogue which shows one cyanide vibration band at 2179 cm-1/66) Using the structure of the Co(ll) analogue as a starting point, the structure of Chapter 3. M[Au(CiV)2j2(a.naiyie/);E and their vapochromic properties 129

i 'I 1 1 1 "i1 i i r—|— ifiiii

P Co JlAJlAJtJi_U~__JLA\_ AJL___A f

O Ni > V^^M^^^A^A^M^^^^J^ •^L. w*V/V

fj^ijl JJAJLAILLAAOUULAL. . /LK. ^AA_ _i l_i i—i i ' * J• &• • 6 7 10.0 10.6

Figure 3.10: Comparison between the powder X-ray diffractogram predicted for Co[Au(CN)2]2(DMF)2 (Co, purple), the diffractogram determined experimentally for Ni[Au(CN)2]2(DMF)2 (Ni, orange) and the diffractogram of its proposed structural model (Ni*, black).

Ni[Au(CN)2]2(DMF)2 was determined from the observed powder X-ray diffractogram (the atomic coordinates are reported in Table A. 12). Figure 3.11 shows the proposed model for Ni[Au(CN)2]2(DMF)2 and Figure 3.10 compares the diffractogram predicted by this model with the diffractogram obtained experimentally. The powder X-ray diffractogram of such compounds is mainly influenced by the atomic position of the heavy atoms (Ni and Au). As such, larger errors are associated with the position and orientation of the DMF molecules, which were mainly estimated from the crystal structure of the Co[Au(CN)2]2(DMF)2 analogous polymer.

As for Co[Au(CN)2]2(DMF)2, the structure contains 2-D square grids of

Ni[Au(CN)2]2 (Figure 3.11A). The DMF molecules are O-bound to the Ni(n) cen ters on each side of the grids, which stack in an offset manner to accommodate the

DMF molecules. The [Au(CN)2]~ units are slightly bent and sit above and below the plane containing the Ni(ll) centers (Figure 3.11B and C). As observed in the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 130

Table 3.9: Unit cell parameters determined for the M[Au(CN)2]2(DMF)2 (M = Co, Ni) and Co[Au(CN)2]2(pyridine)2coordination polymers by either single crystal X-ray diffraction or powder X-ray diffraction.

Co[Au(CN)2]2 Ni[Au(CN)2]2 Co[Au(CN)2] (DMF)2 (DMF)2 (pyridine) 2 Sample crystal" powder powder crystal system monoclinic monoclinic monoclinic space group P2i/c P2i/a P2i/a a, A 8.375(2) 14.7907 15.0425 b, A 14.054(5) 14.1445 14.0712 c, A 15.077(4) 8.5659 8.9334 a, deg 90.0 90.0 90.0 p, deg 92.75(2) 95.011 99.2538 7, deg 90.0 90.0 90.0 Volume, A3 1772.6(9) 1785.2 1866.3 a From reference (66).

Figure 3.11: Structural model proposed for Ni[Au(CN)2]2(DMF)2: A. 2-D square grid array with DMF molecules lying on both sides of the layer; B. and C. side-view of a 2-D layer, showing the positions of the [Au(CN)2]~ units with respect to the Ni(ll) centers. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 131

Co[Au(CN)2]2(DMF)2 structure, the layers stack in pair with aurophilic interactions of ca. 3.3 A present within a pair, whereas Au-Au distances of ca. 5.4 A are observed between pairs.

3.5.3 M[Au(CN)2] 2 (pyridine) 2

Co[Au(CN)2]2(pyridine)2

When a powdered sample of Co(/x-OH2)2[Au(CN)2]2 was immersed for one day in a mixture of water and pyridine, a slight colour change was observed. The FT-IR spectrum of the isolated powder showed only one strong band attributable to cyanide vibration at 2168 cm-1 and pyridine-based bands could also be observed. No feature reminiscent of the Co(/^-OH2)2[Au(CN)2]2 complex was present in the FT-IR spectrum suggesting that a complete conversion has been achieved. The chemical composition of this new product was assessed based on the results from elemental analysis, which were consistent with the chemical for mula being Co[Au(CN)2]2(pyridine)2. The powder X-ray diffractogram of

Co[Au(CN)2]2 (pyridine^ was obtained and is shown in Figure 3.12. It differs slightly from the diffractogram of the Cu-containing analogous complex (discussed in sec tion 3.4.2) and contains a few additional peaks. It also shows similar features to the diffractogram of the Co[Au(CN)2]2(DMF)2 coordination polymer (Figure 3.10).

The diffractogram of Co[Au(CN)2]2(pyridine)2 was indexed and the optimized unit cell parameters are reported in Table 3.9. These are comparable with those of the Ni- and Co-containing M[Au(CN)2]2(DMF)2 complexes (after switching the parameters for the a and c-axis of the Co analogue).

Using a combination of the Cu[Au(CN)2]2(pyridine)2 and Co[Au(CN)2]2(DMF)2 structures as a starting point, a structural model is proposed for r Co[Au(CN)2]2(py idine)2 (the corresponding atomic coordinates are reported in

Table A.13). The structure consists of 2-D square grids of Co[Au(CN)2]2 with pyridine molecules N-bound to the Co(ll) centers on both sides of the layer (Fig ure 3.13A). The structure differs from the Cu(ll) analogue as all the cyanide groups are equidistant from the Co(ll) centers and no Jahn-Teller distortion is observed. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 132

Also, the Co[Au(CN)2]2 layers are not completely flat like the Cu[Au(CN)2]2 layers, - but the [Au(CN)2] units are buckled in an alternate way on each side of the layer as is observed in the related Co[Au(CN)2]2(DMF)2 polymer (Figure 3.13B). This arrangement allows TT-TT between the pyridine rings, but no aurophilic interactions are present as the closest contact between two Au atoms is around 3.7 A. The powder diffractogram predicted by this model is compared to the experimental diffractogram of Co[Au(CN)2]2(pyridine)2 in Figure 3.12. The features (positions and relative intensities) are mostly predicted by the model. The orientation of the pyridine molecules was not refined as the diffractogram is mainly affected by the atomic position of the heavy atoms and slight changes in the position of the pyridine molecules do not cause noticeable changes in the diffractogram. A larger error is then associated with their position. The model is also consistent with the FT-IR spectrum observed for

Co[Au(CN)2]2(pyridine)2. As all the cyanide groups are equivalent to each other, only one cyanide vibration frequency should be observed. The similarities between the structures of the DMF and pyridine Co[Au(CN)2]2(analyte)2 complexes would also explain the similar vibration frequencies (2179 and 2168 cm-1 respectively), the shift being attributable to the differences in ligand field around the Co(ll) centers in the two complexes (M(NC)402 vs M(NC)4N2).

Ni[Au(CN)2]2(pyridine)2 vs Ni[Au(CN)2]2(pyridine)4

Attempts to synthesize the Ni[Au(CN)2]2(pyridine)2 analogous polymer were not suc cessful. Two [Au(CN)2]-containing products seem to be produced in different amounts according to the reaction conditions. The FT-IR spectrum of the reaction products generally shows four different bands attributable to cyanide vibration (2182, 2171, 2153, 2143 cm-1), with different relative intensities for each reaction. When the reaction was performed in a mixture of water (96 %) and pyridine (4 %), it was possible to isolate Ni[Au(CN)2]2(pyridine)4, as confirmed by elemental analysis. The FT-IR spectrum of this product only shows two of the cyanide vibration band (2171 and 2143 cm-1). If four pyridine molecules are attached on each Ni(ll) center, Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 133

i—i—r i—•—] • —'—• ' r™> ' ' -' r

~^MIL, , _A_JUL_ <

Co : 1 A. TV A

Co*- . A. A A • • ' • 1 1 .... 1 1 1 5 6 7 10

Figure 3.12: Comparison between the powder X-ray diffractogram predicted for Cu[Au(CN)2]2(pyridine)2 (Cu, orange), the experimental diffractogram of Co[Au(CN)2]2(pyridine)2 (Co, purple) and the diffractogram predicted by its struc tural model (Co*).

the overall structure can either be molecular with two dangling [Au(CN)2]~ groups 1 or can contain chains of {Ni[Au(CN)2](pyridine^}" " with free [Au(CN)2]~ units in between the chains. Both of these structures would explain the presence of a cyanide vibration band with a frequency similar to what is observed in K[Au(CN)2] (2141 cm-1).

It is proposed that Ni [Au(CN)2] 2 (pyridine) 2 is the other product formed by the reaction. It was not however possible to isolate a pure sample for characterization and to confirm this hypothesis. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 134

Figure 3.13: Structural model proposed for Co[Au(CN)2]2(pyridine)2: A. 2-D square- grid array with pyridine molecules lying on both sides; B. side-view of a 2-D layer, - showing the positions of the [Au(CN)2] units with respect to the Co(n) centers.

3.5.4 Physical properties of Ni- and Co-based polymers

Solid-state UV-vis-NIR spectroscopy

The UV-vis-NIR absorbance spectrum of the Ni and Co-containing M[Au(CN)2]2(analyte)a; polymers were determined and are shown in Figure 3.14. The estimated positions of the absorbance bands are reported in Table 3.10. The spectrum of each Ni(ll)-containing complex shows three absorbance bands.

The Ni[Au(CN)2]2(analyte)2 polymers containing Ni(NC)4(0)2 coordination shells (ie the DMSO and DMF adducts) show absorbance bands with very similar positions.

The spectrum of Ni(/Li-OH2)2[Au(CN)2]2, which contains Ni(NC)2(0)4 centers, display bands shifted to lower energy, whereas the spectrum of Ni[Au(CN)2]2(pyridine)4 is shifted to slightly higher energy. The absorbance bands were assigned to spin-allowed d-d transitions from the 3 8 A2g ground state (Table 3.10). Using the Tanabe-Sugano diagram for d ions in 159 an octahedral field/ ) the crystal field splitting energy, A0, was estimated using the ratio of the transition energies and the energy of the first d-d transition band Chapter 3. M[Au(CN)2]2(a'nalyte)x and their vapochromic properties 135

^ Frequency [10 cm" ] g Frequency [ 10 cm" ] 25 20 15 12.5 10 ' 25 20 15 12.5 10 11 i i i i i i—i—i—i 1 1 1 1 1 M i i • i i i i—i—i—i 1 1 1 r

400 500 600 700 800 900 1000 1100 400 500 600 700 800 900 1000 1100 Wavelength [ nm ] Wavelength [ nin ]

Figure 3.14: A. Solid-state UV-vis-NIR absorbance spectra of Ni[Au(CN)2]2(analyte)2 (water (green), DMF (red), DMSO (purple) and pyridine (blue)); B. Solid-state UV- vis-NIR absorbance spectra of Co[Au(CN)2]2(analyte)2 (water (green), DMF (red), DMSO (purple) and pyridine(blue)).

(Table 3.10).

The values of A0 determined for the Ni[Au(CN)2]2(analyte)2 polymers are similar 2+ 16 to the value reported for [Ni(CH3CN)6] .( °) The general trend observed for the A0 of these complexes:

Ni(NC)2(Npyridine)4 > Ni(NC)4(0)2 > Ni(NC)2(0)4 is consistent with the spectrochemical series, which predicts that O-bound molecules have a weaker ligand field than nitriles.^160)

Despite the differences in coordination spheres (Co(NC)4(0)2 vs Co(NC)4(N)2), the spectra of the Co[Au(CN)2]2(DMSO)2, Co[Au(CN)2]2(DMF)2 and

Co[Au(CN)2]2(pyridine)2 polymers show two similar broad absorbance bands.

The spectrum of the Co(/i-OH2)2[Au(CN)2]2 polymer, containing Co(NC)2(0)4 centers, is however shifted toward lower energy. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 136

-1 Table 3.10: Absorbance maxima (Amax abs., cm ) observed in the solid-state UV-Vis- NIR absorbance spectra of the M[Au(CN)2]2(analyte)a; (M = Co and Ni) coordination 3 polymers and their respective assignment to d-d transitions (ground state: A2g for 4 the Ni(ll) complexes and Tig for the Co(ll) complexes).

1 Ni(ll) complexes '"max abs. (Cm - ) A0 a 3 a -1 T29 Tl9 (F) Tl9 (P) (cm ) Ni(//-OH2)2[Au(CN)2]2 10,750 16,000 25,640 10,750 Ni[Au(CN)2]2(DMSO)2 10,810 16,130 25,975 10,810 Ni[Au(CN)2]2(DMF)2 10,810 16,400 25,975 10,810 Ni[Au(CN)2]2(pyridine)4 11,050 17,390 26,670 11,050 2+ a [Ni(DMSO)6] 7730 12,970 24,040 7730 2+ a [Ni(H20)6] 8500 13,800 25,300 8500 2+ a [Ni(pyridine)6] 10,150 16,500 27,000 10,150 2+ a [Ni(CH3CN)6] 10,700 17,400 27,810 10,700 2+ a [Ni(en)3] 11,700 18,350 29,000 11,700

4 4 4 Co(ll) complexes T2g A2/ or Tl3 Co(/x-OH2)2[Au(CN)2]2 < 9,000 19,050 Co[Au(CN)2]2(DMF)2 9615 20,000; 21,300 Co[Au(CN)2]2(DMSO)2 9660 20,410; 21,370 Co[Au(CN)2]2(pyridine)2 10,000 19,050 (sh); 20,620 2 a [Co(H20)6] + 8100 19,400 2+ a [Co (pyridine) 6] 9800 20,400 2+ a [Co(en)3] 10,100 21,000 a From reference (16°) 6 The absorbance band is usually very weak and no band could be definitively assigned to this transition for the Co[Au(CN)2]2(analyte)2- Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 137

The absorbance bands observed for the Co[Au(CN)2]2(analyte)2 polymers were 4 4 assigned to the spin-allowed d-d transitions from the Tl9 ground state to the T2p 4 and Tlp (P) excited states. Octahedral Co(ll) ions also have a third spin-allowed 4 transition to the A2a state, which should be close in energy to the transition to the 4 160 Ti9 (P) state. However, for most complexes, this absorbance band is very weak^ ^ and is not observed or arises only as a shoulder peak. As a consequence, no band in the spectra of the Co[Au(CN)2]2(analyte)2 polymers was formally attributed to that transition.

The crystal field splitting energy for the Co[Au(CN)2]2(analyte)2 complexes could not be calculated but was estimated from a comparison to other complexes.d The crystal field splitting energy of Co[Au(CN)2]2(DMSO)2 and Co[Au(CN)2]2(DMF)2 is similar to that of [Co(pyridine)6]2+ whereas the crystal field splitting energy of 2+ 16 Co[Au(CN)2]2(pyridine)2 is larger and closer to that of [Co(en)3] .( °) As the lower energy absorbance band of Co(/U-OH2)2[Au(CN)2]2 lies outside the measured energy range (~9000 cm-1), it is hard to compare the crystal field splitting energy of this 2+ complex. It can however be deduced that it is smaller than that of [Co(pyridine)6] 2+ and closer to that of [Co(H20)6] .

The trend in A0 observed for the Co-containing polymers is similar to the one observed for the Ni-containing polymers and also consistent with the spectrochemical series.

Co(NC)4(Npyridine)2 > Co(NC)4(0)2 > Co(NC)2(0)4

Magnetic properties

The magnetic properties of the Ni and Co-containing M[Au(CN)2]2(analyte)a; poly mers were determined and are summarized in Table 3.11. Similarly to the magnetic behaviours of the Cu[Au(CN)2]2(analyte)a; polymers presented in section 3.4.5, the effective magnetic moment for each Ni complex was found to be stable at high tem perature, dropping only at low temperature to reach a minimum at 2 K. In all cases,

Calculations to determine the crystal field splitting energy, A0, using the Tanabe-Sugano diagram for an octahedral d 7 ion require the assignment of all transitions. Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 138 no maximum in susceptibility could be observed down to 2 K. The effective moment observed at room temperature for each Ni(ll)-containing polymer is typical for complexes containing isolated S = 1 centers. (71) The drop at low temperature could be due to either very weak antiferromagnetic coupling or zero-field splitting. The data for each Ni(ll)-containing polymer can be fit to the equation for the zero-field splitting of an isolated S = 1 center (Equation 1.17). The D and g values obtained from the fit to this equation, for each Ni[Au(CN)2]2(analyte)2 polymer, are reported in Table 3.11. The D values obtained for the DMSO- and DMF-containing polymers are larger than the value observed for Ni[Au(CN)2J2(pyridine)4. The smaller D value is consis tent with the presence of more symmetrical Ni(ll) centers in Ni[Au(CN)2J2(pyridine)4

(Ni(N)6) compared to the DMSO- and DMF-containing polymers (Ni(N)4(0)2). Re ported D values for isolated octahedral Ni(ll) systems are usually on the order of a few wavenumbers. ^70^ These large D values suggests that weak antiferromagnetic coupling is also present between the zero-field split Ni(ll) centers. To account for these interactions, a molecular field approximation was introduced to the zero-field splitting equation to estimate the magnetic coupling between the Ni(ll) centers: Xm/=x*/s'{i-i^x24 (3-5) where J is the exchange coupling constant with the z neighboring Ni(ll) centers. (161^ The parameter values obtained with this second fit are also reported in Table 3.11. The coupling constants are relatively small (~ -0.5 cm-1), which is consistent with very weak antiferromagnetic interactions mediated through the [Au(CN)2]~ units. The obtained D values are smaller as a result of the introduction of the zJ pa rameter. In general, the fit was improved by the introduction of this additional parameter, but care has to be taken as overparametrization of very weak interactions (< -1 cm-1) does not allow significant conclusions to be made. It is especially the case for the Ni[Au(CN)2]2(pyndine)4 polymer for which a very small coupling constant is predicted. Reported values of D for octahedral Ni(ll) centers in related [Au(CN)2J- and Chapter 3. M[Au(CN)2]2 (analyte) x and their vapochromic properties 139

Table 3.11: Effective magnetic moment (fJ-eff) determined for the M[Au(CN)2]2(analyte)a; (M = Ni, Co) coordination polymers at higher temper atures (T) and at 2.0 K, along with the D, g and z J parameters obtained from fitting for the Ni(ll) complexes.

Ni(n) complexes T M M, 2.0 K D 9 zJ -1 -1 analyte (K) (»B) (MB) (cm ) (cm ) DMF 300 - 40 3.18(3) 1.97 6.95(7) 2.18(2) 5.80(2)a 2.232(2)a -0.47(2)a DMSO 300 - 60 3.27(3) 2.51 4.28(8) 2.22(2) 2.25(2)a 2.311(2)a -0.546(5)a Pyridine6 300 - 35 3.15(3) 2.76 2.81(4) 2.193(4) 1.45(2)a 2.251(2)a -0.270(3)a Co(ll) complexes analyte DMSO 300 - 100 5.30(3) 3.98 - - - Pyridine 300 - 220 5.00(3) 3.75 - - - DMFC 300 4.92 3.79 - - - a Values were obtained by introducing a molecular field approximation to the zero-field splitting equation (Equation 3.5). 6 For every complex, x = 2 except for the Ni[Au(CN)2]2(pyridine)4 complex for which x = 4. c Reported by Colacio and co-workers. ^66^

[Ag(CN)2]-based polymers with very weak antiferromagnetic interactions, such as in -1 (48) Ni(tren)[Au(CN)2]2, range from 1 to 4 cm . The weak antiferromagnetic interactions between the Ni centers of each square- grid array are most likely mediated by the [Au(CN)2]~~ units. These interactions are consistent with the interactions observed in the analogous CufA^CN^^analyte)^ polymers (section 3.4.5). The magnetic behaviours of the DMSO and pyridine-containing

Co[Au(CN)2]2(analyte)2 polymers are comparable to the behaviour reported for the structurally similar Co[Au(CN)2]2(DMF)2 polymer. Table 3.11 compares the effective magnetic moment observed for each polymer at high and low temperature. For each polymer, the effective moment is stable at high temperature and then drops Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 140 as the temperature is lowered to reach a minimum value at 2 K. No maximum in magnetic susceptibility is observed over the temperature range studied. The observed room temperature values are larger than the spin-only value of an S — | system (3.87 /J,B), but are in the range reported for octahedral high-spin Co(ll) centers (4.1 - 5.2 HB).^2^ The difference is due to the presence of a large first-order orbital contribution, common to high-spin octahedral d7 systems. 53 ) The effective moment of Co(pyrazme)2[Au(CN)2]2^ ' was reported to be 5.21 /iB at 300 K, and decreased to 3.64 \xB at 1.8 K.

3.5.5 Comparing the vapochromic behaviour of Ni- and Co- based polymers

With the characterization of each M[Au(CN)2]2(analyte);r (Ni and Co) coordination polymers in hand, the potential vapochromic behaviour of the M(/i-OH2)2[Au(CN)2J2 polymers (Ni and Co) when exposed to vapour of DMSO, DMF and pyridine was investigated. Each sample was exposed to the specific vapour for several hours (or days) in a sealed container, after which the FT-IR spectrum and the powder X-ray diffractogram of the product was collected. The frequency of the cyanide vibration bands in each product is reported in Table 3.12 and compared with that of the product made from solution.

When Co(/x-OH2)2[Au(CN)2]2 was exposed to vapour of DMF, very faint color changes were observed after several hours but, when investigated by FT-IR spec troscopy, differences could be observed. The product obtained has an FT- IR spectrum and a powder X-ray diffractogram indistinguishable from that of Co[Au(CN)2J2(DMF)2. This suggests that the final product has the same chemical formula and adopts the same structural arrangement as Co[Au(CN)2]2(DMF)2. No band attributable to CO(/J-OH2)2[AU(CN)2]2 could be observed in the FT-IR spec trum, indicating that the conversion was complete.

Upon exposure to DMSO, the Co(/i-OH2)2[Au(CN)2]2 polymer slowly converted and after three days an FT-IR spectrum and a diffractogram identical to those of

Co[Au(CN)2]2(DMSO)2 were obtained. This also indicates that a slow conversion can Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 141

Table 3.12: Cyanide vibration frequencies (VCN, cm-1) for different M[Au(CN)2]2(analyte)x (M = Ni and Co) complexes synthesized in solution and from vapour absorption.

from solution from absorption" Complex UCN vapour V(JN (cm"1) (cm-1)

Co(//-OH2)2[Au(CN)2]2 2204 (s), 2168 (s) b Co[Au(CN)2]2(DMF)2 2182 (s) DMF 2182 (s) Co[Au(CN)2]2(DMSO)2 2177 (s) DMSO 2176 (s) Co[Au(CN)2]2(pyridine)2 2168 (s) pyridine 2168 (s), 2144 (s)

Ni(//-OH2)2[Au(CN)2]2 2214 (s), 2204 (sh), 2170 (s)

Ni[Au(CN)2]2(DMF)2 2189 (s) DMF 2215 (m), 2189 (s), 2171 (m)

Ni[Au(CN)2]2(DMSO)2 2189 (s), 2180 (s) DMSO 2214 (m), 2203 (w), 2181 (s), 2170 (s), 2164 (s) Ni[Au(CN)2]2(pyridine)4 2171 (s), 2143 (s) pyridine 2213 (s), 2202 (m), 2170 (s), 2162 (s), 2141 (s) a Solvent adducts were made from CO(M-OH2)2[AU(CN)2]2 and from Ni(/i-OH2)2[Au(CN)2]2. 6 The value obtained experimentally differs slightly from the value of 2179 cm-1 reported by Colacio and co-workers. (66>

happen from the water to the DMSO containing Co-polymer.

When Co(/Li-OH2)2[Au(CN)2]2 was exposed to pyridine vapour, conversion also occurred, but a product having an FT-IR spectrum and a powder X-ray diffractogram different from the Co[Au(CN)2]2(pyridine)2 polymer is obtained. Simi -1 larly to Co[Au(CN)2]2(pyridine)2, the FT-IR spectrum shows a band at 2168 cm but an additional band attributable to cyanide vibration at 2144 cm-1 is also present. The powder X-ray diffractogram of this product also differs from that of

Co[Au(CN)2]2(pyridine)2. It shows very similar features to those observed in the diffractogram of the Ni[Au(CN)2]2(pyridine)4 complex. This rules out the presence Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 142

of Co[Au(CN)2]2(pyridine)2 mixed with an additional compound in the final product. The diffractogram suggest that the final product, after exposure to pyridine, adopts a structure similar to that of Ni[Au(CN)2]2(pyridine)4 which contains four pyridine molecules instead of two like the Co-based product prepared in solution.

In addition, the Co[Au(CN)2]2(DMSO)2 polymer was found to be stable for a long period of time (months) when exposed to ambient atmospheric conditions. This indicates that replacement of DMSO molecules for water molecules does not oc cur at room temperature for Co[Au(CN)2]2(DMSO)2, which contrasts with the two

Cu[Au(CN)2]2(DMSO)2 polymorphs that readily undergo conversion.

When Ni(/x-OH2)2[Au(CN)2]2 was exposed to the same vapours, the response time was longer and the color changes very subtle. For example, after one week of being exposed to DMF vapour, the FT-IR spectrum of the product showed the presence of the starting Ni(/x-OH2)2[Au(CN)2J2 compound as well as an additional band due partial conversion of the starting material to Ni[Au(CN)2]2(DMF)2. When exposed to DMSO for one week, cyanide vibration bands similar to those observed in the FT-IR spectrum of the Ni(/i-OH2)2[Au(CN)2]2 and

Ni[Au(CN)2]2(DMSO)2 polymers are present as well as an additional band. One -1 of the characteristic bands of Ni[Au(CN)2]2(DMSO)2 (2189 cm ) is however not ob served in FT-IR spectrum of the final product, which could be the result of peak overlap or could indicate that this polymer is not formed.

Upon exposure to pyridine vapour, the Ni(yu-OH2)2[Au(CN)2]2 polymer undergoes a slight color change toward blue over several days. The FT-IR spectrum shows several bands in the cyanide vibration frequency region. Some are similar to those 1 of Ni(^-OH2)2[Au(CN)2]2 (2213, 2202 and 2170 cm" ), some are similar to those of -1 Ni[Au(CN)2]2(pyridine)4 (2170, 2141 cm ) and an additional band is also present (2162 cm-1). This could indicate that exposure to pyridine vapour resulted in a mixture of unreacted Ni(^,-OH2)2[Au(CN)2]2, formed Ni[Au(CN)2]2(pyridine)4 and/or a new compound. These results indicate that conversion upon vapour exposure is not favored for the

Ni(/x-OH2)2[Au(CN)2]2 polymer. Also, in the case of DMSO and pyridine, a different product may be formed by vapour absorption. Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 143

3.6 Discussion

3.6.1 Polymorphism in coordination polymers

One common obstacle in the synthesis of coordination polymers is dealing with the facile formation of supramolecular isomers, or polymorphs: the existence of more than one possible type of superstructure for the same building blocks. (163_167) Factors such as the crystallization solvent, starting reagent, temperature, presence of seed crystals, and concentration all play a role in determining which polymorph will be isolated 165 167 from a given reaction mixture. ( ~ ) For example, crystallizing Ni[Au(CN)2]2(en)2 from the reaction of K[Au(CN)2] with [Ni(en)3]Cl2-2H20 or [Ni(en)2Cl2] generates a molecular complex and a 1-D polymer, respectively. (158) The ability to selectively produce only one polymorph is of vital concern because the materials property targeted usually depends on the three-dimensional solid-state structure. For example, polymorphs of the same material can show different mag netic, (168~170) conducting/171) luminescent ,(172) and zeolitic properties. (173) In this study, the results obtained by X-ray crystallography and elemental anal ysis indicate that the two Cu[Au(CN)2]2(DMSO)2 polymers (green and blue) are polymorphs. They both contain the same building blocks (in the same ratio), but their structural arrangements differ. This contrasts with pseudo-polymorphs that differ by incorporation of varying amounts or identities of cocrystallized solvent 163 164 molecules. ( > ) Pseudo-polymorphs of [Au(CN)2]-containing polymers were not ob served in this work.

In the Cu[Au(CN)2]2(DMSO)2 system, the synthesis of only one polymorph was found to be possible by controlling the concentration of the reaction. If the total concentration of reagents is below 0.2 M, Cu[Au(CN)2]2(DMSO)2 (green) is formed, whereas Cu[Au(CN)2]2(DMSO)2 (blue) is obtained exclusively from more concen trated solutions (> 0.5 M). The concentration-controlled synthesis of structural iso mers of coordination polymers is uncommon relative to the numerous examples with 174 175 molecular systems. ( > ) {Cu[N(CN)2]2(pyrazine)}n is one of the few examples of Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 144 such systems and forms green-blue and blue polymorphs when crystallized from con centrated and dilute solution, respectively. (m) This concentration dependence suggests that Cu[Au(CN)2]2(DMSO)2 (green) is the thermodynamic product, while Cu[Au(CN)2]2(DMSO)2 (blue), which rapidly pre cipitates from more concentrated solutions, is likely a kinetic product. The fact that

Cu(//-OH2)2[Au(CN)2]2 converts exclusively to the Cu[Au(CN)2]2(DMSO)2 (green) polymorph in the presence of DMSO is further evidence that Cu[Au(CN)a]2(DMSO)2 (green) is the most energetically favorable polymorph.

The density of the thermodynamic Cu[Au(CN)2]2(DMSO)2 (green) polymer 3 (2.719 g cm" ) is lower than that of the kinetic Cu[Au(CN)2]2(DMSO)2 (blue) poly mer (2.955 g cm-3). This contrasts with the prediction that the thermodynamic polymorph (at absolute 0 K) should have the largest density of all possible polymor phic forms in order to minimize its packing energy. (176) A few cases of polymorphic systems have been reported in which the thermodynamic polymorph has a lower den 17 sity than the kinetic one. ( °) In the case of the Cu[Au(CN)2]2(DMSO)2 polymorphs, it is conceivable that the formation of shorter Au-Au bonds in Cu[Au(CN)2]2(DMSO)2

(green) relative to Cu[Au(CN)2]2(DMSO)2 (blue) is an important energetic factor that plays a role in determining the energetically favoured polymorph.

The color difference between the two Cu[Au(CN)2]2(DMSO)2 polymorphs can be attributed to a difference in coordination number (5 vs 6), geometry (distorted bipyramidal vs octahedral) and ligand set (3 vs 4 cyanide groups) around the Cu(ll) centers.

Despite having different structural arrangements, the two Cu[Au(CN)2]2(DMSO)2 polymorphs exhibit similar magnetic and vapochromic behaviours.

3.6.2 Solution synthesis vs solvent exchange in the solid state

The Cu[Au(CN)2]2(analyte)-c polymers synthesized from solution can be compared to those generated by exposure to the given solvent vapour. The elemental analysis and thermogravimetric analysis results indicate that the number of analyte molecules x incorporated per Cu(ll) center is always the same for a given analyte, whether the polymer is synthesized from solution or by vapor exposure. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 145

The constant x value for a given analyte can be rationalized by the fact that all incorporated analyte molecules are ligated, through a donor group, to a Cu(n) center in a specific ratio. No additional, loosely trapped analyte molecules are present in the channels/cavities of the polymer (as shown by thermogravimetric analysis, Table 3.6). The Cu(n) to analyte molecules ratio varies as a function of analyte. In most cases, a 1:2 ratio is observed (water, DMSO, acetonitrile, pyridine). However, for DMF, a 1:1 ratio is observed, and a 1:4 ratio is adopted when ammonia is the vapour present. When dioxane is used, a mixture of dioxane and water (1:1) is detected in the final product. The possibility of having a sample containing 50 % pure Cu(/x-

OH2)2[Au(CN)2]2 and 50 % pure Cu[Au(CN)2]2(dioxane)2 was discarded after ex amining the FT-IR spectrum, which does not contain the two bands characteristic for

Cu(/i-OH2)2[Au(CN)2]2.

The constant value of £ observed in this Cu[Au(CN)2]2(analyte);r system contrasts with most porous systems in which guest molecules are included in the cavities/pores and held only by very weak interactions.^9'177"179^ In these cases, the number of guest molecules greatly varies depending on the experimental conditions (temperature, pres sure, guest concentration). Powder X-ray diffraction and FT-IR spectroscopy (Table 3.3) showed indis tinguishable results for a given Cu[Au(CN)2]2(analyte);c, irrespective of the syn thesis route (i.e. from solution or vapour exposure). This indicated that, in addition to an identical chemical formula, the structural arrangement for each

Cu[Au(CN)2]2(analyte)a; polymer synthesized from solution is identical to the polymer generated by analyte exchange.

While identical Cu[Au(CN)2]2(analyte);r polymers are formed for a given ana lyte irrespective of the synthetic conditions, the same is not true when Ni(ll) and Co(ll) are used as the transition metal. Different products, with variable x values, were obtained depending on the synthesis route. For example, at least two different

Ni[Au(CN)2]2(pyridine)a; products were formed in solution. Also, solution synthesis and vapour absorption yielded different Co[Au(CN)2]2(pyridine)^ products. In gen eral, with the analytes chosen in this study, vapour absorption by Ni or Co-containing polymers does not yield single M[Au(CN)2]2(analyte)I products. Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 146

These results suggest that for the Cu-based systems, there exists only one ener getically favoured product for each specific analyte, whereas for the Ni and Co-based systems, more than one stable M[Au(CN)2]2(analyte)2; polymer can be obtained for a given analyte and metal center.

3.6.3 Comparison of structural motifs adopted upon modifi cation of metal center and analyte molecules

Despite the differences between the analyte molecules incorporated, the basic struc tural motif of the M[Au(CN)2]2(analyte)x coordination polymers remained the same.

In almost all the 2-D square-grid network of M[Au(CN)2]2 was obtained from the alternation of M(ll) ions and [Au(CN)2]~ units. The square-grid array is flexible and can adapt to accommodate the system depending on the metal ion and analyte type. Scheme 3.3 summarizes the different types of distortion observed in this study. The major mode of flexibility lies in the fact that the 2-D square-grid can be entirely flat, as in the M[Au(CN)2]2(DMF)2 or M[Au(CN)2]2(pyridine)2 com plexes, or it can buckle to generate a corrugated 2-D array, as observed in several

M[Au(CN)2]2(DMSO)2 (M = Zn, Ni or Cu (blue polymorph)) polymers. In addition, in a flat square-grid, the [Au(CN)2]~ units can lie completely co-planar with the M(ll) centers as in Cu[Au(CN)2]2(pyridine)2 or bent out of the M(ll)-containing plane, as in M[Au(CN)2]2(DMF)2. As expected, due to Jahn-Teller distortions, the structure of the polymers contain ing Cu(ll) ions are more distorted than that of the Ni(ll) and Co(n)-containing poly mers. Jahn-Teller distortions are endemic to octahedral Cu(ll) complexes and yield another mode of flexibility: the equatorial/axial arrangement of the cyanide ligands and analyte molecules. For example, in Cu[Au(CN)2]2(pyridine)2, two cyanide ligands are equatorial and two are axial, leading to significantly different Cu-N(cyanide) bond lengths. This form of structural flexibility is particularly important since substantially different FT-IR signatures in the cyanide vibration region are generated depending on the cyanide bonding arrangement in the system. An additional mode of flexibility lies in the ability of the Cu(n) centers to readily Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 147

Zn/Cu(blue)/Ni Cu (green) DMSO DMSO tofefe

pf pf y* Basic M[Au(CN)J2 4 4 4 square-grid array side view: corrugation fragmentation >*••»

Jahn-Teller effect 1 or 2 analyte JLJUL. \ \ \ molecules per M 4dbC f ^ •* side view: fjj —~$—#- +4- f $ t i i { Cu possible side views: I t t pyridine i _*J_*_U_U- • • I side view: 11. _«—•—&—»—I - „ 1 „ 1 t | Legend: * analyte molecule Ni/Co Co Cu • metal ion DMF DMSO/pyridine DMF ~*~ [Au(CN)2]unit

Scheme 3.3: Different structural models observed for the M[Au(CN)2]2(analyte);E poly mers, all resulting from the distortion of the basic M[Au(CN)2]2 square-grid array. n Chapter 3. M[Au(CN)2]2(& alyte)x and their vapochromic properties 148 alternate between being five- and six-coordinate. Six-coordinate centers are present in both flat Cu[Au(CN)2]2(pyridine)2 and corrugated Cu[Au(CN)2]2(DMSO)2 (blue). A range of five-coordinate geometries can also be accessed: square pyramidal Cu(ll) centers are found in Cu[Au(CN)2]2(DMF) while distorted trigonal bipyramidal Cu(ii) centers are present in Cu[Au(CN)2]2(DMSO)2 (green).

The geometry observed in the green Cu[Au(CN)2]2(DMSO)2 polymorph is be lieved to result from the partial fragmentation of the square-grid array via the break ing of one Cu-N(cyanide) bond. Such fragmentation is probably also present in the

Cu[Au(CN)2]2(NH3)4 complex; each Cu(n) center in Cu[Au(CN)2]2(NH3)4 is likely still octahedral, with two Cu-N(cyanide) bonds (out of four in the fundamental square- grid structure) breaking completely to make way for two additional NH3 ligands, thereby disrupting the 2-D array. For a given analyte, similar structures are usually observed despite the change in metal center. DMSO is the only type of guest that forces corrugation of the layers, whereas DMF and pyridine favour flat square-grid arrays to be obtained.

3.6.4 Responses of the M[Au(CN)2]2(analyte)a; polymers upon exposure to analyte vapours

The vapochromic behaviour of the M[Au(CN)2]2(analyte)x coordination polymers to ward different vapours was investigated. It was previously recognized that a system does not need to be porous in order to undergo guest uptake. (18°) For example, a flexible metal-ligand superstructure can dynamically adapt in order to accommodate a variety of potential guests. (180_186) Analyte exchange seemed to be more favored in the Cu(ll) and Co(ll)-containing systems than in the Ni(ll) system. The Ni(/x-OH2)2[Au(CN)2]2 polymer was found to be rather stable when exposed to different vaporirs and analyte exchange was found to be only partial. A mixture of products was obtained as a result. Ni(ll) is known to be a relatively inert first row transition M(ll) center. (187) Analyte exchange, whether it occurs through an associative or dissociative mechanism, is not favoured due to the high ligand field activation energy of the Ni(ll) centers. This could explain the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 149

very slow and only partial analyte exchange observed for the Ni[Au(CN)2]2(analyte)x system. The Jahn-Teller influenced flexible coordination sphere and the greater lability of Cu(li) compared with other transition metals are likely important features of the

Cu[Au(CN)2]2(analyte)x system. The lability of Cu(ll) is believed to facilitate the re versible exchange of bound analyte molecules without any thermal treatment required. It also likely increases the flexibility of the framework by allowing the breaking and the reformation of Cu-N(cyanide) bonds, thereby adapting to the analyte present.

Aurophilic interactions appear to be present in most of the Cu[Au(CN)2]2(analyte)3: polymers and probably help to stabilize the 3-D networks as analyte exchange takes place. The different modes of structural flexibility mentioned above (section 3.6.3) work in concert to generate the adaptable, dynamic network solid that is ultimately able to bind and sense different donor analytes. The source of the vapochromism in the MbA^CN^^analyte^ coordination polymers presented here differs from that of other Au(l)-containing systems previ ously reported. (145_148) In these systems, the presence of an analyte mainly affects the optical absorption and emission due to the Au-Au bonds, which allows detec tion and identification of the analyte. This is unlikely the principal effect in the

M[Au(CN)2]2(analyte)a; polymers as no luminescence was observed for any of the

M[Au(CN)2]2(analyte)a: polymers. It is possible that the luminescence from the Au- Au interactions is quenched by the transition metal centers (Cu, Ni or Co), which absorb in the UV-vis region.

The M[Au(CN)2]2(analyte);,; polymers show vapochromism in the visible since each analyte molecule that is incorporated binds to the M(n) center and modifies differ ently its crystal field splitting. As a consequence, the color of the vapochromic com pound changes as the d — d absorbance bands shift with each analyte present. The color changes were found to be larger for the Cu(n)-containing polymers. In the

Cu[Au(CN)2]2(analyte)x system, in addition to analyte identity, the resulting coordi nation number (five or six) and the specific geometry adopted by the Cu(ll) center also influence the color of the polymer by altering the splitting of the o?-orbitals.

Identification of the different Co[Au(CN)2]2(analyte)2 polymers purely on the Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 150 basis of colour (see for example Figure 3.14) cannot be accomplished as easily as for the Cu(ll)-containing polymers. The Co(ll) coordination sphere in every Co[Au(CN)2]2(analyte)2 polymer is very similar: an octahedral geometry with four cyanide groups occupying the same plane. The only difference lies in the two donor analytes that are trans to each other. As a result, the crystal field splitting of the

Co(ll) centers in each Co[Au(CN)2]2(analyte)2 is very similar and, hence, so is their colour.

The [Au(CN)2]~ unit is also a key component of these systems. Each

M[Au(CN)2]2(analyte)x polymer has a different FT-IR signature in the cyanide vi bration region since every analyte, in addition to modifying the geometry to some extent, modifies in a different manner the electron density distribution around the M(ll) center. This influences the amount of 7r-back-bonding from the M(ll) center to the cyanide group, which in turn is observed in the FT-IR spectrum due to the change in vibration frequency. (21) Also, the number of cyanide vibration bands observed is a result of the symmetry and coordination number of the M(ll) centers, which help in identifying the analyte present. The use of FT-IR spectroscopy to study the sorption of analytes by vapochromic materials has been previously reported. (151>188>189) For example, in systems containing 2 [Pt(CN)4] ~ units, slight shifts in the vibration frequency of the cyanide groups are observed if hydrogen-bonding between the N-cyanide atoms and the analyte molecules present in the lattice occurs. Analytes cannot be easily differentiated or identified via FT-IR in these cases since vibration frequency shifts of only 0-10 cm-1 are observed. (151'188'189)

In the Cu[Au(CN)2]2(analyte)x system, changes in the number of bands (from one to four) are observed in addition to larger shifts in vibration frequency, vary ing between 10 and 40 cm-1 as analytes are exchanged. The FT-IR signatures are hence unusually diagnostic for a particular analyte present in the Cu(ll)-based sys tem. In the Co[Au(CN)2]2(analyte)2 system, the differences in cyanide vibration frequency between certain analyte adducts also vary by up to 40 cm-1. However,

several Co[Au(CN)2]2(analyte)2 polymers (DMF, DMSO, pyridine) only show one band attributable to cyanide vibration and their frequencies of vibration fall within Chapter 3. M[Au(CN)2]2(a.nalyte)x and their vapochromic properties 151 a 15 cm-1 window. This is consistent with the similar Co(ll) geometry observed in each Co[Au(CN)2]2(analyte)2 polymer and could potentially make analyte iden tification more difficult. Just as it hindered detection of the analyte by color, the minimum changes in coordination shell also seem to hinder detection by FT-IR in the Co(ll)-based system. The FT-IR signatures in the cyanide vibration region for the Ni[Au(CN)2J2(analyte);,; polymers, on the other hand, are different from each other, allowing for analyte identification if analyte exchange actually occurs. In summary, the use of the intense, sensitive FT-IR signature of the coordination polymer framework to report on the nature of the analyte present, in addition to the color differences, sets the Cu[Au(CN)2]2(analyte)a; system apart from other vapochro mic materials. The Co(ll)-based system reacts when exposed to different analytes, but detection and identification of the analyte by FT-IR and UV-vis spectroscopies are more difficult than in the Cu(ll) analogous system. Finally, despite the possibility of discerning the identity of the analyte present, the Ni(ll)-based system is very inert and does not undergo analyte exchange easily, at least with the analyte tested in this study. Also, as mentioned above, only one Cu(ll)-based product seems to be favoured for each analyte, which is advantageous from an application point of view, e.g. in a vapochromic dosimeter or sensor. Detection and identification of the analyte present could become a challenge if several different products were obtained under the same conditions when starting from the same material.

3.7 Conclusions

In this chapter, the preparation and characterization of a series of

M[Au(CN)2]2(analyte),,; coordination polymers was reported. Upon modifica tion of the analyte (DMSO, DMF, pyridine, acetonitrille, dioxane, ammonia) and metal ions (Cu(ll), Ni(ll) and Co(ll)), the basic M[Au(CN)2J2 square-grid structural motif was found to be maintained, except in some cases where breaking of the 2-D array occurred. Different modes of flexibility of the square-grid motif were identified and were found to mainly depend on the metal ion's identity. Chapter 3. M[Au(CN)2]2(saialyte)x and their vapochromic properties 152

It has been illustrated that the family of Cu(n)-based coordination polymers ex hibit a vapochromic behaviour in the presence of different analytes. In addition to differences in colour, each Cu[Au(CN)2]2 (analyte);,; polymer has a very specific FT-IR signature in the cyanide vibration region. The vapochromic responses of Co(ll) and Ni(ll)-containing systems toward a few analytes was also investigated. The Ni[Au(CN)2]2(analyte)2 polymers were found to be much more inert toward analyte exchange compare to the analogous Cu(ll) polymers. The Co[Au(CN)2]2(analyte)2 polymers, on the other hand, were found to undergo analyte exchange, but every polymer was found to have a similar colour and could not be distinguished on the basis of colour only. The FT-IR signatures of the Co[Au(CN)2]2(analyte)2 polymers in the cyanide vibration region were found to be different in some cases, but similar in other cases, making analyte identification slightly more complicated than in the Cu[Au(CN)2]2(analyte)a; system.

The use of [Au(CN)2]~ as a building block is important to the function of these vapochromic coordination polymers. It provides the very sensitive cyanide reporter group that can allow, in most cases, identification of the analyte incorporated in the polymer by FT-IR spectroscopy. Also, Au-Au interactions via the [Au(CN)2]~ units increase the structural connectivity of the system in most cases and probably help provide stabilization points for the flexible M[Au(CN)2]2 framework.

3.8 Future work

3.8.1 Modification of the analytes

The vapochromic behaviour of the M[Au(CN)2]2 (analyte^ coordination polymers to ward other types of analytes could be explored. For example, when exposed to an alyte molecules containing a ketone, alcohol, carboxylic acid, phosphorus- or sulfur- donor group the M[Au(CN)2]2(analyte)x system could show different sensitivity and responses both in the UV-vis and FT-IR spectra.

Quantitative measurements on the sensitivity of the Cu[Au(CN)2]2(analyte)a; sys 19 tem were performed by Dr. T. Ramnial. ( °) For primary amines (NRH2, R = H, Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 153 methyl, propyl, butyl), the sensitivity was determined to range from a few ppm to a few hundred ppm depending on the R group and the response time was found to be on the order of seconds. The sensitivity for these primary amines was found to be much higher than for DMF or DMSO, for which the minimum detectable concen tration is above 1000 ppm and the response time is on the order of several minutes. Given the high sensitivity of the Cu-based system, the responses of the Ni and Co- containing M[Au(CN)2]2(analyte)x polymers to primary amines should also be tested.

It is important to note that the Co[Au(CN)2]2(analyte)2 polymers might be oxidized upon reaction with amines and, as a consequence, this vapochromic sensor may be irreversible. In the study presented in this thesis, the vapochromic behaviour of the Ni(/x-

OH2)2[Au(CN)2]2 polymer was tested, but no attempt was made to test the sensi tivity of the other Ni[Au(CN)2]2(analyte)2 polymers. There is the possibility that the water-containing polymer is the most stable polymer, which would explain the negative results observed. In this regard, the vapochromic behaviour of the other

Ni[Au(CN)2]2(analyte)2 polymers should also be studied.

3.8.2 M[Au(CN)2]2 as sensors

Several other vapochromic systems are initially analyte-free and simply take up the analyte molecules without any exchange required. In these thermal treatment is generally required to regenerate the sensor.

In this light, the behaviour of the analyte-free M[Au(CN)2]2 complexes toward different analyte vapour should be investigated. In particular, the systems that are inert or very slow to undergo analyte exchange when hydrated should be investi gated. The metal ion coordination sphere should be very different in the analyte- free M[Au(CN)2]2 complexes compared to the M[Au(CN)2]2(analyte);c polymers, and larger colour changes may be observed upon coordination of the analyte molecules to the metal ions. Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 154

Figure 3.15: Solid-state structure of Co[Au(CN)4]2(DMSO)4, showing the molecular complexes interacting through N-Au interactions (shown as dashed bonds). DMSO- methyl groups were removed for clarity. Colour scheme: S = yellow, O = red, N = blue, C = gray, Au = gold and Co = pink.

3.8.3 Modification of the building blocks

Use of [Au(CN)4]-

The use of [Au(CN)4]~ as a building block to make vapochromic materials could be attempted. Similarly to [Au(CN)2]~, this building block binds to transition metal ions and can be incorporated into coordination polymers, bridging up to four metal 191 centers.( ) [Au(CN)4]~ is a much weaker donor than [Au(CN)2]~, and this could facilitate the structural rearrangement of the framework upon analyte exchange.

Preliminary studies on the formation of [Au(CN)4]-based complexes in DMSO were performed. As shown in Figure 3.15, the reaction between the KAu(CN)4 and Co(ll) salts in DMSO yielded molecular complexes of Co[Au(CN)4]2(DMSO)4. Each Co(n) ion is surrounded by four DMSO molecules and two trans pendant [Au(CN)4]~ groups. As observed in other [Au(CN)4]-containing complexes, instead of Au-Au interactions,

N-Au interactions are present between the Co[Au(CN)4]2(DMSO)4 complexes. This compound is not polymeric by coordinate bonds, but could be considered as a layered structure through the N-Au interactions. Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 155

Studying the vapochromic properties of this complex and comparing it to the behaviour of the Co[Au(CN)2]2(analyte)2 polymers would be interesting. In addition to studying the effects of the ligand's identity ([Au(CN)2]~~ vs [Au(CN)4]~), the impact of the framework connectivity (1-D vs 2-D) on the vapochromic behaviour will be uncovered.

Use of lanthanide ions

Lanthanide ions are well known for their luminescent properties and a few examples of luminescent lanthanide-containing [Au(CN)2]-based coordination polymers have been reported. The replacement of transition metal ions by lanthanide ions to gen erate Ln[Au(CN)2]a;(analyte)2/ systems could allow the formation of vapoluminescent rather than vapochromic systems. In the solid-state, the photoluminescence spec trum of a compound is usually much sharper than its UV-vis absorption spectrum. This difference could make it easier to detect small changes occurring upon analyte exchange in a vapoluminescent material than in a vapochromic material. As mentioned in Chapter 1, the preparation, solid-state structure and optical properties of the Ln[Au(CN)2]3-2,3H20 (Ln = La, Gd, Sm, Eu, Tb, Dy) polymers have been previously reported. (67~69) In particular, the emission properties of the parent Eu- and Tb-containing polymers have been extensively studied. However, no study on their potential vapoluminescent properties, when exposed to the vapour of different analytes, have been reported and it would be worthy of investigation. Lanthanide ions are usually more oxophilic than the late transition metal ions and analytes containing such a functional group (i.e. DMF, DMSO, etc.) could be tested. These systems could show a preference toward different types of analytes compared to the transition metal-containing systems.

Use of vanadyl units

The use of vanadyl units, (V=0)2+, to replace the transition metal ions and form

VO[Au(CN)2]2(analyte)x, could also be investigated. The vanadium ion in this unit is generally labile and can accommodate three, four or five additional ligands. The Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 156 flexibility and lability of the Cu(n) centers were found to be important to the vapo chromic behaviour of the Cu[Au(CN)2]2(analyte)a; system and this could also be true in the VO[Au(CN)2]2(analyte);E system.

Preliminary attempts to use the vanadyl units to synthesize [Au(CN)2]-based coor dination polymer yielded positive results. Upon mixing of two aqueous solutions, one 2+ containing (V=0) and the second containing [Au(CN)2]~, a reaction occurred and a product with shifted cyanide vibration peaks was obtained. Further analysis on this product should be performed to determine its chemical composition and structure, followed by a study of its potential vapochromic properties, if appropriate.

3.9 Experimental Section

The general procedures concerning the characterization of the M[Au(CN)2]2(analyte)3; complexes are as described in Chapter 2, section 2.7.1, unless otherwise noted.

3.9.1 Synthesis of Cu[Au(CN)2]2(DMSO)2 (green)

A 0.5 mL DMSO solution of Cu(C104)2-6H20 (0.037 g, 0.10 mmol) was added to a 0.5 mL DMSO solution of K[Au(CN)2] (0.057 g, 0.20 mmol). Green crystals of

Cu[Au(CN)2]2(DMSO)2 were obtained by slow evaporation over several days, filtered and air-dried. Yield: 0.050 g, 70 %. Anal. Calcd. for C8Hi2N4Au2Cu02S2: C, 13.39; H, 1.69; N, 7.81. Found: C, 13.43; H, 1.72; N, 7.61. IR (KBr): 3005(w), 2915(w), 2184(B), 2151(m), 1630(w), 1426(w), 1408(w), 1321(w), 1031(m), 993(B), 967(m), 720(w), 473(m) cm"1. The same product can be obtained by absorption of

DMSO by Cu(//-OH2)2[Au(CN)2]2.

3.9.2 Synthesis of Cu[Au(CN)2]2(DMSO)2 (blue)

A 0.2 mL DMSO solution of Cu(C104)2-6H20 (0.037 g, 0.10 mmol) was added to a 0.4 mL DMSO solution of K[Au(CN)2] (0.057 g, 0.20 mmol). Blue needles of

Cu[Au(CN)2]2(DMSO)2 formed after one hour and were filtered and dried under N2.

Yield: 0.057 g, 80 %. Anal. Calcd. for C8H12N4Au2Cu02S2: C, 13.39; H, 1.69; N, Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 157

7.81. Found: C, 13.50; H, 1.76; N, 7.62. IR (KBr): 3010(w), 2918(w), 2206(m), 2194(a), 2176(m), 2162(m), 1631(w), 1407(w), 1316(w), 1299(w), 1022(m), 991 (s), 953(m), 716(w), 458(m) cm"1.

3.9.3 Synthesis of Cu[Au(CN)2]2(DMF)

A 2 mL DMF solution of Cu(C104)2-6H20 (0.037 g, 0.10 mmol) was prepared. This solution was added to a 3 mL DMF solution of K[Au(CN)2] (0.057 g, 0.20 mmol). A dark blue-green mixture of powder and crystals was obtained after several days of slow evaporation and was filtered and air-dried. The composition was found to be consistent with Cu[Au(CN)2]2(DMF). Yield: 0.033 g, 52 %. Anal. Calcd for C7H7N5AU2CUO: C 13.25, H 1.11, N 11.04. Found: C 13.26, H 1.11, N 11.30. IR (KBr): 2927(w), 2871 (w), 2199(s), 2171 (shoulder), 1665(s), 1660(a), 1492(w), 1434(w), 1414(w), 1384(m), 1251(w), 1105(w), 674(w), 516(w), 408(w) cm"1. Single crystals of Cu[Au(CN)2]2(DMF) were obtained by dissolving Cu(/i-OH2)2[Au(CN)2]2 in DMF and allowing the solution to evaporate very slowly. The single crystals and the crystal/powder mixture as prepared above had identical FT-IR spectra. The same product can also be obtained by vapour absorption of DMF by several

Cu[Au(CN)2]2(analyte)a; complexes.

3.9.4 Synthesis of Cu[Au(CN)2]2(pyridine)2

A 10 mL pyridine/water/methanol (5 : 47.5 : 47.5) solution of Cu(C104)2-6H20 (0.111 g, 0.300 mmol) was prepared. This solution was added to a 10 mL pyri dine/water/methanol (5 : 47.5 : 47.5) solution of K[Au(CN)2] (0.171 g, 0.594 mmol). A blue powder was obtained immediately and was filtered and air-dried. The prod uct was determined to be Cu[Au(CN)2]2(pyridme)2. Yield: 0.163 g, 76.3 %. Anal.

Calcd for Ci4Hi0N6Au2Cu: C 23.36, H 1.40, N 11.68. Found: C 23.52, H 1.44, N 11.58. IR (KBr): 3116(w), 3080(w), 2179(s), 2167(s), 2152(s), 2144(m), 1607(m), 1449(m), 1445(B), 1214(m), 1160(w), 1071 (m), 1044(w), 1019(m), 758(B), 690(B), -1 642(m) cm . Single crystals of Cu[Au(CN)2]2(pyridine)2 were obtained by slow eva poration of the remaining solution. The crystals and powder had identical FT-IR Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 158 spectra. The same product can also be obtained by vapour absorption of pyridine by several Cu[Au(CN)2]2(analyte)a; complexes.

3.9.5 Synthesis of Cu[Au(CN)2]2(CH3CN)2

A 1 mL CH3CN solution of Cu(C104)2-6H20 (0.037 g, 0.10 mmol) was prepared and added to a 2 mL CH3CN solution of K[Au(CN)2] (0.057 g, 0.20 mmol). A green powder precipitated immediately along with a white powder. The white powder was assumed to be the KC104 side product, which is insoluble in CH3CN. Upon filtration on a buchner funnel, a colour change was observed from a dark green to a very pale green. The FT-IR of the final pale product was found to be indistinguishable from that of Cu(/i-OH2)2[Au(CN)2]2- To prevent exposure of the precipitate to atmospheric water, almost all of the solvent was removed by decantation. The small remaining amount of solvent was removed using a rotary evaporator. The KCIO4 side product was not removed through washing and filtering. The composition of the dark green powder isolated in this manner was found to be consistent with Cu[Au(CN)2]2(CH3CN)2 mixed with two equivalents of KCIO4. Anal. Calcd for Cu[Au(CN)2]2(CH3CN)2 + 2(KC104)

(C8H6N6Au2Cl2CuK208): C 10.44, H 0.65, N 9.12. Found: C 10.99, H 0.57, N 8.69. IR (KBr): 2297(w), 2269(w), 2192(B), 1600(W), 1445(w), 1369(w), 1088(B), 941 (w), 925(w), 752(w), 695(w), 626(m), 512(w), 468(w), 419(w) cm"1. The same product (without KCIO4) can be obtained by vapor absorption of acetonitrile by

Cu[Au(CN)2]2(DMSO)2 (both polymorphs).

3.9.6 Synthesis of Cu[Au(CN)2]2(dioxane)(H20)

A 2 mL dioxane/water (2:1) solution of Cu(C104)2-6H20 (0.037 g, 0.10 mmol) was prepared. This solution was added to a 4 rnL dioxane/water (2:1) solution of

K[Au(CN)2] (0.057 g, 0.20 mmol). A pale blue-green powder was obtained imme diately and was filtered and air-dried. The composition of the product was found to be Cu[Au(CN)2]2(dioxane)(H20). Yield: 0.057 g, 85 %. The same product can Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 159

be obtained by vapor absorption of dioxane by several Cu[Au(CN)2]2(analyte);c com plexes (the water molecule included in this case is from ambient moisture). Anal.

Calcd for C8Hi0N4Au2CuO3: C 14.39, H 1.51, N 8.39. Found: C 14.31, H 1.21, N 8.43. IR (KBr): 2976(m), 2917(m), 2890(w), 2862(m), 2752(w), 2695(w), 2201(s), 2172(w), 1451(m), 1367(m), 1293(w), 1255(s), 1115(s), 1081(s), 1043(m), 949(w), 1 892(m), 871(s), 705(W), 610(m), 515(m), 428(m) cm" .

3.9.7 Synthesis of Cu[Au(CN)2]2(NH3)4

This product was obtained by vapor absorption of NH3 by several

Cu[Au(CN)2]2(analyte)x complexes. FT-IR spectroscopy and elemental analy sis indicated conversion of the initial complexes to Cu[Au(CN)2]2(NH3)4. Anal.

Calcd. for C4Hi2N8Au2Cu: C 7.63, H 1.92, N 17.80. Found: C 7.56, H 1.98, N 17.71. IR (KBr): 3359(s), 3328(s), 3271(B), 3212(m), 3182(m), 2175(m), 2148(B), 1639(m), 1 1606(m), 1243(B), 685(B), 435(W) cm" .

3.9.8 Synthesis of Ni[Au(CN)2]2(DMSO)2

A 0.75 mL DMSO solution of Ni(N03)2-6H20 (0.057 g, 0.20 mmol) was added to a 1.75 mL DMSO solution of K[Au(CN)2] (0.123 g, 0.427 mmol). Blue turquoise powder started to precipitate after approximately one hour. It was collected by fil tration 24 hours later and air dried. The composition of the final product was de termined to be Ni[Au(CN)2]2(DMSO)2. Yield: 0.101 g, 70.8 %. Anal. Calcd. for

C8H12N4Au2Ni02S2: C, 13.48; H, 1.70; N, 7.86. Found: C, 13.70; H, 1.81; N, 8.10. IR (KBr): 3434(m), 3010(w), 2919(w), 2189(B), 2180(B), 1629(w), 1409(w), 1314(w), 1 1299(w), 1032(m), 1008(B), 1001(B), 956(m), 713(w), 480(w), 430(w) cm" .

3.9.9 Synthesis of Ni[Au(CN)2]2(DMF)2

Ni(/>OH2)2[Au(CN)2]2 (0.115 g, 0.194 mmol) was mixed with 1 mL of DMF and left to sit for 24 hours covered. The blue powder was then isolated by filtration and the composition of the final product was found to be Ni[Au(CN)2]2(DMF)2. Yield: 0.092 Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 160

g, 69 %. Anal. Calcd for Ci0H14N6Au2NiO2: C 17.09, H 2.01, N 11.96. Found: C 16.95, H 2.08, N 12.24. IR (KBr): 3429(m), 2994(w), 2964(w), 2933(m), 2892(w), 2809(w), 2189(B), 1658(B), 1498(W), 1434(m), 1418(w), 1384(m), 1252(m), 1109(a), 1061(m), 689(s), 488(w) cm-1.

3.9.10 Synthesis of Ni[Au(CN)2]2(pyridine)4

A 1.5 mL water/pyridine (96:4) solution of K[Au(CN)2] (0.059 g, 0.21 mmol) was prepared. This solution was added to a 1 mL water/pyridine (96:4) solu tion of Ni(N03)2-6H20 (0.030 g, 0.10 mmol). A purple powder was obtained immediately and was filtered and air-dried. The product was determined to be

Ni[Au(CN)2]2(pyridine)4. Yield: 0.070 g, 78 %. Anal. Calcd for C24H2oN8Au2Ni: C 33.02, H 2.31, N 12.83. Found: C 32.64, H 2.25, N 12.55. IR (KBr): 3095(w), 3070(w), 3041(w), 3023(w), 2993(w), 2924(w), 2855(w), 2171(s), 2143(a), 1602(a), 1574(m), 1488(m), 1449(a), 1443(a), 1237(w), 1216(m), 1151(w), 1068(m), 1042(m), 1011(w), 769(m), 756(m), 699(a), 631 (m), 469(w), 435(m) cm"1.

3.9.11 Synthesis of Co[Au(CN)2]2(DMSO)2

Co(^-OH2)2[Au(CN)2]2 (0.125 g, 0.211 mmol) was mixed with 1 mL of DMSO and let to sit for 72 hours covered. The pink fine powder was then isolated by decantation and left to dry in ambient air. The composition was found to be con- aiatent with Co[Au(CN)2]2(DMSO)2. Yield: 0.121 g, 80.4 %. Anal. Calcd for

C8H12N4Au2Co02S2: C, 13.47; H, 1.70; N, 7.86. Found: C, 13.51; H, 1.73; N, 7.69. IR (KBr): 3021(w), 3009(w), 2920(w), 2177(a), 1421(w), 1408(w), 1316(w), 1298(w), 1031(m), 997(a), 957(m), 938(m), 905(w), 717(w), 467(w), 441 (w) cm"1.

3.9.12 Synthesis of Co[Au(CN)2]2(pyridine)2

Co(//-OH2)2[Au(CN)2]2 (0.052 g, 0.088 mmol) waa mixed with 2.5 mL of wa ter/pyridine (95:5) and let to sit for 24 hours covered. The powder became more orange. It was isolated by filtration and dry in ambient air. The composition was Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 161

found to be consistent with Co[Au(CN)2]2(pyridine)2 by elemental analysis. Yield:

0.049 g, 78 %. Anal. Calcd for Ci4H10N6Au2Co: C 23.51, H 1.41, N 11.75. Found: C 23.67, H 1.55, N 11.68. IR (KBr): 3086(w), 3068(w), 3053(w), 3024(w), 2168(s), 1605(B), 1574(s), 1486(m), 1447(B), 1242(W), 1216(m), 1157(w), 1073(m), 1041 (m), 1016(w), 946(w), 882(w), 759(m), 698(s), 633(m), 460(w), 431 (m) cm"1.

3.9.13 Synthesis of Co[Au(CN)4]2(DMSO)4

Co(N03)2-6H20 (0.029 g, 0.10 mmol) was dissolved in 1 mL of DMSO and added to a 1 mL DMSO solution of K[Au(CN)4] (0.068 g, 0.20 mmol). The solution was left to evaporate, partially uncovered, for several weeks. Pink crystals started to grow when less than 0.5 mL of solution remained. A small amount of crystal was collected by filtration and air dried. The chemical composition and solid-state structure were determined by crystallography. IR (KBr): 3434 (m), 3014(w), 2993 (w), 2919(w), 2217(m), 2192(m), 1426 (m), 1403 (m), 1321 (m), 1024(m), 998(B), 950(B), 902 (w), 715 (w) cm-1.

3.9.14 Details on structure determination of

Cu[Au(CN)2]2(analyte)x

The single crystals chosen for X-ray crystallography were mounted as described earlier except for the crystal of Cu[Au(CN)2]2(DMSO)2 (blue) that was sealed in a glass capillary (0.5 mm in diameter) due to its high sensitivity to moisture. Data were collected and analyzed as explained in Chapter 2 (section 2.7.1), except for the following details. Data for Cu[Au(CN)2]2(DMSO)2 (blue),

Cu[Au(CN)2]2(DMF), Cu[Au(CN)2]2(pyridine)2 and Co[Au(CN)4]2(DMSO)4 were recorded on a Rigaku RAXIS RAPID imaging plate area detector and a numeri 192 cal absorption correction was applied.( ^ For Cu[Au(CN)2]2(DMSO)2 (blue), only the Au, Cu and S atoms were refined anisotropically, whereas the remainder were refined isotropically.

Crystallographic data for compounds Cu[Au(CN)2]2(DMSO)2 (green), Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 162

Cu[Au(CN)2]2(DMSO)2 (blue), Cu[Au(CN)2]2(DMF) and Cu[Au(CN)2]2(pyridine)2 are collected in Table 3.13 and 3.14 while selected bond lengths and angles are reported respectively in Tables 3.1, 3.2, 3.4 and 3.5.

3.9.15 Details on structure determination of

M[Au(CN)2]2(analyte)x (M = Ni and Co)

Ni[Au(CN)2]2(DMF)2 and Co[Au(CN)2]2(pyridine)2

The structures were solved and refined using the DASH software. After indexing the experimental diffractogram, an automatic background fitting and subtraction was performed, followed by peak fitting to extract the intensity of each peak. To solve the crystal structure from a powder diffractogram using DASH, a predefined building block must be used. DASH optimizes the position and orientation of this building block using a simulated annealing algorithm to maximize the agreement with the experimental diffractogram (relative intensity and peak position). Once a structural model is obtained, the coordinates of each atom in the building block can be optimized. (66) The asymmetric units of Co[Au(CN)2]2(DMF)2 and Cu[Au(CN)2]2(pyridine)2 were chosen as the building blocks used to construct the 3-D crystal structure of

Ni[Au(CN)2]2(DMF)2 and Co[Au(CN)2]2(pyridine)2 respectively. The atomic coor dinates for Ni[Au(CN)2]2(DMF)2 and Co[Au(CN)2]2(pyridine)2 are reported in Ta ble A. 12 and A. 13 respectively (Appendix A).

Ni[Au(CN)2]2(DMSO)2

Due to the poor quality of the diffractogram, the structure of Ni[Au(CN)2]2(DMSO)2 could not be determined using the DASH software. The diffractogram could however be indexed using WinPLOTR to a unit cell very similar to that of the analogous Zn- containing polymer. The atomic positions in the Ni[Au(CN)2]2(DMSO)2 structure were then optimized using the coordinates of the Zn[Au(CN)2]2(DMSO)2 structure as a starting point. Chapter 3. M[Au(CN)2]2(^nalyte)x and their vapochromic properties 163

Table 3.13: Crystallographic data and structural refinement details for the two Cu[Au(CN)2]2(DMSO)2 polymorphs.

Cu[Au(CN)2]2(DMSO)2 Cu[Au(CN)2]2(DMSO)2 (green) (blue) Empirical formula C8H12N4Au2Cu02S2 C8H12N4Au2Cu02S2 Fw 717.82 717.82 Colour Green Blue Shape rectangular plate needle Dimension, mm3 0.09 x 0.12 x 0.30 0.11 x 0.11x0.20 Crystal system monoclinic triclinic Space group (#) C2/c (15) PI (2) a, A 11.5449(15) 7.874(7) b, A 14.191(4) 12.761(11) c, A 11.5895(12) 16.207(13) a, deg 90 89.61(7) (3, deg 112.536(9) 82.29(7) 7, deg 90 88.57(7) Volume, A3 1753.8(6) 1613.2(24) Z 4 2 A, A 0.70930 1.54180 Data range, deg 4-55 6.9-136.1 Transmission range 0.0301-0.1726 0.019-0.161 Pealed, g cm"3 2.719 2.955 //, mm-1 18.079 37.500 Reflections, parameters 1231, 93 2026, 205 fli a (I > xa(I))b 0.042 0.062 a b wR2 (I > xa(I)) 0.047 0.082 goodness of fit 2.20 1.38 a — 2 l 2 2 Function minimized X^G-^ol l-^cl) where w = cr {F0) + 0.0001Fo , 2 211/2 R = E \Wo\ - \Fc\\/E\Fo\, Rw = \£w{\F0\ - \Fc\) /EMFo\ 6For the green polymorph, x = 2.5; for the blue polymorph, x Chapter 3. M[Au(CN)2]2(analyte)x and their vapochromic properties 164

Table 3.14: Crystallographic data and structural refinement details for Cu[Au(CN)2]2(DMF) and Cu[Au(CN)2]2(pyridine)2.

Cu[Au(CN)2]2(DMF) Cu[Au(CN)2]2(pyridine)2 Empirical formula C7H7N5Au2CuO Ci4Hi0N6Au2Cu Fw 634.65 719.76 Colour Green-blue Dark blue Shape needle platelet Dimension, mm3 0.09 x 0.09 x 0.15 0.02 x 0.06 x 0.15 Crystal system monoclinic monoclinic Space group (#) C2/c (15) P2l/c (14) a, A 12.8412(10) 7.3438(7) b, A 14.5056(8) 14.1201(10) c, A 13.9932(9) 8.2694(6) a, deg 90 90 P, deg 96.064(3) 94.082(3) 7, deg 90 90 Volume, A3 2591.9(3) 855.34(12) Z 8 4 A, A 1.54180 1.54180 Data range, deg 9.2-144.0 12.0-142.6 Transmission range 0.0070-0.0199 0.3484-0.5826 pealed, g cm-3 3.253 2.794 /j,, mm-1 43.542 33.103 Reflections, parameters 1538, 148 1021, 107 i?i a {I > 3a(7)) 0.032 0.028 a wR2 (I > 3(7(7)) 0.046 0.040 goodness of fit 0.93 1.00 2 1 2 2 "Function minimized ]Pu;(|F0 - |FC|) where w' = cr (F0) + 0.0001Fo , ? 2 2 1 2 R = E\\FO\-\FC\\/Y:\FO\,R - = E^(|i o|-|Fc|) /E'»\Fo\ ] ' - Chapter 3. M[Au(CN)2]2(&nalyte)x and their vapochromic properties 165

Table 3.15: Crystallographic data and structural refinement details for Co[Au(CN)4]2(DMSO)4.

Co[Au(CN)4]2(DMSO)4 Empirical formula C16H12N8Au2Co04S4 Fw 949.36 Colour Pink Shape Dimension, mm3 Crystal system triclinic Space group (#) PI (15) a, A 8.478(3) b, A 8.585(3) c, A 21.615(7) a, deg 82.61(3) P, de§ 81.55(3) 7, deg 88.68(3) Volume, A3 1543.3(9) Z 2 A, A 1.54180 Data range, deg 4.169-68.213 Transmission range - 3 pcalcd, g Cm" 2.03 /i, mm-1 24.508 Reflections, parameters 3340, 240 a Rx (J > 3a(/)) 0.0498 a wR2 (I > 3(T(J)) 0.0531 goodness of fit 1.5191

2 l 2 2 "Function minimized ^iw(|-Fo| - |-FC|) where w = CT (FO)+0.0001FO , r , 1/2 fl = £ll*'o|-|i cll/£l* ol,.Rt ^[J2w(\Fo\-\Fc 7£«W] - Chapter 4

[cation]{M[Au(CN)2]3} polymers: Templating effects of the cationa

4.1 Incorporation of non-coordinating building blocks

The addition of non-coordinating building blocks in the synthesis of a coordination polymer has the potential to influence the 3-D arrangement obtained as well as the re sulting physical properties and has been the subject of considerable interest. (164>193>194) The shape of such building blocks plays an important role in affecting the structure of the framework. Building blocks that can engage in weak interactions with other units

aReprinted in part from: Polyhedron, vol. 26, J. Lefebvre, D. Chartrand, D. B. Leznoff, "Synthesis, structure and magnetic properties of 2-D and 3-D [cation]{M[Au(CN)2]3} (M = Ni, Co) coordination polymers", pp. 2189-2199. Copy right (2007), with permission from Elsevier.

166 Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 167

A. \ B. N 1 ^-N=C-Au-C=N ^C -^ - " c^N v* ^ ^N=C~Au~C=N

Scheme 4.1: Shape and coordination modes of: A. dca ([N(CN)2] ) units and B. _ [Au(CN)2] units. can modify the overall arrangement through hydrogen-bonding or TT — n interactions. For example, counterions that contain phenyl-groups often self-assemble using "phenyl embraces" of ir — n stacking interactions that influence the accompanying framework geometry. (195) In anionic metal oxalate networks, [MM'^C^)]^, the choice of cation influences the formation of 2-D (6-3) sheets or 3-D chiral (10,3) networks. (196>197) Additional physical properties can also arise depending on the choice of cation. For example, cations with nonlinear optical properties, such as stilbazolium derivatives, were included in [MM^C^C^)]31- coordination polymers and, in some cases, the result ing material exhibited second harmonic generation in addition to ferromagnetism. (198^ Multifunctional materials exhibiting superconductivity and ferromagnetism were also prepared using tetrathiafulvalene (TTF) derivatives as counterions. (199>200)

Dicyanamide-containing (dca) polymers of the form [cation]{M(dca)3} have been particularly well-studied in this regard, yielding a wide range of different structural 193 201 204 motifs as a function of the size and shape of the cations. ( > - ) por example, + increasing the size and n — TT stacking ability of the cations such as [EtPh3P] to larger ones such as [Pli4P]+ induces a reduction in structural connectivity as the {Ni(dca)3}~ 3-D system cannot flex to accommodate the bulkier cation/203'204^ The dca anion is a bent bridging unit that can also bind metals via the central nitrogen donor (Scheme 4.1A); the five-atom bridge is not linear but subtends ca. 120-125 ° angles through the central nitrogen and M-M distances range from 7 to 8 A when only the ends are bound (fj,i^-binding)/205) On the other hand, in comparison to dca, very little work has been done on Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 168

, B ^

K+ K VP=N i>

n Scheme 4.2: "Shape" of the different cations: A. K+; B. [PPN]+; C. [ Bu4N]+

the preparation of [cation]{M[Au(CN)2]3} polymers. The [Au(CN)2]~ unit is also a five-atom bridging unit but is essentially linear (C-Au-C angles are generally larger that 175°), leading to M-M distances of 10 to 11 A through this anion. Also, while the central Au(l) atom cannot bind to other transition metals, it can readily form aurophilic interactions that can increase the structural connectivity in the polymer (Scheme 4.IB, see section 1.1.2 for a more detailed explanation). (47^

4.1.1 Research objectives

In this chapter, the impact of the cation's identity on the structure and magnetic properties of [cation]{M[Au(CN)2]3} will be investigated. Three cations with a +1 + n + charge have been chosen for this study: K , tetrabutyl ammonium ([ Bu4N] ) and bis(triphenylphosphoranylidene)ammonium ([PPN]+) (Scheme 4.2). These cations differ in shape, varying from the small K+ to the rigid and large [PPN]+ to the large but flexible [nBu4N]+. These three cations are non-coordinating as well as non hydrogen-bonding. As a consequence, their presence should not disrupt the connec tivity of the network by binding to the metal centers or interacting strongly with the

[Au(CN)2]~ units. Chapter 4. [cation]{M[Au(CN)2J3}: Templating effects of the cation 169

4.2 Modification of the 3-D superstructure through incorporation of K+, [PPN]+ and n [ Bu4N]+ cations

4.2.1 Synthesis and structural characterization of

K{Ni[Au(CN)2]3}

The room temperature reaction of Ni(NC-3)2-6H20 with four equivalents of

K[Au(CN)2] in water afforded an immediate green precipitate, which converted into a blue precipitate after several days of being stirred in the mother liquor. The green and blue compounds formed by this reaction were isolated and investigated by FT-IR spectroscopy, powder X-ray diffraction and elemental analysis.

The reaction of Ni(NC"3)2-6H20 with K[Au(CN)2] in water in a 1:2 ratio was previously discussed in Chapter 2 (Section 2.2). This reaction yielded a green microcrystalline product which was characterized to be the Ni(/i-OH2)2[Au(CN)2]2 co ordination polymer. The FT-IR spectrum, the powder X-ray diffractogram and the elemental analysis of the green powder obtained in this 1:4 reaction were indistin guishable from that of the Ni(/x-OH2)2[Au(CN)2]2 coordination polymer, indicating that the two products were the same. The FT-IR spectrum of the blue powder showed only one band attributable to a cyanide vibration (2170 cm-1). No hydrogen was detected by elemental analysis in the blue powder. The results of elemental analysis were consistent with the chemical composition being K{Ni[Au(CN)2]3}. The presence of only one cyanide vibration mode observed at 2170 cm-1 suggests that all the cyanide groups must be in an identical environment. The shift, from the -1 2141 cm band observed in K[Au(CN)2], toward higher energy indicates the presence of cyanide groups coordinated to the Ni(ll) centers through the N atom. Crystals of the blue powder, suitable for single crystal X-ray diffraction, could not be obtained at room temperature by slow-diffusion methods. In order to obtain ac ceptable single crystals, the reaction of Ni(NOs)2-6H20 with K[Au(CN)2] was carried Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 170

Table 4.1: Selected bond lengths (A) and angles (deg) for K{Ni[Au(CN)2]3}

Bond Lengths Ni(l)-N(l) 2.091(4) K(l)-C(l) 3.175(9) Au(l)-Au(l') 3.307(2) K(l)-N(l) 2.914(8) Angles N(l)-Ni(l)-N(l6) 93.5(4) Ni(l)-N(l)-C(l) 162.4(4) N(l)-Ni(l)-N(V) 86.84(17) C(l)-Au(l)-C(l*) 179.8(7) N(l)-Ni(l)-N(lc) 92.9(4) Symmetry transformations : (') — x + y, — x + 1, z; (*) —y + 1, — x + 1, —z; (°) -x + y, -x, z; (b) x,x-y,-z + l; (c) -x + y,y,-z + l.

out under hydrothermal conditions (125 °C) with a cooling rate of 1 °C per hour. This reaction afforded a mixture of pale blue powder and dark blue hexagonal-shaped crystals. The FT-IR spectrum of the sample prepared hydrothermally showed the same peak in the cyanide vibration region as the FT-IR spectrum of the blue powder obtained at room temperature. The results from elemental analysis were also consistent with the hydrothermal product being K{Ni[Au(CN)2]3}. In addition to comparable elemental analyses and FT-IR spectra, the powder X-ray diffractograms of the two products were found to be superimposable. Thus, the two synthetic routes afforded the same K{Ni[Au(CN)2]3} product, with the same three dimensional arrangement of building blocks. From X-ray diffraction data of a dark blue hexagonal crystal obtained from the hydrothermal sample, the structure of K{Ni[Au(CN)2]3} was determined. Slightly distorted octahedral Ni(li) centers are bridged in all three directions by [Au(CN)2]~ units, via Ni-N bonds of 2.081(13) A, to give rise to a pseudo-cubic array, similar to Prussian Blue (Figure 4.1A, Table 4.1). Two additional identical networks inter penetrate the first one, generating a triply interpenetrated structure (Figure 4.IB). The networks are connected through Au-Au bonds of 3.322(3) A. The potassium countercation occupies half of the cavities formed by the interpenetrated networks, interacting weakly with the electrons of the cyanide 7r-bonds (Figure 4.1C). Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 171

Figure 4.1: A. Solid-state structure of K{Ni[Au(CN)2]3}, showing one Prussian Blue like array (K+ cations were removed for clarity); B. Three interpenetrated networks + of {Ni[Au(CN)2]3}~; C. Environment of the K cation.

A gas adsorption experiment was performed at 77 K to determine if the remaining cavities were accessible to nitrogen (N2). However, no gas uptake was observed under pressure of up to 100 torr, suggesting that the cavities were not easily accessible. The features observed in the FT-IR spectra are consistent with the structure de termined by single-crystal X-ray diffraction as only one crystallographically distinct, metal-bridging cyanide group is present. The powder X-ray diffractogram generated from this single-crystal structure is superimposable with those obtained experimen tally from the bulk samples (prepared at both temperatures). Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 172

4.2.2 Synthesis and structural characterization of

[PPN]{M[Au(CN)2]3} polymers

The reaction of [PPN][Au(CN)2] with M(n) salts (M = Ni and Co) in ethanol im mediately yielded very fine precipitates. The relative amounts of carbon, hydrogen and nitrogen observed by elemental analysis suggested that the chemical composition of the two products was [PPN]{M[Au(CN)2]3} (M = Ni, Co). The FT-IR spectrum of [PPN]{Ni[Au(CN)2]3} showed only one band attributable to a cyanide vibration 1 at a frequency of 2192 cm" . The spectrum of [PPN]{Co[Au(CN)2]3}, on the other hand, contained two overlapping bands, with maxima at frequencies of 2187 and 2175 cm-1. No VQN band at 2140 cm-1 could be observed in the FT-IR spectrum of either [PPN]{M[Au(CN)2]3} complex. -1 The lack of a band corresponding to free [Au(CN)2]~ (2140 cm ) suggests that all [Au(CN)2]~ units are N-bound to the metal centers. The presence of only one band in the FT-IR spectrum of [PPN]{Ni[Au(CN)2]3} indicates that all the cyanide groups are in a similar environment. A slight structural difference among the cyanide groups is believed to exist in [PPN]{Co[Au(CN)2]3} (most likely around the Co(ll) centers). This difference affects the frequency of vibration of the cyanide groups, which results in two slightly different bands observed in the FT-IR spectrum. This likely occurs because of broken symmetry. The small difference between the frequency of vibration (5 cm-1) between the Ni- and Co-containing complexes can be attributed to the different metal centers present in each product. As explained in section 1.4.2, the cyanide vibration frequency is affected by the extent of cr-donation between the N- bound cyanide group and the transition metal, which depends on the electrophilicity of the metal ion.

Pale purple crystals of [PPN]{Ni[Au(CN)2]3} were obtained through slow diffusion of reagents in an H-shaped tube. The crystals were found to have a powder X-ray diffractogram and an FT-IR spectrum identical to those of the powdered sample obtained by fast precipitation, confirming that the two products were the same.

The structure determined for [PPN]{Ni[Au(CN)2]3} is shown in Figure 4.2. The Ni(n) centers have an octahedral geometry and are coordinated to six cyanide groups Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 173

Figure 4.2: Extended structure of [PPN]{Ni[Au(CN)2]3} showing the coordination sphere around the Ni(ll) centers and the Prussian-Blue-type pseudo-cubic array.

through the N atom (Table 4.2). Each [Au(CN)2]~ unit bridges two Ni(n) centers, via Ni-N bonds of 2.072(5) A, to create a 3-D Prussian Blue-type pseudo-cubic array. The [PPN]+ cation occupies the cavity in the center of each cube, preventing the interpenetration of a second network.

Despite many crystallization attempts, no crystal of [PPN]{Co[Au(CN)2]3} suit able for single crystal X-ray diffraction, could be obtained. However, the powder

X-ray diffractograms of [PPN]{Ni[Au(CN)2]3} and [PPN]{Co[Au(CN)2]3} are nearly superimposable (Figure 4.3). Also, the unit cell obtained for [PPN]{Co[Au(CN)2]3} when indexing the powder X-ray diffractogram was found to be very similar to that of the Ni-analogue. The slight differences in dimensions could be attributed to the different size and electronegativity of the metal centers. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 17A

Table 4.2: Selected bond lengths (A) and angles (deg) for [PPN]{Ni[Au(CN)2]3}

Bond Lengths Ni(l)-N(l) 2.072(5) P(l)-N(2) 1.548(3) Au(l)-C(l) 1.988(6) C(l)-N(l) 1.114(8) Angles N(l)-Ni(l)-N(l') 88.9(2) N(l)-Ni(l)-N(") 91.1(2) C(l)-Au(l)-C(l*) 177.3(4) Symmetry operations: (*) x — y, -y, -z + |; (') x - y + |, x + |, —z + §; (") -y + l,x-y,z

The similar unit cells and powder diffractograms indicates that an identical ar rangement of the building blocks is present in the two polymers. This is also consistent with the observed FT-IR spectra being similar.

4.2.3 Synthesis and structural characterization of n [ Bu4N]{M[Au(CN)2]3} polymers

n The reaction of Ni(ll) with [ Bu4N][Au(CN)2]-0.5H2O in ethanol afforded, over a period of several hours, a purple precipitate. The chemical composition was found to n be consistent with [ Bu4N]{Ni[Au(CN)2]3} by elemental analysis.

When the same reaction was performed with Co(C104)2-6H20, a mixture of white and pink products was collected after several days of evaporation. The FT-IR spec trum of the mixture showed the presence of cyanide-containing moieties as well as n + [ Bu4N] and CIO4 ~~ ions. The white product present in the mixture was proposed n to be [ Bu4N][C104], which starts to precipitate at high concentration along with the Co[Au(CN)2]x-containing complex. To eliminate the resulting separation prob lem, the reaction was carried out using Co(/Li-OH2)2[Au(CN)2]2 and one equivalent of n [ Bu4N][Au(CN)2]-0.5H2O. This reaction yielded only one product and the composi n tion was found to be consistent with [ Bu4N]{Co[Au(CN)2]3} by elemental analysis. n n The FT-IR spectra of [ Bu4N]{Ni[Au(CN)2]3} and [ Bu4N]{Co[Au(CN)2]3} show Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 175

1 M=Ni A

111111111111111 I1111 J1 V. A. . . J^ . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 10 11

Figure 4.3: Comparison between the powder X-ray diffractogram of [PPN]{Co[Au(CN)2]3} (top, M = Co) and the diffractogram predicted from the single-crystal structure of [PPN]{Ni[Au(CN)2]3} (bottom, M = Ni). only one band corresponding to a cyanide vibration at 2195 and 2189 cm-1 respec tively. These bands are shifted toward higher energy with respect to the 2146 cm-1 n band found in the spectrum of [ Bu4N][Au(CN)2]-0.5H2O. This shift indicates that all cyanide groups are N-bound to a transition metal. (21) n Dark purple-blue crystals of [ Bu4N]{Ni[Au(CN)2]s} suitable for single crystal X- ray diffraction analysis were obtained by hydrothermal recrystallization at 125 °C. Comparison of the FT-IR spectrum and powder X-ray diffractogram obtained for the crystals prepared hydrothermally with the ones obtained for the powder prepared at room temperature confirmed that the two products were identical. n The solid-state structure determined for [ Bu4N]{Ni[Au(CN)2]3} is shown in Fig ure 4.4. The Ni(n) centers are coordinated to six cyanide groups, with an average Ni-N bond length of 2.04(3) A (Figure 4.4A, Table 4.3), yielding a distorted octahe _ n dral geometry. Each [Au(CN)2] unit in [ Bu4N]{Ni[Au(CN)2]3} bridges two Ni(n) centers, generating a three dimensional network (Figure 4.4C). The structure is built Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 176

n Figure 4.4: A. Coordination sphere around a Ni(n) center in [ Bu4N]{Ni[Au(CN)2]3} (only two cations are shown and butyl groups are simplified for clarity); B. Extended network showing a two dimensional rectangular (4,4) grid and the position of one "Bu4N+ cation; C. Six rectangular grids intersecting at one Ni(il) center (labelled, green): this Ni center occupies an edge position on three grids (shown in black) and a corner position on the other three grids (shown in grey) (C and N atoms were removed for clarity and Au atoms are pale grey and shown smaller). Chapter 4. [cation]{M[Au(CN)2]s}: Templating effects of the cation 177 up by a series of two dimensional (4,4) rectangular grids, each oriented in a different direction. Each rectangular unit in each grid contains six Ni(ll) centers connected by

[Au(CN)2]~ units: one Ni(ll) center at each corner and two Ni(ll) centers along one pair of opposite edges (Figure 4.4B). Each Ni(ll) center occupies the edge position on three intersecting fused grids (shown in black in Figure 4.4C). Additional [Au(CN)2]~ units allow the formation of another set of three grids that intersect at the same Ni(ll) center (edge-position of the first set of grids) which now occupies the corner position in each of this latter set of grids (shown in grey in Figure 4.4C). In other words, six rectangular grids intersect at one Ni(n) center: each Ni(n) center occupies an edge position on three grids and a corner position on the other three grids. No interpenetration of additional networks occurs. The pores generated by this network accommodate the [nBu4N]+ cations. The n + central N atom of the [ Bu4N] group lies above a [Au(CN)2]~ unit while the n- butyl groups spread on both sides to occupy the empty space (Figure 4.4A). Thus, n the structure of [ Bu4N]{Ni[Au(CN)2]3} does not contain any empty cavities. In addition, no aurophilic interactions are present in this system, perhaps a result of the n + "insulating" [ Bu4N] groups that surround and isolate the [Au(CN)2]~ units. n No crystal of [ Bu4N]{Co[Au(CN)2]3} suitable for single crystal X-ray diffrac tion analysis could be obtained. However, the powder diffractogram of

["Bu4N]{Co[Au(CN)2J3} was found to be superimposable on that generated from n the structure of [ Bu4N]{Ni[Au(CN)2]s} (Figure 4.5). When the powder diffrac togram was indexed, a unit cell similar to that of the Ni analogue was obtained (Table 4.10). The dimensions of the cell are slightly larger, consistent with the larger ionic radius and smaller electronegativity of Co(ll) vs Ni(ll). These results suggest n n that [ Bu4N]{Co[Au(CN)2]3} is isostructural with [ Bu4N]{Ni[Au(CN)2]3}. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 178

n Table 4.3: Selected bond lengths (A) and angles (deg) for [ Bu4N]{Ni[Au(CN)2]3}

Bond Lengths Ni(l)-N(l) 2.02(3) Ni(l)-N(4) 2.07(3) Ni(l)-N(2) 2.07(3) Ni(l)-N(5) 2.04(3) Ni(l)-N(3) 2.07(3) Ni(l)-N(6) 2.00(3) Angles N(l)-Ni(l)-N(3) 85.1(10) N(2)-Ni(l)-N(3) 92.4(12) N(l)-Ni(l)-N(4) 94.0(12) N(2)-Ni(l)-N(4) 88.4(12) N(l)-Ni(l)-N(5) 90.9(11) N(2)-Ni(l)-N(5) 88.8(11) N(l)-Ni(l)-N(6) 92.8(11) N(2)-Ni(l)-N(6) 87.5(12) N(3)-Ni(l)-N(5) 93.1(10) N(4)-Ni(l)-N(5) 86.9(12) N(3)-Ni(l)-N(6) 85.9(11) N(4)-Ni(l)-N(6) 94.1(13) C(l)-Au(l)-C(6') 173.1(16) C(3)-Au(3)-C(4*) 175.0(16) C(2)-Au(2)-C(5") 173.5(14)

Symmetry operations: ( ') y, -x - \, z + \• ;(»)y + l,-x,z-\;(*)y, X "•" 2' Z ~ 4

4.2.4 Synthesis and structural characterization of n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] polymer

When water was added to the reaction of pale pink Co(/i-OH2)2[Au(CN)2]2 with n [ Bu4N][Au(CN)2]-0.5H2O in ethanol, a product of a different colour was ob tained. The composition of this dark red-pink product was consistent with n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] from the relative amounts of carbon, hy drogen and nitrogen detected by the elemental analysis. This product differs from n [ Bu4N]{Co[Au(CN)2]3} which does not contain water molecules. n The FT-IR spectrum of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] shows two dif ferent bands corresponding to a cyanide vibration, at 2182 and 2147 cm-1. This could suggest the presence of N-bound [Au(CN)2]~ units as well as free or weakly interact n ing units. As a comparison, the [ Bu4N]{Co[Au(CN)2]3} complex had only one VCN band at 2189 cm"1. n The structure of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] was determined by X- ray diffraction from crystals grown at room temperature over several days. It contains Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 179

S3 <

n Figure 4.5: Powder X-ray diffractograms generated for [ Bu4N]{Ni[Au(CN)2]3} (bottom, M = Ni) from the single crystal diffraction data and measured for n [ Bu4N]{Co[Au(CN)2]3} (top, , M = Co). octahedral Co(li) centers that are coordinated to four cyanide groups (N-bound) in the equatorial plane and two water molecules in the axial sites (Figure 4.6A, Table 4.4).

A 2-D square grid array is formed through the cobalt-bridging [Au(CN)2]~ units (Figure 4.6B) with the water molecules lying above and below the grid.

The layers containing the Co[Au(CN)2]2(H20)2 units are isolated from each other by [™Bu4N][Au(CN)2] layers (Figure 4.6C). In these layers, unbound [Au(CN)2]~ units

are present and lie above and parallel to opposing edges of the Co[Au(CN)2]2 square

array. Each [Au(CN)2]~ unit interacts with the underlying (and overlying) grid both through hydrogen-bonding between the terminal N-cyanide atoms and the hydro gen of the water molecules, and also through aurophilic interactions (Au(l)-Au(2) = n 3.4250(13) A, Figure 4.6A). The orientation of each [ Bu4N][Au(CN)2] layer alternates by 90°, always parallel to a pair of parallel edges of the square arrays (Figure 4.6D). Hence, through aurophilic interactions and hydrogen-bonding, a 3-D network is ob tained. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 180

Figure 4.6: A. Coordination sphere around a Co(ll) center in n n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] (three Bu4N+ units were omitted for clarity); B. 2-D square grid array of Co[Au(CN)2]2(H20)2 viewed down the c-axis; C. Arrangement in one [™Bu4N][Au(CN)2] layer with respect to the position of the Co(ll) ions in the adjacent layer ("Bu groups were simplified for clarity); n D. Alternation between the Co[Au(CN)2]2(H20)2 grids and the [ Bu4N][Au(CN)2] layers along the c-axis. Note the 90° rotation of the [Au(CN)2]~ units in subsequent n n [ Bu4N][Au(CN)2] planes ( Bu groups were omitted for clarity). Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 181

Table 4.4: Selected bond lengths (A) and angles (deg) for n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2]

Bond Lengths Co(l)-N(l) 2.096(12) Au(l)-Au(2) . 3.4250(13) Co(l)-0(l) 2.089(16) 0(1)-H(11)---N(2) 2.898 Angles O(l)-Co(l)-N(l) 85.1(4) C(l)-Au(l)-C(l*) 171.5(10) N(l)-Co(l)-N(l') 90.42(7) C(2)-Au(2)-C(2" ) 179.7(13) Au(l)-Au(2)-Au(l*) 163.30(6) Symmetry operations: (') y, -x + \, -z+\; (*) -y, -x, z; (") -y, -x, -z + 3; (*) x, y, -z + S

The geometry around the Au atom in the bridging [Au(CN)2]~ unit (•^C(l)-Au(l)-C(l*) = 171.5(10)°) differs from that observed in the free unit = (Zc/2)-Au(2)-C(2") 179.7(13)°). A clear bending of the [Au(CN)2]~ units within the 2-D grid can be observed (Figure 4.6B). This unusually large distortion from lin earity around the Au atom could reflect the need to accommodate the large nBu4N+ cation incorporated between the grids.

4.2.5 Solid-state UV-vis-NIR absorption spectroscopy of

[cation] {M[Au(CN)2]3}

The solid-state UV-Vis-NIR absorption spectra of the [cation]{M[Au(CN)2]3} com plexes were determined and are shown in Figure 4.7. The spectra of the

[cation]{Ni[Au(CN)2]3} complexes contain two maxima in the 400-1100 nm range (~ 560-580 and ~ 900 nm). The spectra differ below 400 nm as a broad maximum is observed for the [PPN]{Ni[Au(CN)2]3} complex while two maxima are observed for n the [ Bu4N]{Ni[Au(CN)2]3} and K{Ni[Au(CN)2]3} complexes.

The absorption spectra obtained for the [cation] {Co[Au(CN)2]3} and n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] polymers show two absorption bands in the vicinity of 470 and 1000 nm. Below 400 nm, an increase in absorption is Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 182

A. Frequency [ 10 cm" ] B. Frequency [ 10 cm" ] 30 25 20 15 12.5 10 30 25 20 15 12.5 10 1 i 1

D + Bu4N

[A.U . \ ri PPN

n + Bu4N >anc e ,Bu + o 1 I \ //\\' 4N /water

_/ppN+ 3

i i i i i i > i i i ,.i... i... i. i 7^*r-r-r-r-7 i 300 400 500 600 700 800 900 1000 300 400 500 600 700 800 900 1000 1100 Wavelength [ nm ] Wavelength [ nm ]

Figure 4.7: A. UV-vis-NIR spectra of [cation]{Ni[Au(CN)2]3} (K+ n (green), [PPN]+ (blue) and [ Bu4N]+ (purple)); B. UV-vis-NIR spec n tra of [cation]{Co[Au(CN)2]3} ([PPN]+ (blue) and [ Bu4N]+ (purple)) and {Co[Au(CN)2]2(H20)2}-["Bu4N][Au(CN)2] (green)).

observed for all [cation]{Co[Au(CN)2]3} polymers.

For a given metal center, the absorption spectra of the [cation]{M[Au(CN)2]3} complexes are in general similar. This is consistent with the observation that the geometry and coordination sphere around the metal centers are similar. The absorption bands for each Ni(ll)-containing complex were assigned to spin- 3 allowed d-d transitions from the A2a ground state (Table 4.5). The highest energy + n + band observed for the K and [ Bu4N] -containing polymers could however not be assigned; only three absorption bands are predicted for octahedral Ni(ll) ions and none should be so high in energy.

As was explained in section 3.5.4, the crystal field splitting energy, A0, was es timated using the Tanabe-Sugano diagram for d8 ions in an octahedral field.^159^

For every [cation]{Ni[Au(CN)2]3} complex, A0 was determined to be approximately Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 183

Table 4.5: Absorption maxima (Amax abg., cm ) observed in the solid- state UV-Vis-NIR absorption spectra of the [cation]{M[Au(CN)2]3} and n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] coordination polymers and their re 3 spective assignment to d-d transitions (ground state: A29 for the Ni(ll) complexes 4 and Tl9 for the Co(ll) complexes).

Ni(n) complexes Amax abs. (cm ) A0 a a a -1 T29 Tl9(F) Tl9(P) (cm ) K{Ni[Au(CN)2]3} 10,870 17,040 27,250 28,990 10,900 [PPN]{Ni[Au(CN)2]3> 11,175 17,800 -30,300 11 200 n [ Bu4N]{Ni[Au(CN)2]3} 10,990 17,700 28,170 32,790 11 000 2+ a [Ni (pyridine) 6]' 10,150 16,500 27,000 10,150 2 [Ni(CH3CN)6] + ° 10,700 17,400 27,810 10,700 2 a [Ni(en)3] + 11,700 18,350 29,000 11,700

4 4 4 Co(n) complexes T2ff Tl9 (or A2/) [PPN]{Co[Au(CN)2]3} 10,100 19,610; 21,050 n [ Bu4N]{Co[Au(CN)2]3} 9710 19,420; 21,280 {Co[Au(CN)2]2(H20)2}- 9615 19,420; 20,000 n [ Bu4N][Au(CN)2] 2 a [Co(H20)6] + 8100 19,400 2 a [Co(pyridine)6] + 9800 20,400 2 a [Co(en)3] + 10,100 21,000 a From reference (16°) 6 The absorption band is usually very weak and no band could be definitively assigned to this transition for the Co[Au(CN)2]2(analyte)2-

11,000 cm l. The crystal field splitting energy determined for these Ni(NC)6- 2+ containing complexes falls in between the values reported for [Ni(CH3CN)6] and 2+ 16 [Ni(en)6] .( °)

In the case of the [cation]{Co[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}- n Bu4N[Au(CN)2] complexes, peak overlap occurs in the 440-540 nm range and the absorption bands in this region could not be fully assigned. Octahedral Co(ll) com plexes usually show a band around 500 nm (20,000 cm-1) that is assigned to the 4 16 transition to the Ti9 (P) state. ( °) Also, it is known that the spin-allowed transition Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 184

4 160 to the A2s state is usually very weak^ ^ and the absorption band arises as a shoulder peak (above 500 nm) for several complexes. 4 4 As a consequence, the transition to the A2ff and Tl9 (P) states could not be definitely assigned. Hence, only a comparison to other complexes is made to esti mate the crystal field splitting energy for the Co(n) complexes reported in this chap ter. For [PPN]{Co[Au(CN)2]3}, the crystal field splitting energy is similar to that of 2+ n [Co(en)3] whereas that of [ Bu4N]{Co[Au(CN)2]3} is smaller and closer to that of 2+ (160) [Co(pyridine)3] . n For {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2], which contains Co(NC)4(0H2)2 units, the crystal field splitting is smaller and lies between the values observed for 2+ 2+ 160 [Co(pyridine)e] and [Co(H20)6] .^ ^ This is consistent with the spectrochemical series, in which water has a weaker ligand field than nitriles. (16°) Overall, the solid-state UV-vis-NIR spectral data is consistent with the spin-state and the octahedral coordination sphere of the compounds as determined by X-ray crystallography.

4.2.6 Thermal stability and structural rearrangement

Thermogravimetric analysis was performed for the [cation]{M[Au(CN)2]3} coordina + n + tion polymers between 25 and 500 °C for the K and [ Bu4N] containing complexes and up to 820 °C for the [PPN]+ containing complexes.

The K{Ni[Au(CN)2]3} polymer is stable until 330 °C. Above this temperature, a weight loss is observed (330-380 °C). The weight measured above 380 °C is consistent with the loss of the cyanide groups and the formation of NiO, elemental Au and K20 (Table 4.6).

The [PPN]{M[Au(CN)2]3} complexes start to decompose at 325 and 275 °C for the Ni and Co analogues respectively. A rapid weight loss is first observed followed by a continuous decrease in weight up to 750 and 730 °C for the Ni and Co analogues respectively. Based on previous results obtained for [Au(CN)2]-containing coordina tion polymers, it is suggested that the initial decrease in weight corresponds to the loss of the cyanide groups. Decomposition of the [PPN]+ cation occurs slowly, over Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 185

Table 4.6: Thermal stability of the [cation]{M[Au(CN)2]3} polymers: decomposition temperature and products obtained, with their observed (obs.) and calculated (calc.) relative weight

Complex Temperature % Weight Products °C obs. calc.

K{Ni[Au(CN)2]3} 330 380 85 84.4 NiO + 3 Au + 0.5 K20 [PPN]{Ni[Au(CN)2]3} 325 - 750 55 56.5 Ni(P203) + 3 Au [PPN]{Co[Au(CN)2]3} 275 - 730 57 56.5 Co(P203) + 3 Au ["Bu4N]{Ni[Au(CN)2]3} 295 - 330 64 63.5 NiO + 3 Au n [ Bu4N]{Co[Au(CN)2]3} 260 - 340 63 63.9 CoO + 3 Au {Co[Au(CN)2]2(H20)2}- 100 - 120 96 96.7 {Co[Au(CN)2]2}- n n Bu4N[Au(CN)2] Bu4N[Au(CN)2] 260 - 340 60 61.8 CoO + 3 Au

a large temperature range. According to the weight observed at 820 °C, the final products are suggested to be M(P203) and Au. The difference in final products with 53 previously reported [Au(CN)2]-containing coordination polymers^ ) is attributed to the presence of P atoms in the cation. n The two [ Bu4N]{M[Au(CN)2]3} complexes both decompose with only one weight loss step. This decrease in weight can be attributed to the simultaneous loss of the n + cyanide groups and the [ Bu4N] cation (Table 4.6). The Ni-containing complex has a slightly higher decomposition temperature, being stable until 295 °C, then decomposing in a very narrow temperature range (295-330 °C). The Co-containing complex is less stable and decomposes over a large temperature range (260-340 °C). n For the {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] complex, a first loss is observed between 100 and 120 °C, with a relative weight corresponding to two water molecules (Table 4.6). The remaining complex further decomposes along the same path (and temperature range) as [™Bu4N]{Co[Au(CN)2]3}. Following the thermogravimetric analysis results, a sample of n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] was heated at 150 °C in air for 90 minutes to remove the bound water molecules. The powder X-ray diffractogram and FT-IR spectrum of the dehydrated product were found to be superimposable with Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 186

n those of the independently synthesized [ Bu4N]{Co[Au(CN)2]3}. This suggests that a rearrangement is occurring upon dehydration to generate the water-free 3-D array of ["Bu4N]{Co[Au(CN)2]3}. Upon exposure to water vapour at room temperature (in the solid state), n the [ Bu4N]{Co[Au(CN)2]3} complex remains unchanged and does not convert to n the {Co[Au(CN)2]2(H20)2H Bu4N][Au(CN)2] complex as indicated by FT-IR spec troscopy and diffraction experiments.

4.3 Effect of structural changes on the magnetic behaviour

The magnetization of every [cation]{M[Au(CN)2]3J coordination polymer (prepared at room temperature) reported in this chapter was measured upon cooling from 300 K to 1.8 K in two different applied dc fields, 1 kOe and 100 Oe. Indistinguishable results were obtained for a given complex at the two fields, suggesting that their magnetic behaviour is field independent.

4.3.1 Magnetic properties of [cation]{Ni[Au(CN)2]3}

The magnetization of the K{Ni[Au(CN)2J3} sample prepared at room temperature was determined and the temperature dependence of the effective moment (fieff) is shown in Figure 4.8. The effective magnetic moment was determined to be 3.1 /i# at 300 K. As the temperature is decreased, the effective magnetic moment remains constant between 300 K and 10 K. Below 10 K, a slight decrease to 2.86 /J,B at 1.8 K is observed. The inverse magnetic susceptibility XM shows a linear behaviour with temperature above ~ 50 K (Figure 4.8). The room temperature effective magnetic moment is consistent with isolated 5 = 1 Ni(ll) centers and the temperature dependence of the inverse susceptibility follows the Curie-Weiss law. The drop in the effective moment at low temperature could be due to either zero-field splitting or weak antiferromagnetic interactions mediated through the [Au(CN)2]~ units or a combination of both. The fit to the zero-field Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 187

3.5 ' • I ' ' ' ' I ' ' • ' I ' ' ' ' I • • ' • I

250 SOQQOOOOOOOOQOO €>-^"0-0"0-0">-Cn 3.0 200 ^ 2.5 f 150 § *• 2.0 100 S ./ 1.5 50

1.0 J I—i I—I •—1_ _i—i i—I i—i i—a I • • • i_U Q 0 50 100 150 200 250 300 Temperature [ K ]

Figure 4.8: Temperature dependence of the effective magnetic moment (faff) (°) and 1 the inverse susceptibility (x^ ) (•) for K{Ni[Au(CN)2]3} with a 1 kOe applied mag netic field. The solid line represents the theoretical fit to the zero-field splitting expression (Equation 1.17). splitting equation (Equation 1.17, Figure 4.8) yielded a D value of 2.32(3) cm-1 and a g value of 2.22(1). The D value is comparable to those obtained for the related

Ni[Au(CN)2]2(analyte)a; (analyte = DMSO, pyridine) systems, after accounting for the presence of antiferromagnetic interactions (section 3.5.4). As mentioned in Chapter 3, the magnitude of zero-field splitting in a Ni(ll) com plex, which usually ranges from 0 to 5 cm-1, depends on the extent of distortion from perfect octahedral symmetry/70) As a consequence, the value for a more symmetri cal octahedral Ni(ll) center such as that found in K{Ni[Au(CN)2]3}, with a M(NC)6 coordination sphere, should be on the low side of this range. Hence, this value is rea sonable and would suggest that no interactions are present between the Ni(ll) centers despite their 3-D connectivity.

The magnetic behaviour of the [PPN]{Ni[Au(CN)2]3} (Figure 4.9) n and [ Bu4N]{Ni[Au(CN)2]3} polymers was found to be similar to that of Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 188

Table 4.7: Effective magnetic moment (faff) determined for the n [cation]{M[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] coordi nation polymers at different temperatures (T), along with the zero-field splitting parameters D, g values and coupling constant, zJ, obtained for the Ni complexes.

Ti Veff Heff at D 9 zJ Ni(ii) complexes 1.8 K -1 1 cation (K) (/*B) (VB) (cm ) (cm"" ) K+ 300--10 3.10(3) 2.86 2.32(3) 2.22(1) [PPN]+ 300--20 3.05(3) 2.22 5.0(1) 2.145(4) 2.6(l)a 2.180(2)a -0.52(2)a n [ Bu4N]+ 300--30 3.03(3) 2.22 5.4(1) 2.12(1) 2.7(l)a 2.173(2)a -0.59(2)a Co(ll) complexes cation [PPN]+ 300- 200 5.06(3) 3.69 - - - [™Bu4N]+ 300- 175 4.93(3) 3.51 - - - n [ Bu4N]+ / water 300- 200 4.88(3) 3.57 - - - a Values were obtained by introducing a molecular field approximation to the zero-field splitting equation (Equation 3.5).

K{Ni[Au(CN)2]3}: the effective moment is constant at high temperature and drops at low temperature. Table 4.7 compares the values observed as a function of temperature for the different [cation]{Ni[Au(CN)2]3} polymers. The temperature + n + dependent XMT product of the [PPN] and [ Bu4N] -containing polymers could also be fitted to the zero-field splitting expression (Equation 1.17) and the values obtained are reported in Table 4.7. n The fit of XMT for [PPN]{Ni[Au(CN)2]3} and [ Bu4N]{Ni[Au(CN)2]3} was im proved by introducing a molecular field approximation to the zero-field splitting equa tion to account for the magnetic interactions between the Ni(ll) centers (Equation 3.5). The values obtained for the zero-field splitting parameter and coupling constant are reported in Table 4.7.

As for K{Ni[Au(CN)2]3}, the relatively small D values are consistent with Ni(li) centers in a symmetrical environment. The coupling constants determined for Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 189

5.M I ' I I I I I I > I I I I I I I I I I I I I I I I I I • I 5.25 AAAAAAA 5.00 ,.;J1MAA ' 4.75 4.50 4.25 a;' 4.00 ta 3.75 3.50 3.25 3.00 ixsBxrrooooooooooooooooooo< 2.75 M = Ni 2.50 2.25

2.00 —l—i—I I i I I I i—• ' ' • ' r ' i—i— 50 100 150 200 250 300 Temperature [ K ]

r Figure 4.9: Temperature dependence of the effective magnetic moment (/iejO f°

[PPN]{Ni[Au(CN)2]3} (o) and [PPN]{Co[Au(CN)2]3} (A) with a 1 kOe applied mag netic field. The solid line represents the theoretical fit to the zero-field splitting expression with a molecular field approximation for the Ni(li) analogue (see text).

1 [PPN]{Ni[Au(CN)2]3} and ["Bu4N]{Ni[Au(CN)2]3} (-0.52(2) and -0.59(2) cm" ) in dicate the presence of very weak antiferromagnetic interactions between the zero-field split Ni(ll) centers.

4.3.2 Magnetic properties of [cation]{Co[Au(CN)2]3}

The effective moment determined for the [cation]{Co[Au(CN)2]3} and n {Co[Au(CN)2]2(H20)2H Bu4N][Au(CN)2] polymers at different temperatures are reported in Table 4.7. Figure 4.9 shows the behaviour observed for the [PPN]- containing polymer. At 300 K, the effective moment has a value of 5.06 //#, which remains constant until 200 K and then decreases steadily to reach 3.69 HB at 1.8 K. Despite having different structures, the two [nBu4N]+-containing polymers have a magnetic behaviour similar to [PPN]{Co[Au(CN)2]3} as the temperature is decreased Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 190 from 300 K to 1.8 K (Table 4.7). The observed room temperature value for each Co-containing polymer is larger than the spin-only value of an S = § system (3.87 /J,B), but as mentioned in Chapter 3 (section 3.5.4), it is in the range expected for an octahedral high-spin Co(ll) center 162 (4.1 - 5.2 /j,B) for which a large first-order orbital contribution is present. ( ) The decrease in effective moment at low temperature can be due to a combina tion of single-ion effects and very weak antiferromagnetic interactions. If present, antiferromagnetic interactions are presumed to be very weak as no maximum was observed in the plots of the magnetic susceptibility (XM) as a function of tempera ture. No attempt was made to model this behaviour as any coupling through the [Au(CN)2]~ units would have a very small magnitude and would be difficult to distin guish from the single-ion effects. In related Co(ll)-dca complexes with //15-bridging dca units, any coupling was shown to be below 1 cm"1, with single-ion effects domi nating. (205) n The magnetic behaviour of {Co[Au(CN)2]2(H20)2H Bu4N][Au(CN)2] is also simi lar to the behaviour reported in Chapter 3 for other Co[Au(CN)2]2(analyte)2 coordina tion polymers containing 2D square grid arrays of Co[Au(CN)2]2 (see section 3.5.4).

4.3.3 Magnetic properties of K{Fe[Au(CN)2]3}

(51) The structure of K{Fe[Au(CN)2]3} , which is almost identical to K{Ni[Au(CN)2]3}, was previously reported and this polymer was described as containing low-spin oc tahedral Fe(ii) centers (d6). However, the pale color reported for this complex was inconsistent with the proposed spin-state. This polymer was hence re-synthesised and the magnetic properties investigated. Infrared spectroscopy and powder X-ray diffrac tion were used to confirm that this sample was identical to the previously reported one. At 300 K, the effective moment (faff) was determined to be 5.79 /x#, which is not consistent with a low-spin octahedral d6 system. This value is in fact larger than the expected moment for an S = 2 spin-only system (4.89 ^B), but is as expected for octahedral high-spin Fe(ll) centers which show significant spin-orbit coupling. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 191

6.0 Ooooo OOOOOOOOOOOOOOO 5.5 P < 1 :£—'5> 0

^4.5 o i"4.0 3.5 3.0

2.5 • • •'!• •••!••• • I . • • . i . • . • i . . . . I 1 • ' • ' ' • • 0 50 100 150 200 250 300 -4 -3 -2-101234 Temperature [ K ] Velocity [ mm s'1 ]

Figure 4.10: A. Temperature dependence of the effective magnetic moment (neff) of K{Fe[Au(CN)2]3} in an external field of 1 kOe; B. Mossbauer spectra of

K{Fe[Au(CN)2]3} at 300 K (top) and 5 K (bottom). In both plots, the solid line corresponds to the best fit using a quadrupole split doublet.

The magnetization was then measured upon cooling from 300 K to 1.8 K (Fig ure 4.10A). The effective moment remains relatively constant between 300 and 75 K. Below this temperature, the effective moment drops and reaches 4.03 \LB at 1.8 K. No maximum in susceptibility could be observed for K{Fe[Au(CN)2]3}. Antiferromagnetic interactions could explain the drop in effective moment at low temperature, but the lack of maximum in susceptibility suggests that any cou pling is weak. This is consistent with the behaviour observed for the isostructural

K{Ni[Au(CN)2]3} polymer.

The Mossbauer spectrum of K{Fe[Au(CN)2]3} at 300 K shows a quadrupole pair, with an isomer shift 8 of 1.12(2) mm s_1 and a quadrupole splitting AEQ of 0.44(2) mm s_1 (Figure 4.10B). As the temperature is lowered to 5 K, the quadrupole splitting increases to 0.96(2) mm s-1 while 8 shifts to a value of 1.26(2) mm s-1. The isomer shift observed is consistent with the presence of Fe(ll) centers in a Chapter 4. [cation]{M[Au(CN)2]s}: Templating effects of the cation 192 high-spin state. The small quadrupole splitting observed at high temperature can be explained by the high symmetry of the Fe(NC)6 centers. The quadrupole split ting of several complexes containing high-spin octahedral Fe(ll) centers was reported to have a temperature dependence/206'207^ For example, the related cubic system 1 [Cp^Fe][Fe(N(CN)2)3] has a quadrupole splitting AEQ of 0.60(1) mm s" at 295 K at _1 tributable to the Fe(NC)6 centers (5 = 1.17(1) mm s ), while at 5 K, the quadrupole splitting has a value of 1.99(1) mm s_1. Axial and rhombohedral crystal field distortion as well as spin-orbit coupling are usually associated with high-spin d6 centers (see section 1.3.2 for more details on spin- orbit coupling). The spin-orbit coupling parameter A for Fe(ll)-containing complexes (high-spin) is very similar to that of the free-ion, which has a value of ~-103 cm-1/207) As the electric field gradient associated with each state differs, any change in popu lation distribution, due to temperature variation, can cause a change in the overall electric field gradient surrounding the nucleus. As a consequence, the quadrupole splitting is increased or decreased. Slight distortions of the structure as the temperature is lowered could also be responsible for the larger quadrupole splitting as a more asymmetric coordination sphere induces a more asymmetric electric field gradient. Temperature dependent X- ray diffraction studies could be performed to investigate the possibility of a structural rearrangement.

4.4 Discussion

As presented above, it is possible to incorporate different cations into

[cation]{M[Au(CN)2]3} coordination polymers simply by modifying the [Au(CN)2]~ starting material and choosing the appropriate solvent for the reaction. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 193

4.4.1 Synthesis of M(//-OH2)2[Au(CN)2]2 vs

K{M[Au(CN)2]3}

The room temperature aqueous reaction between Ni(N03)2-6H20 and K[Au(CN)2] generates two different products: Ni(/j-OH2)2[Au(CN)2]2 and K{Ni[Au(CN)2]3}. The green Ni(/>OH2)2[Au(CN)2]2 coordination polymer (investigated in Chapter 2) is be lieved to be a kinetic product that can be trapped due to its fast precipitation.

The Ni(/x-OH2)2[Au(CN)2]2 polymer converts, at room temperature, to the ther modynamic product, K{Ni[Au(CN)2]3}, upon soaking in the mother liquor containing an excess of KAu(CN)2. The conversion is suggested to go to completion as no signs of Ni(^i-OH2)2[Au(CN)2]2 can be detected in the powder X-ray diffractogram or the

FT-IR spectrum of the K{Ni[Au(CN)2]3} sample obtained in this manner.

It is believed that the Ni(/ii-OH2)2[Au(CN)2]2 polymer forms first due to the inert 2+ ness of the starting [Ni(H20)6] material. As mentioned in section 3.6.4, Ni(ll) ions are not kinetically labile compared to other first row transition metals. This most likely favours the product that requires a smaller number of ligand exchanges (i.e.

Ni(H20)4(NC)2 vs Ni(NC)e).

It is suggested that the Ni(/i-OH2)2[Au(CN)2]2 polymer is not completely insol uble in water and that conversion occurs in solution when the building blocks are dissociated into monomeric or oligomeric species. These units can react with the excess KAu(CN)2 and rearrange to form the K{Ni[Au(CN)2]3} polymer. The com plete conversion suggests that K{Ni[Au(CN)2]3} is more stable and less soluble than

Ni(//-OH2)2[Au(CN)2]2. When the same reaction was performed with Fe(ll), only one product,

K{Fe[Au(CN)2]3}, was obtained. It is possible that Fe(//-OH2)2[Au(CN)2]2 is formed as an intermediate, but conversion to K{Fe[Au(CN)2]3} occurs much faster than in the case of the Ni-analogue as no traces of Fe(/i-OH2)2[Au(CN)2]2 can be seen in the isolated product. As presented in Chapter 2, a different procedure is required to favour the formation of the Fe(yLf-OH2)2[Au(CN)2]2 coordination polymer. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 194

4.4.2 Impacts of the cations on the structural arrangement

The structural arrangement of the {M[Au(CN)2]3}~ framework was not found to be sensitive to the identity of the metal center as isostructural polymers could be obtained despite the change in metal center from Ni to Co to Fe. This result is not unexpected as these metal centers all prefer to adopt an octahedral geometry to increase their crystal field stabilization energy. On the other hand, as was observed for other systems, the identity of the countercation was found to affect the structural motif of the {M[Au(CN)2]3}~ coordination polymer framework. When a small cation such as K+ is incorporated, the formation of inter penetrated Prussian Blue-type arrays is favoured. The structural motif ob (51) served in K{Ni[Au(CN)2]3} and K{Fe[Au(CN)2]3} was also reported in other 50 coordination polymers of similar composition, namely K{Co[Au(CN)2]3}( ) and 208 K{Mn[Ag(CN)2]3}.( ) Interpenetration of the {M[Au(CN)2]3}- networks likely help to diminish the total energy of the system by allowing intermolecular interactions. The presence of Au-Au and K-NC interactions between the interpenetrated networks likely help to stabilize the polymer and favour network interpenetration. + + n + The switch of the non-steric K cation for the larger [PPN] and [ Bu4N] cations removed the possibility of network interpenetration as the negatively charged network needs to accommodate them. These structural modifications did not, however, cause a decrease in the network connectivity. The long reach (~ 10-11 A) and flexibility = (^C-Au-C 173-180°) of the [Au(CN)2]~ units is likely responsible for maintaining the connectivity between the transition metal centers. Also, the preference of the metal centers to adopt an octahedral geometry probably forces the 3-D [Au(CN)2]- based connectivity when no other ligands are available to complete the coordination spheres. When water is present in the reaction mixture, hydration of the metal centers can occur, which results in a reduction in metal centers connectivity, as observed in n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2].

The non-interpenetrated network of [PPN]{M[Au(CN)2]3} is strain-free and nearly identical to a single pseudo-cubic network found in the K{M[M'(CN)2]3} polymers Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 195

(M = Ni, Fe, Co, Mn; M' = Au, Ag). The "expanded Prussian Blue" array found in these polymers, with ~10.3 A M-M edges, is comparable to other examples such as 209 210 Fe4[Re6Te8(CN)6]3-xH20( ) and [Ni(en)2]3[Fe(CN)6](PF6)2( ) which have edges of n + 14.1285 and 9.908 A respectively. The incorporation of [ Bu4N] groups forces the system to modify its 3-D framework structure to a great extent to accommodate this differently shaped cation. n The basic 2-D grid structure of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] is sim 51 ilar to that reported for other solvent adducts, such as Mn[Au(CN)2]2(H20)2,( )

Co[Au(CN)2]2(DMF)2 ^ or Ni[Au(CN)2]2(DMF)2 and M[Au(CN)2]2(pyridine)2 (M = Cu, Co) presented in Chapter 3. The packing of the 2-D square grids in n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] differs however due to the presence of the n extra [ Bu4N][Au(CN)2] layer. The significant structural rearrangement that occurs upon dehydration of the 2-

D {Co[Au(CN)2]2(H20)2H"Bu4N][Au(CN)2] polymer in the solid state to generate n the 3-D [ Bu4N]{Co[Au(CN)2]3} coordination polymer is a strong indication of the flexible nature of dicyanoaurate-based coordination polymers.

4.4.3 Impacts of the cations on the physical properties

As the physical properties are often function of the structural arrangement, dif ferent properties, at least some of them, could have been expected for the struc turally different polymers. However, the structural differences between the different

[cation]{M[Au(CN)2]3} coordination polymers did not impact to a large extent the observed physical properties. The polymers have similar UV-vis absorption spectra due to similar geometry and _ ligand field around the M(ll) centers (octahedral M(NC)6). Also, the {M[Au(CN)2]3} frameworks have similar thermal stabilities, being stable until 260 to 325 °C. The Ni- based systems are, however, more stable than the analoguous Co-based systems, most likely due to the inertness of the Ni(ll) centers. (187) These decomposition temperatures 2n of the [cation]{M[Au(CN)2]3} polymers are typical for cyanometallate systems. ( )

The magnetic behaviours of the [cation]{M[Au(CN)2]3} coordination polymers are Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 196 also similar. In each of them, the metal centers are magnetically isolated and the be haviours are mainly dominated by single ion effects intrinsic to the Ni(ll) or Co(ll) cen ters present. Compared to the K{Ni[Au(CN)2]3} polymer, the [PPN]{Ni[Au(CN)2]3} n and [ Bu4N]{Ni[Au(CN)2]3} polymers showed weak antiferromagnetic interactions (J fa —0.5 cm-1). Longer Ni-N distances are present in the K+-containing poly mer (Table 4.1) compared to the other [cation]{Ni[Au(CN)2]3} structures. As longer distances result in poorer orbital overlap, this could explain the lack of magnetic interactions mediated by the [Au(CN)2]~ units in K{Ni[Au(CN)2]3}. Despite the different structural arrangements determined for the n n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] and [ Bu4N]{Co[Au(CN)2]3} coordi nation polymers, their respective observed magnetic behaviours are almost identical. - The increase in connectivity between the Co(ll) centers via [Au(CN)2] units, from 2-D to 3-D, did not affect the magnetic properties. In both polymers, the Co(ll) centers are magnetically isolated, irrespective of the number of possible magnetic pathways. For long-range magnetic order to be present in a system, there must be mag netic interactions in at least two dimensions. Through space dipolar interactions are usually much weaker than interactions mediated through chemical bonds, and their strength decreases rapidly as the distance increases. Hence, magnetic bridges in 3- D are typically incorporated to enhance magnetic coupling, and possibly obtain a magnetically ordered system. However, the magnetic properties determined for the

3-D [cation] {M[Au(CN)2]3} polymers are comparable with those obtained for the 2-D

M[Au(CN)2]2(analyte)a; coordination polymers (Chapter 3). This indicates that the increase in connectivity between the metal centers did not help to improve the overall magnetic coupling. From the results obtained in this study, it can be concluded that the diamagnetic [Au(CN)2]~ unit is a poor mediator of magnetic exchange. This contrasts with 4 the low-spin, diamagnetic [Fe(n)(CN)6] ~ bridges which mediate the key interactions 31 leading to ferromagnetic ordering in Prussian Blue (Tc = 5.6 K).( ) At first sight, 4 both [Au(CN)2]~ and [Fe(ll)(CN)6] ~ units act as five-atom diamagnetic bridges be tween the transition metal centers possessing unpaired electrons. However, electron Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 197

_L_L J_ J_ J_J_X U.lLl| ±1.1.

Fe(m)—NC—Fe(n)—CN—Fe(in) h.s. l.s. h.s.

Scheme 4.3: Magnetic exchange occurring in Prussian Blue between the Fe(lll) centers through the Fe(ll) centers due to partial electron delocalization. Only one electron spin is favoured during electron transfer (shown in red circle) which leads to overall ferromagnetic orientation of the formally Fe(lll) magnetic moment.

delocalization can occur in Prussian Blue due to the facile electron transfer from the 4 [Fe(ll)(CN)6] ~ units to the Fe(lll) units. As shown in Scheme 4.3, when electron transfer takes place, only one spin orientation is favoured as the Fe(lll) centers are 4- already half-filled (S — |). Also, since each [Fe(ll)(CN)6] unit is surrounded by six Fe(lll) units and that electron transfer to any of them has the same probability, the system minimizes its energy by aligning the magnetic moments on the Fe(lll) units in the same direction. (212)

In the [Au(CN)2]-containing polymers, electron transfer is less likely to occur be tween the Au(l) centers and the first row transition M(ll) centers, as the formation of both M(i) and Au(ii) is thermodynamically unfavoured. (162>213) This contrasts with the situation in Prussian Blue in which electron transfer can easily occur between the Fe centers in the 2+ and 3+ oxidation states. Hence, magnetic interactions in

[Au(CN)2]-containing polymers can only be mediated through the overlap or orthog onality of the magnetic orbitals with the molecular orbitals of the [Au(CN)2]~ units. The large number of atoms involved reduces the strength of any possible magnetic interactions, especially compared to one- or two-atom bridges. As a consequence, greater magnetic exchange was expected for the 3-D polymers containing pseudo-cubic arrays, such as K{M[Au(CN)2]3} and [PPN]{M[Au(CN)2]3}, in which better orbital overlap should be present. However, this was not observed n as a similar behaviour was also obtained for the [ Bu4N]{M[Au(CN)2J3} polymers, in Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 198 which less orbital overlap should be present due to the more distorted geometry of the metal centers.

The lack of any significant magnetic interactions in the [Au(CN)2]-bridged poly mers is comparable to the observation that magnetic coupling through the diamagnetic t 164 205 214 A i,5~N(CN)2 bridges is generally very weak. ( > > ) The only compounds for which magnetic ordering was observed contain //ij3-N(CN)2 bridges (i.e. the central amide N-atom binds to the metal center as well as the nitrile N-atoms).^215^

The results observed for the [cation]{M[Au(CN)2]3} polymers differs from those re ported for other 1-D [Au(CN)2]-based coordination polymers in which slightly more significant coupling was observed (J ~ 2 cm"1).'65' These capping-ligand contain ing polymers structurally differ from the [cation]{M[Au(CN)2]3} polymers and no straightforward comparison can be made to explain the differences in magnetic cou pling.

It could be possible that, in the 3-D connected [cation]{M[Au(CN)2]3} sys tems, competing ferromagnetic and antiferromagnetic interactions might contribute to cancel out the overall coupling. This is however unlikely since the same overall cancelling effects would have to be present in every polymer, despite the structural changes and the different identity of the metal center, to explain the identical be haviour observed.

4.5 Conclusions and future work

This study has shown that [cation]{M[Au(CN)2]3} coordination polymers can be read ily prepared. The structures of K{M[Au(CN)2]3} and [PPN]{M[Au(CN)2]3} contain similar 3-D pseudo-cubic anionic frameworks of {M[Au(CN)2]3}~. Network interpenetration is possible when a small countercation (K+) is present but is suppressed when larger cations occupy the network's cavities. When [™Bu4N]+ is included, a distinct 3-D anionic framework of {M[Au(CN)2]3}~ is obtained. Thus the framework structure is sensitive to the cation incorporated. Despite using large cations such as [nBu4N]+ or [PPN]+, a 3-D structure is main tained (except when the metal is solvated), contrary to the reduction in structural Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 199 dimensionality observed in dicyanamide (dca) systems/164'214) The longer reach of linear [Au(CN)2]~ compared to bent dca may account for this difference.

Despite the presence of 3-D networks connected by [Au(CN)2]~ units, no magnetic ordering was detected above 1.8 K for these [cation]{M[Au(CN)2]3} polymers. Indeed, almost no significant coupling through the [Au(CN)2]~ units was observed, indicating 10 that the linear d -[Au(CN)2] bridge is a very poor mediator of magnetic exchange. Further increase in cation bulkiness or introduction of hydrogen-bonding moieties into the cations are likely to eventually result in a significant superstructure rear rangement. (216>217) With this in mind, bulkier [R,4N]+ (R = pentyl, hexyl, heptyl, etc.) cations could be used to verify this hypothesis. 3+ Similarly, the use of a cation with a different charge, such as [Co(NH3)6] , [M(2,2'- 2+ 2+ bipyridine)3] or [M(2,2':6,6"-terpyridine)2] , should be explored (see Scheme 4.4A. and B. for the molecular structures of bipyridine and terpyridine). If a structure similar to that of the [PPN]{M[Au(CN)2]3} polymers is maintained (ie. a cubic-type framework with the cations occupying the cavities), this could lead to the formation of porous networks, since only a fraction of the cavities would be occupied by the cations. Porous Prussian Blue analogues were reported to have high gas adsorption 36 37 :r+ capabilities( > ) and porous (cation )1/a,{M[Au(CN)2]3} (x > 2) polymers are likely to show similar properties. Incorporation of cations having additional material properties could also be pur sued. For example, incorporation of [DAMS]+ (Scheme 4.4) in the channels of the

{M[Au(CN)2]3}~ framework was briefly attempted, but no pure product was obtained (see section 4.6.9 for more details and partial results). Despite the presence of impu rities, the results were promising and modification of the reaction conditions should be attempted to be able to isolate and characterize the pure product. Incorporation of [TTF]n+-type cations (TTF = tetrathiafulvalene, n < 1) could also be attempted, as several examples of polymers containing this countercation were found to be superconductive when the right packing of the [TTF]n+ cations was present. Chapter 4. [cation] {M[Au(CN)2]3}: Templating effects of the cation 200

Scheme 4.4: "Shape" of different ligands and cations: A. 2,2'-bipyridine, B. 2,2':6,6"- terpyridine, C. [DAMS]+ and D. [TTF]n+

4.6 Experimental Section

The general procedures concerning the characterization of the [cation]{M[Au(CN)2]3} complexes are as described in Chapter 2, section 2.7.1, unless otherwise noted.

Porosity measurements

Nitrogen gas adsorption measurements were conducted at 77 K and up to 100 torr using a custom-built (by Prof. Ian Gay, SFU) porosity apparatus, applying a static volumetric method.

4.6.1 Synthesis of K{Ni[Au(CN)2]3}

Room temperature synthesis

A 2 mL aqueous solution of Ni(N03)2-6H20 (0.029 g, 0.10 mmol) was added to a 10 mL aqueous solution of K[Au(CN)2] (0.114 g, 0.396 mmol). A pale green precipitate formed immediately and was left stirring in solution for 4 days. After 4 days, the precipitate had changed to a pale blue powder which was isolated by filtration. The composition was determined to be K{Ni[Au(CN)2]3}. Yield: 0.072 g, 85 %. Anal

Calcd for C6H0N6Au3NiK: C 8.53, H 0.00, N 9.95. Found: C 8.62, H traces, N 9.77. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 201

1 IR (KBr): (uCN) 2170 (s), 473 (m) cm- . UV-Vis-NIR: 345,370, 585, 920 run.

Hydrothernial synthesis

A 1 mL aqueous solution of K[Au(CN)2] (0.114 g, 0.396 mmol) was combined with a

1 mL aqueous solution of Ni(N03)2-6H20 (0.029 g, 0.10 mmol) in a 5 mL ampoule. Water was added to bring the total volume to 3 mL. The ampoule was sealed and loaded into the digestion bomb (as described in section 2.7.1). The bomb was heated to a temperature of 125 °C over a period of two hours, maintained at this temperature for 6 hours, and slowly cooled to 25 °C at a rate of 1 °C per hour. Dark blue crystals of K{Ni[Au(CN)2]3} (0.01-1 mm in diameter) and a pale blue powder were obtained by performing the reaction under these conditions. Anal Calcd for C6H0N6Au3NiK:

C 8.53, H 0.00, N 9.95. Found: C 8.62, H traces, N 9.77. IR (KBr): (uCN) 2169 (s), 473 (m) cm-1.

4.6.2 Synthesis of K{Fe[Au(CN)2]3}

The synthesis of K{Fe[Au(CN)2]3} was modified from that reported in the litera 51 ture/ ) A 0.5 mL aqueous solution of Fe(C104)2-6H20 (0.032 g, 0.088 mmol) was added to a 1.5 mL aqueous solution of K[Au(CN)2] (0.060 g, 0.21 mmol). A pale yellow precipitate formed immediately and was isolated by filtration. The compo sition was determined to be K{Fe[Au(CN)2]3}. Yield: 0.048 g, 81 %. IR (KBr): 2158 (s), 465 (m) cm-1. Comparison between the powder X-ray diffractogram of the product obtained and the diffractogram predicted by the published structure for

K{Fe[Au(CN)2]3} confirmed that the structure of the two compounds was the same and that no impurity was present.

4.6.3 Synthesis of [PPN][Au(CN)2]

A 10 mL aqueous solution of K[Au(CN)2] (0.251 g, 0.872 mmol) was added to a 20 mL mixture of ethanol/water (1:1 v/v) solution of bis (triphenylphosphoranylidene) ammonium chloride (PPNC1) (0.525 g, 0.915 Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 202 mmol) while stirring. An immediate white precipitate formed. After standing for 30 minutes, the solid of [PPN][Au(CN)2] was filtered, washed with water and air-dried overnight. Yield: 0.666 g, 97.0 %. Anal. Calc. for C38H3oN3AuP2: C, 57.95; H, 3.84; N, 5.38. Found: C, 57.71; H, 3.86; N, 5.17. IR (KBr): 3088 (vw), 3075 (w), 3055

(m),3038 (w), 3022 (w), 3008 (vw), 2990 (vw), (uCN) 2145 (s), 1589 (m), 1483 (m), 1456 (m), 1439 (s), 1434 (s), 1285 (s), 1267 (s), 1180 (m), 1160 (w), 1116 (s), 997 (m), 796 (m), 743 (s), 726 (s), 690 (s), 548 (s), 531 (s), 496 (s) cm"1. A different preparation of this salt, via [AuCLj]-, has previously been reported. ^218^

4.6.4 Synthesis of [PPN]{Ni[Au(CN)2]3}

A 5 mL ethanolic solution of [PPN][Au(CN)2] (0.100 g, 0.127 mmol) was added slowly to a 5 mL ethanolic solution of Ni(N03)2-6H20 (0.013 g, 0.044 mmol). A solid immedi ately precipitated from the solution, which changed from blue to completely colorless.

A very fine pale blue powder of [PPN]{Ni[Au(CN)2]3} was obtained upon filtration.

Yield: 0.039 g, 69 %. Anal. Calc. for C42H3oN7Au3NiP2: C, 37.53; H, 2.25; N, 7.29. Found: C, 37.30; H, 2.34; N, 7.19. IR (KBr): 3076 (w), 3055 (m), 3021 (w), 3009 (w),

2998 (vw), 2971 (vw), (vCN) 2192 (s), (UCN) 2151(vw), 1588(W), 1476 (w), 1436 (m), 1383 (s), 1330 (m), 1302 (w), 1272 (w), 1186 (w), 1160 (vw), 1117 (s), 1026 (w), 997 (w), 738 (w), 723 (s), 688 (w), 550 (m), 531 (s), 499 (m) cm"1. UV-Vis-NIR: 335, 562 (broad) and 895 (broad) nm. Crystals suitable for X-ray crystallography were obtained using an H-tube tech nique. On one side of a 55 mL H-shaped tube filled with ethanol, a 2 mL ethanolic solution of [PPN][Au(CN)2] (0.096 g, 0.12 mmol) was delivered at the bottom using a pasteur pipette. On the other side, a 0.5 mL ethanolic solution of Ni(N03)2-6H20 (0.015 g, 0.052 mmol) was delivered in the same way. The H-shaped tube was sealed, and after 10 days, light purple cubic crystals of [PPN]{Ni[Au(CN)2]3} formed. The crystals and the powder had identical powder X-ray diffractograms and IR spectra. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 203

4.6.5 Synthesis of [PPN]{Co[Au(CN)2]3}

A 7 mL ethanolic solution of [PPN][Au(CN)2] (0.150 g, 0.190 mmol) was added slowly to a 7 mL ethanolic solution of Co(C104)2-6H20 (0.023 g, 0.060 mmol). A peach powder immediately precipitated from the solution, which changed from pink to com pletely colorless. A very fine peach powder of [PPN]{Co[Au(CN)2]3} was obtained upon filtration. Yield: 0.040 g, 71 %. Anal. Calc. for C42H3oN7Au3CoP2: C, 37.52; H, 2.25; N, 7.29. Found: C, 37.77; H, 2.15; N, 7.47. IR (KBr): 3145 (vw), 3076 (w),

3055 (m), 3021 (w), 3009 (w), 2989 (vw), (uCN) 2187 (s), {VCN) 2175(sh), 1588(w), 1476 (w), 1438 (m), 1384 (s), 1330 (m), 1320 (m), 1302 (w), 1272 (w), 1186 (w), 1160 (vw), 1117 (s), 1026 (w), 997 (w), 881 (vw), 840 (vw), 800 (vw), 738 (w), 723 (s), 690 (w), 550 (m), 531 (s), 499 (m) cm"1. UV-Vis-NIR: 475 (broad), 490 (sh), 510(sh) and 990 nm.

n 4.6.6 Synthesis of [ Bu4N]{Ni[Au(CN)2]3} Room temperature synthesis

A 5 mL ethanolic solution of Ni(N03)2-6H20 (0.028 g, 0.096 mmol) was added to n a 5 mL ethanolic solution of [ Bu4N][Au(CN)2]-0.5H2O (0.153 g, 0.310 mmol). The solution was left to evaporate for several days. When ca. 5 mL of solvent remained, a purple powder was collected by filtration and air dried. The chemical composition was determined to be ["Bu4N]{Ni[Au(CN)2]3}. Yield: 0.044 g, 44 %. Anal. Calc. for

C22H36N7Au3Ni: C, 25.21; H, 3.46; N, 9.35. Found: C, 25.52; H, 3.59; N, 9.62. IR

(KBr): 2961 (s), 2933 (s), 2887 (m), (uCN) 2195 (s), 1479 (m), 1377 (m), 1150 (w), 1106 (w), 1051 (w), 1024 (w), 880 (m), 800 (w), 736 (m), 491 (s) cm"1. UV-Vis-NIR: 305, 355, 564 (broad) and 910 (broad) nm.

Hydrothernial synthesis

Crystals suitable for single crystal X-ray diffraction analysis were obtained by recrystallization of the purple powder under hydrothermal conditions in a sealed glass n ampoule (see section 2.7.1 for more information). A sample of [ Bu4N]{Ni[Au(CN)2]3} Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 204

(0.090 g in 3 mL of water) was heated to 125 °C over a period of two hours, main tained at this temperature for 6 hours, and slowly cooled to 25 °C at a rate of 1 °C per hour. The crystals and the powder had identical powder X-ray diffractograms and IR spectra.

n 4.6.7 Synthesis of [ Bu4N]{Co[Au(CN)2]3}

n [ Bu4N][Au(CN)2]-0.5H2O (0.040 g, 0.080 mmol) and Co(/u-OH2)2[Au(CN)2]2 (0.031 g, 0.052 mmol) were mixed and completely dissolved in 15 mL of ethanol by stirring for 10 minutes. Once dissolved, the stirring was discontinued and the peach solution was left to evaporate to dryness at room temperature, yielding peach crystals covered with white precipitate. To this solid mixture 3 mL of ethanol was added and the solution was lightly stirred and filtered after 3 minutes. The filtrate was set aside and the pink precipitate was washed with 3 mL of water and 3 mL of cold ethanol. Elemental n analysis was consistent with the chemical formula being [ Bu4N]{Co[Au(CN)2]3}. The separated filtrate was evaporated to dryness, then redissolved in ethanol and subjected to the same treatment to increase the yield. Yield: 0.029 g, 52 %. Anal. Calc. for

C22H36N7Au3Co: C, 25.20; H, 3.46; N, 9.35. Found: C, 25.42.; H, 3.54; N, 9.39. IR

(KBr): 2960 (s), 2933 (m), 2873 (m), {uCN) 2189 (s), (VCN) 2157 (vw), 1479 (m), 1462 (w), 1377 (w), 1175 (vw), 1149 (w), 1104 (vw), 1053 (vw), 1024 (w), 879 (w), 800 (vw), 734 (w), 484 (w) cm-1. UV-Vis-NIR: 470 (broad), 490 (sh), 515 (sh), 1030 nm.

n 4.6.8 Synthesis of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2]

n [ Bu4N][Au(CN)2]-0.5H2O (0.040 g, 0.080 mmol) and Co(/x-OH2)2[Au(CN)2]2 (0.035 g, 0.059 mmol) powders were mixed together, suspended in a 1 mL mixture of ethanol/water (9:1, v/v) and the reaction mixture was sealed with parafilm. The precipitate slowly turned pink, and formed small pink rectangular prism crys tals after three days. The solution was then filtered and the crystals were iso lated. The chemical composition of the crystals was found to be consistent with n {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] by elemental analysis and single crystal

X-ray diffraction. Yield: 0.049 g, 77 %. Anal. Calc. for C22H38N7Au3Co02: C, Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 205

24.37; H, 3.72; N, 9.04. Found: C, 24.55; H, 3.79; N, 9.26. IR (KBr): 3454 (b), 3207

(m), 2960 (s), 2930(m), 2872 (m), (uCN) 2183 (s), (uCN) 2147 (s), 2108 (vw), 1600 (m), 1469 (m), 1380 (w), 1164 (vw), 1104 (vw), 1068 (vw), 1032 (vw), 926 (vw), 881 (w), 731 (vw), 490 (m). UV-Vis-NIR: 470, 500, 1040 nm. n Crystals of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] suitable for single crystal X-ray diffraction analysis were obtained by a different route: A 5 mL ethanol/water n (1:1 v/v) solution of [ Bu4N][Au(CN)2]-0.5H2O (0.105 g, 0.210 mmol) was added slowly to a 5 mL ethanol/water solution of Co(C104)2-6H20 (0.026 g, 0.071 mmol). The pink solution immediately became peach colored, and was left to slowly evaporate. n Pink/red rectangular prism crystals of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] and white [nBu4N](C104) crystals formed as the ethanol evaporated. The solution was n then filtered, and the crystals of {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2] were sep arated by hand; they can be cleaned with ethanol, but are somewhat soluble. Yield 0.025 g, 33 %. The crystals and the powder had identical powder X-ray diffractograms and IR spectra.

4.6.9 Attempts involving [DAMS]+

A 15 mL solution (1:2 water:methanol) of trans-4-[4-(dimethylamino)-styryl]-l- methylpyridinium (DAMS) iodide (0.036 g, 0.10 mmol) was added to a 5 mL methano- n lic solution of [ Bu4N][Au(CN)2]-0.5H2O. To this mixture, a 5 mL methanolic solution of Ni(N03)2-6H20 (0.029 g, 0.10 mmol) was added while stirring. No colour change occurred and no precipitate were formed immediately. The reaction was partly cov ered and left to evaporate overnight. On the next day, a very dark precipitate was collected by filtration. IR (KBr): 3402 (m), 2906 (w), {VCN) 2197 (s), (uCN) 2166 (s), 1644 (w), 1616 (w), 1571 (s), 1530 (m), 1507 (m), 1439 (w), 1376 (m), 1318 (m), 1163 (s), 1041 (w), 970 (w), 823 (w) cm"1. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 206

4.6.10 Details of X-ray structural determinations

Single crystal X-ray diffraction

Data acquisition and general structure solutions refinement was performed as de n scribed in Chapter 2. For [ Bu4N]{Ni[Au(CN)2]3}, the Ni and Au atoms were refined anisotropically while the C and N atoms were kept isotropic due to the limited num n ber of observed data. For {Co[Au(CN)2]2(H20)2}- [ Bu4N][Au(CN)2], the C atoms of the nBu-groups were kept isotropic due to their very large thermal motion.

Crystallographic data for compounds K{Ni[Au(CN)2]3}, [PPN]{Ni[Au(CN)2]3}, n [ Bu4N]{Ni[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}-["Bu4N][Au(CN)2] are collected in Tables 4.8 and 4.9 while selected bond lengths and angles are reported respectively in Tables 4.1, 4.2, 4.3 and 4.4.

Powder X-ray diffraction

+ The powder diffractograms of [cation]{Co[Au(CN)2]3} (cation = [PPN] and [nNBu4]+) were collected on a Rigaku RAXIS-Rapid Auto diffractometer as described in section 2.7.1. Indexing was performed using the WinPLOTR and POWDER CELL software. (141>143) The best unit cell parameters are reported in Table 4.10. Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 207

Table 4.8: Crystallographic data and structural refinement details for

K{Ni[Au(CN)2]3} and [PPN]{Ni[Au(CN)2]3}

K{Ni[Au(CN)2]3} [PPN]{Ni[Au(CN)2]3} Empirical formula C6N6Au3NiK C42H3oN7Au3NiP2 Formula weight 844.82 1344.30 Colour, shape Hexagonal blue plate Pale purple cube Dimension, mm3 0.195x0.156x0.074 0.34 x 0.31 x 0.17 Crystal system Trigonal Rhombohedral Space group (#) P312 (149) R3c (167) a, A 6.786(4) 15.284(3) b, A 6.786 15.284 c, A 7.778(8) 31.530(10) a, deg 90 90 0, deg 90 90 7, deg 120 120 Volume, A3 310.2(4) 6379(2) Z 1 6 A, A 0.70930 0.70930 Data range, deg 4-89 4-55 Transmission range 0.0422 - 0.0639 0.0303 - 0.1154 3 pcaicd, g cm" 4.522 2.100 H, mm-1 37.173 10.867 Reflections, parameters 1309, 30 1172, 87 Riil > 2.5a(I))b 0.0297 0.0303 b wR2(I > 2.5a(I)) 0.0313 0.0411 goodness of fit 0.8792 1.888 a Rx = £ |(|F0| - |Fc|)|/£ \FC\ for observed data (I > 2.5 2.5

Table 4.9: Crystallographic data and structural refinement details for n n [ Bu4N]{Ni[Au(CN)2]3} and {Co[Au(CN)2]2(H20)2}-[ Bu4N][Au(CN)2]

["Bu4N]{Ni[Au(CN)2]3} {Co[Au(CN)2]2(H20)2}- n [ Bu4N][Au(CN)2] Empirical formula C22H36N7Au3Ni C22H40N7Au3CoO2 Formula weight 1048.19 1095.52 Colour, shape Purple block Red/dark pink rectangu lar prism Dimension, mm3 0.42 x 0.42 x 0.28 0.22 x 0.20 x 0.14 Crystal system Tetragonal Tetragonal Space group (#) IAlCd (110) F42/mnm (136) a, A 23.532(4) 14.444(2) b, A 23.532 14.444 c, A 23.181(3) 15.453(2) a, deg 90 90 P, deg 90 90 7, deg 90 90 Volume, A3 12,837(3) 3223.9(6) Z 16 4 A, A 0.70930 0.70930 Data range, deg 4-51 4-48 Transmission range 0.0135 - 0.0409 0.1068-0.1707 pealed, g cm-3 2.169 2.232 H, mm-1 14.270 14.143 Reflections, parameters 1540, 156 689, 87 R^I > 2.ba(I)f 0.0483 0.0420 b wR2(I > 2.ba(I)) 0.0529 0.0351 goodness of fit 1.3778 1.4542 c|)l/E \Fd for observed data (I > 2.5 2.5CT(/)), 2 1 2 GoF = [X>(|F0| |Fc|) /degrees of freedom] / . Chapter 4. [cation]{M[Au(CN)2]3}: Templating effects of the cation 209

Table 4.10: Best unit cell parameters determined for [PPN]{Co[Au(CN)2]3} and n [ Bu4N]{Co[Au(CN)2]3} by powder X-ray diffraction.

[PPN]{Co[Au(CN)2]3} ["Bu4N]{Co[Au(CN)2]3} Crystal system Rhombohedral Tetragonal Space group (#) R3c (149) IAlCd (110) a, A 15.38 23.77 b, A 15.38 23.77 c, A 31.61 23.26 a, deg 90 90 P, deg 90 90 7, deg 120 90 Volume, A3 6478 13,137 Chapter 5

Hydrothermal synthesis of [Au(CN)2J-based polymers and related nanoparticle impurities

5.1 Introduction

5.1.1 Crystallization of coordination polymers

Although the majority of coordination polymers are usually synthesised at room tem perature, the simple mixing of precursor solutions, as was shown in the previous chap ters, often does not yield large single crystals but rather generates microcrystalline powders. As the physical properties of a coordination polymer in the solid-state are expected to be a consequence of its structure, being able to determine the 3-D arrangement is a key step in understanding structure-property relationships. Given

210 Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 211 this importance, the crystallization conditions are often modified to favour the forma tion of single crystals suitable for structure determination. (219>220) For example, dif ferent crystal growth techniques such as slow evaporation, slow diffusion of reagents through H-shaped tubes and gels/221) electrocrystallization/222) and hydrothermal or solvothermal reactions (223'224) are often used to obtain crystals suitable for X-ray crystallography. When using these modified synthesis conditions, care must be taken as different products or polymorphs can also be formed. (163'176) The term solvothermal applies to solution reactions carried out at temperatures beyond the normal boiling point of the solvent, while remaining in the liquid phase. Such conditions can be achieved in a sealed and rigid vessel. When water is used as the solvent, the term "hydrothermal" is applied. Hydrothermal reactions are usu ally carried out between 100 and 250 °C.(225) Hydro- and solvothermal reactions are used a great deal to prepare different types of materials, ^226) including zeolites, ^227^ inorganic solids/228'229' hybrid organic-inorganic materials/230) and molecular clus ters.^225) Such reaction conditions allow one to (1) recrystallize materials that are not soluble under ambient conditions; (2) increase the reactivity of inert building blocks; (3) encourage the formation of thermodynamic or other metastable products; and (4) generate unique solid materials via the in situ hydrothermal synthesis of unusual ligands in the presence of metals. (231>232) However, despite these advantages, the con ditions of hydrothermal reactions can induce decomposition of some building blocks, and allow side reactions to occur.

5.1.2 Research objectives

While investigating the physical properties of the new Prussian-Blue analogue

K{Ni[Au(CN)2]3} (presented in Chapter 4), prepared at room temperature as a pow der or crystallized hydrothermally at 125 °C, some discrepancies were observed in the magnetic behaviour. Upon examination by electron microscopy, small amounts of nanoparticle impurities were found in the samples prepared hydrothermally. The chemical and magnetic properties of the nanoparticle impurities, as deter mined through transmission electron microscopy and SQUID magnetometry studies, Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 212 were investigated and are presented in this chapter. This cautionary narrative serves to illustrate the important lesson that great care must be taken when studying hydrothermally prepared samples to ensure that the observed physical properties are indeed attributable to the sample.

5.2 Characterization of K{Ni[Au(CN)2]3} samples prepared at elevated temperatures

5.2.1 Magnetic behaviour of samples prepared at 125 °C

The magnetic properties of the room temperature synthesized K{Ni[Au(CN)2]3} sam ple were described previously (Chapter 4, section 4.3.1). This sample, prepared at room temperature, will hereafter be referred to as RT.

As mentioned in Chapter 4 (section 4.2.1), the K{Ni[Au(CN)2]3} sample prepared hydrothermally at 125 °C was indistinguishable from the one prepared at room tem perature according to FT-IR spectroscopy, powder X-ray diffraction and elemental analysis. This sample prepared at 125 °C will be referred to as H-125. The same SQUID magnetometry experiments were performed on the presumably pure H-125 sample.

The effective magnetic moment, //ejgr, determined as a function of temperature for sample H-125 is shown in Figure 5.1A and compared to the behaviour of sample RT. The effective magnetic moments of both RT and H-125 are essentially constant between 300 K and 60 K (3.08 - 3.14 /ig). Below 50 K, different behaviours can be observed for the two samples. In sample H-125, an increase to a maximum value of 3.50 HB at 20 K is observed, followed by a decrease to 2.90 HB at 2.0 K. To investigate the differences in magnetic behaviour between the K{Ni[Au(CN)2]a} sample prepared at room temperature (RT) and the sample prepared hydrothermally (H-125), more experiments were performed. The temperature dependence of the magnetization, measured upon warming in a magnetic field of 10 Oe, after being cooled in the absence (Zero-Field Cooled, ZFC) or the presence (Field Cooled, FC) Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 213

i • • • ' i • ' ' ' i { l i i i | i i i i | i i i i | l i • • | i • i c | i l i l | 3.6

£;• 3,2 l0«0008880088M08888d

2.8 k

2.6 k *aao a - ••••'• 1 ' • • ' ' • • • • ' • • • • ' • • ' ' ' • • • • " 0 50 100 150 200 250 300 30 40 50 60 Temperature [ K ] Temperature [ K ]

Figure 5.1: A. Effective magnetic moments determined for the K{Ni[Au(CN)2]3} RT

(•) and H-125 (o) samples. B. ZFC measurements for the K{Ni[Au(CN)2]3} sample RT (A), and FC (•) and ZFC (o) measurements for sample H-125 under an external field of 10 Oe.

of an applied magnetic field was determined. For sample RT, a continuous decrease in magnetization can be observed, irre spective of the cooling conditions (Figure 5.IB shows the ZFC curve). The FC curve (not shown) is superimposable with the ZFC curve. This behaviour is consistent with Curie-Weiss paramagnetism. In the ZFC measurements performed on the H-125 sample (Figure 5.IB), as the temperature is increased from 1.8 K to 10 K, a decrease in the magnetization, similar to that observed for the RT sample, is observed. In contrast however, above 10 K, the magnetization increases and a maximum is reached at a temperature of 20 K, referred to as Tg. Upon further warming, the magnetization decreases and, above 30 K, follows that of the ZFC and FC measurement for the RT sample. When ZFC measurements are performed under different external fields, the position and shape of the maximum vary from a sharp peak at 20.5 K when measured in a field of 5 Oe to Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 214

I II I | I II I J I ITI'I I I IT | II I rT'F'l I l|T'l I I JfT'I I11 ri'T'l I I I'l'l i' '••' ' l ' 1 1 ' ' B. .,...,, T._.

20 o - 0 19 -

o -—. 18 -

LJ 17 o -

16 o -

15 - -

14 o -

I 1 . 1 . , 1 . . ' 10 12 14 16 18 20 22 24 26 28 30 100 200 300 Temperature [ K ] H [Oe]

Figure 5.2: A. ZFC measurements for sample H-125 under different external fields (5, 10, 25, 50, 100, 150, 200 and 300 Oe), from smaller (top, with a solid line) to larger field (bottom, with a solid line). Solid lines are guides for the eyes. B. Field dependence of Tg for sample H-125 a broader peak at 14 K in a field of 300 Oe (Figure 5.2). When the FC measurements were carried out, an initial decrease in magnetiza tion was observed, albeit more slowly, and the data rejoins the ZFC curve at the temperature where a maximum was observed in the ZFC curve for the H-125 sam ple (Figure 5.IB). As the temperature is further increased, the FC curve tracks the ZFC curve. As will be discussed in section 5.3.3, this behaviour is suggestive of a superparamagnetic signature superimposed onto a paramagnetic background. The ac susceptibility measured upon cooling for samples RT and H-125 is shown in Figure 5.3. The shape of the in-phase component, x^, of the ac susceptibility for both samples is similar to that of their respective dc measurements, showing a steady decrease for sample RT and a maximum in the vicinity of 21 K for the sample H- 125. No out-of-phase signal, x'g, was observed for sample RT while a maximum in the out-of-phase signal of the sample H-125 is observed at 20.5 K (Figure 5.3B). Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 215

T" T" Tl-p TTTT Mll| r I • i i i I ' ' ' ' I ' • ' ' I ' ' ' ' I • A. ;• • B. 1.5 D '• -_

D 00 :* \ 1 4> . • : 7o 4 " • - o • a : °» :

• • a • 0.5 D D • 0° D D D D " °3 p ° - •« '• s 8 • s 0.0 null lll.l mi i. 0 5 10 15 20 25 30 35 40 45 50 15.0 17.5 20.0 22.5 25.0 Temperature [ K ] Temperature [ K ]

Figure 5.3: A. In-phase {x'g) and B. out-of-phase (x'g) components of the ac suscep tibility for the RT (filled) and H-125 (empty) K{Ni[Au(CN)2]3} samples (ac field of 5 Oe, driving frequency of 10.0 Hz, and zero dc field).

Clearly, despite the spectroscopic, analytical and diffraction data (presented in Chapter 4, sections 4.2.1 and 4.6.1) that indicates that the two samples are pure and identical, the low temperature magnetic behaviours are different, with the H- 125 system showing indications of some form of magnetic ordering or blocking (see section 5.3.3). This magnetic behaviour was consistently observed for every sample prepared under these conditions.

5.2.2 Transmission electron microscopy studies

Sample H-125 was investigated using transmission electron microscopy (TEM). I was involved in every experiment but the electron microscope was operated by Simon Trudel (Simon Fraser University). The most obvious features observed were hexag onal crystals tens to hundreds of microns in size. Upon higher magnification and careful inspection (Figure 5.4A), ensembles of particles with dimensions between ap proximately 2 and 10 nm were observed (see Figure 5.4D for an histogram of the size Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 216 distribution). The shapes of the nanocrystals ranged from spherical to oblate, with some having more convoluted morphologies. The selected area electron diffraction (SAED) of these nanoparticles was recorded and is shown in Figure 5.4C. Rings were clearly observed in the diffractogram, suggest ing that the inspected area contains a randomly oriented, crystalline material. The electron diffraction pattern observed does not match that which would be expected from the trigonal structure of K{Ni[Au(CN)2]3} as determined by single-crystal X-ray diffraction. The electron diffraction pattern readily indexes to a face centered cubic (fee) structure; the rings were assigned hkl values corresponding to reflections due to the (111), (200), (220), (311), (222), and (331) /cc-planes. A representative image of the particles observed in the H-125 sample under high magnification is presented in Figure 5.4B. Under phase contrast conditions, lattice fringes were clearly observed. The spacing between each fringe was determined to be in the 3.1-3.5 A range. The periodicity of lattices fringes corresponds in theory to that of crystal lattice planes which are perpendicular to the TEM sample's surface. No assignment of chemical identities could be done based on these distances. When the RT sample was investigated by TEM, hexagonal shaped micron size particles could be observed, similarly to the H-125 sample. However, even after extensive investigation, no nanoparticles could be found in the RT sample. Electron diffraction of the large hexagonal crystals showed they were single crystals, and not an agglomerate of nanocrystals. In no area of the RT sample was a SAED diffractogram arising from a collection of fee structured crystals observed. Elemental analysis of each sample was carried out using energy dispersive X-ray spectroscopy (EDXS). Figure 5.5A shows the high angle annular dark field STEM image of a representative area. Figure 5.5B shows the EDXS spectra of selected ar eas indicated in Figure 5.5A. The elemental analysis of the larger hexagonal crystals showed the presence of all the constituent elements of the K{Ni[Au(CN)2]3} coor dination polymer, including clear signals for nitrogen and potassium (Figure 5.5B,

Spectrum ii). This is consistent with the probed region being K{Ni[Au(CN)2J3}. Sig nals for copper, beryllium and silicon are artefacts, and are attributed to the TEM grids, its holder, and the EDXS detector, respectively. They were also observed in all Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 217

o r i i i it i v i i • i i i i i i i i i i i 0 2 4 6 8 10 Diameter of nanoparticles [ nni ]

Figure 5.4: A. Bright field transmission electron microscopy image of nanoparticles present in the H-125 sample of K{Ni[Au(CN)2]3}. Scale bar = 50 nm. B. High reso lution transmission electron microscopy image showing a nanoparticle. Scale bar = 5 nm. Inset shows the fast Fourier transform of sections of the micrograph. C. Selected area electron diffraction of the area shown in A. The rings are indexed to hkl values for a fee structure. D. Histogram showing the size distribution of the nanoparticles in the H-125 sample, based on measurements made on 250 nanoparticles. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 218

:c AU " '- 1 | BeN Cu Lr o aAu 1/ Au : : IT: A Au

K__—A ___AJLAJ : JU . A ... A . iii.:

' i • • • i i i i i '" .I.I • I.I '' 0 1 2 3 4 6 7 8 9 10 11 12 13 14 Energy [ keV ]

Figure 5.5: A. High angle annular dark field scanning transmission electron microscopy image of the H-125 sample. Scale bar = 50 nm. B. Energy dispersive X-ray spectra (EDXS) for areas highlighted in A, showing a region containing: nanoparticles (Spec trum i), K{Ni[Au(CN)2]3} crystals (Spectrum ii) and the carbon-coated copper grid (Spectrum iii).

subsequent EDXS spectra presented. The EDXS spectrum of a region consisting of the nanoparticulate crystals was rich in gold, nickel and oxygen, but no significant amount of nitrogen was observed, and no X-ray line due to potassium was present (Figure 5.5B, Spectrum i). This strongly suggests that the composition of the nanoparticles differs from that of the larger, coordination polymer crystals. In all areas, a strong signal due to the presence of carbon could be observed. It is however impossible to differentiate carbon from the sample or from the supporting grid. The oxygen line could also be due to the supporting grid (Figure 5.5B, Spectrum iii). However, the intensity of the oxygen line (with respect to carbon) in Spectrum i is much higher (by a factor of ~2.5) than the oxygemcarbon line intensities in Spectrum iii. Even though this is not definitive, it suggests that a significant portion of the oxygen signal in Spectrum i arises from the sample. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 219

Hence, the elemental analysis of the H-125 sample by EDXS suggests that the nanoparticles are composed of gold, nickel and oxygen, or mixtures of these elements.

5.2.3 Modification of the reaction conditions

To further investigate the nanoparticle side product present in the H-125 sample, the conditions of the reaction were modified. The temperature and time spent above 125 °C were increased in an attempt to increase the yield of nanoparticles, as is described below.

Sample prepared at 165 °C

A hydrothermal reaction using the same quantities of reagents was carried at 165 °C. A large amount of red-brown powder was obtained along with the dark blue crystals. The bulk product will be referred to as H-165. ZFC and FC magnetization experiments were performed on sample H-165. A magnetic behaviour qualitatively similar to H-125 was observed (Figure 5.6). How ever, the maximum in magnetization reaches a larger value as the maximum temper ature of the reaction was increased, going from 0.245 x 10~2 emu g_1 for H-125 to 8.61 x 10-2 emu g_1 for H-165. Sample H-165 exhibited a primary maximum in magnetization at 22.75 K (compared to 20 K for sample H-125) and, in addition, a shoulder or a second, lesser maximum at approximately 17 K was also observed. The increase in magnetization below 5 K could be a paramagnetic tail, attributable to the

K{Ni[Au(CN)2]3} coordination polymer present in the samples. When investigated by TEM, sample H-165 showed different features than the ones observed in the H-125 sample. Images of the red-brown powder from the H-165 sample showed much larger particles (200-300 nm), along with filament-like particles (Figure 5.7A). The presence of a few smaller particles (few nanometres in diameter) was also detected in the filament-rich regions. They were, however, agglomerated with the filaments, and thus they could not be successfully imaged independently of the filaments. The electron diffraction pattern was collected in regions consisting of the filaments Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 220

0 10 20 30 40 50 60 Temperature [ K ]

Figure 5.6: Comparison of the ZFC magnetic behaviours of samples RT (black, o), H-125 (red, o), H-135 (blue, •) and H-165 (green, A). Lines are guides for the eyes.

(Figure 5.7B). The pattern consisted of diffuse rings, which are consistent with a poorly crystallized material. It is also worth noting that under the effect of electron beam radiation, the probed regions had a tendency to amorphize. The pattern appears to be consistent with an fee structure, but due to the poor intensity and broadness of the rings, not all the reflections expected were clearly visible. Figure 5.8 shows the results obtained from the EDXS measurements on the two separate regions of H-165. It can clearly be determined that the 200-300 nm particles contained primarily gold (Figure 5.8B, Spectrum ii), while the filament-like and small sphere-like particles were composed of nickel and oxygen (Figure 5.8B, Spectrum i). The powder X-ray diffractogram of the red-brown powder from the H-165 sample differs from that predicted for pure K{Ni[Au(CN)2]3} (Figure 5.9). The peaks cor responding to K{Ni[Au(CN)2]3} are present, but they only account for a very small fraction of the sample. The principal pattern could be assigned either to NiO or Au, or a mixture of the two, as both are fee systems with very similar lattice parameters (there is a difference of only 0.1 A in lattice parameter a between Au (4.071 A)(233) Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 221

A. B.

500 nm

Figure 5.7: A. High angle annular dark field scanning transmission electron microscopy image showing the 200-300 nm particles and the filament-like particles present in the red-brown powder from sample H-165. Scale bar = 500 nm. B. Selected area electron diffraction of an area containing only filament-like particles. The rings are indexed to hkl values for an fee structure.

B. 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 • I//I i i i i i i i i i i i i i i i i : Ni Ni 1 \ : Au H '

n : - Si - 1 : O • Cu : Au a :_, 1\J • c Au Cu : Al Au : • 1 Au '• • ON1 i Si Ni Culi .. •

ki . . . i . . L^ i , , , ,_y/Ljj_1 • •'•••••'•••'• 0 1 2 3 4 6 7 8 9 10 11 12 13 14 Energy [ keV ]

Figure 5.8: High angle annular dark field scanning transmission electron microscopy image of the H-165 sample. Scale bar = 100 nm. B. Energy dispersive X-ray spectra (EDXS) for areas highlighted in A: filament-like NiO particles (Spectrum i) and Au particles (Spectrum ii). Chapter 5. Hydrothermal synthesis and related nanoparticle impurities Til

ttI I1 1l\ — .... III. J I 1 i 1 J i. . A 1 B. T~

.1 1 1 C.

,1 1 1 D. • ' ' ' ' ' *^""» • • i i i i i i 12 3 4 5 6 7 8

Figure 5.9: Comparison of the measured X-ray diffractogram of H-165 (A.) with the diffractograms predicted for K{Ni[Au(CN)2]3} (B.), Au (C.) and NiO (D.). andNiO (4.178 A) (234)).

Sample prepared at 135 °C

To further investigate the system, a reaction was also carried at 135 °C for 65 hours, using the same quantities of reagents. The bulk product was visually similar to H- 125, but a small amount of red-brown powder could be observed mixed with the pale blue powder and dark blue crystals. This bulk product will be referred to as H-135. The same ZFC and FC magnetization experiments were also performed on sample H-135 (Figure 5.6). These showed that sample H-135 had a magnetic behaviour similar to H-165, with a maximum (at 22.5 K) and a subsidiary peak at lower tem perature. The magnetization Mg at the primary peak in the ZFC measurement had a magnitude intermediate between that of the H-125 and H-165 sample. Similarly to sample H-125, sample H-135 showed a mixture of large crystal blocks and spherical nanosized particles with diameters ranging from 5 to 10 nm when investi gated by TEM. The large crystals were single crystalline, by SAED, and contained, by

EDXS, carbon, nitrogen, potassium, nickel and gold, indicative of K{Ni[Au(CN)2]3}, as for H-125. Chapter 5. Hydro-thermal synthesis and related nanoparticle impurities 223

i • i • i • i • i ' i • i ' i ' i

3 4 6 7 89 10 11 12 13 14 Energy [ keV ]

D. i i i i i/o—i i i r^—i • i • i '—i • i—T

Ni Ol Au Si NiCu

Au 1/ Au Au iii. J Ni _. C i Si o

IV.

I I I I I I • I I//I I I 0 1 2 3 4 6 7 8 9 10 11 12 13 14 Energy [ keV ]

Figure 5.10: A. High angle annular dark field scanning transmission electron mi croscopy image of the H-135 sample. Scale bar = 20 nm. B. Energy dispersive X-ray spectra (EDXS) for areas highlighted in A., showing a region rich in nickel and oxy gen (Spectrum i) and a region rich in gold (Spectrum ii). C. Image obtained in a different location on the same TEM grid as A. Scale bar = 50 nm. D. EDXS for areas highlighted in C, showing a region rich in gold, nickel and oxygen (Spectrum iii) and a region only rich in nickel and oxygen (Spectrum iv). Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 224

Various regions of H-135 containing nanosized particles were examined by EDXS. The EDXS data of these different regions are shown in Figure 5.10. No potassium was observed in the regions containing nanoparticles. In most areas, strong signals for nickel and oxygen were seen in the EDXS spectra (Figure 5.10B, spectra i, hi, iv). Varying amounts of gold could also be observed in these regions. In some other areas, gold was the major component of the spectrum (Figure 5.10B, spectrum ii). These results would be consistent with the presence of various amounts of two types of nanoparticles, namely NiO, and presumably metallic Au. Despite this chemical difference observed by EDXS, the regions are visually very similar when observed by TEM (Figure 5.10A). One difference is that in some of the nickel-oxygen rich regions, filament-like structures similar to those observed in sample H-165 could also be imaged, in addition to spherical particles (Figure 5.IOC).

5.3 Discussion

5.3.1 Chemical identity of the nanoparticles

What is the chemical identity of the nanoparticles?

This question can be partially addressed by examining the (S)TEM images, EDXS and SAED data. This ensemble of data (compare, for example, the (S)TEM images of H-125, H-135, and H-165 in Figures 5.4, 5.10 and 5.7) makes it clear that different nanoparticulate products are generated at each of the three hydrothermal temperature regimes used. However, all the SAED data for the nanoparticle samples can be indexed to essentially the same simple fee structure, which at first glance appears to suggest that the chemical composition does not change. All EDXS spectra in the nanoparticle regions of H-125, H-135 and H-165 lack potassium signals and do not show a proportionate nitrogen signal for K{Ni[Au(CN)2]3}. As such, none of the nanoparticle products are nano-crystals of the primary coordination polymer

K{Ni[Au(CN)2]3}. The results obtained for the H-165 sample clearly showed that two types of very different nanosize products were present (Figure 5.7): large Au particles (200-300 Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 225 nm) and much smaller NiO filament-like particles, with their compositions clearly identified by EDXS (Figure 5.8). The powder X-ray diffractogram of the bulk H-165 sample confirms the presence of at least one of these candidates (Figure 5.9). Unfortunately, under our experimental conditions, X-ray diffraction methods could not discriminate between NiO and Au only based on peaks positions. Nevertheless, the relative intensities of the peaks observed between 1.0 and 2.5 A suggest the presence of Au in large quantities in the H-165 sample. The presence of NiO as a minor component cannot, however, be ruled out. It must also be stressed that the NiO filaments were rather poorly crystallized, as observed by SAED, and would thus contribute less to the powder X-ray diffractogram of the H-165 sample. Focusing on the nanoparticles in the H-125 sample, a strong signal for gold, as well as weaker signals for nickel and oxygen, were observed in the EDXS spectrum in all nanoparticulate regions. In light of the chemical identity of the particles observed in H-165, it is suggested that a mixture of Au and NiO nanoparticles is formed in the reaction at 125 °C. The structure of both NiO and Au is /cc/233'234^ as was the structure of the nanoparticles when observed in the SAED. Note that at this temperature, in contrast with the product obtained at 165 °C, the Au and the NiO particles formed have similar spherical/oblate morphologies and could not readily be distinguished on the basis of shape or size. The large size distribution (2-10 nm) could result from the presence of chemically different particles. Different chemical species may be expected to have different growth kinetics at a given temperature.

The EDXS data could also suggest that AuxNii_x alloy nanoparticles were formed, rather than the chemically-segregated Au and NiO units as described above. Au;rNii_x alloys have previously been reported in the literature, and their structures were also determined to be fee by X-ray diffraction for bulk^235) and nanoparticle samples. (236) The presence of pure Ni nanoparticles amongst the products of any of these high temperature reactions is unlikely, since the reactions were carried out in water and the samples were handled in ambient atmosphere. Indeed, it has been reported that metallic Ni nanoparticles readily oxidise in air to form Ni-NiO core-shell nanoparti cles. (237) No core-shell structures were observed in any of our TEM investigations. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 226

In contrast with the H-125 sample, the inhomogeneous chemical composition of the H-135 nanoparticles, as determined by EDXS, is a clear indication that at least two nanoparticulate products, likely Au and NiO, were formed under these conditions. As in sample H-125, despite a different chemical composition, both types of particles could not be distinguished on the basis of shape and average diameter. In addition to the spherical particles, a few filament-like particles were present in some of the NiO-rich regions, a feature that was not observed in H-125, but was present in large quantities in H-165. Changes in the reaction conditions are known to affect the growth of nanoparticles, yielding different particle sizes, size distributions, and morphologies. (238^ This could readily explain the differences observed between these hydrothermal samples. At lower temperatures, small, approximately spherical particles, which are likely composed of NiO and Au, are produced. Reactions at higher temperatures (accompanied by longer reaction times at high temperature) result in the growth of these particles. It is apparent that the Au and NiO particles grow in different ways. In the case of the Au nanoparticles, an Ostwald-ripening type process results in the growth of larger particles from the dissolution of smaller sized particles. (239~241) For NiO, the aspect ratio of the particles is more affected than their diameters, yielding filament- like particles which are more and more present as the reaction temperature and time is increased. The concentration of the metal ions, which will ultimately form the nanoparticles, is also likely to be a function of the reaction temperature, and is likely to affect the size and morphology of the nanoparticles.

5.3.2 Nanoparticle formation route

The formation of Au nanoparticles requires the reduction of Au(i) metal ions. As no reducing agent was added to the reaction, the cyanide groups likely acted as internal reductants for the metal ions. Cyanide containing coordination complexes and poly mers are known to release cyanogen, (CN)2, at elevated temperatures via oxidation of the cyanide groups and reduction of the metal centers. (211>242) In particular, cyanogels, Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 227 which are amorphous Prussian-Blue analogues, have been investigated for their appli cations as single-source precursors to binary and ternary transition metal alloys and intermetallics. (243^ Superparamagnetic nanoparticle alloys of NiFe^244) and PdCo^245) have also been synthesised from the high temperature autoreduction of preformed cyanide-containing nanoparticles under an argon atmosphere. A recent report showed that the thermal decomposition of nanoscale 246 Co3[Co(CN)6]2 yields Co304 nanoparticles/ ) In view of this, it may also be possible that K{Ni[Au(CN)2]3} coordination polymer nanoparticles are formed, and decom pose to yield NiO and Au nanoparticles. Several methods to prepare nanoparticles of NiO have been reported, including 247 248 the thermal treatment of Ni(OH)2 gels/ ' ) Under our experimental conditions, it is also possible that Ni(OH)2 is formed as an intermediate toward the synthesis of the observed NiO nanoparticles. It is clear that the formation of nanoparticles is enhanced with higher reaction temperatures and times. This was evidenced with successively larger contributions of the superparamagnetic impurity to the ZFC measurements (Figure 5.6).

5.3.3 Magnetic behaviours

As discussed in Chapter 4 (section 4.3.1), the magnetic behaviour observed for the

RT sample of K{Ni[Au(CN)2]3} is typical of a paramagnetic system that does not order spontaneously over the temperature range studied here (Figure 5.1). The magnetic properties of the samples prepared under hydrothermal conditions show a remarkable difference to the simple, paramagnetic-like behaviour of the sample RT. The maximum observed in the ZFC measurements for the H-125, H-135, and H-165 samples is suggestive of a superparamagnetic signature superimposed onto a paramagnetic background, as will be detailed below. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 228

Superparamagnetic Impurities

Superparamagnetism is a magnetic state in which the large magnetic moment of individual nanoparticles fluctuate freely, due to thermal randomization, in a para magnetic fashion. Superparamagnetism is a finite scale effect, and is encountered in small particles (~20 nm diameter or less, depending on the material). (249^ A material below its Curie temperature will be magnetically ordered. In the case of a nanopar ticle, its size is small enough that its energy will be minimized when it constitutes a single magnetic domain. (25°) As such, the magnetic moment of a nanoparticle will be on the order of a few thousand Bohr magnetons. The anisotropy energy EA, which describes the energy barrier for the relaxation of the magnetization vector of a quasi- spherical nanoparticle in the absence of interparticle interactions, has been described by Stoner and Wohrfarth^251) as:

2 EA = KVsm 6 (5.1) in which K is the anisotropy constant (which is material dependent), V is the volume of the nanoparticles, and 9 is the angle between the magnetization vector and its preferred orientation within the crystal structure of the nanoparticle, also known as the easy axis. When thermal energy is high enough, the anisotropy energy barrier KV is over come, and the magnetization vector will freely fluctuate (Scheme 5.1A). If the time scale of the measurement is much greater than the relaxation time of the magnetization vector, the measurement will only see a freely fluctuating magnetization vector, ie the material will appear to be paramagnetic. The distinction between a superparamag netic and a paramagnetic magnetization vector is the magnitude of its magnetization. While a paramagnetic ion only has a magnetic moment of a few Bohr magnetons, a superparamagnetic nanoparticle will have a magnetic moment of a few thousand Bohr magnetons. When the thermal energy is much lower than the anisotropy energy barrier, the magnetization vector will not be able to flip orientation, and will be in the blocked state (Scheme 5.IB). This occurs below a particular temperature called the blocking temperature Tg. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 229

A. lh

Scheme 5.1: A. Magnetic nanoparticles in the superparamagnetic state. B. Mag netic nanoparticles in the blocked state (after being cooled in zero field). Red lines indicate the easy axis, and black arrows indicate the magnetization vector of each nanoparticles. In the superparamagnetic state (A.), the vectors fluctuate freely over all orientations in space (depicted as double-headed arrows), while in the blocked state (B.) the moments are pinned, aligned with their easy axes.

A non-oriented, isotropic sample has an equally isotropic distribution of easy-axis orientations. When such a sample is cooled in the absence of an applied magnetic field (ZFC), the magnetization vector of each nanoparticle will align itself with its easy axis, and the overall magnetization of the sample will be close to zero. Upon warming in an applied magnetic field, the overall magnetization will increase, as the individual magnetization vectors will align with the poling field (instead of the easy axis). Upon passing through the blocking temperatiire TB, the thermal energy overcomes the poling effect of the magnetic field, and randomizes the orientation of the nanoparticles' magnetization vector. In the case of a FC measurement, the magnetization vectors are oriented by the applied field upon cooling. As such, the measured magnetization at low temperature Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 230

^ 18

12 J • i i i I i i i i I i i • ' * i i_.-t.~j L__J—i _ 0 10 20 30 40 H-3 [Oe ]

Figure 5.11: Field dependence of the blocking temperature TB determined from ZFC 2 measurements for sample H-125. The line follows the Hs de Almeida-Thouless law. is high. As the temperature is increased, thermal energy randomizes the orientations of the magnetization vectors, and a decrease in magnetization is observed. Above the blocking temperature, the response is paramagnetic for both the FC and ZFC measurements. With this in mind, an examination of the ZFC and FC magnetization measure ments shown in Figure 5.IB (H-125) indicate that they contain two components: (i) a continuously decreasing, Curie-type paramagnetic component, and (ii) a super paramagnetic component passing through its blocking temperature. The Curie-type contribution arises from the major product, K{Ni[Au(CN)2]3}, as is evidenced in the Curie-type magnetic behaviour of the nanoparticles-free sample prepared at room temperature (Figure 5.IB). As is shown in Figure 5.11, the blocking temperatures TB determined for sample H-125 under various applied fields H follow the de Almeida- 2

Thouless law, which states that the TB of superparamagnetic nanoparticles has a H 3 dependence. (252) This suggests that this magnetic component is due to superparamagnetism. No nanoparticles could be found by TEM in the sample synthesized at room tem perature, while for every hydrothermally prepared sample examined by TEM, collec tions of nanoscopic particles (ranging from 5 to 200 nm, depending on the sample) could readily be observed, in all regions of the supporting grid (Figures 5.4, 5.10 and Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 231

5.7). Thus, the superparamagnetic component, which is only observed in the samples prepared under hydrothermal conditions, is almost certainly due to the presence of the nanoparticles observed by TEM. Furthermore, the shape, position, and magnitude of the maximum in the ZFC curves, which strongly depend on the temperature at which the hydrothermal reaction was carried out (Figure 5.6), are consistent with the formation of different products at different temperatures (as seen by TEM). Nanoparticles obtained in the samples prepared under hydrothermal conditions, whether they consist of NiO filaments or spherical particles could be responsible for the observed superparamagnetic behaviour. While bulk NiO is known to be an antiferromagnet (TN = 525 K), NiO nanoparticles with average diameters of 3 and 7 nm were found to be superparamagnetic, with blocking temperatures of 10 and 15 K, respectively, determined from ZFC measurements at 100 Oe. ^253' It is also worth noting that, above a given nickel composition (~ 50 at. % Ni), bulk Au^Nii-^ alloys are ferromagnetic. (254>255) More importantly, Auo.45Nio.55 nanoparticles with an aver age size of 12.0 nm were found to behave as a superparamagnetic material, with a blocking temperature of 8 K under an applied field of 500 Oe; (236^ different sizes and

Au:Ni ratios would certainly yield slightly different TB values. It should also be noted that recent reports have shown that in nanosized particles of gold, superparamagnetism^256,257) and even room-temperature ferromagnetism^258^ have been observed. However, this only occurs in very specific circumstances; (257>258) it is very unlikely that the observed magnetic behaviour in any of our samples is due to superparamagnetic gold nanoparticles. To verify this possibility, a control exper iment was performed with a solution of K[Au(CN)2], using the same hydrothermal conditions as H-125. The resulting product, after evaporation of the solvent, showed no maximum in the ZFC magnetization curve and only a diamagnetic response. The presence of two peaks in the ZFC measurements for H-135 and H-165 (Fig ure 5.6) can have three different origins: (i) size distribution, (ii) shape, and (iii) chemical identity. According to Equation 5.1, different nanoparticle size distributions (V) and chemical identities (K) would yield different anisotropy energy barriers. In a uniaxial structure, such as a filament, an additional shape anisotropy term may Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 232 also contribute to jf.(259>260) These three effects would thus lead to different block ing temperatures. Thus, either the H-135 and H-165 samples contain two distinct superparamagnetic chemical species, or there is only one superparamagnetic species, but with a bimodal size distribution, or different aspect ratios. We currently have no specific control over size, shape, or composition of the nanoparticles in these synthetic procedures. As such, all are equally likely expla nations, in principle. However, given the data described above, the two superpara magnetic species can most likely be attributed to the two differently shaped forms of NiO observed (spherical vs filament-like), which only occurred in H-135 and H- 165. In H-125, where only NiO spherical particles were observed, only one peak was observed in the ZFC curve. The anisotropic shape of filament-like particles can be responsible for the increase in blocking temperature, from 20 K for the spherical particles in sample H-125 to 22.75 K for the filament-like particles in sample H-165.

5.3.4 Shortcomings of conventional analytical methods

Clearly, the presence of even a trace amount of superparamagnetic nanoparticles can have a huge impact on the measured magnetic properties. For example in H-125, a convolution of the properties of the target coordination polymer K{Ni[Au(CN)2]3} and the superparamagnetic nanoparticles is observed. It must be stressed that stan dard characterization techniques (such as FT-IR spectroscopy and elemental analy sis), as well as the commonly accepted checks to ensure sample purity and identity (such as the comparison of predicted and obtained X-ray diffractograms) failed to uncover the presence of the nanoparticle impurity in H-125. For example, samples of K{Ni[Au(CN)2]3} containing up to 4.2 % (by weight) of a 1:1 mixture of NiO and Au nanoparticles would still show elemental analysis results consistent with a pure sample within standard accepted tolerances. Although powder X-ray diffraction can be used as an analytical tool to quantify the composition of mixtures, the typical limit of detection for this method is ca. 5 %, for well crystallized samples with well resolved peaks. For a nanocrystalline sample, Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 233

Scherrer broadening of the linewidths tends to decrease the intensity of the reflection peaks, (261) and, in such a case, the nanoparticles could be undetectable. Also, if the impurity is poorly crystallized (as was the case for the NiO filaments formed at 165 °C), it will be harder to detect. The detection may be further complicated by the substantial peak overlap with the powder diffractogram of K{Ni[Au(CN)2]3} in this case. Such small amounts of NiO and Au would also be very difficult to observe in the FT-IR spectra in the 400-4000 cm-1 range. As such, these analytical methods are inappropriate to detect small impurity levels of nanocrystalline material. However, such small amounts of magnetic nanoparticle impurity are more than sufficient to dominate the overall observed magnetic response. In the case of

K{Ni[Au(CN)2]3} presented here, the comparison of the magnetic properties from samples prepared at room temperature and by hydrothermal methods suggested that a (much) closer look at the samples was imperative. However, in other cases where the products obtained by hydrothermal methods are unique and thereby inaccessible by more conventional synthetic routes (such as an analogous room-temperature-based preparation), a comparison with "pure" samples may be difficult, if possible at all. In these cases, the danger of erroneously attributing the observed magnetic (or indeed any other) properties to the "pure" bulk material cannot be underestimated.

5.3.5 Presence of nanoparticles in other systems

The formation of nanoparticles as a side product is not unique to the hydrother mal reaction of Ni(N03)2-6H20 with K[Au(CN)2]. Hydrothermal recrystallization of

K{Ni[Au(CN)2]3} (which excludes the presence of [N03]"), at 165 °C, afforded a bulk product with a similar magnetic behaviour to that shown in Figure 5.6 for the H-165 sample. This would suggests that [N03]~ ions are not playing a role in the formation of nanoparticles. n Also, magnetic studies on hydrothermally prepared [ Bu4N]{Ni[Au(CN)2]3} sam n ples (using Ni(N03)2-6H20 and [ Bu4N][Au(CN)2]-0.5H2O as starting materials) showed a maximum at 20 K in the ZFC curves. No investigation on these samples by Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 234 electron microscopy has been performed, but it is suggested that similar superpara magnetic nanoparticles were likely formed.

5.4 Conclusions

Hydrothermal synthesis of K{Ni[Au(CN)2]3} yielded a smattering of nanoparticle im purities which dominated the observed magnetic properties, even though conventional analytical methods were inadequate in highlighting their presence. It should be clear that great care must be taken in the characterization of sam ples prepared under hydrothermal conditions. While it is obviously a highly valuable synthetic tool, especially for obtaining good quality single crystals for structural de termination and by offering an alternative synthetic route to synthesizing unique materials, one must be aware that trace amounts of nanoparticulate impurities may also be generated as a result of the thermal treatment. Although hydrothermal reactions were used to obtain single crystals of several coordination polymers reported in this thesis, the physical properties of these polymers were always determined on samples prepared at room temperature and, thus, free of nanoparticles. Observing coordination polymer samples by electron microscopy is not a common practice in the field, but in this case it was invaluable to determine the full nature of the material, thereby avoiding attributing the magnetic properties of the sample as intrinsic properties of the K{Ni[Au(CN)2]3} coordination polymer. Although this narrative has revolved around the discussion of magnetic properties, the presence of amounts of nanoparticles small enough to be undetected by conven tional analytical methods may impact other physical properties of a sample (such as conductivity, dielectric constant, optical or gas adsorption properties) which are often the motivation behind the synthesis of coordination polymers. In short, such nanoscale materials often have physical properties that could inadvertently masquer ade as properties of the overlying coordination polymer product. As such, it is crucial to consider this possibility when interpreting data, whether it is from the laboratory Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 235 or from published literature. Finally, this work confirms that, given the right condi tions, the hydrothermal treatment of cyanometallate polymers can generate a range of nanoparticle materials.

5.5 Experimental Section

The general procedures concerning the synthesis and characterization of the complexes presented in this chapter are as described in Chapter 2, section 2.7.1, unless otherwise noted.

5.5.1 Hydrothermal synthesis of K{Ni[Au(CN)2]3} at 125 °C

The preparation of K{Ni[Au(CN)2J3} under hydrothermal conditions (125 °C) is re ported in Chapter 4, section 4.6.1. The bulk sample obtained by this reaction will be referred to as H-125.

5.5.2 Hydrothermal synthesis of K{Ni[Au(CN)2]3} at 135 and 165 °C

For each reaction, a 1 mL aqueous solution of K[Au(CN)2] (ca. 0.114 g, 0.396 mmol) was combined with a 1 mL aqueous solution of Ni(N03)2-6H20 (ca. 0.029 g, 0.10 mmol) in a 5 mL ampoule. Water was added to bring the total volume to ~ 3 mL. The ampoule was sealed and loaded into the reaction vessel. In the first reaction (135 °C), the vessel was heated in a furnace from 25 to 135 °C over a period of two hours, maintained at 135 °C for 65 hours, and then rapidly cooled back to 25 °C at a rate of 55 °C per hour to generate the bulk sample H-135. In the second reaction (165 °C), the maximum temperature of 165 °C was reached after two hours, the temperature was maintained for 6 hours and then slowly cooled at a rate of 1.4 °C per hour to generate the bulk sample H-165. Chapter 5. Hydrothermal synthesis and related nanoparticle impurities 236

5.5.3 Control experiment with K[Au(CN)2]

A 3 mL aqueous solution of K[Au(CN)2] (0.094 g, 0.33 mmol) was sealed in a 5 mL ampoule, which was loaded into the reaction vessel. The vessel was heated in a furnace from 25 to 125 °C over a period of two hours, maintained at 125 °C for 6 hours, and then slowly cooled back to 25 °C at a rate of 1.4 °C per hour.

5.5.4 SQUID Magnetometry

Direct current (dc) magnetization was measured upon cooling from 300 to 1.8 K under an applied dc field of 1 kOe. Zero-field cooled (ZFC) and field cooled (FC) magnetization measurements were performed upon warming from 1.8 K to 300 K under different external fields (1, 10, 50, 100 and 300 Oe). For the ZFC measurements, the magnetic field was reset to zero before the sample was cooled below 100 K. For the FC measurements, the sample was cooled from 100 K in the same magnetic field as the one used for the measurements performed upon warming. The alternating current (ac) susceptibility of both samples was also determined in zero applied dc field as a function of temperature (100 to 1.8 K). The amplitude and frequency of the ac field were 5 Oe and 10.0 Hz, respectively.

5.5.5 Transmission electron microscopy

Samples were prepared by evaporating an aqueous suspension of each sample onto a carbon-coated copper grid. Imaging was carried out by Simon Trudel (Simon Fraser University) using a FEI Tecnai 20 scanning transmission electron microscope (STEM) operating at 200 kV and equipped with a CCD camera. Bright field (BF) and high- resolution (HR) images, as well as selected area electron diffractograms (SAED), were acquired in TEM mode. High-angle annular dark field (HAADF) and energy dispersive X-ray spectroscopy (EDXS) was carried out in STEM mode. EDXS spectra were acquired using an EDAX unit, for a duration of 60 seconds. Analysis of the images was carried out using ImageJ software. (262*> Chapter 6

Global conclusions and perspectives

The goal of this thesis was to examine the preparation of capping ligand-free coor dination polymers containing the [Au(CN)2]~ building block. A variety of structural motifs have been identified for these coordination polymers and the resulting physical properties were investigated.

Structural motifs

Reactions performed in donor solvents in the presence of two equivalents of [Au(CN)2]~ units per metal center yielded the formation of coordination poly mers of the form M[Au(CN)2]2(analyte)a; (analyte = water, dimethylsulfoxide, N,N- dimethylformamide, pyridine, ammonia; M — Cu, Ni, Co). In these 1-D or 2-D systems, solvent molecules are incorporated and coordinate to the transition metal ions. In the M(/i-OH2)2[Au(CN)2]2 polymers, an uncommon double-aqua bridge mo tif, which leads to the formation of chains, was identified.

In the presence of three equivalents of [Au(CN)2]~ units per metal center, no + solvent molecules were incorporated and the [cation]{M[Au(CN)2]3} (cation = K , + n + [PPN] , [ Bu4N] ; M = Ni, Co) coordination polymers were obtained. In these 3-D

237 Chapter 6. Global conclusions and perspectives 238 polymers, the countercation present was found to play an important structural role.

Optical properties, vapochromism and sensing applications

Amongst the properties investigated, the Cu[Au(CN)2]2(analyte);r system was shown to be vapochromic. Exchange of analyte molecules was found to be possible, in the solid-state and under ambient conditions. Identification of the analyte present is possible due to the specific color and different infrared signature in the cyanide stretching frequency region for each Cu[Au(CN)2]2(analyte)a; polymer. Vapochromic materials capable of sensing environmentally hazardous vapours and gases in very low concentrations are of great interest. ^ The M[Au(CN)2]2(analyte)rr systems were not tested toward analytes such as CO, S02 or NO^ and this should be the subject of further investigations. As mentioned in Chapter 3, modification of the systems could also be attempted to optimize the response (time required, concentra tion detected and output) and broaden the types of analyte detected.

Although several [Au(CN)2]-containing compounds are luminescent due to the presence of aurophilic interactions, (56>263>264) preliminary studies have shown that none of the coordination polymers presented in this thesis were significantly luminescent.

Incorporation of metal centers such as Zn(ll) in the M[Au(CN)2]2(analyte)3; systems to generate vapoluminescent materials would especially be worth pursuing. The dif ferent impact of each analyte on the framework structure would likely modify the luminescence of the aurophilic interactions, which could allow analyte identification. The vapour sensing experiments performed in this thesis were done on bulk sam ples. Diffusion of analyte vapour through the materials was most likely a determining factor in the response time observed. In principle, the larger the fraction of converted material is, the more pronounced the change in colour should be and it should be easier to identify the analyte present.

In that respect, the preparation of films of M[Au(CN)2]2(analyte);r should improve the response time as the ratio of surface to bulk is greatly increased. In addition, the integration of the coordination polymer into a device would be easier as a thin film than as a powder. The use of smaller amounts of material (especially gold-containing Chapter 6. Global conclusions and perspectives 239 materials) is also an advantage with respect to cost.

Magnetism

The bridging [Au(CN)2]~ unit was found to be a very poor mediator of magnetic interactions, irrespective of whether metal centers were connected in two or three di mensions. The lack of interaction mediated by this diamagnetic bridge contrasts with 4 what had been observed with the diamagnetic low spin [Fe(CN)6] ~ building block, which mediates ferromagnetic interactions (leading to long-range order) in Prussian Blue.^31)

The [Au(CN)2]~ unit was proven to be useful to increase the framework connectiv ity, but a better mediator of magnetic exchange is required to obtain a material that magnetically orders. To achieve this goal, an additional bridging building block should be incorporated in the coordination polymer framework. The M(/J-OH2)2[AU(CN)2]2 coordination polymers illustrate that point. The water molecules in these polymers act as bridging ligands and mediate magnetic interactions between the metal centers, leading to long-range magnetic order in the Cu(ll)-containing system, although at low temperature. As competition could occur between an anionic bridging building block and the [Au(CN)2]~ unit, incorporation of neutral bridging ligands should be targeted first. Ligands such as pyrazine, 2,2'-bipyrimidine and 1,2,4-triazole, which are known to mediate magnetic interactions between metal centers, should be examined.

Other magnetic applications, not requiring magnetic order, for which [Au(CN)2]- based coordination polymers would be more appropriate include the development of spin-crossover polymers. Several Fe(ll)-containing molecular complexes are known to undergo a high-spin state (HS, S = 2) to low-spin state (LS, S = 0) transition upon cooling. (265'266) In order for such a transition to occur, the crystal field splitting energy has to be very close to the pairing energy, and such a situation is usually observed in complexes containing an [Fe(N)6] coordination sphere. This spin-state transition changes the magnetic properties drastically, along with the color and, to some extent, the first coordination shell bond length. In some complexes, the transition is sharp and even hysteretic (THS^LS ¥" TLS^HS), allowing two discrete states to exist at a Chapter 6. Global conclusions and perspectives 240 given temperature. Some spin-crossover polymers have been studied and it was found that the structural connectivity increases the spin-transition temperature and the temper ature difference between the two reverse transitions (i.e. spin-transition hys teresis), increasing the temperature range over which both the HS and LS states can be obtained. A few examples of [M(CN)2]-based coordination poly mers (M = Ag, Au), such as [Fe(pyrimidine)-(H20)(M(CN)2)2]-H20 and [Fe(3-

CNpyridine)2(CH3OH)2/3[Au(CN)2]2], with transition temperatures up to 223 K, have been reported. (267>268) Modifications of the systems by incorporating other types of ligands are worth pursuing. To extend the work presented in this thesis on vapochromic sensors, the devel opment of an Fe[Au(CN)2]2(analyte)a; system that undergoes a spin-transition upon analyte exchange should be attempted. It was shown that the crystal field splitting energy was affected by the analyte molecules surrounding the metal center, and the same principle should be transferable to Fe(ll)-containing polymers. For example, the

Fe(ii) centers in Fe(^-OH2)2[Au(CN)2]2 were found to be high-spin (with [Fe(N)2(0)4] coordination sphere), but upon replacement of the water molecules by ammonia or pyridine molecules (which are stronger field ligands), a [Fe(N)6] coordination shell would be obtained, and a spin-state transition could be observed. In addition to a modification of the magnetic properties, an important color change should also ac company the spin-transition, which would allow analyte identification by more than one method.

To conclude

Even though capping ligands can be useful in obtaining coordination polymers with functionality, this research has clearly shown that they are not essential to obtain polymers with useful properties. In addition to the physical properties surveyed in this thesis, it is likely that capping ligand-free [Au(CN)2]-containing coordination polymers will display other unique behaviours of interest, including porosity for gas storage applications. Appendix A

Summary of Crystallographic Data

A.l Fractional atomic coordinates for the M(/z-OH2)2[Au(CN)2]2 coordination polymers

Table A.l: Fractional atomic coordinates and equivalent isotropic thermal parameters 2 (U(iso) in A ) for Ni(//-OH2)2[Au(CN)2]2

2 Atom X y z U(iso) (A ) Occupancy Type Au(l) 0.0000 1.0000 0.7500 0.0293 1.0000 Uani Ni(l) 0.0000 1.0000 1.0000 0.0185 1.0000 Uani N(l) 0.0000 1.0000 0.901845(7) 0.0296(16) 1.0000 Uiso C(l) 0.0000 1.0000 0.846518(5) 0.0266(16) 1.0000 Uiso N(ll) 0.0000 1.0000 0.598201(7) 0.0297(16) 1.0000 Uiso C(ll) 0.0000 1.0000 0.653529(5) 0.0266(16) 1.0000 Uiso 0(1) 0.216(3) 0.5000 1.0000 0.020(3) 1.0000 Uiso H(H) 0.3002 0.5000 1.0370 0.020(3) 1.0000 Uiso

241 Appendix A. Summary of Crystallographic Data 242

'/t-

I WOA>*«JAJw^ uLx.

I I I i r I I I i I I I 1 '* 1.5 2.0 2.5 3,5 4.0 4.5 5.0 5.5 6.0 10.0 10.5

Figure A.l: Comparison of the powder diffractogram obtained experimentally for Co(/i-OH2)2[Au(CN)2]2 (i., purple) and the diffractogram predicted by the proposed structural model based on the structure of Ni(/i-OH2)2[Au(CN)2]2 (ii-, black).

1 W^^_^_JVA_A'\A_X.

Figure A.2: Comparison of the powder diffractogram obtained experimentally for Fe(//-OH2)2[Au(CN)2]2 (L, green) and the diffractogram predicted by the proposed structural model based on the structure of Ni(/u-OH2)2[Au(CN)2]2 (ii., black). Appendix A. Summary of Crystallographic Data 243

...... ii. u ILLJLUL JL -j—•—i—•—L_J—t i i 1 i— _J I I I L_l I L_l I I I I YA-L_/ • • < 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 10.0 10.5 «UA]

Figure A.3: Comparison of the powder diffractogram obtained experimentally for

Mn(/x-OH2)2[Au(CN)2]2 (i-, red) and the diffractogram predicted by the proposed structural model based on the structure of Ni(//-OH2)2[Au(CN)2]2 (ii-, black).

1.5 2.0 2.5 3.0 10.0 10.5

Figure A.4: Comparison of the powder diffractogram obtained experimentally for Fe(/i-OH2)(^-OH)[Au(CN)2]2 (i-, green) and the diffractogram predicted by the pro posed structural model based on the structure of Ni(/x-OH2)2[Au(CN)2]2 in which the transition metal and oxygen atoms have a half occupance (ii., black). Appendix A. Summary of Crystallographic Data 244

Table A.2: Fractional atomic coordinates for Cu(/i-0H2)2[Au(CN)2]2.

Atom X y z Occupancy Au(l) 0.000 0.250 0.000 1 Au(2) 0.500 0.250 0.500 1 Cu(l) 0.000 0.000 0.000 1 Cu(2) 0.500 0.500 0.500 1 N(l) 0.000 0.903 0.000 1 N(2) 0.000 0.403 0.000 1 N(3) 0.500 0.903 0.500 1 N(4) 0.500 0.403 0.500 1 C(l) 0.000 0.848 0.000 1 C(2) 0.000 0.348 0.000 1 C(3) 0.500 0.848 0.500 1 C(4) 0.500 0.348 0.500 1 0(1) 0.219 0.000 0.500 1 0(2) 0.719 0.500 1.000 1

Table A.3: Fractional atomic coordinates for Co(/i-OH2)2[Au(CN)2]2.

Atom X y z Occupancy Au(l) 0.0000 0.2500 0.0000 1 Au(2) 0.5000 0.7500 0.5000 1 Co(l) 0.0000 0.0000 0.0000 1 Co(2) 0.5000 0.5000 0.5000 1 C(l) 0.0000 0.8475 0.0000 1 N(l) 0.0000 0.0972 0.0000 1 C(2) 0.0000 0.6524 0.0000 1 N(2) 0.0000 0.5972 0.0000 1 C(3) 0.5000 0.1525 0.5000 1 N(3) 0.5000 0.0972 0.5000 1 C(4) 0.5000 0.6524 0.5000 1 N(4) 0.5000 0.5972 0.5000 1 0(1) 0.2403 0.0000 0.5330 1 0(2) 0.7392 0.5000 0.9426 1 Appendix A. Summary of Crystallographic Data

Table A.4: Fractional atomic coordinates for Fe(/>OH2)2[Au(CN)2]2.

Atom X y z Occupancy Au(l) 0.0000 0.2456 0.0000 1 Au(2) 0.5000 0.2510 0.5000 1 Fe(l) 0.0000 0.0000 0.0000 1 Fe(2) 0.5000 0.5000 0.5000 1 C(l) 0.0000 0.8475 0.0000 N(l) 0.0000 0.0972 0.0000 C(2) 0.0000 0.6524 0.0000 N(2) 0.0000 0.5972 0.0000 C(3) 0.5000 0.1525 0.5000 N(3) 0.5000 0.0972 0.5000 C(4) 0.5000 0.6524 0.5000 N(4) 0.5000 0.5972 0.5000 0(1) 0.2193 0.0000 0.5000 1 0(2) 0.7193 0.5000 1.0000 1

Table A.5: Fractional atomic coordinates for Mn(/i-0H2)2[Au(CN)2]2.

Atom X y z Occupancy Au(l) 0.0000 0.2500 0.0000 1 Au(2) 0.5000 0.2500 0.5000 1 Fe(l) 0.0000 0.0000 0.0000 1 Fe(2) 0.5000 0.5000 0.5000 1 C(l) 0.0000 0.8475 0.0000 N(l) 0.0000 0.9028 0.0000 C(2) 0.0000 0.3476 0.0000 N(2) 0.0000 0.4028 0.0000 C(3) 0.5000 0.8475 0.5000 N(3) 0.5000 0.9028 0.5000 C(4) 0.5000 0.3476 0.5000 N(4) 0.5000 0.4028 0.5000 0(1) 0.8207 0.0000 0.5000 1 0(2) 0.3357 0.5000 1.0000 1 Appendix A. Summary of Crystallographic Data 246

Table A.6: Fractional atomic coordinates for Fe(//-OH2)(/>OH)[Au(CN)2]2-

Atom x y z Occupancy Au(l) 0.0000 0.0000 0.5000 1 Fe(l) 0.0000 0.0000 0.0000 0.5 C(l) 0.0000 0.0000 0.7000 1 N(l) 0.0000 0.0000 0.8200 1 0(1) 0.3000 0.0000 0.0000 0.5 Appendix A. Summary of Crystallographic Data 247

A.2 Fractional atomic coordinates for the

M [Au(CN)2] 2 (analyte)x coordination polymers

Table A.7: Fractional atomic coordinates and equi valent isotropic thermal parameters (U(iso) in A2) for

Cu[Au(CN)2]2(DMSO)2 (blue)

2 Atom X y z U(iso) (A ) Occupancy Type Au(l) 0.1300(3) 0.2948(2) 0.32205(13) 0.0406 1.0000 Uani Au(2) -0.0372(3) 0.4698(2) 0.76578(13) 0.0396 1.0000 Uani Au(3) -0.9037(3) 0.2102(2) 0.83479(13) 0.0419 1.0000 Uani Au(4) 0.9841(3) 0.0247(2) 0.27305(13) 0.0419 1.0000 Uani Cu(l) -0.3607(8) 0.2401(6) 0.5909(4) 0.0302 1.0000 Uani Cu(2) 0.6505(8) 0.2626(5) 0.0956(4) 0.0277 1.0000 Uani S(l) -0.6553(14) 0.416(1) 0.6151(7) 0.0394 1.0000 Uani S(2) -0.5844(16) 0.1695(11) 0.4605(7) 0.0412 1.0000 Uani S(3) 0.4128(15) 0.333(1) -0.0339(8) 0.0363 1.0000 Uani S(4) 0.3425(17) 0.0754(11) 0.1159(7) 0.0432 1.0000 Uani 0(1) -0.501(4) 0.364(2) 0.5584(17) 0.036(8) 1.0000 Uiso 0(2) -0.454(4) 0.141(2) 0.5195(17) 0.033(8) 1.0000 Uiso 0(3) 0.540(4) 0.362(3) 0.0246(19) 0.048(9) 1.0000 Uiso 0(4) 0.491(4) 0.125(3) 0.0625(18) 0.040(9) 1.0000 Uiso N(ll) -0.159(5) 0.283(3) 0.470(2) 0.04(1) 1.0000 Uiso N(12) 0.450(5) 0.298(3) 0.192(2) 0.046(11) 1.0000 Uiso N(21) -0.262(5) 0.341(3) 0.662(2) 0.04(1) 1.0000 Uiso N(22) 0.820(6) 0.399(4) 0.136(3) 0.061(13) 1.0000 Uiso N(31) -0.590(6) 0.197(4) 0.700(3) 0.054(13) 1.0000 Uiso N(32) 0.820(6) 0.234(4) -0.008(3) 0.058(13) 1.0000 Uiso Continued on next page 00 CD CM o o o 0 0 0 0 O 0 O 0 0 O O O O 0 O O 0 0 0 O O O O 0 CO CO CO CO CO CO a CO CO CO CO co CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 03 & P ^ p P & p £> Z> £ p ^ & P P £3 £3 & £3 z> p i^ Z> & P & ^ Q, H So

PI o o 0 O 0 0 0 O O 0 0 0 O O 0 0 O 0 0 0 O O 0 O 0 0 CI o o 0 O 0 0 0 O O 0 0 0 O O 0 0 O 0 0 0 O O 0 O 0 0 o o 0 O 0 0 0 O O 0 0 0 O O 0 0 O 0 0 0 O O 0 O 0 0 d

000 0 0 0 O O 0 O 0 0 o pi o o O 0 0 0 O O 0 0 0 O O 0 0 O 0 0 t3 o 1—1 I—1 7—1 I—1 I—1 T—1 1—t 1—1 1—1 T—1 T—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 1—1 r-H CD CD 3 O •S CD +3 toO d ce co" CO" CM" ^?" I—1 LO" 00" 00" T—1 T—1 CM" P_ ~< to" to" 1—1 1—1 a 1—1 T—1 T—1 T—1 I—1 T—1 T—1 1—1 i—l 1—1 1—1 LO LO LO LO LO LO LO LO LO CO CM" CM" T—1 pi CM b- LO Ci LO C2 00 O 1—1 0 O O 00 OO 00 1—1 1—1 i—l o o 00 co 1—1 co O CM 1—( LO CO CO CM CO LO CM 0 1—1 r-H 1—1 0 O i—l 1—1 LO LO 0 0 co 0 i-H O O O O O 0 O o o 0 O 0 0 0 0 0 O O 0 O O O > 0.029(9 ) 0 0 0 O CD O O O O O 0 O S-i o o 0 O 0 0 0 o,

Data CO co

fro m co CO CO co CM co co co CM CO CO^ CO CO co co co co CO co CO^ co T—1 00" to" c5" 0O" u CM to" T—1 06" 06" to" LO 1—1 1—1 b^ b^t b^ co LO 06" 00" TJ O b- co 00 O -CJ CD ^ CO co CO CM co 1—1 CO co CO O b- co LO LO O T—1 LO 626(2 ) T—1 LO CO co O T—1 CM CO T—1 b- OS CO CM LO LO CO CO LO LO Pi 0 O 0 O O O 0 0 O O O O O O O 0 O 0 0 0 0 0 O O 8- o o o 1 1 & ont i

s3 o SO LO ?3 co LO LO LO co co co co b- 06" tS tS 5^ co" ts 06" co" 06" T—1 CM 7—1 to" b^ CM LO b^ CM co CM" co" CM to" 1—1 00" CM 1—1 h ?3> 0 LO co <^ LO LO co O Oi O 00 00 O CM CO 1—1 00 00 CO CO CO 0

126(3 ) 0 3 T—1 I—1 O O CM CM CO CM CM 0 1—1 co LO LO LO 1—1 0 1—1 <+H CD Q o 0 O O 0 O O 0 0 O O O O O 0 0 0 0 0 0 0 0 0 0 0 rO o o 1 ^ 03 03 H

m to" to" to" to" to" to" P" b^" to" 00" 00" "P" to" to" to" b^ 02" CO b^ CM b^ 1—1 CM CM '0' CO 1—1 1—1 a3 to" 00" CM co LO to" to" co b- "co" oo T—1 b- T—1 1—1 H CO 0 CO co 0 r-1 LO co O b- CM CO CO LO co co CO LO b- r—l 1—1 b- LO 00 b- 00 b- 00 LO CM CM 0 co b- O 00 00 00 CO 0 1—J 0 0 0 CO 0 0 0 0 0 O 0 O O O 0 0 O O O O O 0 O -0.208(5 ) o 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 < 1 X •X} C3 i—i CM" 1—1 CM i—l CM 1—1 CM 1—1 CM 1—1 CM CM i—l CM co 1—1 CM co CD O T—1 CM CO ^ LO CO b- 00 1—1 T—1 CM CM co co 1—1 1—1 1—1 CM CM CM co co co a, OH 3 iz; Z 0 0 0 0 O 0 O 0 0 O O O O O O O tn X X ffl ffi ffi ffi ffi ffi

Table A. 7 - continued from previous page

2 Atom X y U(iso) (A ) Occupancy Type H(41) -0.415(7 ) 0.120(4) 0.338(3) 0.11(5) 1.0000 Uiso H(42) -0.495(7 ) 0.022(4) 0.383(3) 0.11(5) 1.0000 Uiso H(43) -0.605(7 ) 0.095(4) 0.333(3) 0.11(5) 1.0000 Uiso H(51) 0.181(7) 0.377(4) 0.049(3) 0.10(5) 1.0000 Uiso H(52) 0.269(7) 0.477(4) 0.011(3) 0.10(5) 1.0000 Uiso H(53) 0.164(7) 0.409(4) -0.042(3) 0.10(5) 1.0000 Uiso H(61) 0.595(7) 0.369(4) -0.153(3) 0.10(5) 1.0000 Uiso H(62) 0.513(7) 0.470(4) -0.110(3) 0.10(5) 1.0000 Uiso H(63) 0.408(7) 0.402(4) -0.163(3) 0.10(5) 1.0000 Uiso H(71) 0.148(5) 0.106(4) 0.029(2) 0.08(5) 1.0000 Uiso H(72) 0.264(5) 0.011(4) -0.003(2) 0.08(5) 1.0000 Uiso H(73) 0.119(5) -0.003(4) 0.071(2) 0.08(5) 1.0000 Uiso H(81) 0.499(8) -0.053(5) 0.169(4) 0.14(5) 1.0000 Uiso H(82) 0.471(8) -0.085(5) 0.080(4) 0.14(5) 1.0000 Uiso H(83) 0.325(8) -0.099(5) 0.153(4) 0.14(5) 1.0000 Uiso CO o S-4 LO 0? o o o -4-= Pi Pi Pi Pi CM Pi Pi Pi Pi Pi Pi Pi Pi Pi 172 c? c? g CD cC c3 cd cd cd cd cd cd cd cd cd a cd Jii •ii .S £> p p p ^ & P P ;-> P p p p \5 \D Z) Z> cd 5 P S-l cd o o o o o o o o o o o o o o o o o o o 13 o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o S-l o o o o o o o o o o o o o o o CD o o o o i—1 i—l ^H 1—t 1—1 ^H 1—1 I—1 ^H 1—1 1—1 1—1 1—1 l—l l—l l—l l—l 1—1 l—l 4-3

o (9 ) (9 ) S-H (9 ) (9 ) CM 05 t- 4^> 00 t— l>- oo 05 oo t- co oo ^ to^to^ O XjH >>- t- Tt< 02 1—1 t> 1—1 CM t- CM CO t- LO oo

c3 40 4 CD - co ^H t^ o ^ r~ OT> o ^ o 5% <£> CM o ^f t- CO co -tf r~ Oi ->cf 1—1 CM ^f LO 00 CO • L48 4 LO 1—1 CM 1—1 1—i 00 00 i-i O 1—1 oo o o o o I LO C5 CO C5 CM 1—1 1—1 l—l CO 1328 8 ^ *? L-071 6 1—1 1—5 i—i l—l 1—1 1 ! 1 1—1 I—1 1 I—1 l—l 1 1 1 1—1 1 r-5 o o 1 o o o o 1 i 1 1 1 4^ 1 1 1 1 1 1 cd Id pi s 1—\ Pi ^ •p ^f 05 ^f ^ ^ 05 05 OJ a o u o o o i-H o o 1—1 1—1 l—l l—l l—l l—l T—l l—l I—1 -.—I o o o CM o o o o 4-=> (-1 o o LO o oo l—l (M (M t~ -tf 1-1 CO H o o o o o ^ o LOO CLMO LO LO LO o LO LO CI o 03 o oo l—l 00 LO CM cd !643 l ^ 1—1 LO 05 LO CM ^f K.N •>* ^f o o o o o CO o o o o o co 1—1 CM o l—l o CM CN CO •< o o o o o o o o o o o

~~< r^TcsT 1—1 r-l CM co 1—1 "pf^ ^ s ^ ^ -^ < <: o ~ZD cf izf 2f 2f o o o o o" ffi' K XX x*x Appendix A. Summary of Crystallographic Data 251~~

~~Table A.9: Fractional atomic coordinates and equivalent isotropic thermal parameters 2 (U(iso) in A ) for Cu[Au(CN)2]2(DMF)~~

Atom x y U(iso) Occupancy Type (A2) Au(l) -0.38582(5) 0.02052(4) 0.70548(4) 0.0331 1.0000 Uani Au(2) -0.37817(5) -0.51955(4) 0.24851(4) 0.0346 1.0000 Uani Cu(l) -0.40687(15) -0.25013(11) 0.47374(12) 0.0243 1.0000 Uani 0(1) -0.5767(9) -0.2616(7) 0.4329(9) 0.0562 1.0000 Uani N(l) -0.3810(9) 0.1702(7) 0.8660(8) 0.0334 1.0000 Uani N(2) -0.4082(10) -0.1401(7) 0.5586(8) 0.0375 1.0000 Uani N(3) -0.3769(10) -0.3570(7) 0.3953(8) 0.0348 1.0000 Uani N(4) -0.3972(9) -0.6692(7) 0.0886(8) 0.0331 1.0000 Uani N(5) -0.7304(12) -0.3402(9) 0.4185(10) 0.0510 1.0000 Uani C(l) -0.3779(11) 0.1153(10) 0.8054(10) 0.0347 1.0000 Uani C(2) -0.3987(13) -0.0798(10) 0.6116(11) 0.0435 1.0000 Uani C(3) -0.3732(12) -0.4197(10) 0.3437(10) 0.0402 1.0000 Uani C(4) -0.3926(13) -0.6141(10) 0.1479(11) 0.0448 1.0000 Uani C(5) -0.6315(15) -0.3236(10) 0.4552(11) 0.0459 1.0000 Uani C(6) -0.7852(15) -0.4153(11) 0.4559(14) 0.0716 1.0000 Uani C(7) -0.7806(15) -0.2846(13) 0.3430(14) 0.0685 1.0000 Uani H(51) -0.6018(15) -0.3648(10) 0.5034(11) 0.12(9) 1.0000 Uiso H(61) -0.8540(15) -0.4187(11) 0.4235(14) 0.22(7) 1.0000 Uiso H(62) -0.7891(15) -0.4062(11) 0.5227(14) 0.22(7) 1.0000 Uiso H(63) -0.7489(15) -0.4711(11) 0.4464(14) 0.22(7) 1.0000 Uiso H(71) -0.8500(15) -0.3061(13) 0.3264(14) 0.22(7) 1.0000 Uiso H(72) -0.7825(15) -0.2222(13) 0.3633(14) 0.22(7) 1.0000 Uiso H(73) -0.7424(15) -0.2888(13) 0.2885(14) 0.22(7) 1.0000 Uiso Appendix A. Summary of Crystallographic Data 252

~~Table A. 10: Fractional atomic coordinates and equivalent isotropic thermal para 2 meters (U(iso) in A ) for Cu[Au(CN)2]2(pyridine)2~~

2 Atom x y U(iso) (A ) Occupancy Type Au(l) 0.95611(6) 0.24833(3) 0.73311(4) 0.0354 1.0000 Uani Cu(l) 0.5000 0.0000 0.5000 0.0363 1.0000 Uani N(l 0.6612(13) 0.1078(6) 0.5822(9) 0.0387 1.0000 Uani N(2 1.2640(14) 0.3812(7) 0.8907(11) 0.0483 1.0000 Uani N(3 0.3841(11) -0.0036(5) 0.7125(8) 0.0357 1.0000 Uani C(l 0.7730(15) 0.1592(7) 0.6361(10) 0.0373 1.0000 Uani C(2 1.1473(16) 0.3341(8) 0.8331(12) 0.0472 1.0000 Uani C(3 0.3118(14) 0.0743(7) 0.7756(12) 0.0422 1.0000 Uani C(4 0.2384(16) 0.0731(8) 0.9245(13) 0.0509 1.0000 Uani C(5 0.2364(14) -0.0079(8) 1.0123(11) 0.0448 1.0000 Uani C(6 0.3143(15) -0.0868(8) 0.9492(12) 0.0412 1.0000 Uani C(7 0.3835(15) -0.0834(8) .7967(13) 0.0476 1.0000 Uani H(31) 0.3131 0.1320 0.7166 0.057(16) 1.0000 Uiso H(41) 0.1862 0.1291 0.9649 0.067(16) 1.0000 Uiso H(51) 0.1846 -0.0099 1.1143 0.059(16) 1.0000 Uiso H(61) 0.3201 -0.1441 1.0096 0.055(16) 1.0000 Uiso H(71) 0.4321 -0.1394 0.7528 0.063(16) 1.0000 Uiso Appendix A. Summary of Crystallographic Data 253

~~Table A.ll: Fractional atomic coordinates for Ni[Au(CN)2]2(DMSO)2~~

Atom X y z Occupancy Au(l) 2.1066 0.7873 0.6662 1 Au(2) 1.9793 0.9854 0.2306 1 Ni(l) 1.6355 0.7440 0.4026 1 S(l) 1.3400 0.9250 0.3844 1 S(2) 1.4014 0.6647 0.5340 1 0(1) 1.4898 0.8707 0.4380 1 0(2) 1.5259 0.6385 0.4749 1 N(l) 1.8198 0.7738 0.5121 1 N(2) 2.4262 0.7892 0.8037 1 N(3) 1.7592 0.8520 0.3316 1 N(4) 2.2114 1.1261 0.1381 1 C(l) 1.9216 0.7785 0.5701 C(2) 2.3077 0.7924 0.7558 C(3) 1.8354 0.9002 0.2938 C(4) 2.1243 1.0752 0.1712 C(5) 1.1953 0.9591 0.4545 C(6) 1.4162 1.0323 0.3670 C(7) 1.2193 0.5867 0.5006 C(8) 1.4841 0.5976 0.6255 Appendix A. Summary of Crystallographic Data

~~Table A.12: Fractional atomic coordinates for Ni[Au(CN)2]2(DMF)2~~

Atom X y z Occupancy Au(l) -0.0290 0.2259 0.2354 1 Au(2) 0.5052 -0.2393 0.1372 1 Ni(l) 0.2379 -0.0059 0.1894 1 0(1) 0.2094 0.0332 -0.0446 1 0(2) 0.2608 -0.0527 0.4213 1 N(l) 0.1293 0.0807 0.2384 1 N(2) 0.3506 -0.0901 0.1418 1 N(3) 0.3269 0.1077 0.2324 1 N(4) 0.1496 -0.1199 0.1418 1 N(5) 0.1777 0.1241 -0.2558 1 N(6) 0.3459 -0.0797 0.6479 1 C(l) 0.0684 0.1315 0.2393 1 C(2) 0.4087 -0.1450 0.1370 1 C(3) 0.3834 0.1651 0.2567 1 C(4) 0.0920 -0.1766 0.1284 1 C(5) 0.2229 0.1064 -0.1237 1 C(6) 0.1043 0.0645 -0.3221 1 C(7) 0.2045 0.2042 -0.3506 1 C(8) 0.3271 -0.0630 0.5134 1 C(9) 0.2902 -0.1570 0.6986 1 C(10) 0.4228 -0.0485 0.7489 1 Appendix A. Summary of Crystallographic Data

~~Table A.13: Fractional atomic coordinates for Co[Au(CN)2]2(pyridme)2~~

Atom x y z Occupancy Au(l) -0.0358 0.2188 0.2471 1 Au(2) 0.5018 -0.2322 0.1489 1 Co(l) 0.2447 0.0012 0.1777 1 N(10) 0.2144 0.0292 -0.0735 1 N(H) 0.2715 -0.0197 0.3744 1 N(l) 0.1293 0.0807 0.2384 1 N(2) 0.3506 -0.0901 0.1418 1 N(3) 0.3269 0.1077 0.2324 1 N(4) 0.1496 -0.1199 0.1418 1 C(l) 0.0684 0.1315 0.2393 1 C(2) 0.4087 -0.1450 0.1370 1 C(3) 0.3834 0.1651 0.2567 1 C(4) 0.0920 -0.1766 0.1284 1 C(10) 0.2514 -0.1028 0.4348 1 C(ll) 0.2740 -0.1213 0.5877 1 C(12) 0.3174 -0.0552 0.6829 1 C(13) 0.3398 0.0287 0.6202 1 C(14) 0.3141 0.0458 0.4657 1 C(20) -0.3245 -0.4627 -0.1702 1 C(21) -0.3411 -0.4777 -0.3247 1 C(22) -0.3187 -0.5613 -0.3841 1 C(23) -0.2770 -0.6283 -0.2852 1 C(24) -0.2632 -0.6114 -0.1295 1 ^ cicy o o o o o o _Q _o Ô ô Ô Ô O Ô ^ ^ ^ ^ ^ ^ ^ j2 ô o o o CO CO CO CO O O > > t3 1 1 ^ i— i— h-» 1—* 1 1 1 'tf^'co "to 1—' o o CD 'to'to 'to "to 'oo "as" 'cn ^ 'co 'to R 'to "To "to 'to H- I— i— H ^ 'co "to 1—' ) 1.3222 1 ) 1.1156 1 ) 0.5637 1 ) 0.8274 1 ^^ ^J 1 4^ CO to h-' 4^ CO to 1—I 4^ CO to I— ^ CO to 'to 1—» to i—i P H P g- 1 ^ O I— I—1 p p -1 1 1 1 1 t p 1 I— I— O o h-' >—' I— O o I— ^ p J- p p p p p p p b p p p *- to b Cn ^ 1—> o b Cn I—" CO OS 4^ Cn CO Cn OS O 4^ I-1 O bo b cn to1 co co co bo co to co to ^ 1 4^ o 00 O Cn b ~q CO O I— -J 00 b CO Cn CO CO CO Cn I— CTJ Oi Cn Cn > co co -a ot- to 1 o oo as cn co as CO i—» Cn CO 4^ Cn -

~~1 0.6764 1 0.8179 1 0.3775 1 I— 1.0822 1~~

0. 2 0. 4 0. 9 i—» 0. 7 0. 7 0. 4 0. 6 0. 5 0. 5 0. 8 0. 2 0. 5 0. 9 0. 6 0. 4 O 1. 0 1. 2 o p p o p p -1 I—1 I—1 rt- 3 t p b p b 4^ i—' O CO 4^- O J—» CO co I—1 05 o Cn 4^ 00 '-I-J o 1 1 oo as o Os 4^ K 00 C5 Cn OS Os i— 4^ OS h-> -j Cn Cl C5 4^ i— h-' CO Cn p co 1 to oo o to oo -a to to oo -a oo h-' as to Os I— oo co oo to CO as co OS co to C5 C5 t*^ to CO Ol Cn CO b O oo o «£ p O v"~~N ,-~—"v ^ -v ^"~--

~~i—» coo r I—1 0748 ( 9697 ( 6222 ( 5834 ( 5355 ( 6317 ( p o p o O 1324 ( p p p p p p765 5 ( O 7335 ( p O p75773(3 ) 74060(3 ) p p p p o p o p p p p o o o h-' O~~

48 2 45 6 43 2 C94 6 n Cn Cn 91 7 Cn 62 5 52 3 35 8 36 5 86 4 54 0 26 3 74 7 24 9 25 1 cn ^ OS 73 5 CO ^ bo ^ bo ^ as bo ^ b11 8 b bo ^ b Cn OS CO O b Cn O Cn 4^ Cn h-* o O ^ Cn cn 05 0 o a. rs 1 1 1 1 1 as as o t—' i— i— i— i—> i— i— i—1 i—1 o o P 00 OO i—1 h-' i—1 'to 'to

(6 ) co (8 ) (6 ) (9 ) (s ) U ) (6 ) (8 ) U ) (6 ) U ) to (6 ) Cn co co (6 ) (8 ) (9 ) (9 ) (s ) P to co co rt- ?o r^- Ca>O CO t?> O O o o o o o o o o o o o o o o oi-i p p o o O p p o o o o p p p p p p p o o o i—» p p O b4^ 4o^ Cn 4^ co b*^ cbo b4^ bOS bOi b-q cbo boo b05 bCT> cbo to co CO CO ObS b 0b0 ObS 0b0 0b0 b-j CO tl^ CO 4^ Cbn Cbn Cbn Cbn b4^ 4b^ Cbn b£» b b cbo cbo o O co oo oo oo o o 05 to oo co co 1 co oo Cn co to Os o as 4^ O CO as 4^ CO oo Cn 4^ i— i—' '^'4^ 'cn' CT>C O CO -j "CO ^j -a to 4^ O - 'cn ^ ^ 'cn ^ co to cn ~a 'w 'cn 'cn cn 'cn 'co co co H p O ^

\—> 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 o t—» I—' i—> i— h-' i—> h-' i— h-' i— h-' t— M i—•> i— h-» I— i— I— •—' i— i—' t—' i— I— i— t—' i—' i—» h-' H-'- h—' I— i— i— o o o O o o o o O o O o o o O o o o o O o o O o o o o O o b b o o o o o o o o o o o o o o O o o o o o o o o o o o o o O o o o o o o o o o o o o o o p o o o o o O o o o o o o o o o o o o o o o o o o o o o o o o 'a o o O o o o o o p o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o P CO o O a a c< d c| a c| a addaaaaa a C a a a c! a a a a a a C C C C adda CD CO - rz~ p p p p £• £• £;• p p p p p p p p CO CO CO ™ CO s* p p p p CO CO p p co P p p p p P p co co p p p P P p p p & & p o o o 8 O O B O O O g O O s to s s s s Cn OS Appendix A. Summary of Crystallographic Data 257

~~A.3 Fractional atomic coordinates for the~~

~~[cation]{M[Au(CN)2]3} coordination polymers~~

~~Table A. 15: Fractional atomic coordinates and equivalent isotropic thermal para 2 meters (U(iso) in A ) for K{Ni[Au(CN)2]3}~~

2 Atom X V z U(iso) (A ) Occupancy Type Au(l) 0.49576 0.50424(5) 0.0000 0.0125 1.0000 Uani Ni(l) 0.0000 0.0000 0.5000 0.0082 1.0000 Uani K(l) 0.6667 0.3333 0.5000 0.0159 1.0000 Uani N(l) 0.2439(11) 0.2453(11) 0.3364(5) 0.0133 1.0000 Uani C(l) 0.3378(13) 0.3458(13) 0.2153(6) 0.0137 1.0000 Uani Appendix A. Summary of Crystallographic Data 258

~~Table A. 16: Fractional atomic coordinates and equivalent isotropic thermal para 2 meters (U(iso) in A ) for [PPN]{Ni[Au(CN)2]3}~~

2 Atom X V z U(iso) (A ) Occupancy Type Au(l) 0.47104(3) 0.0000 0.2500 0.0433 1.0000 Uani Ni(l) 0.6667 0.3333 0.3333 0.0224 1.0000 Uani P(l) 0.6667 0.3333 0.13239(11) 0.0477 1.0000 Uani N(l) 0.5822(4) 0.2068(4) 0.2961(2) 0.0395 1.0000 Uani N(2) 0.6667 0.3333 0.0833 0.1024 1.0000 Uani C(l) 0.5407(5) 0.1329(5) 0.2792(2) 0.0419 1.0000 Uani C(2) 0.6694(7) 0.2249(6) 0.1524(3) 0.0542 1.0000 Uani C(3) 0.7292(8) 0.2322(8) 0.1860(4) 0.0759 1.0000 Uani C(4) 0.7291(12) 0.1478(12) 0.2008(6) 0.1257 1.0000 Uani C(5) 0.6679(14) 0.0546(12) 0.1809(7) 0.1260 1.0000 Uani C(6) 0.6084(15) 0.0502(11) 0.1483(6) 0.1137 1.0000 Uani C(7) 0.6114(10) 0.1346(9) 0.1339(4) 0.0874 1.0000 Uani H(31) 0.7671 0.2938 0.1996 0.08(2) 1.0000 Uiso H(41) 0.7728 0.1537 0.2225 0.13(2) 1.0000 Uiso H(51) 0.6665 -0.0023 0.1920 0.12(2) 1.0000 Uiso H(61) 0.5688 -0.0128 0.1362 0.14(2) 1.0000 Uiso H(71) 0.5725 0.1293 0.1104 0.10(2) 1.0000 Uiso Appendix A. Summary of Crystallographic Data 259

~~Table A. 17: Fractional atomic coordinates and equi valent isotropic thermal parameters (U(iso) in A2) for~~

~~["Bu4N]{Ni[Au(CN)2]3}~~

2 Atom x y U(iso) (A ) Occupancy Type Au(l) -0.13516(7) -0.36017(7) 0.2767(3) 0.0694 1.0000 Uani Au(2) 0.13446(8) -0.16087(8) -0.0046(3) 0.0807 1.0000 Uani Au(3) -0.10511(7) -0.40584(7) 0.0168(3) 0.0708 1.0000 Uani Ni(l) 0.0020(2) -0.25941(16) 0.1391(3) 0.0445 1.0000 Uani N(l -0.0525(11) -0.2901(11) 0.1985(11) 0.047(7) 1.0000 Uiso N(2 0.0574(13) -0.2310(13) 0.0759(14) 0.066(8) 1.0000 Uiso N(3 -0.0249(11) -0.3267(11) 0.0889(11) 0.053(7) 1.0000 Uiso N(4 0.0280(13) -0.1924(14) 0.1902(14) 0.073(9) 1.0000 Uiso N(5 0.0672(12) -0.3049(11) 0.1744(13) 0.053(7) 1.0000 Uiso N(6 -0.0588(13) -0.2144(13) 0.0995(14) 0.069(9) 1.0000 Uiso N(7 0.0259(12) -0.2804(13) 0.3897(13) 0.08(1) 1.0000 Uiso C(l) -0.0839(16) -0.3162(16) 0.2264(16) 0.066(11) 1.0000 Uiso C(2) 0.0866(14) -0.2091(14) 0.0427(15) 0.055(9) 1.0000 Uiso C(3) -0.0516(16) -0.3590(16) 0.0619(16) 0.07(1) 1.0000 Uiso C(4) 0.0525(15) -0.1600(16) 0.2165(15) 0.06(1) 1.0000 Uiso C(5) 0.1046(16) -0.3233(15) 0.1970(16) 0.06(1) 1.0000 Uiso C(6) -0.0892(16) -0.1839(16) 0.0696(16) 0.06(1) 1.0000 Uiso C(10) 0.0838(13) -0.2679(19) 0.4130(16) 0.087(13) 1.0000 Uiso C(ll) 0.1285(18) -0.252(2) 0.3670(17) 0.117(17) 1.0000 Uiso C(12) 0.189(2) -0.240(4) 0.393(3) 0.22(3) 1.0000 Uiso C(13) 0.220(3) -0.222(4) 0.342(4) 0.26(5) 1.0000 Uiso C(20) -0.0120(15) -0.3016(19) 0.4368(16) 0.104(16) 1.0000 Uiso C(21) -0.0770(15) -0.302(2) 0.426(2) 0.110(16) 1.0000 Uiso C(22) -0.100(2) -0.340(2) 0.476(2) 0.15(2) 1.0000 Uiso Continued on next page ^ ffi K w ffi ffi ffi ffi a tn ffi ffi w K a ffi ffi ffi ffi o o o Q Q Q O o o > ^ 1 1 1 1 1 bo to to to to bO to to to K I— h-> I— 1—1 I— I— I— I—1 4^ 4^ 4^ 4^ co CO co co to O t3 1 1 CO 1—1 I— CO bO I—1 1—> I—1 O CO bO i— 0) co co 1 to to o o co co to o o 1 co to o co CO bO I— bO I—1 I—1 bO 1—» bO 1—» bO h-' 1—1 bO H fcs to co to Q-

1 1 1 1 S © p p o p i > 1 1 O p p p © p © o o o o o o o o I— I— b b p o o p p p o p b b b bO M h-» 00 i—i 1—1 1—1 o o to to to co CO o 1—" b b 00 CO 1—» b b b b b b b CO ~'9(3 ) - J '3(3 ) OS ~J '6(2 ) CO '9(2 ) - a i8(15 ) Oi 3(15 ) - bO :7(15 ) - 0 4^ Oi5(15 ) - 0 3(3 ) - o to C7(3 ) n 07(2 ) - 0 o 05(18 ) - 5 8(18 ) - o3(13 ) - 0 ~1(13 ) - 0 J ~2(3 ) a 42(4 ) - ^ Cn b8(2 ) - 0 O 18(3 ) - to C.8(3 ) - O '6(3 ) - o o C,9(3 ) n co o 4^ o 4^ c 'co 'co 'bO 'co "CO 5 ar y Ta b S

~~1 o 1 1 1 p p CD o o p p p p p bO co p p p p p p p to co p o co co p p p bO p co co co co co co to -J co h-' to to bO to to bo CO o 4^ 4^ oo to h-' bO co > rys 1 b f-» 1 r> i—> -J -~~

~a cn oo ;i9 ) o ;i9 ) o -a co o co ;i9 ) o ;i9 ) o to o c15 ) 0 o co to co 1 O o p £ rt- ^ p o E!-C! P o o o p o p o 4^ 4^ p o p o o CO co 4^ 4^ p p o co o p o p o ^ 4^ 4^ Cn 4^ -J 4^ CO 4^ 4^ 4^ CO CO CO 4^ CO CO 4^ 03 £r lf^ 1 co 4^ co co1 4^ 1 co co 1 co co C3 bO co I— -J CO to o co to I— Cn to I— ~NJ o co co to I— OJ Ol co as CO 05 C3 P- o 4^ 1—> CO Cn 4^ 4^ H Cn 1—* 1—> -J Cn -J K CO 4^ 4^

bo oo to oo o co o to -17 ) 0 a o 1 1 1 tf to to bO to I— co co p—» I— I— CO CO bO co co co bO t3 v!b sfe o Eo ^b ^b -!b C5 © ^b © © ^b in- B fa X3 i-i Ct) O o p p p p p p p p p p p o p p p p p p p p i—> p p p p <; 1 1 1 1 1 1 1 CO co co bO to t—' h-' I— I— co CO co to bO 1—» I— I— I— bO bO I— o to bO I— 1—1 bo o 4^ 4^ 4^ Cn Cn 05 © bO ~J -J -Jl -J 4^ Cn Cn h-' Cn OJ Cn

o o to to to to 4^ ;i5 ) to is CO 4^ 4^ 4^ 4^ 4^ 4^ 4^ £- 4^ 4^ 4^ 4^ 4^ 4^ 4^ J** 4^ 4^ rfs- 4^ bO 4^ co to CO 4^ >° V EC 9 oq S3 fD e-t- S' O c 1 1 1 1 1 1 1 1 1 1 1 1 1 o CD I— 1—' K 1—» I— I— 1—' 1—* 1—1 1—» I— I— I— h-» H I— 1—> t—L I— I— I— >—' I— h-' I— h-' i— o o o o o o o o o o o o o o o o o o o o o o o o O o o O o o o o o o o o o o o o o o o o o o o o o o o O o o o o o o © o o o o o o o o o o o o o o o o o o o o o o o p § © o o o o o o o o o o o o o o o o o o o o o o o o o o o e-t- a a Cl c a a c a a a a a a a a a a a a c a a a a a a a CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO co CO CO CO CO CO CO CO to CD o o o o o o o o o o o o o o o o o o o o O O O O o O OS o o Appendix A. Summary of Crystallographic Data 261

~~Table A. 17 - continued from previous page~~

2 Atom x U(iso) (A ) Occupancy Type H(301) -0.034(3) -0.2377(18) 0.347(2) 0.24(4) 1.0000 Uiso H(302) 0.027(3) -0.2202(18) 0.327(2) 0.24(4) 1.0000 Uiso H(311) -0.027(3) -0.182(2) 0.428(3) 0.21(4) 1.0000 Uiso H(312) 0.031(3) -0.161(2) 0.404(3) 0.21(4) 1.0000 Uiso H(321) -0.077(3) -0.129(3) 0.361(3) 0.28(4) 1.0000 Uiso H(322) -0.023(3) -0.117(3) 0.325(3) 0.28(4) 1.0000 Uiso H(331) -0.045(3) -0.035(3) 0.376(3) 0.28(4) 1.0000 Uiso H(332) -0.043(3) -0.075(3) 0.429(3) 0.28(4) 1.0000 Uiso H(333) 0.012(3) -0.064(3) 0.393(3) 0.28(4) 1.0000 Uiso H(401) -0.010(2) -0.3330(15) 0.3327(17) 0.15(4) 1.0000 Uiso H(402) 0.051(2) -0.3160(15) 0.3160(17) 0.15(4) 1.0000 Uiso H(411) 0.027(3) -0.3985(18) 0.399(2) 0.23(4) 1.0000 Uiso H(412) 0.088(3) -0.3832(18) 0.378(2) 0.23(4) 1.0000 Uiso H(421) 0.003(4) -0.439(3) 0.310(3) 0.30(4) 1.0000 Uiso H(422) 0.057(4) -0.413(3) 0.281(3) 0.30(4) 1.0000 Uiso H(431) 0.068(3) -0.508(3) 0.296(3) 0.29(4) 1.0000 Uiso H(432) 0.057(3) -0.498(3) 0.362(3) 0.29(4) 1.0000 Uiso H(433) 0.111(3) -0.473(3) 0.333(3) 0.29(4) 1.0000 Uiso ffi ffi a W a ffl o o o o o o o o o o 3 ^ S! o > > Q

~~1 Cn 4^ 4^ co co h—» t—» co 00 -q ai en 4^ co to I— co to 1—1 I—1 £ p o 1 1 I— to I—» to I— ' 3 'to I—1 h-'~~

~~0.4279 1 p~~

o p p o 50000 1 p o o p p p p 4^ CO to 4^ co to p p p p p p 1 1 p p p 1 1 p ^«-, <: H CO 1—> h-1 I— I— en 4^ 4^ CO I— to en I— 00 p oo1 en 4^ 4^ 4^ CO to h-1 I— C3 co to 00 o ? co co co co1 o -a1 oo o en H c to CO 1—1 4^ CO I— C5 I— to Ci CO Cn CO en -a to n a> cr 1 M oo co o o o P i— to OS (10 ) to -q CD o -a h-» o > Cn OS o oo en to co co en co en co co co H K h-» CO o oo h-> 00 C5 "UT P fa v ^ co' > ' ^ 3 " O O h-' oo 1 t 1 1 1 1 '^ 1 p o 1 1 i 1 1 1 1 1 to i 1 1 i 1 1 o « p p1—» p pI—1 p p to p p p p p p p p to io p.253 1 co io o O 1 o 1 o ^1 o o o p o H I— to to to 1—» 00 to I— K CO to to I—» b bCO 0to0 4^ 4^ to 00 en o to oo ccS ffi rl- p en o oo 1 o c^ co o oo ~q to b4^ 05 to to I— CO 4^ 05 to to P" CD ? o oH^ 1—1 co co oo s—*, o oo oo en o Cn CO •*—-s to —q 1000 ' oo co 1 1 1 1 > c-t- to en to oo Cn 00 'co 'co co 'en CO 'oo K I— I— I— 1—» o i-i o

^ ^ h-' Cn (8 ) 'en to soo 11 oo o B p —' ' ^^~J^ ^ 'i-^-' p p * ^ " ' . . 1 ' 1—> 1—1 h-' I—1 I—1 I—1 p 1 f 1—> H K h-» I— 1—1 Cd p 1 1 1 to 1 1 to co 1 to 1 1—1 h-' I— o K I— b b I— H- I— I— I— i-( tr to t—» 1—» -q OS P o 1 p CO 4^ 4^ ~J ~oq 0o5 o b4^ o en to en I— oo oo to o o o en cn o !N! b Cbn CbO CO b to 4^ 4^ C5 to 1—" 4^ b -q to -q B o o I—1 co o o o o o o en B Ol 4^ 1(7 ) 0 1(7 ) 0 o H^ CD o 00 00 CoOo o CO o ,-—^ o I—1 o o I—1 I—1 o o Co o o 'co co o 1 o o o o o -^ s ^ I— > CD s o o (8 ) >-S n ^J CO O o i-i 2 rl Ht 1 1 s o p p p p o p o p o p I— p I— p p p p p p p p p "—-^ 1—» CO i—1 h-1 1—» to to 1—> C5 pu co b i>o io rf^ b b b b b b b b b CO Oi en 00 00 00 CO Cn to O 4^ en Oi CO CT> en co o -a 1 co oo o -a en ^ CO O CD 1 I— OS en co CO 4^ 4^ 4^ 4^ 4^ 4^ co co 4^ to 4^ I— to to co en 1 oo -—' O CO I— to co o co Cn CO 9 PH- . P P fcs >° P- trt- ts3 S' CD 1 1 1 1 1 1 1 1 o C t—» p p p \—» I— p H-> p h-» p M p I— K I— I— I— h-* I— H- I— 1—' o CD Cn Cn Cn o P. CL b en bi en b b b b b en b b b b b b b b b b c? O o o o o o o o o o o o o o o o o o o o o o o o p 1 13 o o o o o o o o o o o o o o o o o o o o o o o p o o o o o o o o o o o o o o o o o o o o o o o p o

~~P C a a a a a a a cl a c a c cl c a cl a a a cl a a p P £0 p So p p p p co co co CO co CO CO CO CO CO CO CO p p p p p 0 p ts p O o o o o o o o o o o CD to o c o Oi to~~

CD HH pH pH ffi ffi pH ffi HH KH pH MH PH PH > I—1 I—1 I—1 CO CO ~4 -~4 CT> Cn c-t- 1 oo oo en en 1 O O O to t- i—» h-L I— O co to i—' s—' "-^ to to co to to *' o O o O o o o o o o o o o 00 I—1 I—' o H o h-' to I—1 to Cn Cn Oi 4^ -.1 i—1 -4 to I—1 h-' I—1 4^ I—1 c co co co 1 co Oi to cn CO oo Cn co ~4 •£- CO 4^ I— co Cn h-' -4 co oo co i—» cn 1—» I—> O 4^ o cr 3

O O > o o o o o o o o o o i—1 o o o M o o I—> to to co to to to oo o c» -J cn CO 4^ o o I—1 41* 4^ oo -4 «££ f 00 CT> --J 4^ --J 4^ 4^ Oi -4 I 80 co -J Con CcOo H ~J 4^ 0o0 i—- to cn to o to o i—^ o ts ci ts' Pi to to to to i—> O h—» O i—1 O i—» o CO ft) Oi CO CO -4 -4 i—1 4^ 4^ iSi P- co o cn Oi to 1 h—i 4^ o to co CT> (X) ->4 •4^ o ~j I— 1—> co 1—> -4 co -4 00 Cn Cn Cn oo CT> 00 Cn O So 9=

~~>-j O O O O O O O O O O O O O < co co 4^ to to to 1—> o' to to to 1 to I—1 l—> i—» 1—> h-» I— ~4 -4 -4 ~4 cn pi to co CO 4^ 4^ 4^ 4^ 4^ 4^ 4^ 4^ 4^ 4^ 4^ tt^ 4^ •—^ O~~

(jq o as popppppp^pppo o CncnCnCncnCnCnCnOCncnCnCn o OOOOOOOOOOOOO OOOOOOOOOOOOO ss OOOOOOOOOOOOO o daaacdaccaccc; co cn co co co co co co co co co co O O O O O O O CO to o o o o o o on co References

P. Cassoux, L. Valade, H. Kobayashi, A. Kobayashi, R. A. Clark, and A. E. Un derbill, "Molecular metals and superconductors derived from metal complexes of l,3-dithiol-2-thione-4,5-dithiolate (dmit)," Coordination Chemistry Reviews, vol. 110, pp. 115-160, 1991.

X. Marguerettaz, G. Redmond, S. N. Rao, and D. Fitzmaurice, "Hetero- supramolecular chemistry: An approach to modulating function in molecular devices," Chemistry: A European Journal, vol. 2, pp. 420-428, 1996.

~~P. Chandrasekhar, Conducting polymers, fundamentals and applications : a practical approach. Boston, London: Kluwer Academic, 1999.~~

~~J. S. Miller, "Organometallic- and organic-based magnets: New chemistry and new materials for the new millennium," Inorganic Chemistry, vol. 39, pp. 4392- 4408, 2000.~~

~~O. Kahn, Magnetism: A Supramolecular Function. Dordretch: Kluwer Aca demic Publishers, 1996.~~

~~D. M. Roundhill and J. P. J. Fackler, Optoelectronic Properties of Inorganic Compounds. New York: Plenum Press, 1999.~~

T. J. Marks and M. A. Ratner, "Design, synthesis, and properties of molecule- based assemblies with large second-order optical nonlinearities," Angewandte Chemie International Edition, vol. 34, pp. 155-173, 1995.

~~B. J. Coe, "Molecular materials possessing switchable quadratic nonlinear op tical properties," Chemistry: A European Journal, vol. 5, pp. 2464-2471, 1999.~~

M. Eddaoudi, D. B. Moler, H. Li, B. Chen, T. M. Reineke, M. O'Keeffe, and O. M. Yaghi, "Modular chemistry: Secondary building units as a basis for the design of highly porous and robust metal-organic carboxylate frameworks," Accounts of Chemical Research, vol. 34, no. 4, pp. 319-330, 2001.

~~264 References 265~~

~~[10] T. Iwamoto, Inclusion Compounds, vol. 5, p. 177. London: Oxford University Press, 1991.~~

~~[11] M. J. Zaworotko, "Nanoporous structures by design," Angewandte Chemie In ternational Edition, vol. 39, pp. 3052-3054, 2000.~~

[12] S. Kitagawa, R. Kitaura, and S.-I. Noro, "Functional porous coordination poly mers," Angewandte Chemie, International Edition, vol. 43, pp. 2334-2375, 2004. [13] C. Janiak, "Engineering coordination polymers towards applications," Journal of the Chemical Society, Dalton Transactions: Inorganic Chemistry, vol. 14, pp. 2781-2804, 2003.

[14] B. J. Holliday and C. A. Mirkin, "Strategies for the construction of supramolecular compounds through coordination chemistry," Angewandte Chemie International Edition, vol. 40, pp. 2022-2043, 2001.

[15] S. L. James, "Metal-organic frameworks," Chemical Society Reviews, vol. 32, pp. 276-288, 2003. [16] D. J. Chesnut, D. Hagrman, P. J. Zapf, R. P. Hammond, R. LaDuca, R. C. Haushalter, and J. Zubieta, "Organic/inorganic composite materials: the roles of organoamine ligands in the design of inorganic solids," Coordination Chem istry Reviews, vol. 190-192, pp. 737-769, 1999.

[17] H. W. Roesky and M. Andruh, "The interplay of coordinative, hydrogen bond ing and 7T7T stacking interactions in sustaining supramolecular solid-state archi tectures. A study case of bis(4-pyridyl)- and bis(4-pyridyl-n-oxide) tectons," Coordination Chemistry Reviews, vol. 236, no. 1-2, pp. 91-119, 2003. [18] O. M. Yaghi, H. Li, C. Davis, D. Richardson, and T. L. Groy, "Synthetic strate gies, structure patterns, and emerging properties in the chemistry of modular porous solids," Accounts of Chemical Research, vol. 31, pp. 474-484, 1998.

[19] T. Iwamoto, "Supramolecular chemistry in cyanometallate systems," in Com prehensive Supramolecular Chemistry (J. Lehn, J. L. Atwood, J. E. D. Davies, D. D. MacNicol, F. Vogtle, and G. Alberti, eds.), vol. 6, pp. 643-690, Pergamon, 1996. [20] J. Cernak, M. Orendac, I. Potocnak, J. Chomic, A. Orendacova, J. Skorsepa, and A. Feher, "Cyanocomplexes with one-dimensional structures: preparations, crystal structures and magnetic properties," Coordination Chemistry Reviews, vol. 224, no. 1-2, pp. 51-66, 2002. References 266

[21] K. R. Dunbar and R. A. Heintz, "Chemistry of transition metal cyanide compounds: modern perspectives," Progress in Inorganic Chemistry, vol. 45, pp. 283-391, 1997. [22] M. Ohba and H. Okawa, "Synthesis and magnetism of multi-dimensional cyanide-bridged bimetallic assemblies," Coordination Chemistry Reviews, vol. 198, pp. 313-328, 2000. [23] M. Verdaguer, A. Bleuzen, V. Marvaud, J. Vaissermann, M. Seuleiman, C. De- splanches, A. Scuiller, C. Train, R. Garde, G. Gelly, C. Lomenech, I. Rosen- man, P. Veillet, C. Cartier, and F. Villain, "Molecules to build solids: high Tc molecule-based magnets by design and recent revival of cyano complexes chem istry," Coordination Chemistry Reviews, vol. 190-192, pp. 1023-1047, 1999. [24] H. Vahrenkamp, A. Geiss, and G. N. Richardson, "Cyanide-bridged oligonuclear complexes: features and attractions," Journal of the Chemical Society, Dalton Transactions: Inorganic Chemistry, pp. 3643-3651, 1997. [25] A. G. Sharpe, "Comprehensive coordination chemistry," in Comprehensive Co ordination Chemistry (G. Wilkinson, R. D. Gillard, and J. A. McCleverty, eds.), vol. 2, pp. 7-14, New York: Pergamon, 1987. [26] A. G. Sharpe, The Chemistry of Cyano Complexes of the Transition Metals. New York: Academic Press: London, 1976. [27] J. Lefebvre and D. B. Leznoff, "Structural diversity, physical properties, and applications of cyanometalate coordination polymers," in Metal-Coordination Polymers (A. S. Abd-El-Aziz, C. E. Carraher Jr., C. U. Pittman Jr., and M. Zeldin, eds.), vol. 5 of Macromolecules Containing Metal and Metal-Like Elements, pp. 155-208, Wiley-Interscience, 2005.

[28] Anonymous, "Notitia coerulei berolinenfis nuper inventi," Miscellanea Beroli- nensia ad Incrementum Scientiarum (Berlin), vol. 1, pp. 377-378, 1710. [29] J. F. Keggin and F. D. Miles, "Structures and formulas of the Prussian blues and related compounds," Nature (London), vol. 137, pp. 577-578, 1936.

[30] H. J. Buser, A. Ludi, W. Petter, and Schwarzenbach, "Single-crystal study of Prussian blue: Fe4[Fe(CN)6]2-14H20," Journal of the Chemical Society, Chem ical Communications, p. 1299, 1972. [31] F. Herren, P. Fisher, A. Ludi, and W. Halg, "Neutron diffraction study of Prussian blue, Fe4[Fe(CN)6]3 • a:H20. Location of water molecules and long- range magnetic order," Inorganic Chemistry, vol. 19, pp. 956-959, 1980. References 267

~~[32] J. S. Miller, "Three-dimensional network-structured cyanide-based magnets," MRS Bulletin, vol. 25, no. 11, pp. 60-64, 2000.~~

[33] E. Ruiz, A. Rodriguez-Fortea, S. Alvarez, and M. Verdaguer, "Is it possible to get high Tc magnets with Prussian Blue analogues? A theoretical prospect," Chemistry: A European Journal, vol. 11, no. 7, pp. 2135-2144, 2005.

~~[34] S. M. Holmes and G. S. Girolami, "Sol-gel synthesis of~~

~~KV(ll)[Cr(lll)(CN)6]-2H20: A crystalline molecule-based magnet with a magnetic ordering temperature above 100 °C," Journal of the American Chemical Society, vol. 121, pp. 5593-5594, 1999.~~

~~[35] M. Verdaguer and G. Girolami, Magnetism: Molecules to Materials, vol. 5, ch. Magnetic Prussian Blue analogs, pp. 283-346. Weinheim: WileyVCH, 2005.~~

~~[36] S. S. Kaye and J. R. Long, "Hydrogen storage in the dehydrated prussian blue analogues," Journal of the American Chemical Society, vol. 127, pp. 6506-6507, 2005.~~

~~[37] K. W. Chapman, P. D. Southon, C. L. Weeks, and C. Kepert, "Reversible hydrogen gas uptake in nanoporous Prussian Blue analogues," Chemical Com munications, pp. 3322-3323, 2005.~~

[38] L. G. Beauvais, M. P. Shores, and J. R. Long, "Cyano-bridged ReeQs (Q = S, Se) cluster-cobalt(il) framework materials: Versatile solid chemical sensors," Journal of The American Chemical Society, vol. 122, pp. 2763-2772, 2000.

[39] J. H. Rayner and H. M. Powell, "Structure of molecular compounds. Part X. Crystal structure of the compound of benzene with an ammonia-nickel cyanide complex," Journal of the Chemical Society, pp. 319-328, 1952.

[40] T. Iwamoto, S. Nishikiori, Y. A. Dyadin, and D. V. Soldatov, "Structure of the catena-bis(trimethylamine)cadmium(ll) tetra-m-cyanonickelate(ll) complex and contact self-stabilization of molecules," Journal of Structural Chemistry, vol. 40, no. 5, pp. 750-756, 2000.

[41] B. F. Hoskins and R. Robson, "Design and construction of a new class of scaffolding-like materials comprising infinite polymeric frame works of 3D-linked molecular rods, a reappraisal of the zinc cyanide and cadmium cyanide structures and the synthesis and structure of the

diamond-related frameworks [N(CH3)4][Cu(i)Zn(ii)(CN)4] and Cul[4,4',4",4"'- tetracyanotetraphenylmethane]BF4 • XC6H5NO2," Journal of the American Chemical Society, vol. 112, no. 4, pp. 1546-1554, 1990. References 268

[42] C. A. Mir kin and B. J. Holliday, "Strategies for the construction of supramolecular compounds through coordination chemistry," Angewandte Chemie, International Edition, vol. 40, pp. 2022-2043, 2001. [43] L. Ouahab, "Organic/inorganic supramolecular assemblies and synergy between physical properties," Chemistry of Materials, vol. 9, pp. 1909-1926, 1997.

~~[44] D. Braga, F. Grepioni, and A. G. Orpen, Crystal Engineering: From Molecules and Crystals to Materials. Kluwer Academic Publishers: Dordretch, 1999.~~

~~[45] D. Braga, "Inorganic crystal engineering: a personal perspective," Journal of the Chemical Society, Dalton Transactions: Inorganic Chemistry, pp. 3705- 3713, 2000.~~

~~[46] D. Braga, F. Grepioni, and G. R. Desiraju, "Crystal engineering and organometallic architecture," Chemical Reviews, vol. 98, p. 1375, 1998.~~

~~[47] D. B. Leznoff and J. Lefebvre, "Coordination polymers with cyanoaurate build ing blocks: Potential new industrial applications for gold," Gold Bulletin, vol. 38, pp. 47-56, 2005.~~

[48] C. J. Shorrock, B.-Y. Xue, P. B. Kim, R. J. Batchelor, B. O. Patrick, and D. B. Leznoff, "Heterobimetallic coordination polymers incorporating [M(CN)2]~ (M = Cu, Ag) and [Ag2(CN)3]~ units: Increasing structural dimensionality via M- M' and M-• • NC Interactions," Inorganic Chemistry, vol. 41, pp. 6743-6753, 2002.

[49] B. F. Hoskins, R. Robson, and N. V. Y. Scarlett, "Six interpenetrating quartz like nets in the structure of ZnAu2(CN)4," Angewandte Chemie, International Edition in English, vol. 34, no. 11, pp. 1203-1204, 1995.

[50] S. C. Abrahams, J. L. Bernstein, and R. Liminga, "Piezoelectric KCo[Au(CN)2]3: Room temperature crystal structure of a cobalt-hardened gold electrodeposition process component," Journal of Chemical Physics, vol. 73, no. 9, pp. 4585-4590, 1980.

[51] W. Dong, L.-N. Zhu, Y.-Q. Sun, M. Liang, Z.-Q. Liu, D.-Z. Liao, Z.-H. Jiang, S.-P. Yan, and P. Cheng, "3D porous and 3D interpenetrating triple framework structures constructed by aurophilicity-coordination interplay in {Mn[Au(CN)2]2(H20)2}n and {KFe[Au(CN)2]3}„," Chemical Communications, pp. 2544-2545, 2003. References 269

[52] I. K. Chu, I. P. Y. Shek, K. W. M. Siu, W.-T. Wong, J.-L. Zuo, and T.-C. Lau, "Synthesis, crystal structure and electrospray ionisation mass spectrometry of a novel one-dimensional cyano-bridged Ni(ll)-Au(l) polymer," New Journal of Chemistry, vol. 24, pp. 765-769, 2000.

[53] D. B. Leznoff, B.-Y. Xue, C. L. Stevens, A. Storr, R. C. Thompson, and B. 0. Patrick, "Synthesis, structure and magnetic properties of 3D interpenetrating nets of M(pyrazine)[Au(CN)2]2 (M = Cu, Ni, Co) supported by aurophilic in teractions," Polyhedron, vol. 20, pp. 1247-1254, 2001. [54] H. Schmidbaur, "The aurophilicity phenomenon: A decade of experimental find ings, theoretical concepts and emerging applications," Gold Bulletin, vol. 33, pp. 3-10, 2000. [55] H. Schmidbaur, "Ludwig Mond Lecture: High-carat gold compounds," Chemi cal Society Reviews, vol. 24, pp. 391-400, 1995. [56] M. Stender, M. M. Olmstead, A. L. Balch, D. Rios, and S. Attar, "Cation and hydrogen bonding effects on the self-association and luminescence of the dicyanoaurate ion, [Au(CN)2]~," Dalton Transactions, no. 22, pp. 4282-4287, 2003. [57] D. B. Leznoff, B.-Y. Xue, R. J. Batchelor, F. W. B. Einstein, and B. O. Patrick, "Gold-gold interactions as crystal engineering design elements in heterobimetallic coordination polymers," Inorganic Chemistry, vol. 40, pp. 6026-6034, 2001. [58] F. Scherbaum, A. Grohmann, B. Huber, C. Krger, and H. Schmid baur, ""Aurophilicity" as a consequence of relativistic effects: The hex- 2+ akis(triphenylphosphaneaurio)methane dication [(Ph3PAu)6C] ," Angewandte Chemie International Edition, vol. 27, pp. 1544-1546, 1988. [59] M. Bardaji and A. Laguna, "Gold chemistry: Aurophilic attractions," Journal of Chemical Education, vol. 76, no. 2, pp. 201-203, 1999. [60] D. D. Heinrich, J.-C. Wang, and J. P. Fackler Jnr, "Structure of Au2[S2CN(C2H5)2]2, Bis(diethyldithiocarbamato)digold(i)," Acta Crystallo- graphica Section C: Crystal Structure Communications, vol. C46, pp. 1444-1447, 1990. [61] P. Pyykko, "Theoretical chemistry of gold," Angewandte Chemie, International Edition, vol. 43, pp. 4412-4456, 2004. [62] G. R. Desiraju and T. Steiner, The weak hydrogen bond: In structural chemistry and biology. Oxford: Oxford University Press, 1999. References 270

[63] M. A. Rawashdeh-Omary, M. A. Omary, and H. H. Patterson, "Oligomeriza- tion of Au(CN)^" and Ag(CN)^" ions in solution via ground-state aurophilic and argentophilic bonding," Journal of the American Chemical Society, vol. 122, pp. 10371-10380, 2000.

[64] M. A. Rawashdeh-Omary, M. A. Omary, H. H. Patterson, and J. P. J. Fack- ler, "Excited-state interactions for [Au(CN)^"]n and [Ag(CN)^]ra oligomers in solution. Formation of luminescent gold-gold bonded excimers and exciplexes," Journal of the American Chemical Society, vol. 123, pp. 11237-11247, 2001.

[65] D. B. Leznoff, B.-Y. Xue, B. O. Patrick, V. Sanchez, and R. C. Thomp son, "An aurophilicity-determined 3-D bimetallic coordination polymer: using [Au(CN)2]~ to increase structural dimensionality through gold- • • gold bonds in (tmeda)Cu[Au(CN)2]2," Chemical Communications, pp. 259-260, 2001. [66] E. Colacio, F. Lloret, R. Kivekaes, J. Ruiz, J. Suarez-Varela, and M. R. Sund- berg, "Aurophilicity as a cofactor in crystal engineering. Dicyanoaurate(l) anion as a building block in a novel Co(ll)-Au(l) bimetallic assembly," Chemical Com munications, pp. 592-593, 2002.

[67] J. C. F. Colis, C. Larochelle, R. Staples, R. Herbst-Irmer, and H. Patterson, "Structural studies of lanthanide ion complexes of pure gold, pure silver and mixed metal (gold-silver) dicyanides," Journal of the Chemical Society, Dalton Transactions: Inorganic Chemistry, pp. 675-679, 2005.

[68] A. Stier and K.-J. Range, "Dicyanometallate, VII [1] - Darstellung und Kristallstruktur von Gadolinium-tris-dicyanoaurat(I), Gd(Au(CN)2)3-2,3H20," Zeitschrift fuer Naturforschung, B: Chemical Sciences, vol. 51, no. 5, pp. 698- 702, 1996. [69] Z. Assefa, G. Shankle, H. H. Patterson, and R. Reynolds, "Photoluminescence studies of lanthanide ion complexes of gold and silver dicyanides: A new low- dimensional solid-state class for nonradiative excited-state energy transfer," In organic Chemistry, vol. 33, pp. 2187-2195, 1994.

~~[70] R. L. Carlin, Magnetochemistry. Berlin Heidelberg, Germany: Springer-Verlag, 1986. [71] O. Kahn, Molecular Magnetism. VCH: Weinheim, 1993.~~

~~[72] J. H. Van Vleck, The theory of electric and magnetic susceptibilities. Oxford: Oxford University Press, 1932. References 271~~

~~[73] P. W. Atkins, Quanta - A handbook of concepts. New York, United State: Oxford University Press, 1991.~~

~~[74] B. I. Swanson, "Aspects of the structure and bonding in Prussian Blues. A~~

~~single-crystal raman study of Mn3[Co(CN)6]2-xH20 and Cd3[Co(CN)6]2-a;H20," Inorganic Chemistry, vol. 15, pp. 253-259, 1976.~~

[75] M. J. Katz, P. M. Aguiar, R. J. Batchelor, A. A. Bokov, Z.-G. Ye, S. Kroeker, and D. B. Leznoff, "Structure and multinuclear solid-state NMR of a highly birefringent lead-gold cyanide coordination polymer," Journal of the American Chemical Society, vol. 128, pp. 3669-3676, 2006.

~~[76] Ocean Optics, Inc., User's manual, 2005.~~

~~[77] P. Giitlich, R. Link, and A. Trautwein, Mossbauer Spectroscopy and transition metal chemistry. New York: Springer-Verlag, 1978.~~

~~[78] G. M. Bancroft, Mossbauer Spectroscopy - An introduction for inorganic chemists and geochemists. UK: McGraw-Hill Book Company Limited, 1973.~~

~~[79] B. E. Warren, X-ray diffraction. New York: Dover publications, Inc., 1st edi tion ed., 1990.~~

~~[80] R. Young, The Rietveld method. New York: Oxford University Press, 1993.~~

~~[81] A. Schenck, Muon Spin Rotation Spectroscopy: Principles and Applications in Solid State Physics. Bristol, England: Adam Hilger Ltd, 1985.~~

[82] G. M. Kalvius, D. R. Noakes, and O. Hartmann, "/iSR studies of rare-earth and actinide magnetic materials," in Handbook on the Physics and Chemistry of Rare Earths (K. A. Gschneidner, L. Eyring, and G. H. Lander, eds.), vol. 32, p. 55, Elsevier, 2001.

[83] T. Lancaster, S. J. Blundell, F. L. Pratt, M. L. Brooks, J. L. Manson, E. K. Brechin, C. Cadiou, D. Low, E. J. L. Mclnnes, and R. E. P. Winpenny, "Muons as a probe of magnetism in molecule-based low dimensional magnets," Journal of Physics: Condensed Matter, vol. 16, pp. S4563-S4582, 2004.

~~[84] J. L. Manson, T. Lancaster, L. C. Chapon, S. J. Blundell, J. A. Schlueter, M. L.~~

Brooks, F. L. Pratt, L. C. Nygren, and J. S. Quails, "Cu(HC02)2(pyrimidine) low-dimensional magnetic behaviour and long-range ordering in a quantum-spin lattice," Inorganic Chemistry, vol. 44, no. 4, pp. 989-995, 2005. References 272

[85] T. Lancaster, S. J. Blundell, F. L. Pratt, E. Coronado, and J. R. Galan- Mascaros, "Magnetic order and local field distribution in the hybrid magnets [FeCp*2][MnCr(ox)3] and [CoCp*2][FeFe(ox)3]: a muon spin relaxation study," Journal of Materials Chemistry, vol. 14, pp. 1518-1520, 2004. [86] T. Jestadt, M. Kurmoo, S. J. Blundell, F. L. Pratt, C. J. Kepert, K. Prassides, B. W. Lovett, I. M. Marshall, A. Husmann, K. H. Chow, R. M. Valladares, C. M. Brown, and A. Lappas, "Muon-spin-rotation and magetization study of metal-organic magnets based on the dicyanamide anion," Journal of Physics: Condensed Matter, vol. 13, pp. 2263-2270, 2001.

[87] G. Maruta, K. Nishiyama, and S. Takeda, "^SR studies on the dimensionality of layered compounds Cu2(OH)3X," Polyhedron, vol. 20, pp. 1185-1188, 2001. [88] S. C. Abrahams, L. E. Zyontz, and J. L. Bernstein, "Cobalt cyanoaurate: Crys tal structure of a component from cobalt-hardened gold electroplating baths," Journal of Chemical Physics, vol. 76, no. 11, pp. 5458-5462, 1982.

[89] F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry: A Comprehen sive Text, ch. 29-D. Manganese, 29-E. Iron, 29-F. Cobalt, pp. 834-878. New York: Interscience Publishers, 2nd edition ed., 1966.

~~[90] F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry: A Comprehen sive Text, ch. 29-E. Iron, pp. 847-862. New York: Interscience Publishers, 2nd edition ed., 1966.~~

[91] E. Dubler, G. Hanggi, and H. Schmalle, "Synthesis and structure of dimeric metal complexes with N(3)/N(9)-chelating hypoxanthine ligands and with bridging water molecules: [M2(/i-hyxan)2(S04)2(A*-H20)2(H20)2] (M = cop per, cadmium, zinc; hyxan = hypoxanthine)," Inorganic Chemistry, vol. 29, pp. 2518-2523, 1990.

[92] O. Svajlenova, J. Vanco, and J. Marek, "Di-/i-aqua-bis[(N-salicylidene-/3- alaninato-K30, N, O')copper(n)] urea disolvate," Acta Crystallographica, Sec tion C: Crystal Structure Communications, vol. C60, pp. m275-m277, 2004. [93] P. J. van Koningsbruggen, J. W. van Hal, E. Miiller, and R. A. G. de Graaff, "A novel type of twisted antiparallel double-chain structure with stacking between the two strands. Structure, synthesis and magnetic properties of [[Cu3L2(dien)(H20)2]-3H20][L = lH-l,2,4-triazole-3,5-dicarboxylate(3), dien = 3-azapentane-l,5-diamine]," Journal of the Chemical Society, Dalton Transac tions: Inorganic Chemistry, pp. 1371-1375, 1993. References 273

[94] D. Ghoshal, T. K. Maji, G. Mostafa, S. Sain, T.-H. Lu, J. Ribas, E. Zangrando, and N. R. Chaudhuri, "Polymeric networks of copper (n) using succinate and aromatic NN donor ligands: synthesis, crystal structure, magnetic behaviour and the effect of weak interactions on their crystal packing," Dalton Transac tions, pp. 1687-1695, 2004.

[95] D. Lee, R-L. Hung, B. Spingler, and S. J. Lippard, "Sterically hindered carboxylate ligands support water-bridged dimetallic centers that model features of metallohydrolase active sites," Inorganic Chemistry, vol. 41, pp. 521-531, 2002.

[96] P. King, R. Clerac, C. E. Anson, and A. K. Powell, "The building block ap proach to extended solids: 3,5-pyrazoledicarboxylate coordination compounds of increasing dimensionality," Dalton Transactions, pp. 852-861, 2004.

[97] A. M. Barrios and S. J. Lippard, "Interaction of urea with a hydroxide-bridged dinuclear nickel center: An alternative model for the mechanism of urease," Journal of the American Chemical Society, vol. 122, pp. 9172-9177, 2000.

[98] A. M. Barrios and S. J. Lippard, "Amide hydrolysis effected by a hydroxo- bridged dinickel(ll) complex: Insights into the mechanism of urease," Journal of the American Chemical Society, vol. 121, pp. 11751-11757, 1999.

[99] F. A. Cotton and B. H. C. Winquist, "Crystal and molecular struc tures of a toluene-containing form of di-/x-aquo-di-//3-hydroxodecakis( 1,1,1- trifluoropentane-2,4-dionato)hexanickel(ll)," Inorganic Chemistry, vol. 8, pp. 1304-1312, 1969. [100] X. Meng, H. Hou, G. Li, B. Ye, T. Ge, Y. Fan, Y. Zhu, and H. Sakiyama, "Tetrametallic macrocyclic frameworks constructed from ferrocenedicarboxy- lato and 2,2'-bipyridine: synthesis, molecular structures and characteristics," Journal of Organometallic Chemistry, vol. 689, pp. 1218-1229, 2004.

[101] D. Brown, W. Glass, N. Fitzpatrick, T. Kemp, W. Errington, H. Miiller-Bunz, A. Hussein, and H. Nimir, "Mononuclear and dinuclear model hydrolases of nickel and cobalt," Inorganica Chimica Acta, vol. 358, pp. 2755-2762, 2005. [102] D. Concernvanis, R. A. Reynolds III, and W. R. Dunham, "Synthesis and charac terization of a new class of asymmetric aqua-acetate bridged dimers. Solid state molecular structures of the [M2(/U-H20)(/U-OAc)2(OAc)3(Py)2]~ anions (M = Mn(ll), Fe(ll), Co(ll)). A structural model for the Fe2 site in methane monooxy- genase," Journal of the American Chemical Society, vol. 117, pp. 7570-7571, 1995. References 274

~~[103] S. Yoon and S. J. Lippard, "Water-dependent reactions of diiron(n) carboxylate complexes," Journal of the American Chemical Society, vol. 126, pp. 16692- 16693, 2004.~~

[104] S. K. Ghosh, J. Ribas, M. S. El Fallah, and P. K. Bharadwaj, "Mn(n) staircase structures stitched by water clusters to a 3D metal-organic open framework: X-ray structural and magnetic studies," Inorganic Chemistry, vol. 44, pp. 3856- 3862, 2005.

[105] H. O. Davies, R. D. Gillard, M. B. Hursthouse, and A. Lehmann, "A linear polymer of copper(ll) with water and the carboxy group of L-alanine as bridges," Journal of the Chemical Society, Chemical Communications, pp. 1137-1139, 1993.

~~[106] M. Riiegg, A. Ludi, and K. Rieder, "The crystal structure of manganese(n) hex- acyanoruthenate (n) octahydrate, Mn2[Ru(CN)6]-8H20," Inorganic Chemistry, vol. 10, pp. 1773-1777, 1971.~~

[107] I. V. Korendovych, S. V. Kryatov, W. M. Reiff, and E. V. Rybak-Akimova, "Di- iron(ll) /i-aqua-/i-hydroxo model for non-heme iron sites in proteins," Inorganic Chemistry, vol. 44, pp. 8656-8658, 2005.

[108] G. A. Baker, G. S. Rushbrooke, and H. E. Gilbert, "High-temperature series expansions for the spin-1/2 Heisenberg model by the method of irreducible representations of the symmetric group," Physical Review, vol. 135, pp. A1272- 1277, 1964.

[109] R. S. Hayano, Y. J. Uemura, J. Imazato, N. Nishida, T. Yamazaki, and R. Kubo, "Zero-and low-field spin relaxation studied by positive muons," Physical Review B: Condensed Matter, vol. 20, pp. 850-859, 1979. [110] Y. J. Uemura, T. Yamazaki, D. R. Harshmann, M. Senba, and E. J. Ansaldo, "Muon-spin relaxation in AuFe and CuMn spin glasses," Physical Review B: Condensed Matter, vol. 31, pp. 546-563, 1985.

~~[Ill] F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry: A Comprehen sive Text, ch. 29-F. Cobalt, pp. 863-878. New York: Interscience Publishers, 2nd edition ed., 1966.~~

~~[112] M. E. Fisher, "Magnetism in one-dimensional systems - the Heisenberg model for infinite spin," American Journal of Physics, vol. 32, pp. 343-346, 1964. References 275~~

[113] C. Desplanches, E. Ruiz, A. Rodriguez-Fortea, and S. Alvarez, "Exchange cou pling of transition-metal ions through hydrogen bonding: A theoretical inves tigation," Journal of the American Chemical Society, vol. 124, pp. 5197-5205, 2002.

[114] W. Plass, A. Pohlmann, and J. Rautengarten, "Magnetic interactions as supramolecular function: Magnetic properties of H-bridged dinuclear copper(n) complexes," Angewandte Chemie, International Edition in English, vol. 40, pp. 4207-4210, 2001. [115] M. Ardon, A. Bino, K. Michelsen, E. Pedersen, and R. C. Thompson, "Chain- structured cis-hydroxoaquachromium(lll) complexese and their magnetic prop erties," Inorganic Chemistry, vol. 36, pp. 4147-4150, 1997.

[116] C. Rancurel, N. Daro, O. Benedi Borobia, E. Herdtweck, and J.-P. Sutter, "H20 as a chemical link for ferromagnetic interactions between aminoxyl units," European Journal of Organic Chemistry, pp. 167-171, 2003. [117] D. Luneau, P. Rey, J. Laugier, E. Belorizky, and A. Cogne, "Ferromagnetic be haviour of nickel(il)-imino nitroxide derivatives," Inorganic Chemistry, vol. 31, pp. 3578-3584, 1992.

[118] G. V. R. Chandramouli, T. K. Kundu, and P. T. Manoharan, "Magneto- structural correlation in di-/i-oxo bridged dicopper complexes - Predictability of isotropic exchange coupling constant from structure," Australian Journal of Chemistry, vol. 56, pp. 1239-1248, 2003. [119] S. Sain, T. K. Maji, G. Mostafa, T.-H. Lu, J. Ribas, X. Tercero, and N. R. Chaudhuri, "Magneto structural correlations of a 'stepped- wise' discrete tetranuclear unit of copper(n), [Cu4(//2-OH)2(/Li3-OH)2(2,2'- bipy)4Cl2]Cl2-6H20," Polyhedron, vol. 22, pp. 625-631, 2003. [120] G. A. van Albada, I. Mutikainen, O. Roubeau, U. Turpeinen, and J. Reedijk, "Ferromagnetic trinuclear carbonato-bridged and tetranuclear hydroxo-bridged Cu(ll) compounds with 4,4'-dimethyl-2,2'-bipyridine as ligand. X-ray structure, spectroscopy and magnetism," Inorganica Chimica Acta, vol. 331, pp. 208-215, 2002.

[121] J.-P. Costes, F. Dahan, and J.-P. Laurent, "A further example of a dinuclear copper(ll) complex involving monoatomic acetate bridges. Synthesis, crystal struc ture, and spectroscopic and magnetic properties of bis(/x-acetato)bis(7-amino-4- methyl-5-aza-3-hepten-2-onato(l-)) dicopper(il)," Inorganic Chemistry, vol. 24, pp. 1018-1022, 1985. References 276

[122] M. Rodriguez, A. Llobet, and M. Corbella, "A theoretical analysis of how geo metrical distortions on CufjU-C^Cu dimers influence their electronic and mag netic properties," Polyhedron, vol. 19, pp. 2483-2491, 2000. [123] M. Rodriguez, A. Llobet, M. Corbella, A. E. Martell, and J. Reibenspies, "Synthesis, structure, and magnetic properties of a new chloro-bridged dimer [Cu2(dpt)2Cl2]Cl2 with an unusual structure and ferromagnetic coupling," In organic Chemistry, vol. 38, pp. 2328-2334, 1999. See also references therein.

[124] M. J. Katz, C. J. Shorrock, R. J. Batchelor, and D. B. Leznoff, "[Au(CN)4]- as both an intramolecular and intermolecular bidentate ligand with [(tmeda)Cu(/i- OH)] dimers: from antiferro- to ferromagnetic coupling in polymorphs," Inor ganic Chemistry, vol. 45, pp. 1757-1765, 2006. [125] E. Ruiz, P. Alemany, S. Alvarez, and J. Cano, "Toward the prediction of mag netic coupling in molecular systems: hydroxo- and alkoxo-bridged Cu(n) binu- clear complexes," Journal of the American Chemical Society, vol. 119, pp. 1297- 1303, 1997. [126] C. Yasuda, S. Todo, K. Hukushima, F. Alet, M. Keller, M. Troyer, and H. Takayama, "Neel temperature of quasi-low-dimensional Heisenberg anti- ferromagnets," Physical Review Letters, vol. 94, p. 217201, 2005. [127] J. S. Miller and A. J. Epstein, "Organic and organometallic molecular magnetic materials - Designer Magnets," Angewandte Chemie, International Edition in English, vol. 33, pp. 385-415, 1994. [128] P. S. Mukherjee, S. Dalai, E. Zangrando, F. Lloret, and N. R. Chaudhuri, "The first metamagnetic one-dimensional molecular material with nickel(n) and end- to-end azido bridges," Chemical Communications, pp. 1444-1445, 2001. [129] M. Monfort, I. Resino, J. Ribas, and H. Stoeckli-Evans, "A metamagnetic two- dimensional molecular material with nickel(ll) and azide," Angewandte Chemie, International Edition in English, vol. 39, pp. 191-193, 2000. [130] X. Hao, Y. Wei, and S. Zhang, "Crystal structure and metamagnetic prop // erty of a 2-D layered complex, [Fe (N3)2(pyz)]n (pyz = pyrazine)," Chemical Communications, pp. 2271-2272, 2000. [131] T.-F. Liu, H.-L. Sun, S. Gao, S.-W. Zhang, and T.-C. Lau, "Ferromag netic ordering and metamagnetism in malonate bridged 3D diamond-like and honeycomb-like networks: [Cu(mal)(DMF)]n and {[Cu(mal)(0.5pyz)]-H2O}n (mal = malonate dianion, DMF = N,N-dimethylformamide, pyz = pyrazine)," Inorganic Chemistry, vol. 42, pp. 4792-4794, 2003. References 277

~~[132] J. H. Yoon, H. C. Kim, and C. S. Hong, "Cyanide-bridged W(v)-Mn(n) bimetal lic double-zigzag chains with a metamagnetic nature," Inorganic Chemistry, vol. 44, pp. 7714-7716, 2005.~~

~~[133] H.-Z. Kou, S. Gao, B.-W. Sun, and J. Zhang, "Metamagnetism of the first cyano- bridged two-dimensional brick-wall-like 4f-3d array," Chemistry of Materials, vol. 13, pp. 1431-1433, 2001.~~

[134] H.-Z. Kou, W.-M. Bu, S. Gao, D.-Z. Liao, Z.-H. Jiang, S.-P. Yan, Y.-G. Fan, and G.-L. Wang, "Reactions between macrocyclic nickel(ll) complexes and ferricyanide. Metamagnetic properties of a two-dimensional honeycomb assembly [NiL3]3[Fe(CN)6]2-8H20 (L3 = 3,10-dihydroxyethyl-l,3,6,8,10,12- hexaazacyclotetradecane)," Journal of the Chemical Society, Dalton Transac tions: Inorganic Chemistry, pp. 2996-3000, 2000.

[135] C. Aoki, T. Ishida, and T. Nogami, "Molecular metamagnet [Ni(4ImNNH)2(N03)2] (4ImNNH = 4-imidazolyl nitronyl nitroxide) and the related compounds showing supramolecular H-bonding interactions," Inorganic Chemistry, vol. 42, pp. 7616-7625, 2003.

~~[136] E. Gabe, P. White, and G. Enright, DIFRAC A Fortran 77 Control Routine for 4-Circle Diffractometers. Ottawa: N.R.C., 1995.~~

~~[137] E. Gabe, Y. LePage, J.-P. Charland, F. Lee, and P. White, "NRCVAX - an in teractive program system for structure analysis," Journal of Applied Crystallo graphy, vol. 22, pp. 384-387, 1989.~~

[138] P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, and D. J. Watkin, "Crystals version 12: software for guided crystal structure analysis," Journal of Applied Crystallography, vol. 36, p. 1487, 2003.

~~[139] L. J. Farrugia, "Ortep-3 for windows - a version of ortep-III with a graphical user interface (gui)," Journal of Applied Crystallography, vol. 30, p. 565, 1997.~~

[140] T. Fenn, D. Ringe, and G. Petsko, "Persistence of vision raytracing: http://www.povray.org," Journal of Applied Crystallography, vol. 36, pp. 944- 947, 2003. [141] T. Roisnel and J. Rodriguez-Carvajal, "WinPLOTR: a Windows tool for pow der diffraction patterns analysis," in Proceedings of the Seventh European Pow der Diffraction Conference (EPDIC 7), (Switzerland), pp. 118-123, Trans Tech Publications, 2000. References 278

[142] W. I. F. David, K. Shankland, J. van de Streek, E. Pidcock, W. D. S. Motherwell, and J. C. Cole, "DASH: a program for crystal structure determination from powder diffraction data," Journal of Applied Crystallography, vol. 39, pp. 910- 915, Dec 2006.

[143] W. Kraus and G. Nolzeb, "POWDER CELL - a program for the representation and manipulation of crystal structures and calculation of the resulting X-ray powder patterns," Journal of Applied Crystallography, vol. 29, pp. 301-303, 1996.

~~[144] M. K. Ehlert, S. J. Rettig, A. Storr, R. C. Thompson, and J. Trotter, "Metal pyrazolate polymers. Part 1. Synthesis, structure, and magnetic properties of~~

~~the [Cu(pz)2]s polymer," Canadian Journal of Chemistry, vol. 67, pp. 1970- 1974, 1989.~~

~~[145] R. L. White-Morris, M. M. Olmstead, F. Jiang, D. S. Tinti, and A. L. Balch, "Remarkable variations in the luminescence of frozen solutions of~~

[Au{C(NHMe)2}2](PF6)-0.5(acetone). Structural and spectroscopic studies of the effects of anions and solvents on gold(l) carbene complexes," Journal of the American Chemical Society, vol. 124, pp. 2327-2336, 2002.

[146] M. A. Mansour, W. B. Connick, R. J. Lachicotte, H. J. Gysling, and R. Eisen- berg, "Linear chain Au(l) dimer compounds as environmental sensors: A lumi nescent switch for the detection of volatile organic compounds," Journal of the American Chemical Society, vol. 120, pp. 1329-1330, 1998.

[147] M. A. Rawashdeh-Omary, M. A. Omary, J. Fackler, John P., R. Galassi, B. R. Pietroni, and A. Burini, "Chemistry and optoelectronic properties of stacked supramolecular entities of trinuclear gold(l) complexes sandwiching small or ganic acids," Journal of the American Chemical Society, vol. 123, pp. 9689-9691, 2001.

~~[148] J. Fernandez Eduardo, M. Lopez-De-Luzuriaga Jose, M. Monge, M. E. 01- mos, J. Perez, A. Laguna, A. Mohamed Ahmed, and J. Fackler John, P.,~~

~~"(Tl[Au(C6Cl5)2])n: A vapochromic complex," Journal of the American Chem ical Society, vol. 125, pp. 2022-2023, 2003.~~

~~[149] C. L. Exstrom, J. Sowa, John R., C. A. Daws, D. Janzen, K. R. Mann, G. A. Moore, and F. F. Stewart, "Inclusion of organic vapors by crystalline, solva-~~

~~tochromic [Pt(aryl isonitrile)4][Pd(CN)4] compounds. "Vapochromic" environ mental sensors," Chemistry of Materials, vol. 7, pp. 15-17, 1995. References 279~~

[150] S. M. Drew, D. E. Janzen, C. E. Buss, D. I. MacEwan, K. M. Dublin, and K. R. Mann, "An electronic nose transducer array of vapoluminescent platinum(n) double salts," Journal of The American Chemical Society, vol. 123, pp. 8414- 8415, 2001.

~~[151] C. E. Buss, C. E. Anderson, M. K. Pomije, C. M. Lutz, D. Britton, and K. R. Mann, "Structural investigations of vapochromic behavior. X-ray single-crystal~~

~~and powder diffraction studies of [Pt(CN-iso-C3H7)4][M(CN)4] for M = Pt or Pd," Journal of the American Chemical Society, vol. 120, pp. 7783-7790, 1998.~~

[152] C. Bariain, I. R. Matias, I. Romeo, J. Garrido, and M. Laguna, "Detection of volatile organic compound vapors by using a vapochromic material on a tapered optical fiber," Applied Physics Letters, vol. 77, pp. 2274-2276, 2000.

~~[153] Y. Kunugi, K. R. Mann, L. L. Miller, and C. L. Exstrom, "A vapochromic LED," Journal of the American Chemical Society, vol. 120, pp. 589-590, 1998.~~

~~[154] Y. Kunugi, L. L. Miller, K. R. Mann, and M. K. Pomije, "A vapochromic photodiode," Chemistry of Materials, vol. 10, pp. 1487-1489, 1998.~~

[155] A. W. Addison, T. N. Rao, J. Reedijk, J. Van Rijn, and G. C. Verschoor, "Syn thesis, structure, and spectroscopic properties of copper(ll) compounds con taining nitrogen-sulphur donor ligands; the crystal and molecular structure of aqua[l,7-bis(N-methylbenzimidazol-2-yl)-2,6-dithiaheptane]copper(ll) perchlorate," Journal of the Chemical Society, Dalton Transactions: Inorganic Chem istry, pp. 1349-1356, 1984.

~~[156] M. J. Katz and D. B. Leznoff , unpublished results, 2006.~~

~~[157] M. E. Lines, "The quadratic-layer antiferromagnet," Journal of Physics and Chemsitry of Solids, vol. 31, pp. 101-116, 1970.~~

[158] E. Colacio, F. Lloret, R. Kivekaes, J. Suarez-Varela, M. R. Sundberg, and R. Uggla, "Aurophilicity-coordination interplay in the design of cyano-bridged nickel(ll)-gold(l) bimetallic assemblies: Structural and computational studies of the gold(l)-gold(i) interactions," Inorganic Chemistry, vol. 42, pp. 560-565, 2003.

~~[159] D. Shriver and P. Atkins, Inorganic Chemistry. New York: W.H. Freeman and Company, 3rd edition ed., 1999.~~

~~[160] A. Lever, Inorganic electronic spectroscopy. Amsterdam, The Netherlands: El sevier Science Publisher B.V., 2nd edition ed., 1984. References 280~~

[161] R. L. Carlin, C. J. O'Connor, and S. N. Bhatia, "Large zero field splitting and subcritical antiferromagnetic interactions in [N^CsHsNO^](0104)2," Journal of the American Chemical Society, vol. 98, pp. 3523-3525, 1976.

~~[162] F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry: A Comprehen sive Text. New York: Interscience Publishers, 3rd edition ed., 1972.~~

~~[163] B. Moulton and M. J. Zaworotko, "From molecules to crystal engineering: Supramolecular isomerism and polymorphism in network solids," Chemical Re views, vol. 101, pp. 1629-1658, 2001.~~

~~[164] S. R. Batten and K. S. Murray, "Malleable coordination networks," Australian Journal of Chemistry, vol. 54, no. 9 & 10, pp. 605-609, 2001.~~

~~[165] J. D. Dunitz and J. Bernstein, "Disappearing polymorphs," Accounts of Chem ical Research, vol. 28, pp. 193-200, 1995.~~

~~[166] J. L. Bernstein, Polymorphism in Molecular Crystals. Oxford: Oxford Univer sity Press, 2002.~~

~~[167] D. Braga and F. Grepioni, "Organometallic polymorphism and phase transi tions," Chemical Society Reviews, vol. 29, pp. 229-238, 2000.~~

[168] D. B. Leznoff, C. Rancurel, J.-P. Sutter, S. J. Rettig, M. Pink, C. Paulsen, and O. Kahn, "Ferromagnetic interactions and polymorphism in radical-substituted gold phosphine complexes," Journal of the Chemical Society, Dalton Transac tions: Inorganic Chemistry, pp. 3593-3599, 1999.

[169] W. E. Broderick, D. M. Eichhorn, X. Liu, P. M. Toscano, S. M. Owens, and + B. M. Hoffman, "Three phases of [Fe(C5Me5)2] [TCNQ]-: Ferromagnetism in a new structural phase," Journal of the American Chemical Society, vol. 117, pp. 3641-3642, 1995.

[170] R. Heintz, H. Zhao, X. Ouyang, G. Grandinetti, J. Cowen, and K. R. Dunbar, "New insight into the nature of Cu(TCNQ): Solution routes to two distinct poly morphs and their relationship to crystalline films that display bistable switching behaviour," Inorganic Chemistry, vol. 38, pp. 144-156, 1999. [171] K. Nakasuji, M. Sasaki, T. Kotani, I. Murata, T. Enoki, K. Imaeda, H. Inokuchi, A. Kawamoto, and J. Tanaka, "Methylthio- and ethanediyldithio-substituted 1,6-dithiapyrenes and their charge-transfer complexes: new organic molecular metals," Journal of the American Chemical Society, vol. 109, pp. 6970-6975, 1987. References 281

[172] R. L. White-Morris, M. M. Olmstead, and A. L. Balch, "Aurophilic interactions in cationic gold complexes with two isocyanide ligands. Polymorphic yellow 7 and colorless forms of [(cyclohexyl isocyanide)2Au ](PF6) with distinct lumi nescence," Journal of the American Chemical Society, vol. 125, pp. 1033-1040, 2003.

[173] D. V. Soldatov and J. A. Ripmeester, "Novel 4-vinylpyridine-extended metal- dibenzoylmethanate host frameworks: structure, polymorphism, and inclusion properties," Chemistry: A European Journal, vol. 7, pp. 2979-2994, 2001.

[174] P. Jensen, S. R. Batten, G. D. Fallon, D. C. R. Hockless, B. Moubaraki, K. S. Murray, and R. Robson, "Synthesis, structural isomerism, and magnetism of extended framework compounds of type [Cu(dca)2(pyz)]n, where dca = dicyanamide (N(CN)^") and pyz = pyrazine," Journal of Solid State Chemistry, vol. 145, pp. 387-393, 1999.

~~[175] S. A. Barnett, A. J. Blake, N. R. Champness, and C. Wilson, "Struc tural isomerism in CuSCN coordination polymers," Chemical Communications, pp. 1640-1641, 2002.~~

~~[176] J. Bernstein, R. Davey, and J.-O. Henck, "Concomitant polymorphs," Ange- wandte Chemie, International Edition, vol. 38, pp. 3440-3461, 1999.~~

[177] K. Uemura, S. Kitagawa, M. Kondo, K. Fukui, R. Kitaura, H.-C. Chang, and T. Mizutani, "Novel flexible frameworks of porous cobalt(n) coordination poly mers that show selective guest adsorption based on the switching of hydrogen- bond pairs of amide groups," Chemistry: A European Journal, vol. 8, no. 16, pp. 3586-3600, 2002.

[178] S.-i. Noro, R. Kitaura, M. Kondo, S. Kitagawa, T. Ishii, H. Matsuzaka, and M. Yamashita, "Framework engineering by anions and porous functionalities of Cu(ll)/4,4'-bpy coordination polymers," Journal of the American Chemical Society, vol. 124, no. 11, pp. 2568-2583, 2002.

[179] K. Seki, "Design of an adsorbent with an ideal pore structure for methane adsorption using metal complexes," Chemical Communications, pp. 1496-1497, 2001. [180] J. L. Atwood, L. J. Barbour, A. Jerga, and B. L. Schottel, "Guest transport in a nonporous organic solid via dynamic van der Waals cooperativity," Science, vol. 298, no. 5595, pp. 1000-1002, 2002. References 282

[181] A. P. Cote and G. K. H. Shimizu, "The first example of a functional pillared metal sulfonate network," Chemical Communications, pp. 251-252, 2001. [182] M. Edgar, R. Mitchell, A. M. Slawin, P. Lightfoot, and P. A. Wright, "Solid-state transformations of zinc 1,4-benzenedicarboxylates mediated by hydrogen-bond- forming molecules," Chemistry: A European Journal, vol. 7, pp. 5168-5175, 2001. [183] P. W. Miller, M. Nieuwenhuyzen, X. Xu, and S. L. James, "Assembly of a co ordination cage with four aromatic channel receptors on the outside," Chemical Communications, pp. 2008-2009, 2002. [184] E. Lozano, M. Nieuwenhuyzen, and S. L. James, "Ring-opening polymerisation of silver-diphosphine [M2L3] coordination cages to give [M2L3]oo coordination polymers," Chemistry: A European Journal, vol. 7, pp. 2644-2651, 2001. [185] R. Kitaura, K. Fujimoto, S.-i. Noro, M. Kondo, and S. Kitagawa, "A pillared-layer coordination polymer network displaying hysteretic sorp tion: [Cu2(pzdc)2(dpyg)]n (pzdc= pyrazine-2,3-dicarboxylate; dpyg=l,2-di(4- pyridyl)glycol)," Angewandte Chemie, International Edition, vol. 41, pp. 133- 135, 2002. [186] A. P. Cote, M. J. Ferguson, K. A. Khan, G. D. Enright, A. D. Kulynych, S. A. Dalrymple, and G. K. H. Shimizu, "Intercalation of alcohols in Ag sulfonates: Topotactic behavior despite flexible layers," Inorganic Chemistry, vol. 41, pp. 287-292, 2002. [187] G. L. Miessler and D. A. Tarr, Inorganic Chemistry. New Jersey: Pearson Prentice Hall, 3rd edition ed., 2004. [188] C. A. Daws, C. L. Exstrom, J. Sowa, John R., and K. R. Mann, ""Vapochro- mic" compounds as environmental sensors. 2. Synthesis and near-infrared and infrared spectroscopy studies of [Pt(arylisocyanide)4][Pt(CN)4] upon exposure to volatile organic compound vapors," Chemistry of Materials, vol. 9, pp. 363- 368, 1997. [189] C. L. Exstrom, M. K. Pomije, and K. R. Mann, "Infrared spectroscopy studies of platinum salts containing tctracyanoplatinate(n). Evidence for strong hydrogen- bonding interactions in "vapochromic" environmental sensor materials," Chem istry of Materials, vol. 10, pp. 942-945, 1998. [190] T. Ramnial, H.-Z. Yu, and D. B. Leznoff, "Sensitive detection of volatile or ganic compounds using a vapochromic Cu(ll)-Au(l) coordination polymer: An ammonia/amines case study." Manuscript in preparation, 2008. References 283

[191] C. J. Shorrock, H. Jong, R. J. Batchelor, and D. B. Leznoff, "[Au(CN)4]- as a supramolecular building block for heterobimetallic coordination polymers," Inorganic Chemistry, vol. 42, pp. 3917-3924, 2003.

~~[192] T. Higashi, Program for Absorption Correction. Tokyo, Japan: Rigaku Corpo ration, 1999.~~

[193] J. A. Schlueter, J. L. Manson, and U. Geiser, "Structural and magnetic diversity in tetraalkylammonium salts of anionic M[N(CN)2]^ (M = Mn and Ni) three- dimensional coordination polymers," Inorganic Chemistry, vol. 44, pp. 3194- 3202, 2005.

[194] A. Yuan, J. Zou, B. Li, Z. Zha, C. Duan, Y. Liu, and Z. Xu, "An unusual three- dimensional porous framework complex {[Cu(en)2][KCr(CN)6]} (en = ethylenediamine) from a template self-assembly reaction," Chemical Communications, pp. 1297-1298, 2000.

[195] M. Scudder and I. Dance, "Crystal supramolecular motifs. Ladders, layers and + labyrinths of Ph4P cations engaged in fourfold phenyl embraces," Journal of the Chemical Society, Dalton Transactions, pp. 3155-3166, 1998.

~~[196] P. Day, "Co-ordination of complexes as organic-inorganic layer magnets," Jour nal of the Chemical Society, Dalton Transactions, pp. 701-706, 1997.~~

[197] S. Decurtins, H. Schmalle, and R. Pellaux, "Polymeric two- and three- dimensional transition-metal complexes comprising supramolecular host-guest systems," New Journal of Chemistry, vol. 22, pp. 117-121, 1998.

[198] P. G. Lacroix, "Organic nonlinear optical chromophores in an inorganic envi ronment. Toward an interplay between magnetism, conductivity, and nonlinear optics," Chemistry of Materials, vol. 13, pp. 3495-3506, 2001.

[199] E. Coronado, J. R. Galan-Mascaro s, C. J. Gomez-Garcia, and V. Laukhin, "Coexistence of ferromagnetism and metallic conductivity in a molecule-based layered compound," Nature, vol. 408, pp. 447-449, 2000.

~~[200] E. Coronado and P. Day, "Magnetic molecular conductors," Chemical Reviews, vol. 104, pp. 5419-5448, 2004.~~

~~[201] S. R. Batten, P. Jensen, B. Moubaraki, and K. S. Murray, "Anionic metal dicyanamide networks with paramagnetic counter-cations," Chemical Commu nications, pp. 2331-2332, 2000. References 284~~

[202] P. M. Van der Werff, E. Martinez-Ferrero, S. R. Batten, P. Jensen, C. Ruiz- Perez, M. Almeida, J. C. Waerenborgh, J. D. Cashion, B. Moubaraki, J. R. Galan-Mascaros, J. M. Martinez-Agudo, E. Coronado, and K. S. Murray, "Hy brid materials containing organometallic cations and 3-D anionic metal dicyanamide networks of type [Cp*2M][M(dca)3]," Dalton transactions, no. 2, pp. 285-290, 2005.

[203] P. M. van der Werff, S. R. Batten, P. Jensen, B. Moubaraki, K. S. Murray, and J. D. Cashion, "Structure and magnetism of 3D anionic metal dicyanamide (MePh3P)[M(dca)3] (M = Fe, Co, Ni) and (EtPh3P)[M(dca)3] (M = Mn, Co, Ni) networks," Crystal Growth & Design, vol. 4, pp. 503-508, 2003. [204] P. M. van der Werff, S. R. Batten, P. Jensen, B. Moubaraki, and K. S. Murray, "Cation templation of anionic metal dicyanamide networks," Inorganic Chem istry, vol. 40, pp. 1718-1722, 2001.

[205] D. Armentano, G. De Munno, F. Guerra, M. Julve, and F. Lloret, "Lig- and effects on the structures of extended networks of dicyanamide-containing transition-metal ions," Inorganic Chemistry, vol. 45, pp. 4626-4636, 2006.

[206] R. Ingalls, "Electric-field gradient tensor in ferrous compounds," Physical Re view, vol. 133, pp. A787-A795, 1964. [207] J. R. Sams and T. B. Tsin, "Orbital ground states and crystal field splittings in some octahedrally coordinated high-spin ferrous complexes," Inorganic Chem istry, vol. 14, pp. 1573-1579, 1975.

[208] W. Dong, Q.-L. Wang, S.-F. Si, D.-Z. Liao, Z.-H. Jiang, S.-P. Yan, and P. Cheng, "Two novel three-dimensional coordination polymers with [Ag(CN)2]~ as bridging ligands: synthesis and structural characterization of {KMn[Ag(CN)2]3(H20)}n and {Mn[Ag(CN)2]2(bpy)2}n (bpy=4,4'-bipyridine)," Inorganic Chemistry Communications, vol. 6, pp. 873-876, 2003.

[209] M. P. Shores, L. G. Beauvais, and J. R. Long, "Cluster-expanded Prussian Blue analogues," Journal of the American Chemical Society, vol. 121, pp. 775-779, 1999. [210] N. Fukita, M. Ohba, H. Okawa, K. Matsuda, and H. Iwamura, "Bimetallic assemblies [Ni(L)2]3[Fe(CN)6]X2 (L = ethylenediamine, trimethylenediamine; X = PFjT, ClO^) with a three-dimensional network extended through Fe/7-CN- Ni77 linkages," Inorganic Chemistry, vol. 37, pp. 842-848, 1998. References 285

[211] J. Chomic and J. Cernak, "Thermal-properties of some bis-(l,2-diamirioethane) and tris-(l,2-diaminoethane) metal(ll) bis-(dicyanosilver) complexes," Ther- mochimica Acta, vol. 93, pp. 93-96, 1985.

[212] B. Mayoh and P. Day, "Charge transfer in mixed-valence solids. Part vii. Con tribution of valence delocalisation to the ferromagnetism of Prussian Blue," Journal of the Chemical Society, Dalton Transactions, pp. 1483-1486, 1976.

~~[213] A. Laguna and M. Laguna, "Coordination chemistry of gold(ll) complexes," Coordination Chemistry Reviews, vol. 193-195, pp. 837-856, 1999.~~

~~[214] S. R. Batten and K. S. Murray, "Structure and magnetism of coordination poly mers containing dicyanamide and tricyanomethanide," Coordination Chemistry Reviews, vol. 246, pp. 103-130, 2003.~~

[215] J. L. Manson, Magnetism: Molecules to Materials, vol. 5, ch. Cooperative Mag netic Behaviour in Metal-Dicyanamide Complexes, pp. 71-104. Weinheim: Wi ley VCH, 2005. [216] N. D. Draper, R. J. Batchelor, and D. B. Leznoff, "Tuning the structures of mercury cyanide-based coordination polymers with transition metal cations," Crystal Growth & Design, vol. 4, pp. 621-632, 2004.

[217] N. D. Draper, R. J. Batchelor, P. M. Aguiar, S. Kroeker, and D. B. Leznoff, "Factors affecting the solid-state structure and dimensionality of mercury cyanide/chloride double salts, and NMR characterization of coordination ge ometries," Inorganic Chemistry, vol. 43, pp. 6557-6567, 2004.

~~[218] R.-Y. Liau, H. Ehlich, A. Schier, and H. Schmidbaur, "Bis(triphenylphosphoranylidene)ammonium dicyanoaurate(l)," Zeitschrift fur Naturforschung. B, vol. 57b, no. 10, pp. 1085-1089, 2002.~~

~~[219] I. V. Markov, Crystal growth for beginners : fundamentals of nucleation, crystal growth, and epitaxy. New Jersey: World Scientific, 1995.~~

[220] J. Hulliger, "Chemistry and crystal growth," Angewandte Chemie, International Edition in English, vol. 33, pp. 143-162, 1994. [221] O. M. Yaghi, G. Li, and H. Li, "Crystal growth of extended solids by nonaque ous gel diffusion," Chemistry of Materials, vol. 9, pp. 1074-1076, 1997. and references therein. References 286

[222] P. Batail, K. Boubekeur, M. Fourmigu, and J.-C. P. Gabriel, "Electrocrystalliza- tion, an invaluable tool for the construction of ordered, electroactive molecular solids," Chemistry of Materials, vol. 10, pp. 3005-3015, 1998. [223] G. Demazeau, "Solvothermal processes: a route to the stabilization of new materials," Journal of Materials Chemistry, vol. 9, pp. 15-18, 1999.

[224] K. Byrappa and M. Yoshimura, Handbook of hydrothermal technology - Tech nology for crystal growth and materials processing. Materials Science and Pro cessing Technologies Series, New Jersey, U.S.A: Noyes Publications, 2001.

~~[225] R. H. Laye and E. J. Mclnnes, "Solvothermal synthesis of paramagnetic molec ular clusters," European Journal of Inorganic Chemistry, vol. 2004, pp. 2811— 2818, 2004.~~

~~[226] S. Feng and R. Xu, "New materials in hydrothermal synthesis," Accounts of Chemical Research, vol. 34, pp. 239-247, 2001.~~

~~[227] C. S. Cundy and P. A. Cox, "The hydrothermal synthesis of zeolites: History and development from the earliest days to the present time," Chemical Reviews, vol. 103, pp. 663-702, 2003.~~

[228] R. Murugavel, M. G. Walawalkar, M. Dan, H. W. Roesky, and C. N. R. Rao, "Transformations of molecules and secondary building units to materials: A bottom-up approach," Accounts of Chemical Research, vol. 37, pp. 763-774, 2004.

~~[229] J. Livage, "Synthesis of polyoxovanadates via 'chimie douce'," Coordination Chemistry Reviews, vol. 178-180, pp. 999-1018, 1998.~~

~~[230] A. K. Cheetham, C. N. R. Rao, and R. K. Feller, "Structural diversity and chemical trends in hybrid inorganic-organic framework materials," Chemical Communications, pp. 4780-4795, 2006.~~

[231] X.-M. Chen and M.-L. Tong, "Solvothermal in situ metal/ligand reactions: A new bridge between coordination chemistry and organic synthetic chemistry," Accounts of Chemical Research, vol. 40, pp. 162-170, 2007. [232] X.-M. Zhang, "Hydro(solvo)thermal in situ ligand syntheses," Coordination Chemistry Reviews, vol. 249, pp. 1201-1219, 2005.

~~[233] E. R. Jette and F. Foote, "Precision determination of lattice constants," Journal of Chemical Physics, vol. 3, pp. 605-616, 1935. References 287~~

[234] S. Sasaki, K. Fujino, and Y. Takeuchi, "X-ray determination of electron-density distributions in oxides, MgO, MnO, CoO, and NiO, and atomic scattering fac tors of their constituent atoms.," Proceedings of the Japan Academy, vol. 55, pp. 43-48, 1979.

~~[235] M. Bienzle, T. Oishi, and F. Sommer, "Thermodynamics and local atomic ar rangements of gold-nickel alloys," Journal of Alloys and Compounds, vol. 220, pp. 182-188, 1995.~~

[236] R. Brayner, T. Coradin, M.-J. Vaulay, C. Mangeney, J. Livage, and F. Fievet, "Preparation and characterization of metal (Au)- and bimetallic alloys (AuNi)- gelatin nanocomposites," Colloids and Surfaces A: Physicochemical and Engi neering Aspects, vol. 256, pp. 191-197, 2005.

[237] B. Rellinghaus, S. Stappert, E. Wassermann, H. Sauer, and B. Splietho, "The effect of oxidation on the structure of nickel nanoparticles," European Physical Journal D, vol. 16, pp. 249-252, 2001.

[238] Y.-w. Jun, J.-H. Lee, J.-s. Choi, and J. Cheon, "Symmetry-controlled colloidal nanocrystals: Nonhydrolytic chemical synthesis and shape determining para meters," Journal of Physical Chemistry B, vol. 109, pp. 14795-14806, 2005. and references therein.

~~[239] E. Matijeviec, "Preparation and properties of uniform size colloids," Chemistry of Materials, vol. 5, pp. 412-426, 1993.~~

[240] D. V. Talapin, A. L. Rogach, M. Haase, and H. Weller, "Evolution of an en semble of nanoparticles in a colloidal solution: Theoretical study," Journal of Physical Chemistry B, vol. 105, pp. 12278-12285, 2001.

[241] M. B. Mohamed, Z. L. Wang, and M. A. El-Sayed, "Temperature-dependent size-controlled nucleation and growth of gold nanoclusters," Journal of Physical Chemistry A, vol. 103, pp. 10255-10259, 1999.

[242] P. K. Gallagher and J. P. Luongo, "The thermal decomposition of K2Pt(CN)4Br2 and K2Pt(CN)4," Thermochimica Acta, vol. 12, no. 2, pp. 159- 164, 1975. [243] M. Vondrova, C. M. Burgess, and A. B. Bocarsly, "Cyanogel coordination poly mers as precursors to transition metal alloys and intermetallics - from tradi tional heating to microwave processing," Chemistry of Materials, vol. 19, no. 9, pp. 2203-2212, 2007. References 288

[244] B. Folch, J. Larionova, Y. Guari, L. Datas, and C. Gurin, "A coordination poly mer precursor approach to the synthesis of NiFe bimetallic nanoparticles within hybrid mesoporous silica," Journal of Materials Chemistry, vol. 16, pp. 4435- 4442, 2006. [245] M. Vondrova, T. Klimczuk, V. L. Miller, B. W. Kirby, N. Yao, R. J. Cava, and A. B. Bocarsly, "Supported superparamagnetic Pd/Co alloy nanoparti cles prepared from a silica/cyanogel Co-gel," Chemistry of Materials, vol. 17, pp. 6216-6218, 2005.

[246] D. H. M. Buchold and C. Feldmann, "Synthesis of nanoscale Co3[Co(CN)6]2 in reverse microemulsions," Chemistry of Materials, vol. 19, pp. 3376-3380, 2007. [247] C. L. Carnes, J. Stipp, and K. J. Klabunde, "Synthesis, characterization and adsorption studies of nanocrystalline copper oxide and nickel oxide," Langmuir, vol. 18, pp. 1352-1359, 2002. [248] S. A. Makhlouf, F. T. Parker, F. E. Spada, and A. E. Berkowitz, "Magnetic anomalies in NiO nanoparticles," Journal of Applied Physics, vol. 81, pp. 5561- 5563, 1997. [249] C. P. Bean and J. D. Livingston, "Superparamagnetism," Journal of Applied Physics, vol. 30, pp. S120-S129, 1959. [250] C. Kittel, "Theory of the structure of ferromagnetic domains in films and small particles," Physical Review, vol. 70, pp. 965-971, 1946. [251] E. C. Stoner and E. P. Wohlfarth, "A mechanism of magnetic hysteresis in het erogeneous alloys," Philosophical Transactions of the Royal Society of London, Series A: Physical Sciences and Engineering, vol. 240, pp. 599-642, 1948.

~~[252] J. R. L. de Almeida and D. J. Thouless, "Stability of the Sherrington- Kirkpatrick solution of a spin glass model," Journal of Physics A: Mathematical and General, vol. 11, pp. 983-990, 1978.~~

~~[253] M. Ghosh, K. Biswas, A. Sundaresan, and C. N. R. Rao, "MnO and NiO nanoparticles: synthesis and magnetic properties," Journal of Materials Chem istry, vol. 16, pp. 106-111, 2006.~~

[254] R. Kuentzler and J. P. Kappler, "Heat capacity and magnetization of gold-nickel alloys," Journal of Physics F: Metal Physics, vol. 9, pp. 195-206, 1979. [255] A. R. Kaufman, S. T. Pan, and J. R. Clark, "Magnetization of gold-iron and gold-nickel solid solutions," Reviews of Modern Physics, vol. 17, pp. 87-92, 1945. References 289

[256] H. Hori, T. Teranishi, M. Taki, S. Yamada, M. Miyake, and Y. Yamamoto, "Magnetic properties of nano-paraticles of Au, Pd and Pd/Ni alloys," Journal of Magnetism and Magnetic Materials, vol. 226-230, pp. 1910-1911, 2001.

[257] P. Dutta, S. Pal, M. S. Seehra, M. Anand, and C. B. Roberts, "Magnetism in dodecanethiol-capped gold nanoparticles: Role of size and capping agent," Applied Physics Letters, vol. 90, p. 213102, 2007.

[258] P. Crespo, R. Litran, T. C. Rojas, M. Multigner, J. M. de la Fuente, J. C. Sanchez-Lopez, M. A. Garcia, A. Hernando, S. Penades, and A. Fernandez, "Permanent magnetism, magnetic anisotropy, and hysteresis of thiol-capped gold nanoparticles," Physical Review Letters, vol. 93, p. 087204, 2004.

[259] J. L. Dormann, L. Bessais, and D. Fiorani, "A dynamic study of small interact ing particles: superparamagnetic model and spin-glass laws," Journal of Physics C: Solid State Physics, vol. 21, pp. 2015-2034, 1988.

[260] F. Dumestre, B. Chaudret, C. Amiens, M.-C. Fromen, M.-J. Casanove, P. Re- naud, and P. Zurcher, "Shape control of thermodynamically stable cobalt nanorods through organometallic chemistry," Angewandte Chemie, Interna tional Edition in English, vol. 41, pp. 4286-4289, 2002.

~~[261] P. Scherrer, "Gottinger nachrichten," in Kolloidchemie (R. Zsigmondy, ed.), p. 394, 3rd edition 1920 ed., 1918.~~

~~[262] M. Abramoff, P. Magelhaes, and S. Ram, "Image processing with ImageJ," Biophotonics International, vol. 11, no. 7, pp. 36-42, 2004.~~

~~[263] Z. Assefa, M. A. Omary, B. G. McBurnett, A. A. Mohamed, H. H. Patterson, R. J. Staples, and J. Fackler, John P., "Syntheses, structure, and photolumines-~~

~~cence properties of the 1-dimensional chain compounds [(TPA)2Au][Au(CN)2] and (TPA)AuCl (TPA = l,3,5-triaza-7-phosphaadamantane)," Inorganic Chemistry, vol. 41, no. 24, pp. 6274-6280, 2002.~~

~~[264] M. Stender, R. L. White-Morris, M. M. Olmstead, and A. L. Balch, "New structural features of unsupported chains of metal ions in luminescent~~

~~[(NH3)4Pt][Au(CN)2]2-1.5(H20) and related salts," Inorganic chemistry, vol. 42, no. 15, pp. 4504-4506, 2003.~~

~~[265] P. Giitlich, Y. Garcia, and H. A. Goodwin, "Spin crossover phenomena in Fe(ll) complexes," Chemical Society Reviews, vol. 29, pp. 419-427, 2000. References 290~~

[266] J. A. Real, A. B. Gaspar, V. Niel, and M. C. Mufioz, "Communication between iron(ll) building blocks in cooperative spin transition phenomena," Coordination Chemistry Reviews, vol. 236, pp. 121-141, 2003.

[267] V. Niel, A. L. Thompson, M. C. Mufioz, A. Galet, A. E. Goeta, and J. A. Real, "Crystalline-state reaction with allosteric effect in spin-crossover, inter penetrated networks with magnetic and optical bistability," Angewandte Chemie International Edition, vol. 42, pp. 3760-3763, 2003. [268] A. Galet, M. C. Mufioz, and J. A. Real, "Coordination polymers undergoing spin crossover and reversible ligand exchange in the solid," Chemical Commu nications, pp. 4321-4323, 2006. Index

n [ Bu4N][Au(CN)2]-0.5H2O Experimental details, 96 Synthesis, 89 Experimental setup, 29 ["Bu4N]{Co[Au(CN)2]3} Principles, 28, 29 Magnetism, 189 {Co[Au(CN)2]2(H20)2}- n Structure, 174 [ Bu4N][Au(CN)2] Synthesis, 204 Magnetism, 189 UV-vis-NIR spectroscopy, 182 Structure, 178 ["Bu4N]{Ni[Au(CN)2]3} Synthesis, 204 Magnetism, 187 UV-vis-NIR spectroscopy, 182 Structure, 174 Synthesis, 203 Aurophilicity, 7 UV-vis-NIR spectroscopy, 181 Baker model [PPN][Au(CN)2] General equation, 56 Synthesis, 201 Brillouin function, 11 [PPN]{Co[Au(CN)2]3} Magnetism, 189 Co(/>OH2)2[Au(CN)2]2 Structure, 172 /^SR results, 72 Synthesis, 203 SQUID results, 67-68 UV-vis-NIR spectroscopy, 182 Structure, 45 Synthesis, 91 [PPN]{Ni[Au(CN)2]3} Magnetism, 187 Co[Au(CN)2]2(DMSO)2 Structure, 172 Magnetism, 137 Synthesis, 202 Structure, 125-128 UV-vis-NIR spectroscopy, 181 Synthesis, 160 /xSR UV-vis-NIR spectroscopy, 135 Co[Au(CN) ] (pyridine) Co(^-OH2)2[Au(CN)2]2, 72 2 2 2 Magnetism, 137 Cu[Au(CN)2]2, 59-61 Structure, 131-132 Cu(Ai-OH2)2[Au(CN)2]2, 57-59 Synthesis, 160 Fe(M-OH2)(>-OH)[Au(CN)2]2, 72 UV-vis-NIR spectroscopy, 135 Fe(/x-OH2)2[Au(CN)2]2, 72 Co[Au(CN) ] (DMSO) Mn(/^-OH2)2[Au(CN)2]2, 72 4 2 4 Ni(^-OH2)2[Au(CN)2]2, 65-67 Structure, 154 Synthesis, 161

~~291 Index 292~~

Crystallographic data Experimental details, 87 [cation]{M[Au(CN)2]3}, 206 Principles, 19 M[Au(CN)2]2(analyte):c, 161-162 FeGu-OH2)2[Au(CN)2]2 Ni(/U-OH2)2[Au(CN)2]2, 96 Atomic coordinates, 241-263 /xSR results, 72 SQUID results, 69-71 Cu(/x-OH2)2[Au(CN)2]2 //SR results, 57-59 Structure, 45 SQUID results, 55-57 Synthesis, 91 Structure, 44 Fe(/x-OH2)(A*-OH)[Au(CN)2]2 Synthesis, 90 /uSR results, 72 SQUID results, 71 Cu[Au(CN)2]2 /iSR results, 59-61 Structure, 48 Synthesis, 90 Synthesis, 92 Fisher 1-D model, 69 Cu[Au(CN)2]2(CH3CN)2 Synthesis, 158 Hydrothermal synthesis Cu[Au(CN)2]2(dioxane)(H20) Experimental details, 87 Magnetism, 122 General procedure, 87 Synthesis, 158 Principles, 211 Cu[Au(CN)2]2(DMF) Magnetism, 122 Infrared spectroscopy Structure, 113 Experimental details, 87 Synthesis, 157 Principles, 19 Cu[Au(CN)2]2(DMSO)2 (blue) Magnetism, 122 K{Fe[Au(CN)2]3} Structure, 102 Mossbauer spectroscopy, 191 Synthesis, 156 Magnetism, 190 Cu[Au(CN) ] (DMSO) (green) Synthesis, 201 2 2 2 K{Ni[Au(CN) ] } Magnetism, 122 2 3 Structure, 101 Magnetism, 186 Synthesis, 156 Structure, 169 Synthesis, 200 Cu[Au(CN) ] (NH ) 2 2 3 4 UV-vis-NIR spectroscopy, 181 Magnetism, 122 Synthesis, 159 Mossbauer spectroscopy Cu[Au(CN)2]2(pyridine)2 Fe(^-OH2)(/x-OH)[Au(CN)2]2, 47 Magnetism, 122 Fe(/i-OH2)2[Au(CN)2]2, 36 Structure, 116 K{Fe[Au(CN)2]3}, 191 Synthesis, 157 Experimental Details, 95 Curie-Weiss equation, 13 Principles, 21 Magnetic properties Elemental analysis [cation]{M[Au(CN)2]3}, 186-192 Index 293

Cu[Au(CN)2]2(analyte)x, 122-124 Ni[Au(CN)2]2(pyridine)4 M[Au(CN)2]2(analyte);c, 137 Synthesis, 132, 160 M(/i-OH2)2[Au(CN)2]2, 54-72 Hydrothermal product (125 °C), 212 Spin-orbit coupling Hydrothermal product (135 °C), 222 Principle, 18 Hydrothermal product (165 °C), 219 Superparamagnetism Magnetism Anisotropy energy, 228 Experimental details, 89 de Almeida-Thouless law, 230 Introduction, 9 Principles, 227 Principles of SQUID Magnetometry, Synthesis 26 [cation]{M[Au(CN)2]3}, 200-205 M[Au(CN) ] (analyte) , 156-161 Mn(/Li-OH2)2[Au(CN)2]2 2 2 a; /xSR results, 72 M(//-OH2)2[Au(CN)2]2, 90-92 SQUID results, 69-71 Hydrothermal products, 235 Structure, 45 Thermal stability Synthesis, 92 Cu[Au(CN)2]2(DMSO)2, 107 Nanoparticles CuiAuiCNJ^Canalyte)*, 120 Chemical identity, 224 M(At-OH2)2[Au(CN)2]2, 49 Formation route, 226 [cation] {M[Au(CN)2]3}, 184-186 Hydrothermal product (125 °C), 215 Thermogravimetric analysis Hydrothermal product (135 °C), 222 Experimental details, 87 Hydrothermal product (165 °C), 219 Principles, 19 Transmission electron microscopy Ni(/x-OH2)2[Au(CN)2]2 //SR results, 65-67 Experimental details, 236 SQUID results, 61-65 Hydrothermal product (125 °C), 215 Structure, 40 Hydrothermal product (135 °C), 222 Synthesis, 90 Hydrothermal product (165 °C), 219 Ni[Ag2(CN)3][Ag(CN)2] UV-vis-NIR spectroscopy Synthesis, 93 [cation]{M[Au(CN)2]3}, 181-184 Ni[Au(CN)2]2(DMF)2 Co[Au(CN)2]2(analyte)2, 135-137 Magnetism, 137 Cu[Au(CN)2]2(analyte)x, no Structure, 128-131 Ni[Au(CN)2]2(analyte)2, 134-135 Synthesis, 159 Experimental details, 87 UV-vis-NIR spectroscopy, 134 Principles, 20 Ni[Au(CN)2]2(DMSO)2 Magnetism, 137 Vapochromic behaviour Structure, 124-125 Co[Au(CN)2]2(analyte)2, 140 Synthesis, 159 Cu[Au(CN)2]2(analyte)a:, 109-113 UV-vis-NIR spectroscopy, 134 Ni[Au(CN)2]2(analyte)2, 142 Index 294

~~X-ray diffraction Experimental details (powder), 88 Experimental details (single crystal), 88 Principles (powder), 25 Principles (single crystal), 25~~

~~Zero-field splitting General equation, 16 Principle, 16~~