SYNTHESIS AND CHARACTERIZATION OF TRIDECAMERIC GROUP 13 HYDROXIDE CLUSTERS by ZACHARY LEE MENSINGER A DISSERTATION Presented to the Department of Chemistry and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor ofPhilosophy September 2010 11 University of Oregon Graduate School Confirmation of Approval and Acceptance of Dissertation prepared by: Zachary Mensinger Title: "Synthesis and Characterization of Tridecameric Group 13 Hydroxide Clusters" This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Chemistry by: Victoria DeRose, Chairperson, Chemistry Darren Johnson, Member, Chemistry James Hutchison, Member, Chemistry Michael Haley, Member, Chemistry Raghuveer Parthasarathy, Outside Member, Physics and Richard Linton, Vice President for Research and Graduate Studies/Dean of the Graduate School for the University of Oregon. September 4,2010 Original approval signatures are on file with the Graduate School and the University of Oregon Libraries. © 2010 Zachary Lee Mensinger 111 IV An Abstract of the Dissertation of Zachary Lee Mensinger for the degree of Doctor ofPhilosophy in the Department of Chemistry to be taken September 2010 Title: SYNTHESIS AND CHARACTERIZATION OF TRIDECAMERIC GROUP 13 HYDROXIDE CLUSTERS Approved: Professor Darren W. Johnson In the research area of Group 13 hydroxide clusters, progress is often hampered by difficult and inefficient synthetic procedures. This has greatly limited the numerous potential applications of Group 13 hydroxide compounds, many ofwhich require large amounts ofmaterial. Most relevant to this dissertation is their application as precursors for high quality amorphous metal oxide thin films. Addressing this issue, this dissertation presents a series of Group 13 containing hydroxide compounds of general formula [M13(u3-0H)6(u-OH)18(H20h4](N03)15 which are generated through an efficient, scalable synthetic procedure. Throughout this dissertation, the compounds are generally referred to by their metal content, i.e. [Ga13(u3-0H)6(u-OH)18(H20h4](N03)15 is designated as Ga13. Chapter I reviews the literature of inorganic and ligand-supported vGroup 13 hydroxide compounds with the aim of identifying common structural trends in metal composition and coordinating ligands. This summary is limited to clusters of aluminum, gallium, and indium. Chapter II describes in detail the synthesis and characterization of one such cluster, AI l3 . Following this in Chapter III is the description of the first heterometallic Group 13 hydroxide compound, Ga7In6, which along with Gal3 was used as a precursor material for metal oxide thin films in collaboration with Professor Doug Keszler at Oregon State University. Chapter IV describes a series of six GalIn compounds, as well as two AllIn compounds. Included in this chapter is an analysis of the heat-induced decomposition properties of the GalIn clusters. Understanding such thermal decomposition is particularly relevant for the use of these compounds as precursor materials, as an annealing step is used to condense the films. Chapter V addresses the potential for post-synthetic modification of the compounds through metal and ligand exchange reactions, an area that also addresses the issue of solution stability ofthe structures. Chapter VI describes the synthesis and characterization of related Group 13 compounds, including two infinite chain structures and additional heterometallic compounds. Lastly, Chapter VII concludes this dissertation and discusses potential areas of future research. This dissertation includes co-authored material and previously published results. VI CURRICULUM VITAE NAME OF AUTHOR: Zachary Lee Mensinger GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene, Oregon Macalester College, Saint Paul, Minnesota DEGREES AWARDED: Doctor of Philosophy, Inorganic Chemistry, 2010, University of Oregon Bachelors of Arts, Chemistry, Biology (Minor), 2005, Macalester College AREAS OF SPECIAL INTEREST: Inorganic Metal Hydroxide Chemistry Supramolecular Chemistry Acetylene Chemistry PROFESSIONAL EXPERIENCE: Graduate Research Assistant, Prof. DalTen W. Johnson, University of Oregon, June 2006 - September 2010 Visiting Research Assistant, Prof. Pablo Ballester, Institute of Chemical Research of Catalonia, TalTagona, Spain, March - June 2009 Graduate Teaching Fellow, University of Oregon, September 2005 - June 2006 Chemical Research and Development Intern, Dr. Peter G.M. Wuts, Pfizer, Inc., May - August 2005 Research Assistant, Prof. Ronald G. Brisbois, Macalester College, September 2003 - May 2005 Vll Research Experience for Undergraduates (REU) Fellow, Prof. Paul A. Gl;eco, Montana State University, May - August 2003 GRANTS, AWARDS AND HONORS: IGERT Traineeship, University of Oregon, 2008-2010 PUBLICAnONS: Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Synthesis and Crystallization of Infinite Indium and Gallium Acetate 1D Chain Structures and Concomitant Ethyl Acetate Hydrolysis" Inorg. Chern. 2009,48,3505-3507. Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Kesz1er, D. A.; Johnson, D. W. "Heterometallic Group 13 Nanoc1uster Synthesis and Inks for Oxide TFTs" Angew. Chern. Int. Ed. 2008,47,9484-9486. (Highlighted as UO press release and in Popular Mechanics - New Zealand) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric AIl3 Nanoc1uster [All3(u3-0H)6(j12- OH)18(H20h4](N03)IS" Inorg. Chern. 2008, 47(4), 1267-1269. Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Triammonium hexahydroxido- octadecaoxidohexamolybdogallate(III) heptahydrate" Acta Cryst. 2008, E64(2), i8-i9. V111 ACKNOWLEDGMENTS It would be easy to double the length of this dissertation by thanking all the people who have contributed to making the past five years such a memorable time. They have been a time of immense personal growth, intellectual curiosity, and invaluable experiences. First and foremost, Prof. DalTen W. Johnson is thoroughly acknowledged for an excellent graduate school experience. His endless creativity and excitement for science is inspirational while maintaining a rich life outside of the lab, a balance even more impressive and rare. DmTen is also a barbecue master, and such a debt of delicious meat could never be repaid. I thank all the members of the Johnson lab, past and present, who have been there through many ups and downs. In particular, Tim Carter, Nate Lindquist, and Aaron Sather, were great labmates and friends. I have had the pleasure of working with wonderful colleagues; Dr. Lev Zakharov, Dr. Jason Gatlin, Maisha Kamunde-Devonish, Matt Baseman, Sharon Betterton, Dr. Stephen Meyers, Laura Thompson, Anna Oliveri, Kurtis Fairley, Prof. Doug Keszler, and many others. I express sincere gratitude to those who proofread material for this dissertation. Past advisors and mentors Prof. Ron Brisbois, Dr. Peter Wuts, Prof. Gautam Bhattacharyya, Prof. Paul Grieco, and Prof. Pablo Ballester cannot be thanked enough for their role in my education and development. The members of my committee are acknowledged for expecting nothing but my best. I appreciate everyone at the ICIQ in Tarragona, Spain, where I spent three wonderful months. I felt welcomed into your lab, country, culture, and community; I loved them and miss them all. I also thank the funding agencies that have IX provided generous financial support for this research: the National Science Foundation for an IGERT traineeship (including an IGERT-Intemational supplemental grant to support my time at the ICIQ in Tarragona, Spain), as well as the Center for Green Materials Chemistry. The Army Research Lab also provided funding for this research. The University of Oregon is graciously acknowledged for startup funds, a GTF appointment, and world-class research facilities. Looking forward, I thank my future advisors, Prof. Jeff Long, Dr. Stan Prusiner, and Prof. Holger Wille. They have offered me the opportunity to work on a challenging project at the forefront of prion research at some of the best chemistry facilities in the world. I look forward to the challenge. On a more personal level, I am grateful to all the people who have enriched life outside the lab. Erich Chapman has been a friend and confidant since the beginning, as well as the other great members of my class. For Duck sports, Max's trivia, and photography, I can always tum to Adam Hopkins. Friends like Paula Matano, Silke Friedrich, Elin McIlhattan, Jeff Mack, Mike Jespersen, Nathan Lien, Teresa Roark, McCailin Wunder, and many more have enriched my life through shared adventures. Sean Fontenot would be on my short list of people to find in the event of the Zombie Apocalypse. I thank the city of Eugene and the state of Oregon for providing a beautiful, unique place to study, learn, and grow as a person. I'm glad that Ashlee Trueb decided to take a chance on Zack the Chemist; if Dr. Sweeney asked again, I hope I was worth it. Lastly, I thank my parents who have offered non-stop love and support, even when they might not quite understand what I'm working on. I know I don't have to worry, but I hope to keep making you proud. I dedicate this dissertation to my friends and family; without your support I could never have become the person I am proud to be today. x Xl TABLE OF CONTENTS Chapter Page I. OLIGOMERIC GROUP 13 HYDROXIDE COMPOUNDS - A RARE BUT VARIED CLASS OF MOLECULES 1 Introduction............................................................................................................ 1 Aluminum Clusters in the Environment 2 Gallium and Indium......................................................................................... 4 Inorganic Hydroxide Clusters 5 Inorganic Aluminum Clusters.......................................................................... 6 Inorganic Gallium Clusters 10 Ligand Bound Hydroxide Clusters 13 Ligand Bound Aluminum Clusters.................................................................. 13 Ligand Bound Gallium Clusters 23 Ligand Bound Indium Clusters......................................... 27 Heterometallic Clusters.......................................................................................... 28 Discussion 30 Bridge to Chapter II 33 II. FACILE SYNTHESIS OF THE TRIDECAMERIC Al 13 NANOCLUSTER [AI 13(u3-0H)6(u-OH)18(H20h4] (N03) 15 34 Introduction............................................................................................................ 34 Related Group 13 Hydroxide Clusters................................................................... 36 Synthetic Procedure for Tridecameric Al Hydroxide 37 Structural Description of Tridecameric Al Hydroxide 40 Conclusion............................................................................................................. 41 Crystallographic Methods and Summary of Data.................................................. 42 Bridge to Chapter III.............................................................................................. 44 Chapter xii Page III. SYNTHESIS OF HETEROMETALLIC GROUP 13 NANOCLUSTERS AS INKS FOR OXIDE THIN-FILM TRANSISTORS..................................................... 45 Introduction............................................................................................................ 45 Materials Precursor Interest....................... 48 Synthetic Procedures for Tridecameric Clusters.................................................... 48 Application as Precursor Inks for Thin-Film Transistors 51 Conclusion............................................................................................................. 54 Crystallographic Methods and Summary of Data.................................................. 55 Bridge to Chapter IV.............. 57 IV. PREPARATION OF A SERIES OF NOVEL HETEROMETALLIC TRIDECAMERIC GROUP 13 HYDROXIDE CLUSTERS 58 Introduction............................................................................................................ 58 Experimental Details.................................................... 62 Synthesis of GalIn Heterometallic Compounds............................................... 63 Synthesis of AllIn Heterometallic Compounds 64 Description of Characterization Methods 65 Description of Synthesized Compounds................................................................ 66 Structural Description of GalIn Compounds 69 Structural Description of AllIn Compounds 72 Variable-temperature Powder X-ray Diffraction 73 Discussion of Cluster Formation 78 Conclusion............................................................................................................. 81 Blidge to Chapter V 82 V. POST-SYNTHETIC MODIFICATION AND POTENTIAL CHEMICAL APPLICATIONS OF CLUSTER COMPOUNDS 83 Introduction............................................................................................................ 83 Summary of Experimental Approach 86 Chapter X111 Page Metal Exchange Background 88 Metal Exchange Procedure.............................................................. 90 Potential Route to New Cluster Compositions 92 Ligand Exchange Background 94 Heterogeneous Ligand Exchange 97 Homogeneous Ligand Exchange 100 Biphasic Ligand Exchange 102 Conclusion 103 Bridge to Chapter VI.............................................................................................. 105 VI. ISOLATION OF TWO ADDITIONAL CLASSES OF GROUP 13- CONTAINING MOLECULES 106 Introduction............................................................................................................ 106 Interest in Infinite Group 13 Networks 107 Synthetic Procedure for 1-D Chains........................ 109 Structural Description of Chain Molecules......................... 110 Hydrolysis of Ethyl Acetate 112 Conclusion to Group 13 Chain Compounds...... 113 Summary of Crystallographic Data 114 Description of Anderson Clusters.......................................................................... 114 Synthetic Preparation of Substituted Anderson Clusters 116 Crystallographic Details............................. 117 Conclusion to Ga- and AI-substituted Anderson Molybdates 120 Bridge to Chapter VII............................................................................................ 121 VII. CONCLUSIONS AND FUTURE DIRECTIONS 122 Introduction 122 Research SUlnlnary 123 Future Work 124 Chapter XIV Page Oxidation Chemistry........................................................................................ 124 Metal Exchange 125 Anion Exchange 126 Nanocomposite Formation 127 Ligand Exchange 127 Additional EXAFS Studies 128 Interaction ofM13 Clusters with Nucleic Acids 129 REFERENCES 13] xv LIST OF FIGURES CHAPTER I 1.1 Depiction of the crystal structure of A18 7 1.2 Structural depiction of the crystal structure of Ah3 8 1.3 Structural depiction of the crystal structure of Gal3 11 1.4 Structural depiction of the crystal structure of Abheidh 14 1.5 Depictions of the crystal structures of Abntaz and Abidaz.... 16 1.6 Dimeric aluminum species bound by ethyl acetate ligands 16 1.7 Dimeric aluminum hydroxide complex with hbo ligands............................ 18 1.8 Depictions of two different trimeric structures with citrate......................... 20 1.9 Tetrameric aluminum structure.................................................................... 21 1.10 Tridecameric aluminum cluster bound by heidi ligands 22 1.11 Dimeric gallium compound bound by heidi ligands 24 1.12 Ligands bound to an octameric gallium hydroxide core............................ 25 1.13 Highest nuclearity gallium hydroxide ligand-supported structure............. 26 1.14 Heterometallic compound consisting of gallium and indium atoms 29 1.15 Summary ofligands in ligand-supported compounds 31 CHAPTER II 2.1 Comparison figure of the "flat" A1 l3 and s-Keggin AI l3 .............................. 36 2.2 Comparison of post-reaction conditions 39 2.3 Depiction of hydrogen bonding 42 CHAPTER III 3.1 Structural representation of tridecameric Group 13 hydroxide clusters...... 46 3.2 Representative characteristics for a bottom-gate IGO-channel TFT 52 XVI ~~ P~ 3.3 Thin-film XRD pattern and SEM image of an Ino.92Ga1.Og03 film............... 53 3.4 Process to deposit IGO films from heterometallic precursors 54 CHAPTER IV 4.1 Top and side view of a tridecameric cluster 60 4.2 Three different metal centers are found in the clusters................................ 70 4.3 Different potential arrangements of the outer ring metal centers 72 4.4 Summary of tridecameric clusters 73 4.5 X-ray diffraction traces for Gal3 75 4.6 X-ray diffraction traces for GalOIn3 76 4.7 X-ray diffraction traces for Ga9In4............................................................... 76 4.8 X-ray diffraction traces for GagIns............................................................... 77 4.9 X-ray diffraction traces for Ga7In6............................................................... 77 4.10 28 values of five strong ~-Ga203 peaks 78 4.11 NMR spectra of di-n-butylnitrosamine...................................................... 79 CHAPTER V 5.1 Representation of Keggin-AI 13, Anderson, and M 13 clusters....................... 89 5.2 Depiction of metal exchange process 91 5.3 Preliminary EXAFS data 93 5.4 Functionalization ofM13 compound with organic ligands 96 5.5 NMR spectra from ligand exchange reactions............................................. 99 CHAPTER VI 6.1 Ortep visualization (50 % probability) ofIn-acetate I-D chain 110 6.2 Ortep representation (50 % probability) of the gallium I-D chain III 6.3 ChemDraw representation of the GaM06 anion 116 6.4 One of the two symmetrically independent GaM06 clusters 118 Figure XVll Page 6.5 The second polyanionic cluster present in GaMo6 •..... 119 6.6 PyMol representation of the crystal structure of GaMo6 121 XVlll LIST OF TABLES Table Page Chapter IV 4.1 Characterization data for cluster compounds............................................... 68 XIX LIST OF GRAPHS Graph Page Chapter IV 4.1 Graph used for GalIn ratio prediction of missing clusters 67 xx LIST OF SCHEMES Scheme CHAPTER II Page 2.1 Synthetic scheme depicting the synthetic preparation of the Al 13 compound............................................................................................................ 35 CHAPTER III 3.1 Synthetic preparation of Gal3 and Ga7In6 by two different methods 49 CHAPTER IV 4.1 Starting material ratios for GalIn clusters 64 CHAPTER VI 6.1 General synthetic route to 1-D acetate chain compounds............................ 108 1CHAPTER I OLIGOMERIC GROUP 13 HYDROXIDE COMPOUNDS - A RARE BUT VARIED CLASS OF MOLECULES Introduction Trivalent Group 13 metal ions such as aluminum, gallium, and indium exist as a wide and dynamic range of species in solution where they have been said to possess a "bewildering variety of behavior". 1 Aluminum has been by far the most widely studied of the group, with gallium and indium receiving less scrutiny. Much of the work in this area has focused on solution studies, with several recent reviews on this subject.2-5 In contrast to the large body of work aimed at studying the complex solution speciation, the number of solid state examples of Group 13 hydroxides is relatively limited. This chapter seeks to provide a thorough summary of solid state Group 13 (aluminum, gallium, and indium) structures containing only octahedral metal ions bridged by hydroxide ligands. Both ligand-supported and fully inorganic clusters will be discussed. Ligand-supported compounds are those comprised of an inorganic metal hydroxide core with organic ligands (often aminocarboxylates) filling the peripheral coordination sites of the cluster ions.6-8 The fully inorganic clusters are exactly that: metal hydroxide clusters with water most commonly filling the coordination sites not saturated by bridging 2hydroxides. As means of providing context, several examples of cluster compounds containing mixtures of octahedral and tetrahedral metals are mentioned as well, particularly the much studied Keggin-AI 13 ions.2 The compounds discussed will not be limited solely to those isolated from aqueous solutions, as a significant number of recent compounds have been isolated from organic solutions as well. Solid state structures are the focus of this review, and therefore the primary means of characterization is single-crystal X-ray diffraction (XRD). The necessity to isolate crystals suitable for single-crystal diffraction studies has in all likelihood limited the number of structures reported, as our own experience has shown that successful crystallization of these compounds can sometimes be elusive. While characterization of these compounds by solution means is often difficult or inconclusive, methods such as NMR, powder X-ray diffraction, mass spectrometry, and extended X-ray absorption fine- edge spectroscopy have been used to garner both solution phase and solid state data on metal hydroxide clusters. Aluminum Clusters in the Environment There are many reasons for interest in Group 13 hydroxide compounds. For example, aluminum is highly environmentally relevant, due to its high natural abundance (the most abundant metal in the biosphere) and geochemical diversity.2 Aluminum itself has no known biological role, largely due to a lack ofbioavailability of its natural mineral forms. 9 However, as human activity has begun to produce large quantities of 3bioavailable aluminum, various environmental problems such as poisoning of fish in lakes and the dreaded "Waldsterben" in forests exposed to acid rain have been attributed to aluminum.9,10 In fact, several neutral Al(III) complexes with organic ligands have shown to have much higher toxicity than inorganic aluminum salts. I I Biological effects have been attributed to enzyme inhibition, nucleotide destruction, or mineral formation. To better understand these biological effects, research on aluminum aggregates has examined the similar hydrolytic behavior of iron and aluminum, seeking to establish whether Al (III) might be incorporated into biological systems through similar pathways to Fe(III). For instance, can aluminum be incorporated into biomolecules like ferritin?9 Synthetic aluminum complexes may also provide insight into iron-containing metalloenzymes. Also of environmental relevance, aluminum sols have been shown to transport heavy metal ions out of mining areas. The sols bind to heavy metal ions and the complexes remain soluble due to the acidic conditions of the mining areas. As the sol- containing water leaves mining areas, the pH rises upon contact with native waters, rendering the aluminum complexes insoluble, and floes precipitate out of solution with the heavy metal ions. 12 A similar process is used in a positive light to sequester heavy metals and other contaminants from drinking water supplies where precipitates are easily filtered off and removed. 13 Beyond their use in water purification, aluminum compounds have found applications in many other areas as well. Examples include the formation of ceramics (through clay pillaring agents) and antiperspirants where aluminum chloride salts are a common ingredient. 14 4Beyond these direct applications of aluminum complexes, another body of work has focused on the use of synthetic aluminum compounds as mineral mimics.2 The structure and reactivity ofmany minerals is poorly understood because the surface of bulk samples is often difficult to characterize.2 The use of synthetic small molecule analogues for these structures enables the prediction of some of the natural processes taking place at the surface of natural minerals, such as exchange of water ligands and oxygen atoms. The controlled nature of a laboratory and the smaller sizes of the cluster compounds can provide a clearer understanding of the more complicated processes in nature. One compound that fits into all of the above categories is the Keggin-AI 13 structure. 15 It has been used as a sequestration agent, clay pillaring agent, and mineral mimic. It has even been shown to form naturally in aqueous environments where it can be prevalent in high concentrations. 12 In this compound, the central tetrahedral aluminum metal ion can be exchanged, with Ga or Ge taking its place. 13,15-18 But aside from this prominent example, few Group 13 oxo/hydroxo compounds are known, with even fewer Group 13 hydroxides. Gallium and Indium Moving down the group, gallium and indium are far less abundant than aluminum and have been used and studied to a lesser extent. However, both have seen increasing use in the semiconductor industry, where they are key components in IUN semiconductors. Additionally, some state of the art solar cell technology utilizes copper- 5indium-gallium selenide (eIGS) as the photovoltaic layer. Devices using this technology currently show great promise for photovoltaic efficiency, with systems demonstrating nearly 20% efficiency having been prepared. 19 Similarly, the oxides of these materials, such as gallium oxide, indium oxide, and indium-gallium oxide, have been shown to be effective components of amorphous thin film semiconductors.2o,21 Recent efforts have demonstrated the use of Group 13 hydroxide clusters as precursors for materials of this type,21 A better understanding of the chemistry of hydroxide compounds ofgallium and indium may allow easier and more efficient isolation of important materials precursors.20 Inorganic Hydroxide Clusters It has generally proven difficult to isolate purely inorganic Group 13 metal hydroxide clusters. As higher nuclearity structures form, high charge densities build up, and compounds are progressively less stable. In the case of a tridecameric Group 13 hydroxide cluster with formula "M 13(OH)24(H20)24", a +15 charge is present, compared to the Keggin-AI13 clusters with only a +7 charge (oxide ligands lessen the charge). Also, the water ligands that generally fill the coordination spheres of some metal ions in inorganic structures exchange rapidly in solution, decreasing the stability of the complexes in solution. Despite this, some lovely structures have been isolated, and are described in the following sections. 6Inorganic Aluminum Clusters The first example of an aluminum hydroxide cluster was a simple edge-shared dimer bridged by two hydroxides.22 This structure, described in 1962 by Johansson, was proposed as a less basic salt intermediate compound to the larger Keggin-AI 13 • The compound was isolated as both the sulfate and selenate salt, and each possesses four H20 ligands that fill the coordination sites of the octahedral aluminum in addition to two waters of hydration to give structures of [Alz(u-OH)z(H20)s] (S04)z·2H20 and [Alz(u- OH)2(H20)s](Se04)z·2H20 (Ab). The sulfate was prepared by dissolving aluminum metal in sulfuric acid in the presence of a small amount of mercury catalyst, followed by filtration and crystallization of the resulting solution. The selenate was prepared by dissolving aluminum hydroxide in selenic acid. Individual clusters in each crystal structure hydrogen bond to each other through the oxygen atoms of the anions and water molecules. Following the isolation of this compound, it was another 36 years before another inorganic aluminum hydroxide oligomer was described in the solid state.23 The next larger oligomeric inorganic aluminum cluster to be isolated was [AIs(u3- OH)z(u-OH)12(H20hs](S04)s,16H20 (AIs).24 This complex is isolated by Casey et al. in a similar manner to the Alz structure mentioned above, however, a longer crystallization time eventually yields Als in low isolated yield. In this impressive complex, both edge- shared and comer-shared octahedra are found, with three different aluminum environments observed (Figure 1,1). The inner four aluminum ions are edge-sharing octahedra, bonded to each other through two ,u3-0H and two ,u-OH bridging ligands (four total ,u-OH ligands bridge the inner core). This inner core of four ions is bonded to four 7Figure 1.1. Depiction of the crystal structure of All!. An inner core of four edge- shared aluminum ions is found with four additional corner-shared aluminum ions capped by water ligands. Aluminum is teal, oxygen is red, and hydrogen is white; image generated with PyMol. corner-shared aluminum ions through eight p-OH bridges. The remaining coordination sites are filled by water ligands, one each on two of the edge-shared aluminum ions, and four water molecules bonded to each of the corner-shared aluminum ions. This structure is palticularly interesting as it represents the only non-ligand supported intermediate between dimeric Ab and tridecameric M 13 (M = AI or Ga) structures. which has been isolated as both the cr and NO)' salt (Figure 1.2).23.25 The original isolation of this structure occurred when an aqueous solution of basic aluminum chloride 8Figure 1.2. Structural depiction of the crystal structure of AID. This is the highest nuclearity inorganic aluminum hydroxide cluster. The heptameric central hydroxide core forms a planar center, while six additional aluminum ions are bridged by hydroxides and alternate positions above and below the central plane. Aluminum is teal, oxygen is red, and hydrogen is white; image generated with PyMol. was slowly evaporated, affording crystalline solid that contained mostly large crystals of AICI 3·6H20 in addition to a small amount of the AI 13 cluster. To synthesize this compound, AICi}·6H 20 was partially thermally decomposed at 180°C in a whirling air stream. Crystals suitable tor single crystal X-ray diffraction were isolated after tour months of evaporation open to atmosphere. 2J This cluster had been previously suggested to exist based on powder X-ray diffraction data, but absolute assignment had not been 9possible prior to the report by Seichter et al.26 Initially, positions of all hydrogen atoms could not be established, so the location of hydroxide vs. water ligands was inconclusive. The authors used charge balance to draw their conclusions: the thirteen Al(III) ions have a +39 charge, which was countered by 15 cr anions. This gave a remaining charge of +24, which is balanced by bridging hydroxide ligands. There are three different types of octahedral aluminum atoms present in this structure. The central aluminum atom is bonded to six ,u3-0H ligands. There are then six more aluminum atoms edge-shared with the central metal; each of these are bonded to two ,u3-0H bridges and four ,u-OH bridges. Beyond the planar core of seven edge-shared aluminum atoms, there are six additional aluminum atoms, which are bonded to the core by two ,u-OH bridges each. These outer shell metal ions alternate positions above and below the planar core of Ah(,u3-0H)6(,u- OHk The remaining coordination sites of the six exterior A13+ ions are occupied by four water molecules per aluminum ion, giving a total of 24 coordinated water molecules, with thirteen additional waters of hydration present.23 The cluster isolated as the nitrate salt by Gatlin et al. is isostructural to that reported by Seichter et aI., but the conditions for isolation are quite different. 25 AI(N03k9H20 was used as received, with no thennal decomposition procedure used. The aluminum salt was dissolved in methanol along with either di-n-butylnitrosamine or nitrosobenzene. A small amount of base (KOH, NaOH, NH40H, or AI(OH)3) was added to the reaction vial as well. After several weeks, crystals of [AI13Vt3-0H)6(u- OH)18(H20h4](N03)15 were isolated. The crystalline product is generally a mixture containing other inorganic nitrate salts, such as Al(N03h. In the case where NH40H is 10 used as the base, an additional positive charge is present from a NH4+ cation with an additional N03- present to balance the charge. This isolation represented the first intentional synthesis of this structure, and was achieved in a much faster time frame. Inorganic Gallium Clusters Moving down the periodic table, inorganic clusters of Ga(IIl) and In(IlI) become quite limited. The only example of an inorganic gallium hydroxide structure was isolated by two separate methods. The first isolation was by Rather et al. in a method similar to that of the second described Al 13 synthesis: Ga(N03h'nH20 is dissolved in methanol with either nitrosobenzene or di-n-butylnitrosamine.25.27 These solutions are then left to evaporate open to air for several weeks, at which point single-crystalline [Gal3(u3- OH)6(u-OH)18(H20)24](N03)15 (Gal3) is isolated in 60% yield with nitrosobenzene and greater than 80% yield with di-n-butylnitrosamine (Figure 1.3).21,25 This compound is isostructural with the AI l3 hydroxide structure, but falls in a different crystallographic space group. It is rationalized that no base is required in this reaction because gallium is more Lewis acidic than aluminum (Ga(IlI) pKa = 2.6, AI(IlI) pKa = 5.0).28 In the synthesis using nitrosobenzene, it was suspected that the nitroso functionality is oxidized to nitrobenzene, which works to consume nitrate and drive nucleation of the higher nuclearity species. However, experiments with di-n-butylnitrosamine revealed no evidence for the same oxidation (NMR and LC/MS suggest only starting material is present), while the clusters formed in higher yields. The use of di-n-butylnitrosamine I J also allows isolation ofGau in much larger quantities, with gram-scale quantities synthesized within several weeks. 21 Figure 1.3. Structural depiction of the crystal structure of Ga 13. The tridecameric gallium hydroxide is isostructural with the Al u cluster. There is the planar structural core with six additional gallium ions bridged to the core by hydroxides. These gallium atoms alternate positions above and below the plane of the core. Nearly concurrently, Gerasko et a!. reported the same Gau structure. 2() In this CQse, no organic Ditroso con1pound "V8S present. Instead, Ga(i'~OJ)]·81 120 Vv'8S dissolved in water with cucurbit[6]uril. A small amount of pyridine as added dropwise to adjust the pH to 1.8. Without this step, no higher nuclearity clusters were isolated. Initially, a 12 wonderful higher nuclearity [Ga32(u4-0)d,Ll3-0)s(,u-Oh(u-OH)39(H20ho]3+ oxo/hydroxo compound crystallizes. The remaining liquid was separated and left for an additional two weeks, at which point more crystals of Ga32 were isolated along with crystals of Gal3. The role of cucurbit[6]uril in the isolation of Gal3 is uncertain, and attempts to alter multiple reaction variables such as Ga(III) concentration, pH, and temperature failed to produce the cluster in its absence, so while the role of cucurbit[6]uril remains mysterious, it appears necessary nonetheless. This cluster has also been isolated in the absence of nitroso compounds or cucurbit[6]uril. In several cases, I have isolated Gal3 simply from either methanolic solutions of Ga(N03h or from methanolic solutions of Ga(N03)3 with EtOAc added prior to solvent evaporation. In both cases, yields of the crystalline material appears similar to cases with di-n-butylnitrosamine present, however, the crystallization times are generally longer and product isolation is less consistent. Partially due to the wide range of conditions under which Gal3 has been isolated, its mechanism of formation (and the role of the nitroso additive) is still unknown. This observation also underscores a quizzical concept: why had GaI3 not been previously isolated? Perhaps prior attempts to recrystallize Ga(N03knH20 have unknowingly produced GaI3 and it had just not been characterized as such. 13 Ligand Bound Hydroxide Clusters Several of the above clusters have been isolated in ligand-supported fon11S as well. Ligand-supported clusters are often more readily isolated due to the mitigation of molecular charge by the organic ligands.2 Inorganic clusters often build up extensive charge character, such as Gal3 and AI l3 which each have +15 charges.21 ,25,27 Complexation by organic ligands such as aminocarboxylates not only mitigates this charge, as in the case of a hydroxyethyliminodiacetic acid (H3heidi) ligand which when complexed and deprotonated calTies a -3 charge, but also fills the coordination sites on the peripheral metal ions. These would otherwise be occupied by water molecules which exchange much more rapidly. For these reasons, a larger variety ofligand-supported Group 13 hydroxides have been isolated.9 Presence ofligands can also affect the shape and size of the species that fonn and crystallize by directing fOll11ation through ligand coordination geometry or steric bulk. With a wide variety of available ligands, it is perhaps not surprising that a wider variety of elegant compounds have been isolated by a number of research groups utilizing coordinating ligands. Ligand Bound Aluminum Clusters The simplest version of a ligand-supported aluminum structure is the aforementioned dimeric Ab structure.22 They represent the largest variety of octahedral Group 13 hydroxide structures encountered, potentially because as higher nuclearities are generated, it becomes more difficult to walk the line between oxo/hydroxo and 14 oligomeric/polymeric species. This structural core has been observed in complexes isolated with several different ligand classes of varying coordination number. The structure with the greatest deviation from the inorganic dimer is that with heidi ligands reported by Heath et aI., [AI(heidi)(H20)h'2I-~hO (AI 2heidi 2, Figure 1.4).'> In this case, Figure 1.4. Structural depiction of the crystal strllchue of Aizheidh. The heid i ligand can be seen binding to the aluminum atoms in place of water and hydroxide ligands. One water ligand per aluminum remains bound (shown without hydrogens). Aluminum is teal, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMol. the hydroxyl bridges are actually replaced by the ethoxy groups of the heidi ligand. The aluminum ions bind to two ethoxy groups, two carboxylate oxygen atoms, one amino nitrogen atom, and one water ligand. This compound was isolated from a solution with a 2: 1:4 ratio of H3heidi:AI(N03)3'9H20:pyridine (pH = 4.3). Using AICI,·6H 20 in place of 15 Al(N03k9H20 affords the same compound with cr counteranions. While this compound actually lacks any hydroxide bridges, it is included because of the structural similarity to the inorganic Ah structure, and because it represents a potential "intermediate" to higher order structures, as will be discussed in the case of ligand- d 11 ' 822supporte ga mm structures. ' Similar compounds, but with different carboxylate groups, have also been characterized in the solid state. In one case, nitrilotriacetic acid (H3nta) is bound giving a [Al(H20h] [Ah(u-OHh(ntah]OH'3H20 formulation (Figure 1.5).11,30 To isolate this compound, Al(OH)3 was prepared and mixed with H3nta in water with heating. Further treatment with acetone and ethanol gave a white powder, which when recrystallized from water by slow evaporation gave the described product. II H3nta has been used as a detergent builder; its use has raised concerns about mobilization of toxic metal ions from deposits in lakes and rivers, driving research to detern1ine what species it may be forming with metal ions. I I Additionally, iminodiacetic acid (H2ida) has given a similar planar dimeric dihydroxide structure, [Ah(p-OHh(H20h(ida)2], but because ida only possesses two ethylcarboxylate functionalities, the resultant compound still has the bound water molecule found in the inorganic structure (Figure 1.5).22.3u2 This compound was isolated by dissolving AI(N03)3,9H20, H2ida, and NaOH in water in a I: 1:2 ratio. The solution was diluted with ethanol and evaporated over several weeks.7 Another example of a dimeric ligand-supported hydroxide structure was isolated by a less conventional method, obtained from non-polar organic solvents. A dimeric 16 isolated from a 1:2 ratio of ethyl acetate:AICI, suspended in n-hexane. 33 The initial product was recrystallized in CH 2Ch in the presence of moisture from the atmosphere Figure t.5. Depictions of the crystal structures of Alzntaz and Alzidaz. [n these structures, the ligands only have nitrogen and carboxylate coordinating groups. Because of this, there are no hydroxide ligands replaced, as with structures isolated with heidi ligands. Aluminum is teal, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMol. Figuloe 1.6. Dimeric aluminum species bound by ethyl acetate ligands. This structure has one hydroxide and two acetate bridging ligands. A total of six ethyl acetate ligands are bound to the aluminum atoms. Aluminum is teal, oxygen is red, carbon is black, and hydrogen is white; image generated with PyMol. 17 (originally introduced accidentally when solvent dissolved the grease used to seal the vessel). The colorless crystals with the stated fonnula were obtained more directly by exposing the recrystallization solution to air for one hour. This stmcture has potential interest in the area of polymerization catalysts, specifically for Ziegler-Natta, in order to better understand the interactions between ester molecules and the different catalyst . h" I I" I 33components, m t IS case t le organoa ummum cocata ysts. Another dimeric aluminum hydroxide compound is that fonned by the reaction of AI(N03k9H20 with three equivalents of2-(2'-hydroxyphenyl)-2-benzoxazole (Hhbo) and 3.5 equivalents ofNaOH in an 80:1 mixture ofMeOH:H20 (Figure 1.7).34 The phenolic group of the ligand is deprotonated and binds to the aluminum ion along with the nitrogen atom of the oxazoline. A total of four hbo ligands bind to the complex, along with two ,u-OH bridges to give the fonnula [AI(OH)(hbo)2h·4MeOH·H20 and a neutral compound. 27Al NMR studies were used to show that the solution stmcture was the same as the solid state. Similar reactions with Ga(N03)3'nH20 give different stmctures with no hydroxide bridges. 34 Interest in these compounds was stimulated by the desire to synthesize trivalent metal containing analogs of metalloenzymes containing polynuclear Fe(III) centers and carboxylato and/or hydroxo bridging ligands.34 The dimeric Ah core has also been isolated as a complex with another organic molecule: cucurbit[6]uri1. 35 In this example, the stmcture of the inorganic compound remains intact ([Ah(u-OH)2(H20)S](N03)4) and fonns an extended network with cucurbit[6]uril through hydrogen bonds only. The water ligands on the aluminum ions 18 ( Figure 1.7. Dimeric aluminum hyd.·oxide complex with hbo ligands. In this structure, a more rigid ligand is bound to aluminum. The aluminum atoms are bound to four oxygen atoms and two nitrogen atoms. Aluminum is teal, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMol. hydrogen bond to the carbonyl oxygen atoms of the cllcurbit[6]uril molecules. Each Ab dimer forms six hydrogen bonds to cucurbit[6]uril carbonyl oxygen atol11s. The compound was isolated by evaporation of a solution of AI(NO)b'9I-hO and cucurbit[6]uril, with pyridine added until a pH value of 3.1 was achieved. Pyridine molecules used to basicify the solution are bound as guests within the cavity of the cucurbit[6]uril host. This system forms an infinite chain of hydrogen bonded molecules. Waters of hydration, nitrate counteranions, Clnci Clciciitional clicurhit[6]uri! molecu!es are fOllnd between these hydrogen bonded chains.)) 19 Beyond these varied cases of Al dimers, a trimeric aluminum complex with citrate has been isolated.36 This compound, (NH4)5[A1}(cith(u-OH)(H20)](N03)'6H20 (Figure 1.8), is proposed as a possible species that fonns in the interaction of aluminum with blood plasma. The structure was isolated by slow evaporation of an aqueous ammonia solution of aluminum nitrate and citrate. Each of the three aluminum ions in the molecule has a different coordination environment. The first is bonded only to citrate oxygen atoms, the second is bonded to five citrate oxygen atoms and a hydroxide bridge, and the third is bonded to four citrate oxygen atoms, one hydroxide bridge, and one water molecule. The space between the molecules is filled by a network of hydrogen bonded water molecules, ammonium cations, and a nitrate anion. The presence of an additional cation is similar to the synthesis of inorganic Al 13 using NH40H as a base, where an extra NH4+ cation was present and charge balanced by an additional nitrate counteranion.25 Another aluminum citrate trimer was isolated by Malone et al.37 The structure is far more symmetric than the previously discussed citrate trimer. Two AI(III) ions fonn a motif similar to that found in the AI(Ill) dimer species, with two ,u-OH bridges shared between them. These aluminum ions are also coordinated by two carboxylate oxygen atoms from a citrate, an alkoxide from a citrate, and a water ligand. The third aluminum ion is coordinated through an alkoxide and carboxylate group from each of the two citrate molecules, in addition to two water ligands. The structure crystallizes as a double salt, with every two [(A1}(cithC,u-OHh(H20)4r anions neutralized by one [Al(H20)6]3+ and one N03-, giving a solid state complex with fonnula [Al(H20)6][(A1}(cithC,u- OHh(H20h]2(N03)-6H20 (Figure 1.8).37 This structure has two particular points of 20 interest: 1) the pH levels used for the synthesis (pH = 1.6) would normally give only monomeric citrate complexes, 2) the interesting appearance of monomeric hydrated aluminum ion. The second point is particularly interesting due to the presence of hydrolyzed Al(II1) ions in the main section of the molecule, and is assumed to occur due to the excess of this species present at the low pH values of the synthesis.37 Figure 1.8. Depictions of two different t.-imeric structures with citrate. On the left, the structure only has one hydroxide bridge and one water ligand. The structure on the right has two hydroxide bridges and four bound water ligands. The structure reported by Malone et al..\7 is much more symmetric than the structure from Feng et al. 31> Aluminum is teal, oxygen is red, carbon is black, and hydrogen is white; image generated with PyMol. A tetrameric aluminum complex with hydroxypropanediaminotetraacetic acid (H~hpdta) has also been isolated, (enI-b)[AI4(,lI-OHMhpdtah]·7.5H20 (en = ethylenediamine).7 This compound was synthesized by dissolving H.dlpdta and i\lCl~'6!-I20 in \V3ter, follc\.ved by ~lovv'ly ruising the pH to 5.8 through addition of ethylenediamine. Formation of this cluster species is thought to occur through a planar tetranuclear M40(OH)(hpdtah(H 20k species that forms upon the condensation of two 21 dimeric species. This species represents a fully dehydrated cluster (completely lacking water ligands, Figure 1.9) which is proposed to be particularly stable because H20 ligands normally have a rapid exchange rate. 7 The aluminum atoms occupy the corners of a tetrahedron. Four edges of the tetrahedron are formed by hydroxide bridges, the other two by alkoxide groups from hpdta. Figure 1.9. TetnlJneric aluminum structure. The four aluminum atoms sit roughly on corners ofa tetrahedron with the ligands forming the six edges of the tetrahedron. The hpdta ligand has the highest coordination number of the ligands encountered in a ligand- supported structure and each binds two different aluminum atoms. Aluminum is teal, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMoJ. After the tetrameric species, there is a fairly large gap until the next ligand-bound 1.10).5 This compound was isolated in a similar procedure to Ahheidi 2, with a 1:2 ratio of H3heidi:AI(N03h9H20. Instead of pyridine to adjust the pH (pH 5 gives the best 22 yields), I M NaOH was used. As with Al2heidh, using AICly6l-hO in place of AI(N03h-9l-bO affords the same compound with cr counteranions.5 This complex has the same metal ion arrangement as the inorganic Al u compound discussed earlier in this review.23.25 A planar core of seven aluminum atoms is observed, all octahedral and Figure l.IO. Tridecameric aluminulll cluster bound by heidi ligands. This compound is roughly isostructural with the inorganic AI 13 compound, with each heidi ligand replacing three water ligands and one hydroxide bridge between the outer aluminum atoms and the central core. Aluminum is teal, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMoI. bound to p-OH ligands. Next are six additional aluminum atoms, alternating above and belo'vV the inner core of seven ions. The::;e uuier ::;ix aluminum ions are bound to the inner core through p-OH bridges and the ethoxide group of the heidi ligand. The remaining four coordination sites of the octahedral ions are filled by two carboxylate groups, the 23 amino nitrogen of the heidi ligand, and one water molecule. The heidi ligands are again proposed to stabilize the structure against hydrolysis upon dilution. This compound has been examined as an alternative pillaring agent for clays to Keggin-AI 13 , which are used as adsorptive media and catalysts. Al 13heidi6 cations are inserted between layers of the clay and upon calcination yield oxide pillars that have two-dimensional porosity, h·· ~ 1 350 2 -I 63Sac levmg surlace area va ues up to mg. ,- Ligand Bound Gallium Clusters Moving down the group to gallium, an isostructural compound to the Abheidi2 compound has been isolated by Goodwin et aI., [Ga(heidi)(H20)h (Ga2heidi2' Figure 1.11).S This structure possesses the same geometry and morphology of the Abheidh structure, with gallium ions replacing the aluminum ions. Isolation is achieved by reaction of a 2:1:8 ratio of hydrated Ga(N03)3, H3heidi, and pyridine in aqueous solution. Evaporation of this solution yields the Ga2heidb complex, as well as [Gas(u3-0Hh(U- OH)s(heidi)4(H20)4(CsHsN)2f+, described in the next paragraph. If a 1: 1 ratio of gallium nitrate and H3heidi are dissolved with two equivalents of sodium hydroxide, Ga2heidb is isolated exclusively. structure was isolated along with the Ga2heidh compound. This is a fascinating example of multiple discrete compounds being isolated from the same reaction mixture. Similar to the inorganic Als structure isolated by Casey et aI., Gasheidi4 contains a tetrameric core 24 of edge-shared gallium ions, [Ga4(P3-0Hh(p-OHht+, which is bound to tour corner- shared [Ga(heidi)(H20)t units. Each heidi ligand replaces three water ligands on the corner-shared gallium atoms (when compared to the inorganic Alx structure) and a ,Ll-OH ligand between the edge-shared ions and corner-shared core. 24 Two pyridine ligands also replace two water ligands that are bound to the edge-shared Al(lII) ions in the Alx structure (Figure 1.1 and 1.12). Figlll'e 1.11. Dimeric gallium compound bound by heidi ligands. Gallium is bound by heidi ligands in the same manner as the Al 2heidi compound. The cthoxide group replaces a hydroxide bridge, with three additional coordination sites filled by the heidi ligand. Gallium is dark blue, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMoL After the initial isolation of the Gaxheidi4 structure, fascinating behavior was observed. X From the solid state, the compound redissolves and within six weeks recrystallized as needles of a higher nuclearity, [Ga13Cheidi)(,Vt,,-OH)6(p-OI-I)dH20)6]]-I- (Gal]heidi6) complex (Figure 1.13). This compound is isostructural with the Al 1]heidi6 25 compound, featuring the same planar core of seven metals with an additional ring of six gallium ions alternating above and below the plane formed by the inner seven. The heidi ligand has the same connectivity as well, with three water ligands replaced by two Figure 1.12. Heidi and pyridine ligands bound to an octamedc gallium hydroxide cOI'e. In the Gaxheidi4 complex, a similar motif is seen where heidi replaces three water molecules and one hydroxide bridge as compared to the inorganic Al x compound. The central core is composed off-our edge-shared gallium octahedra. Gallium is dark blue, oxygen lS red, carbon is black, nitrogen is blue, and hydrogen is white~ image generated with PyMol. carboxylate groups and the amino nitrogen. The ethoxide ligand again replaces the ,u-OH bridges in between the core and outer shell. The compound can also be isolated directly from a 3: I ratio of [-[Jheidi:Ga(NOJ)J by adjusting the pH of the solution to 5 with pyridine. This represents an interesting reversal of base addition from the Abheidi2 and 26 AI d1eidi(, structures where pyridine is used for the dimeric structure and NaOH for the tridecameric structure. Figul'e 1. 13. The highest Ilucleadty gallium hydroxide ligalld-suppOl·ted structure. AU six exterior gallium atoms are bound to heidi ligands, with the same coordination mode as seen previously. The central planar heptameric core remains intact. Gallium is dark blue, oxygen is red, carbon is black, nitrogen is blue, and hydrogen is white; image generated with PyMol. This report was unique because three different, but related clusters were isolated from the same reaction mixture. This was important because it allowed formation of the first conclusions about the growth of these cluster compounds based on observed "'intermediates". Ga2heidi2 and GaKheidi4. x It had previously been suggested that a seven metal core was a likely intermediate on the way to higher nuclearity structures such as the tridecameric clusters, but in conjunction with the inorganic AI K compound, it seems likely 27 that in the case of Group 13 metals, this complete shell of the brucite lattice does not need to fonn prior to addition of other metal ions. To isolate structures of increasing nuclearity with no pH alteration is a fascinating and unexpected result. Ligand Bound Indium Clusters I have come across only two examples of indium hydroxide compounds, both slight variations of a tetrameric structural core. In these tetrameric [(tacn)4In4CU- OH)6](S206)3-4H20 and [(tacn)4In4Cu-OH)6](Cl04k6H20 compounds (tacn = 1,4,7- triazacyclononane), the indium atoms occupy the comers of a tetrahedron and the six hydroxide bridges lie above the edges of the tetrahedron.39 Each indium atom is bound to three hydroxide bridges and three nitrogen atoms. There are waters of crystallization as well, which have significant mobility in the structure at room temperature. These compounds were isolated from aqueous solutions ofInBr3, NaOH (pH ~ 8), and the corresponding sodium salts. Colorless crystals of the product fom1 in good yields. The authors reported that other counterions can be used as well, suggesting perhaps that with indium compounds, the counterion may playa smaller role, unlike gallium compounds which have usually used N03- counteranions. 28 Heterometallic Clusters Beyond the homometallic inorganic and ligand-supported clusters described above, our group has been able to isolate a series of inorganic heterometallic clusters.4o These clusters are the first of their kind. As mentioned in the introduction, previous reports have described the substitution of the central tetrahedral metal in Keggin-Al l3 compounds (with either Ga or Ge) but there are no reports ofheterometallic Group 13 compounds with all hydroxide bridges and all octahedral metals. 16,17 The initial compounds of this type we were able to isolate contained mixtures of gallium and indium. We isolated and characterized a series of six compounds where indium takes the place of gallium in the outer shell sites of the original GaB cluster (Ga7+xIn6-x, x = 0 - 5). These clusters possess the same structural motif observed in the GaB compound with a planar core of seven Ga(III) atoms always maintained (Figure 1.14).27,40 The same synthetic methods used to isolate GaB were utilized to isolate these GalIn heterometallic compounds. Instead of starting with a solution containing only Ga(N03)3 as with the GaB synthesis, various ratios of Ga(N03)3 and In(N03)3 are dissolved in methanol with di-n-butylnitrosamine. No pH adjustment is performed, and solutions are evaporated open to atmosphere. In the extremes, a 1:12 ratio ofGa(N03)3:In(N03)3 gives Ga7In6 and a 5:1 ratio of Ga(N03)3:In(N03)3 gives Ga12In. With larger quantities of these compounds available, the Ga7In6 compound has been used as a precursor material for solution-processed amorphous indium-gallium oxide thin films. 21 ,40 29 Figure 1. 14. Heterometallic compouud consisting of gallium and indium atoms. The heterometallic Group 13 hydroxide structures are isostructural with the AII~ and Gal.1 homometallic compounds. Indium atoms are found in the exterior. corner-shared positions. Gallium is blue, indium is green, oxygen is red, and hydrogen is white; image generated with PyMol. In addition to the heterometallic Galin hydroxide clusters, we also isolated several heterometallic AllIn hydroxide clusters. These structures are again isostructural to the previously described inorganic tridecameric structures. In this case, as with the AI 13 synthesis, base addition is required to successfully isolate crystalline product of the clusters. As with the other heterometallic clusters, substitution of indium for aluminum only takes place at the exterior metal sites. Unlike the Galin compounds however, only two heterometallic AllIn compounds have been isolated, AIl{lns and All)In4. For AIRln.'i, 30 Al(N03)3:In(N03h ratios of 6:7 and 7:6 both afforded the compound. For A19In4, 9:4 and 8:5 ratios of Al(N03h:In(N03h were utilized to isolate the compound. Discussion Summarizing the structures presented in this review, there are three primary observed structure types for purely inorganic hydroxide compounds: dimeric, octameric, and tridecameric. When ligand-supported compounds are considered, there is wider structural variety: dimers, trimers, tetramers, octamers, and tridecamers have all been observed bound by a variety of ligands (Figure 1.15), and when structures with some oxo bridges are considered, even more structure types are observed.HI Most of the described compounds were isolated from aqueous solutions, but a wide variety of recently reported tridecamers have been isolated from methanolic solutions, suggesting this may be an avenue for other researchers to explore. It is proposed that the use of methanol may help attenuate the hydrolysis of the metal ions and prevent precipitation of unspecified 1· . d 42o 19omenc pro ucts. It has been suggested that the fonnation of these hydroxide clusters follows the fonnation of a brucite type lattice.2.8 Brucite, Mg(OHh, fonns planar sheets of hydroxide ligands with octahedral holes. 8 In the case ofmany of the higher nuclearity clusters discussed in this review, the compounds are partial sections ofa brucite lattice with aluminum, gallium, or indium filling the octahedral holes. Group 13 hydroxide structures with a brucite structure type have not been observed with nuc1earity above thirteen metal 31 HOl ° ° 0 0y--OH h y--OH H'Ny--OHN HO N }-OH }-OH >-OH 0 0 0 H3heidi H3nta H2 ida Hhbo o OH 0 HO~-J-OH HO 0 HN~ ( NHN~ H taen Hshpdta Figure 1.15. Summary of ligands in ligand-supported compounds. As stated, amino carboxylates are often used as coordinating ligands for Group 13 metal hydroxide structures. Compounds containing Hhbo, H4citrate, and tacn were also summarized. ions (though there are examples of higher nuclearity compounds with partial brucite lattice structure containing oxo ligands), while in the case ofMn and Fe, higher nuclearity brucite type structures have been isolated, such as Mnl9 and Fe19.7 The closest example to such a structure with Group 13 is the elegant [Ahs(u-O)4(u3-0H)6(u- OH)14(hpdta)4t cluster isolated by Powell et aI., where the exterior eight metal ions no longer lie in a plane and four of the exterior hydroxide bridges have been deprotonated to 32 [Olm oxide bridges.41 By analyzing the known structures, we can gain a better understanding of the stable fragments and intermediate speciation along the path to higher nuclearity brucite type structures. What we ourselves, and we assume others, have learned through the years of working with Group 13 metal hydroxide compounds is that you can never fully predict their behavior. While the heterometallic GalIn structures form quite readily, we have yet to isolate any strictly indium-based structure, the closest being an indium acetate hydroxide 1-D chain stmcture, which could to some extent be considered a ligand- supported structure, though was excluded because of its infinite nature.43 Indium seems more likely to form extended structures such as those observed with benzene carboxylate ligands, many examples of which exist in the literature.44-48 Similarly, while many ligand-supported Fe(IH) complexes have been isolated, we have yet to isolate any inorganic Fe(III) hydroxide compounds, certainly not for lack of trying. In considering Fe(III) and Mn(H or III) compounds, ligand-supported versions of Fe? and Mn? cores have been isolated.49.5o Heptameric cores of this type were thought to be the building block for larger structures such as the tridecameric Gal3 and All3. However, with Group 13 metals, octameric structures are the highest nuclearity structures observed prior to the tridecameric compounds (with no reports of heptameric Group 13 hydroxides), suggesting these may instead be intermediates to the larger structures. Goodwin et al. provided structural evidence for this through their report of Ga8heidi4 redissolving and recrystallizing as Ga13heidi6.8 In regard to the heidi ligand, it is also interesting to note that in all ligand-supported structures where heidi ligands are present, 33 the same binding motif is observed, with the ethoxy group of the heidi ligand replacing a hydroxide bridge while an H20 ligand remains. It seems that alkoxide ligands will take the place of the hydroxide bridges, while carboxylate groups bind in place of water ligands (as compared to the inorganic congeners). This more firm structural incorporation of the heidi ligand suggests that it may be difficult to achieve post-synthetic ligand exchange for such a ligand onto inorganic compounds, such as Ga!3. Comparing reports summarized in this review, a number of additional "missing" structures can be highlighted as potential synthetic targets, such as the inorganic Gas isomer of inorganic AIs, the Al isomer of Gasheidi4, and gallium dimers, trimers, and tetramers isostructural to the aluminum structures summarized here. 5,7,s,24,30 Many potential indium compounds may be worth pursuing as well, as compounds containing indium remain quite limited. The "bewildering variety of behavior" of these compounds can be both inspirational and infuriating; it seems sometimes certain compounds exist on a knife's edge of pH range, and before you know, slip away into the abyss of chemical speciation (or the amorphous precipitate on the bottom of your flask). Bridge to Chapter II The following chapters summarize the work I have been involved with in the area oftridecameric Group 13 hydroxide structures, such as homometallic AI!3 and Gal3 and a host ofheterometallic compounds. In Chapter II, the reader will be treated to a thorough discussion of the tridecameric cluster at the top of the chart, AI!3. 34 CHAPTER II FACILE SYNTHESIS OF THE TRIDECAMERIC AI13 NANOCLUSTER [AI13Gu3-0H)6Gu-OH)ls(HzO)z4]eN03)15 Introduction This chapter discusses the synthesis and isolation of a tridecameric aluminum hydroxide cluster. This work represents the first intentional synthesis of this compound. I contributed to this work through the development of the synthetic method utilizing an alternative reagent, di-n-butylnitrosamine, which allowed superior yields and easier isolation ofthe title compound and has enabled much of the progress detailed in the following chapters. I also contributed to the writing of drafts and extensive editorial work. Dr. Jason T. Gatlin initially isolated the compound and was the primary author of the initial draft of the manuscript. Dr. David MacInnes was a visiting faculty member from Guilford College who worked on yield determinations and some editorial aspects. Dr. Lev N. Zakharov was responsible for collecting most, and solving all, ofthe X-ray crystal structures presented in this chapter, and wrote the X-ray details. Prof. Darren W. Johnson was the principle investigator for this work and provided editorial assistance. This work was published in volume 47 of Inorganic Chemistry, a publication of the American Chemical Society, in January 2008.1 35 Aluminum is the third most abundant element and the most abundant metal in the earth's crust found in many minerals and ores. Aluminum complexes 20(1)]: R1 = 0.0479, wR2 = 0.1267, GOF = 1.069. Crystal datafor H9r0II3N170106 (Aln"NH4N03): Mr = 2381.68, colorless block, 0.08 x 0.08 x 0.05 mm, ttic1inic, space group P-1 (no. 2), a = 12.623(3), b = 13.251(3), c = 13.597(3) A, a = 74.877(4), P= 72.419(4), y = 86.790(4)°, V= 2092.4(2) A3, Z = 1, rcalcd = 1.890 g'cm-3, f.1 = 0.326 mm- I , F(OOO) = 1232, 20max = 50.0°, 15065 reflections collected, 7308 unique [Rint= 0.0700], R indices [I> 20(1)]: R1 = 0.0836, wR2 = 0.1961, GOF = 1.051. 44 Bridge to Chapter III Two different homometallic tridecameric hydroxide clusters have now been synthesized, Gal3 and AIl3. The next step down the periodic table was to try and synthesize an In13 compound, but this compound has not been isolated to date. However, in the process of attempting to synthesize an Inl3 compound, binary mixtures of metal nitrate starting material salts were examined, which has led to the isolation of the heterometallic tridecameric hydroxide compounds that will be described in the following three chapters. 45 CHAPTER III SYNTHESIS OF HETEROMETALLIC GROUP 13 NANOCLUSTERS AS INKS FOR OXIDE THIN-FILM TRANSISTORS Introduction I contributed to the work described in this chapter through the identification of the di-n-butylnitrosamine compound which allowed the isolation of these compounds in higher yields and larger quantities. Without this discovery, sufficient quantities of the Ga7In6 compound could not have been prepared. I also performed extensive rewriting of the paper after a direction shift, complete with final editing and addressing reviewer comments. Dr. Jason T. Gatlin originally synthesized the Ga7In6 cluster described in the chapter and wrote initial drafts of the paper prior to the direction shift. Dr. Stephen T. Meyers performed all TFT related studies (including X-ray powder diffraction) and wrote the relevant sections of the paper. Dr. Lev N. Zakharov performed all single-crystal X- ray diffraction studies and wrote the relevant X-ray data section of the paper. Prof. Douglas A. Keszler and Prof. Darren W. Johnson were the principle investigators for this work and provided editorial assistance. The results of this work were published in 2008 in Angewandte Chemie, International Edition, a publication ofWiley-VCH Verlag GmbH & Co. KGaA, volume 47 pages 9484-9486. 1 46 We have recently reported high-yielding syntheses of two inorganic Group 13 OH)6M(,(P-OH)6] central fragment of these clusters forms a planar core with six additional M(H:>O)4 groups bound to the core by two ,LI-OH bridges. The outer metal ions alternate above and beloyv the plane formed by the central seven metal ions. Prior Figure 3.1. Structural representation of tridecameric Group 13 hyd."oxide clusters. On the left is a representation of the crystal structure of [Gal3(PJ-OH)(,(p- OH)IS(H:>Ob](NOJ)15 (GaLl) and on the right is a representation of[Ga71n6(PJ-0l-I)6(p- OH)IS(H:>0):>4](NOJ)15 (Ga7In6). Both are structurally similar, \vith Ga71n6 having indium atoms (shown in green) in place of the outer ring gallium atoms of GaIJ. Images generated with PyMol. synthetic preparation of Group IJ metal hydroxide cOinp0unds sudl as these has proven difficult. Their syntheses often require caustic or acidic conditions, elevated temperatures and pressures, and only provide clusters in low yields (sometimes only a 47 few single crystals). Crystallization periods of months or even years are not uncommon. 4 - 7 Owing to these difficulties, relatively few discrete Group 13 metal hydroxide clusters have been synthesized, though several striking examples of aluminum4,S.7,8 and gallium3,6,9,lO complexes have been reported. Both hydrated clusters and those stabilized by organic ligands are known, with a larger variety of ligand- supported clusters having been isolated, owing to enhanced stability resulting from lower charge density due to charge balance provided by the ligands.4 In the case of these inorganic and ligand-supported compounds, neither heterometallic nor indium-containing clusters are known. However, in the case of Keggin-AI 13 clusters, the central tetrahedral metal can be substituted, forming M 1AI 12 structures (compounds with M=AI, Ga, or Ge are known, with others suggested). I 1-14 To our knowledge though, no heterometallic Group 13 metal hydroxide clusters with multiple substitutions have been repOlied. Furthermore, the low-yielding, challenging syntheses often associated with these clusters have prevented attempts to explore applications requiring large quantities of such compounds. In order to address synthetic difficulties and to explore the use of these clusters as precursors for materials, I have developed an improved synthesis that affords heterometallic Group 13 nanoclusters of the type reported herein. Presented in this chapter is a new heterometallic gallium-indium cluster, [Ga7In6(u3-0HMjt- OH)18(H20)24(N03)IS] (Ga7In6, Figure 3.1, right). This compound can be synthesized reliably, in yields ranging from 25 % to 95 %, by utilizing two different nitroso additives.2.3 Gal3 and Ga7In6 can both be prepared in gram-scale quantities which enables the unprecedented use of these nanoclusters as single-source solution precursors 48 for the deposition of oxide semiconductor thin films. I 5,16 Such deposition provides a new route to the fabrication of high-performance thin-film transistors (TFTs) comprising spin- coated Ino.92GaI.OS03 semiconductor layers. Materials Precursor Interest There has been recent interest in the use of nanoscale cluster precursors to synthesize new materials. 17-ZO The difficult syntheses of Group 13 metal hydroxide clusters have mostly prevented their use in these applications. Most solution precursors for printed oxide films involve controlled hydrolysis of metal-organic compounds and the condensation ofmetal-hydroxo sols that are then pyrolyzed to form the oxide. Such films are beset by a variety of density, defect, and segregation issues relating to the inhomogeneous nature of the sol, retention of significant organic components, or oxygen nonstoichiometry associated with organic burnout. From this perspective, soluble, all- inorganic, heterometallic hydroxide clusters, such as the M 13 system, provide model precursors and an entirely inorganic, rapid, low-volume-loss condensation pathway, eliminating the aforementioned detrimental effects of organic moieties. Synthetic Procedures for Tridecameric Clusters Scheme 3.1 depicts synthetic routes to structures GaB and Ga7In6 using two different nitroso compound additives. Applying the previously reported procedure of 49 adding nitrosobenzene to a 1: 12 ratio of Ga(N03)3 and In(N03)3 affords single crystals of cluster Ga7In6, which are isolated in 25 % yield. 3 However, the manual separation of crystals from the tar-like product mixture limited the amount of material that could be isolated. To address the problems of difficult isolation and limited reaction scale, I sought alternatives to nitrosobenzene. The use of di-n-butylnitrosamine affords GaB and Ga7In6 in superior yields of 85 % and 95 %, respectively.2 The reaction with di-n- butylnitrosamine produces a mixture of a transparent oil (which can be reused in future x Ga(N03h y In(N03h NO 6~~ MaOH ~N~ I NO "Ga13" 3: x =13, Y =0; 65% "Ga7In6" 4: x = 1, Y = 12; 25% "Ga13" 3: x =13, Y =0; 85% "Ga7In6" 4: x = 1, Y = 12; 94% Scheme 3.1. Synthetic preparation of Ga]3 and Ga7In6 by two different methods. The top route depicts the synthesis of GaB and Ga7In6 using nitrosobenzene as the organic additive. The bottom method using di-n-butylnitrosamine affords the same compounds in much higher yields and larger scales. In both cases, a large excess of In(N03)3 is required to fonn the fully substituted Ga7In6 compound. syntheses) and gram-scale quantities of single-crystalline GaB and Ga7In6 products. Multiple characterization techniques confinned that the bulk crystals isolated from the reaction and the single crystals reported herein are the same. This robust synthetic strategy enables the use of these clusters as precursors for bulk materials. 50 Optimized synthesis of[Ga 13(fl3-0H)6(fl-OHhs(H20)24](N03) 15 (Ga13) using an alternate organic additive. Di-n-butylnitrosamine (1.15 g, 7.26 mmol, 24 equiv.) was added to gallium(III) nitrate hydrate (1.0 g, 3.91 mmol, 13 equiv.) in 10 mL of methanol, as a homogeneous solution. The mixture was evaporated over 10 days at which point the methanol and nitrosamine were no longer miscible and crystals began to form on the sides and bottom ofthe reaction vessel. After several more days the methanol completely evaporated giving a single yellow tinted liquid phase. The remaining oil was decanted and single crystals of Ga13 were washed with cold ethyl acetate (3 x 10 mL) and dried open to the atmosphere, providing the identical Ga13 cluster isolated by the previous nitrosobenzene route in 85 % yield with respect to Ga(N03)3. Preparation 0/[Ga7Indfl3-0H)dfl-OH)18(H20)24](N03)15 (Ga7/1l6). Nitrosobenzene was utilized as the organic additive. Gallium(III) nitrate (1.73 mg, 0.006 mmol, 1 equiv.) and indium(III) nitrate (24.7 mg, 0.082 mmol, 12 equiv.) were dissolved in 5 mL of methanol. Nitrosobenzene (17.6 mg, 0.165 mmol, 24 equiv.) was dissolved in 2 mL of methanol, and the solutions were mixed together. The mixture was evaporated at room temperature over 10-12 days, yielding large single crystals of Ga7In6 in 25 % yield with respect to gallium, although the crystals were embedded within the tar-like product mixture of decomposed nitrosobenzene (Figure 2.2). Optimized preparation 0/[Ga7Indfl3-0H)6(fl-OH) Is(H20)24] (N03) 15 (Ga7/n6)' This method uses di-n-buty1nitrosamine instead of nitrosobenzene. Di-n- butylnitrosamine (0.93 g, 5.9 mmol, 24 equiv.) was added to a solution of Ga(N03)3 (0.068 g, 0.267 mmol, 1 equiv.) and In(N03)3 (0.872 g, 2.97 mmol, 12 equiv.) in 10 mL 51 methanol, and fonned a homogenous solution. The mixture was evaporated at room temperature over 2 weeks, affording single crystals of Ga7In6 in 94 % yield with respect to gallium (longer crystallization times generally increase the yield of the reaction). Di- n-butylnitrosamine was removed via syringe, and the remaining crystals were washed with cold ethyl acetate (3 x 10 mL) and dried in air. Application as Precursor Inks for Thin-Film Transistors To demonstrate the utility of the improved synthesis, we explored the use of these clusters in the fabrication of electronic devices, which is driven by a rising interest in printed macroelectronics and the high carrier mobilities recently reported in Group 13 and other p-block amorphous oxide semiconductors. 15.21,22 Our collaborators, the Keszler lab at Oregon State University, have recently described all-inorganic metal hydroxide cation condensation routes to dense, high-quality oxide dielectric films. 23 .24 On the basis of these results, the discrete metal hydroxide clusters Ga 13 and Ga7In6 were immediately recognized as potential oxide precursors operating on similar principles. Ga7In6 is of particular interest because of the large indium fraction and the excellent perfonnance of In203-based semiconductors.2l ,22 Initial device characteristics of a TFT with an amorphous Ino.92GaI.Og03 (IGO) semiconductor derived from a spin-coated aqueous solution of Ga7In6 are presented in Figure 3.2; Von = - 6 V and limited hysteresis is observed. On-to-off current ratios are 52 > 106: 1 for aU devices fabricated on 100 nm thermally grown Si02 dielectric surfaces. Field-effect mobilities for these bottom-gate devices are approximately 9 cm 2y-l s -1 after annealing at 600°C. Characterization of similar, thicker films by X-ray diffraction results in the pattern depicted in Figure 3.3. A single broad reflection centered near 2(1 = 33° is consistent with previously reported amorphous IGO films. l ) A cross-sectional image ofthe same film recorded by scanning electron microscopy (Figure 3.3, top) shows a generally dense morphology with small inhomogeneities « I0 nm) possibly resulting from agglomeration of subcolloidal species during spin-coating. 10.2 10-3 Vos = 30 V W/L = 17 10-4 L = 86 IJrn 10.5 -« <",2 - 10.6 _0 E- o 10.7 - 1 10.8 0 0 10 20 30 40 10.9 VDS (A) 10.10 ·10 0 10 20 30 40·20 VGS (V) Figure 3.2. Representative transfel- and (inset) output characteristics for a bottom- gate IGO-channel TFT with a 100 nm thermally grown SiOz dielecta-ic surface. JIGS in the output curve is stepped from 0 - 40 Y in lOY steps. 10 = drain current; Jlos = drain-to-source voltage; JIGS = gate-source voltage; L = channel length; W = channel width. 53 10 15 20 25 30 35 40 45 29 (0) 50 55 60 Figure 3.3. Thin-film XRD pattern (bottom) and cross-sectional SEM image of an Ino.92Gal.OS03 film (top). The film was fabricated by spin-coating an aqueous solution of Ga7In6 followed by annealing at 600°C for 1 hour in air. The broad peak seen in the XRD pattern indicates that the film is still amorphous in nature. The direct deposition of such high-perfonnance semiconductors from aqueous solutions is unprecedented and represents an important step toward printed macroelectronics. Thin-film X-ray diffraction was performed on a Rigaku RAPID diffractometer with Cu Ka radiation. TFTs were fabricated by dissolving Ga7In6 in deionized water with an initial resistance near 18 Mn (Figure 3.4). This ink was deposited by spin-coating on p-type Si substrates capped with 100 nm ofthermally grown Si02• Semiconductor/dielectric film stacks were then annealed for 1 h at 600°C in air. Aluminum source and drain electrodes were evaporated through a shadowmask to 54 complete the device fabrication. TFTs were characterized in the dark using a Hewlett- Packard 4156C semiconductor parameter analyzer. / Figure 3.4. Representation of process to deposit IGO films from heterometaJlic precursors. Single-crystalline product isolated by the described method was dissolved in water and spin-coated onto Si02 substrates, followed by an annealing step at 600 °C for I hour. High quality, robust thin film metal oxides were generated with this method. Conclusion In summary, I have devised a synthetic strategy for making gallium clusters which utilizes a reusable reagent and proceeds relatively quickly, providing high yields of 55 the cluster at ambient temperature. 1 I expanded this strategy and showed general utility by synthesizing aluminum clusters and heterometallic gallium/indium congeners. l ,2 Insofar as these molecules might hold promise as single-source precursors for novel materials (as demonstrated by the IGO thin films reported herein), developing an efficient synthetic method is highly important. This work also perhaps sheds light on the mechanism of cluster growth. For example, previous work suggests that dimeric and octameric fragments might form initially.6 This synthetic method led to no structures with varying compositions of the inner seven metal atoms, suggesting the M7 core might be particularly stable. A recent report of an analogous ligand-stabilized Fe7 cluster h · h h' 25supports t IS ypot eSIS. Crystallographic Methods and Summary of Data Experiments were carried out on a Bruker Smart Apex diffractometer at 153 K using MoKa radiation (A = 0.71073 A). Absorption corrections were applied by SADABS (Tmin/Tmax = 0.767). Crystals of Ga7In6 are hexagonal, R 1" (no. 148). The Ga7In6 cation is on a 1" axes, analogous to the related Ga13 cluster. Ga7In6 is isostructural with Ga13, possessing the same "Ga(,u3-0H)6G~(j.L-OHV'core, but with indium atoms in the exterior positions of the cluster connected via ,u-OH bridges (Figure 3.1). The Ga-O distances range from 1.907 Ato 2.159 A, while In-O distances range from 2.082 A to 2.167 A. Two N03- anions (in general positions) provide twelve N03- anions per Ga7In6 cation. Three other N03- anions and solvent methanol molecules (in 56 general positions as well) are highly disordered and randomly fill six other possible positions around the Ga7In6 cation. Highly disordered N03- anions and solvent methanol molecules were treated by SQUEEZE.26 Corrections of the X-ray data by SQUEEZE (638 electrons/cell) are close to the required value of603 electrons/cell for 9 N03- anions and 18 methanol molecules in the full unit cell. All non-H atoms were refined with anisotropic thermal parameters. Hydrogen atoms have not been taken into consideration. Refinements of the crystal structures of Ga7In6 without symmetry restrictions on occupation factors for the gallium and indium atoms show that the refined occupation factors of the Ga( 1), Ga(2), and In( 1) atoms are very close to those based on the crystal symmetry. The Ga( 1)-0 and Ga(2)-0 distances in this structure are comparable and close to the distances found in the related GaB cation. This indicates that in all of these structures the central M7 cores of the M 13 cations are composed of gallium atoms only. The average In-O(H20) distance in Ga7In6, 2.166(4) A, is close to the distances found before in complexes with an In-O(H20) bond: for example, 2.156 and 1.158 A were reported in catena-[(u-oxalato-O,O',0',Olll)-bis(u-O',0",Olll)-tetraaqua-diindium dihydrato].27 H72Ga7In6N15093 (Ga7In6)-6CH30H. Mr = 3139.94,0.21 x 0.18 x 0.12 mm, hexagonal, R j (no. 148), a = 20.6974(14) A, b = 20.6974(14) A, c = 18.256(3) A, a = 90°, fJ = 90°, y = 120°, V = 6773(1) A3, Z = 3, pealed = 2.310 g cm-3,,u = 3.704 mm-1, F(OOO) = 4620, 2Bmax = 54.00°, T= 153(2) K, 16375 reflections measured, 3290 reflections independent [Rint = 0.0187], Rl = 0.0246, wR2 = 0.0721 for 3290 reflections 57 (165 parameters) with I> 2fJ(l), and R1 = 0.0256, wR2 = 0.0727, and GOF = 1.102 for all 3290 data, maximin residual electron density +1.034/-0.406 e k 3. Bridge to Chapter IV We have now successfully demonstrated the utility of our synthetic procedure in making two homometallic M 13 compounds (AIl3 and GaB) as well as a heterometallic compound (Ga7In6) described in Chapter III. In Chapter IV, a full series of heterometallic GalIn compounds with varying indium content will be described, as well as the first examples of AllIn heterometallic compounds. Additional characterization details are also provided for these compounds. 58 CHAPTER IV PREPARATION OF A SERIES OF NOVEL HETEROMETALLIC TRIDECAMERIC GROUP 13 HYDROXIDE CLUSTERS Introduction I contributed to this work through the isolation of several of the new GalIn clusters, as well as the AIglns cluster. I also did the primary writing of the manuscript that is to be submitted for publication. Sharon A. Betterton performed the variable- temperature powder X-ray diffraction experiments and analyzed the resultant data. Maisha K. Kamunde-Devonish isolated the Al9In4 structure. Dr. Jason T. Gatlin isolated GalIn clusters, performed the electron probe microanalysis characterization, and sent samples for elemental analysis. Dr. Lev N. Zakharov is responsible for all single-crystal X-ray data. Prof. Darren W. Johnson and Prof. Douglas A. Keszler were the principle investigators for this work and provided editorial assistance. 59 The Group 13 metals, particularly gallium and aluminum, possess a "bewildering variety of behavior" and often exhibit similar hydrolytic behavior.1 In solution, the Group 13 metals exist in a dynamic series of species from water-ligated monomers to hydroxide oligomers, dependent upon pH and concentration variables.2-10 In contrast to this, surprisingly few discrete Group 13 metal hydroxide structures have been isolated and characterized. It is likely that their "bewildering variety of behavior" is what has hindered the successful isolation of discrete structures, making isolation of one major product difficult. Aluminum has received much attention due to its environmental prevalence and effects such as its implication in the transport of heavy metals downstream from mining areas, in addition to observed phytotoxicity of some aluminum oligomers themselves.25.7,8,11-14 Gallium and indium clusters have also been studied, albeit to a lesser extent.4,6.9.l5 I have described hydroxide structures containing each of these metals in the preceding chapters; Ga13, AlB, and a heterometallic Ga7In6.8-1O,15 We sometimes refer to these structures as "flat" to differentiate them from the more widely studied Keggin-AI13 structures.7,16,17 However, this chapter will continue to use abbreviated nomenclature referring to the metal content of the clusters only; i.e. [Gal3(u3- OH)6(u-OH)18(H20)24](N03)15 or "flat-Gal 3" will be designated as GaB (Figure 4.1).15 Until our original report of the GaB structure, the only Group 13 metal hydroxide structures that had been isolated as purely inorganic structures were aluminum-based (Ah, A18, and AlB hydroxides were known), with the lack of structures likely owing testament to the difficulty of their synthesis and isolation.2,5,6,18 Such compounds are usually isolated through hydrolysis ofmetals salts, in combination with heating or other 60 steps. Fedin and co-workers also reported the isolation of GaD utilizing curcurbit[6]uril and pyridine in the crystallization process, though the exact role of cucurbit[6]uril is unknown.!> The judicious choice of supporting ligands bound to the exterior metal sites in place of water molecules has enabled the synthesis of a wider variety of Group 13 hydroxide structures.3.4.12.19.20 Aminocarboxylate ligands such as N-(2- hydroxyethyl)iminodiacetic acid (I-I 1 heidi) have been used to synthesize Jigand-suPPOlted versions of AID and GaJ:l, as well as several lower nuclearity structures.4.21.22 It is suggested that bound ligands act to balance charge and control hydrolysis activity.12.2o A summary of relevant clusters (both inorganic and ligand-supported) is found in Chapter 1. Figure 4.1. Top and side view of a tddecameric cluster. On the left can be a top view oftridecameric structure. The profile view showing the up-and-down alternation of the outer ring of metal centers is seen on the right. Metal ions are blue, oxygen red, and hydrogen white. Jmage generated with PyMol. One problem with many of these preparations (for both the inorganic and ligand- supported clusters) has been the lack of wide utility. Many procedures result in the 61 isolation of only one or two structures, making a more general synthetic strategy highly desirable. Another common pitfall is the limited scale and low isolated yields for many of these metal hydroxides. The syntheses often provide only a few crystals of the reported compounds.2 Such issues have largely hindered the application of these clusters in areas such as materials precursors. Perhaps more notably, in both the purely inorganic and the ligand-supported structures, one cluster type has been absent: a heterometallic hydroxide structure with a mixture of Group 13 metal ions. I described in Chapter III the first example I have encountered of this heterometallic hydroxide structure type, and also described the use of this compound as a single-source precursor for the fabrication of solution processed amorphous metal oxide thin films. 9 This expanded on our previous isolations ofGa13 and AI13 and began to show a broader applicability of this synthetic method. Three distinct metal coordination environments are seen in these tridecameric metal hydroxide clusters (Figure 4.1). The metal ions are bridged to each other by hydroxide groups, with water ligands filling the four remaining coordination sites of the exterior metal ion sites. The central metal ion is bound to six ,u3-OH ligands while the six interior ring metals are bound to two ,u3-OH and four ,u-OH ligands each. The six exterior metal ions are coordinated to two ,u-OH ligands and four water molecules. More comprehensive descriptions of the metal center coordination environments are given in the "Structural Description" sections of this chapter. Our synthesis of these clusters proceeds through an unusual course, whereby Group 13 metal nitrate salts are dissolved in methanol with an organic nitroso compound 62 as an additive. These solutions are then evaporated open to air at room temperature; within two weeks crystals begin to fonn on the bottom of the scintillation vials. With the Gal3 and Ga7In6 clusters, yields up to 85 % and 95 %, respectively, were achieved using the di-n-butylnitrosamine additive.9 Using a slightly altered procedure with added base (NaOH, NH40H, or AI(OH)3), I was able to fonn the Al13 cluster as well, albeit in lower yields and with the fonnation of other inorganic salts.s To isolate the Ga7In6 cluster, a large excess of indium nitrate salt is required to drive fonnation of the fully indium- substituted cluster: a one-to-twelve ratio of gallium nitrate to indium nitrate was used. The natural curiosity was whether varying equivalents of indium could be incorporated into the heterometallic clusters as well. In follow-up to the previous chapters, Chapter IV describes a full series ofheterometallic Ga7+xIn6-x (x = 1 - 5) hydroxide structures which are isolated in similar yields to the previously discussed Ga7In6 heterometallic cluster, in addition to the first heterometallic AllIn hydroxide clusters: AIs1ns and A19In4. Experimental Details General Methods. All reagents were purchased from commercial sources and used as received. Gallium and indium nitrate salts were purchased from Strem Chemicals. Aluminum nitrate was obtained from our reuse facility and originally manufactured by Baker and Adamson. Di-n-butylnitrosamine was purchased from TCI AmeIica. Methanol was also used as received, with no drying attempted. Unless 63 specified, all reactions were conducted in standard 20 mL scintillation vials, open to atmosphere and at room temperature. Synthesis ofGalIn Heterometallic Compounds Synthesis of[Ga7+xIn6-x(fl3-0H)dfl-OH) 1s(H20h4}(N03) 15 (Ga7In6, GasIn5' Ga9In4' GalO1n3, GaIllIl], GaI2In). Methanolic solutions of Ga(N03)3 and In(N03)3 were prepared. Di-n-butylnitrosamine was added to these solutions via a syringe. The solutions were shaken vigorously, uncapped, and evaporated at room temperature. After several weeks, clear, colorless, cubic single crystals of the heterometallic GalIn clusters fonned on the bottom of the vials in > 90 % yield, as calculated with Ga(N03)3 as the limiting reagent and assuming six H20 ligands per Ga(N03)3 equivalent. Crystals were isolated by decanting the remaining pale yellow liquid or removing it with a syringe, then rinsing the product crystals with ethyl acetate. Alternately, an acetone wash also effectively washes away residual nitroso compound. The synthetic details are the same for all of the heterometallic GalIn clusters, with the only difference being the starting Ga(N03)3:In(N03)3 ratios which are as follows: Ga71n6 requires a 1: 12 ratio, Gaslns requires a 2: 11 ratio, Ga91n4 requires a 1:2 ratio, GalOln3 requires a 7:6 ratio, Gall1n2 requires a 2:1 ratio, and Gal2ln requires a 5:1 ratio (Scheme 4.1). x Ga(N03h y In(N03b ~N~ NO MeOH ... Ga71nG: x = 1, Y = 12 Gae1n5: X =2, Y = 11 Gagln,: X =1, Y =2 Ga101n3: X = 7, Y=6 Ga111n2: x =2, Y=1 Ga121nf x =5, Y = 1 64 Scheme 4.1. Starting material ratios for GalIn clusters. The number of Ga(N03h equivalents are presented as the x values, while the In(N03h equivalents are presented as the y values. Synthesis ofHeterometallic AllIn Compounds Al(N03)3 (0.251 g, 8 equiv.) and In(N03)3 (0.126 g, 5 equiv.) was prepared in 3.2 mL of a 0.26 M NaOH solution in MeOH. Di-n-butylnitrosamine (0.318 g, 24 equiv.) was then added to the solution via a syringe. The solution was shaken vigorously until all solids were dissolved, uncapped, and the solvent evaporated over several weeks. If a large amount of solid has precipitated, the mixture can be filtered. A mixture of crystalline material is obtained, with larger crystals on the sides of the vial that are potentially M(N03)3 salts, sodium salts, Al(OHh In(OHh or other inorganic impurities. On the bottom of the vial, smaller crystals form which contain the desired product. As with the Galin clusters, the remaining liquid was decanted or removed with a syringe, then crystals were rinsed with ethyl acetate to remove the residual di-n-butylnitrosamine. The same cluster has been isolated using an 9:4 ratio ofAl(N03)3 to In(N03h as well. 65 Synthesis of[Alsln5(/13-0H)d/1-0H) ls(H20)24}(N03) 15 (A/sIns). The same synthetic procedure used to isolate A19In4 was used for Als1ns except either 6:7 or 7:6 ratios ofAl(N03h to In(N03)3 were applied in the synthetic procedure. Description of Characterization Methods Single-crystal X-ray Diffraction (XRD). XRD experiments were carried out on a Bruker Smart Apex diffractometer at 153 K and 173 K using Mo Ka radiation (J, = 0.71073 A). Absorption corrections were applied by SADABS. The structures were solved by direct methods, completed by subsequent difference Fourier syntheses, and refined by full-matrix least-squares procedures on F 2 . Highly disordered N03- anions and solvent methanol molecules were treated by SQUEEZE.23 In the case of Ga7In6' corrections of the X-ray data by SQUEEZE (638 electrons/cell) are close to the required value of 603 electrons/cell for 9 N03- anions and 18 methanol molecules in the full unit cell. All non-hydrogen atoms were refined with anisotropic thermal parameters. Hydrogen atoms have not been taken into consideration. Refinements of the crystal structure of Ga7In6 without symmetry restrictions on occupation factors for the gallium and indium atoms show that the refined occupation factors of the Ga(1), Ga(2), and In(1) atoms are very close to those based on the crystal symmetry. The data from the other heterometallic structures were analyzed by the same method. Electron probe microanalysis. Electron probe microanalysis (EPMA) measurements were carried out on a Cameca SXl 00 and analyzed with Probe for EPMA 66 software. The measurements were nonnalized to the number of oxygen atoms, and the number of gallium and indium atoms calculated relative to this. Elemental Analysis. Elemental analysis (EA) on compounds was conducted by Desert Analytics. Samples of bulk amounts of single-crystalline material were sent and used for analysis. Powder X-ray DifFaction. Powder X-ray diffraction data were collected on a Rigaku Miniflex diffractometer with eu Ka radiation. Description of Synthesized Compounds Based on the successful isolation of the Gal3 and AlB structures, we also examined binary metal nitrate mixtures, first those containing gallium and indium. The first heterometallic cluster we chose to target was one with full indium substitution in the six peripheral metal positions, [Ga7In6(u3-0HMjl-OH)18(H20)24](N03)lS, (Ga7In6). A 7:6 ratio of Ga(N03)3 to In(N03)3 was used in the reaction. Analysis of the single- crystalline product showed the actual gallium-to-indium ratio to be 10:3 (GaI01n3). Based on this observation, it was suspected that an excess ofIn(N03)3 would be required to achieve substitution of all six gallium centers. To verify this, a 1:12 Ga(N03)3-to- In(N03)3 ratio was used, which successfully afforded the Ga7In6 cluster. Following this, a 2: 11 ratio of Ga(N03)3 to In(N03)3 was tested and afforded a GasIns cluster. Graphing these three GalIn heterometallic compounds along with GaB using the number of gallium atoms in the product cluster as the y-axis values and number of starting gallium 67 equivalents (out of 13 total metal equivalents) as the x-axis values provided a linear fit (Graph 4.1). This fit was used to predict the starting material ratio for the missing GalIn structures. Calculating missing y values of9, 1J, and 12 (the number of gallium atoms in the desired product cl uster for the Gai)ln4, Ga Illnz, and Ga 121n compounds, respectively), we were able to predict the required stal1ing material ratios to be roughly 1:2 (4.71 :8.29),2: 1 (8.88:4. I2), and 5: I (10.97:2.03) ofGa(NOJ}l to In(NOJ)J for Ga91n4, Ga1111lz, and Gadn, respectively (Graph 4.1). Applying these ratios, we successfully synthesized the remaining clusters, as confirmed by single-crystal X-ray diffraction experiments. Bulk characterization techniques have confirmed the metal ratios in these clusters. Both EDS and EA were performed on the isolated bulk solids, and showed only slight variation fi'om the compositions determined by single-crystal XRD in the case of the GalOln] and Ga7lnC, compounds (Table 4.1). 13 12 II III <) H 7 II 2 J 4 5 n 7 H '} J(l II 12 13 Graph 4.1. G."aph used for GalIn ratio prediction of missing clusters. X axis values represent the number of Ga(NOJh equivalents used (Ollt of J 3 total). Y axis values are gallium equivalents in the product cluster. The blue triangles represent the known clusters; red squares represent the predicted values. 68 Table 4.1. Characterization data for cluster compounds. The table summarizes data gathered by single-crystal XRD, EA, and EPMA. The data for the bulk characterization methods (EA and EPMA) matches the data collected by single-crystal XRD, except in the case of Ga101n3, where EA gave a ratio closer to Ga9In4, and Ga7In6, where both EA and EPMA determine a ratio closer to Ga6In7. Data collected by Dr. Jason Gatlin. Starting Material Characterization XRD EA EPMA Gallium Indium Gallium Indium Gallium Indium Gallium Indium Gao 13 0 13 0 13 0 - - Gauln 5 I 12 I 12 1 12 1 Gallln2 2 1 1J 2 1I 2 11 2 Ga1oln3 7 6 10 3 9 4 10 3 Ga91n4 1 2 9 4 - - - - GaMins 2 11 8 5 - - - - Ga,In6 1 12 7 6 6 7 6 7 I have observed one other case of such a high degree of control over metal content in heterometallic clusters. Cronin and co-workers were able to demonstrate metal ratio control in dodecanuclear Ni/Co clusters using both starting material stoichiometry and solution pH.24 Utilizing a carbonate core template, cis, trans-l ,3,5-triaminocyclohexane, and acetate directing ligands, they were able to isolate both the homometallic Ni l2 and C012 complexes, as well as a series ofheterometallic clusters. In our system, starting material stoichiometry is exclusively used to control the metal ratio. While the similar size and electron count of Co and Ni prevented differentiation of the metals in their system, the relative position of gallium and indium atoms are determined in our clusters, however the exact location of the indium atoms and their arrangement within the outer ring metal position cannot be determined, as an average of the six positions is seen crystallographically. This is discussed in more detail in the following section. 69 Structural Description ofGalIn Compounds Characterization of the heterometallic GalIn metal hydroxide clusters by single- crystal X-ray diffraction shows a structure similar to that observed in the previously reported Ga13, AlB, and Ga7In6 clusters. As an example, crystals of Ga7In6 are hexagonal, R"3 (no. 148). The Ga7In6 cation is on a"3 axes, analogous to the related GaB cluster. Two N03- anions (in general positions) provide twelve N03- anions per Ga7In6 cation. Three other N03- anions and solvent methanol molecules (in general positions as well) are highly disordered and randomly fill six other possible positions around the Ga7In6 cation. In addition to characterization by full data collections, we are able to roughly characterize the different compounds by unit cell volumes-the unit cell volume increases as more indium atoms are incorporated, from ~6494 A3 for GaB to 6774 A3 for Ga7In6. Three different metal positions can be assigned in these compounds; Ml, M2, and M3 (Figure 4.2). All compounds feature a conserved planar core of seven gallium atoms (Ml and M2) bridged by f.13- and f.1-0H ligands. Ml is the central metal atom, bridged via f.13 hydroxyl ligands to a ring of six more metal atoms (M2). M2 metal centers are bound to each other through six f.1-0H bridges. Six more metal atoms (M3) are bound to the M2 centers through twelve f.1-0H bridges. In the case of Ga7In6, the six bond distances between Ml and the f.13 oxygen atoms average to 1.966 A. The average bond length between the f.13 oxygen atoms and the six M2 metal centers is 2.113 A. Six f.1-0Hs bridge the six M2 gallium atoms, with an average bond distance of 1.915 A. The six M3 metal atoms are bound to the core by f.1-0H bridges and alternate above and below the 70 plane of the central atoms. The average M3-,l1-01-l distance is 2.086 A. It is in this outer ring of metal ions only that indium atoms are found, ranging hom one indium to six. Each outer ring metal atom (either Ga(IlI) or In(IlI)) tills its coordination sphere with four water ligands. The average In-O(I-I~O)bond length is 2. 166 A. This is close to the distances found before in complexes with an In-O(1-I 20) bond: for example, 2.156 and 2. 158 A were reported in catena-[(p-oxalato-O,O',O',O"')-bis(p-O',O'''O'")-tetraaqua- d iind ium dihydrato]. 25 Figure 4.2. Three different metal centers are found in the clusters. M 1 is the central metal site, M2 are the ring of six metal sites that form the remainder of the planar core, and M3 metal centers occupy the exterior metal sites and alternate above and below the central planar core. These clusters crystallize in R3 with Ml on the '3 special position, so only one M2 and M3 from each of the next six-metal rings is unique. Because of this, the electron count for the outer metal position is an averaged number representative of the gallium-to- indium ratio. Solving for this ratio determines the number of indium atoms. In some --------------------- ----- -------~----- - ~------------ --------- 71 cases, partial atom values are obtained for gallium and indium stoichiometries, such as GalO.3In2.7. Such a ratio could result from a crystal structure containing a mixture of GalOIn3 and GallIn2 clusters, with a Galo.3In2.7 stoichiometry representing a 70:30 ratio of the clusters in the solid state, though a more broad product distribution is possible, even likely. We have not ruled out the possibility that other metal stoichiometries exist within the same crystal, and crystals characterized as GalOIn3 contain molecules with Ga9In4' Ga IOIn3, and Ga uIn2 stoichiometries, if not even more. Because only one M3 site is observed and the averaged electron count is used for quantification, we cannot determine the precise location of indium atoms in relation to each other. For example, there are three potential arrangements for M3 indium atoms in GalOIn3: all indium atoms next to each other, two next to each other with the third separated by a gallium atom, or all three altemating with gallium atoms (analogous to the 1,2,3; 1,2,4; or 1,3,5 substitution pattems in benzene, Figure 4.3). I have postulated that the arrangement with altemating gallium and indium atoms is most likely, as it affords the most space between the larger indium atoms and would be the most symmetric of the species. However, a mixture ofthe possible "isomers" is likely present and any claims of regioselectivity represent conjecture at this stage. 72 Figure 4.3. Different potential arrangements of the outer ring gallium and indium metal centers. Three different arrangements are possible for the six outer ring metal centers. A staggered pattern is depicted on the left; an arrangement with two neighboring indium centers is in the middle, and on the right is an arrangement with all indium and gallium atoms neighboring each other. Gallium atoms are blue; indium atoms are green. Images generated with PyMol. Slructural Descnj)lioJ1 oIAI/IJ1 COI11!7ouml.\· Single-crystal X-ray diffraction was used to characterize these compounds and their structures were, as expected, very similar to the other triclecameric Group 13 metal hydroxides described in this dissertation. Unlike the A11:I structure which is triclinic, these aluminum containing structures are rhombohedral, as are Gal:l and all the GalIn heterometallic structures. Similar to the synthesis of All :I however, the addition of base is required in order to produce these clusters. We propose the same rationale as that case- the Lewis acidity of aluminum is less than that of gallium or indium, and therefore to promote the formation of hydroxide bridges, base must be added to aid the deprotonation of water ligands.K.26 In both AllIn compounds, there are again three different mctal environments, though the M3 position contains both aluminum and indium. Using AIKIns as a example, the AI-O bond distance from Ml to I':l-OH is 1.897 A. The average I':l-OH 73 to M2 distance is 2.075 A. The average bond length from M2 aluminum atoms to the p- OH ligands bridging the M2 sites is 1.859 A. The average distance between M3 and the p-OH ligands that bridge M3 to the core is 2.083 A. Lastly, the average In-O(H;>O) distance is 2.156 A. Figure 4.4 pictorially summarizes all the cluster compounds that have been synthesized to date. = l\Iuminum • ::::Gallium = Indiurn Figure 4.4. Sumnutry of tridccameric cJnstcl·s. To date ten different compounds have been isolated and characterized. Three aluminum-based compounds have been isolated (AID, AI,)ln4, and Alliin:;) and seven gallium-based compounds (GaD, Gantn, Galllnz, GalOJn;l, Ga91n4, Gallln:;, and Ga7Inc;). In the case where multiple regioisomers are possible, the highest symmetry isomer is shown. Images generated with PyMol. Variable-temperature Powder X-ray Diffraction To understand the effect varying Ga:ln ratios have on the decomposition properties of Ga7+x!n(>-" clusters, bulk samples were calcined and studied via XRD. As illustrated in Figure 4.5, dehydration of crystalline GaD produces a white amorphous 74 solid that persists to ~600 0c. By 700 °e, reflections indicative of monoclinic gallium oxide W-Ga203) are evident. The decomposition behavior of Gat01n3 is similar (Figure 4.6), demonstrating a loss of crystallinity upon dehydration. Here, the material (also white) remains amorphous up to ~800 °e. At 900 °e, peaks attributable to ~-Ga203 are clearly evident, and intensities increase with heating. Of note in these patterns is the shift of Ga203 peaks to lower 28 values with increasing indium content, indicating an increase in cell volume. Similar behavior is observed during the heating of Ga9In4 (Figure 4.7). The white powder remains amorphous to ~800 °e, and at ~900 °e, ~-Ga203 peaks emerge, again shifted to lower 28 values from the reference peaks. The results of calcining the indium-rich Gas1ns (Figure 4.8) and Ga7In6 (Figure 4.9) are distinct from those of the previous compounds. There is no amorphous phase formed upon heating either sample to 200°C; rather, peaks consistent with the cubic In203 structure type are present, and the powders are yellow. This phase is stable to ~800 °e, and at ~900 °e, ~-Ga203 peaks become apparent. The material remains a mixture of both phases to 1100 °C. The incorporation of indium into the ~-Ga203 structure has been previously studied, and it is known that indium preferentially substitutes 011 Ga sites, forming a ~­ Galn03 structure.27 ,28 These studies have demonstrated that the substitution of the larger indium results in an increase in the a, b, and c lattice parameters and a decrease in the ~ parameter, changes that ostensibly produce a shift of ~-Ga203 peaks to lower 28 values, which we observed in our studies.28 These peak shifts, as a function of indium content, are detailed in Figure 4.10. As In203 peaks emerge in the oxides formed fi'om Gas1ns 75 and Ga7In6, and we observe no significant peak shifts above x = 0.62, our system appears to have a saturation limit between x = 0.62 (32 mol %) and x = 0.77 (38 mol %). This value is smaller than that reported by Edwards et al. (42 mol %),27 but falls within the range reported by Shannon and Prewitt (33-50 mol %).28 P·Ga20 3 - 1100°C ::> on I J{] Figure 4.11. NMR spectra of di-II-butylnitrosamine. The top spectrum of di-n- butyl nitrosamine was taken before a GaD synthesis reaction. The bottom spectrum was taken after a Gal:l synthesis, from di-n-butylnitrosamine recovered from the scintillation vial. compounds, by simply dissolving Ga(NO,h or a mixture of Ga(NO,), and In(N03h in MeOH and evaporating the solution. In these cases, the crystallization generally takes much longer and the syntheses are less reliable, however yields have been shown to be comparabie, with yieids of 80-t- % having been achieved. AdditionaUy, when a methanolic Ga(NO,h solution is placed in a vapor diffusion chamber with ethyl acetate, similar to the method used to produce the Ga- and In-acetate chains I will describe in 80 Chapter VI, GaB was again isolated. 10 Crystals formed on the bottom of the vial were confirmed as Gal3, with a yield of 83 %. Aside from our own work, Fedin and coworkers have also been able to isolate Ga13 in the absence of a nitroso compound, using cucurbit[6]uril and pyridine instead.6 The role of the cucurbit[6]uril is also unclear however, and the authors were unable to isolate the Ga13 compound in its absence through variations in Ga(III) concentrations, temperature, and pH. With multiple examples of isolation in the absence of a nitroso compound, what is the role played by the nitroso compound in the reaction? One postulate is that the nitroso compounds are slightly basic, as the oxygen of the nitroso functionality can be protonated. This slight increase in basicity may aid in the formation of the hydroxide bridges needed to produce the clusters. Another suggestion involves solubility: Ga(N03)3 and Gal3 are not soluble in di-n-butylnitrosamine, but are readily methanol soluble. The methanol evaporates much more quickly than nitrosamine, creating a system similar to vapor diffusion. The solubility properties of the solution steadily change to those of the nitrosamine, encouraging the crystallization of Ga13. The slow evaporation that results leads to reliable isolation of the clusters. It has recently been suggested by Prof. Bill Casey that using methanol as the solvent is key as well, by slowing down the hydrolysis activity and controlling aggregation by limiting the amount of H20 available to form the hydroxide bridges.29 This agrees with our findings that the clusters can form in absence of a nitroso compound, but the nitroso functionality clearly plays a role, speeding up crystallization and making for a more reliable, easily isolated product. We believe the use ofMeOH as 81 solvent is a key element to this synthesis, compared to most syntheses of Group 13 hydroxide structures where water is the solvent. As stated, this is believed to slow down the hydrolysis of the Ga(III) ions, and better facilitate aggregation and crystallization of the tridecameric cluster over more random products. Conclusion In conclusion, we have been able to demonstrate control over the metal ratios in a series ofheterometallic Ga/In and AllIn tridecameric metal hydroxide clusters. These represent the first such heterometallic clusters of their kind, and have implications for the use of such clusters as single-source precursors for the fabrication of indium-gallium oxide and gallium oxide thin film semiconductors. Single-crystalline product is generated with tuned gallium-to-indium ratios in high-yielding, gram-scale syntheses. Such a wide range of potential materials would enable highly specific control of the material content of thin films made from these precursor compounds. Variable- temperature powder X-ray diffraction studies repOlied in this chapter suggest interesting decomposition behavior of the different potential precursor materials, particularly because we may have a possible route to prepare indium-gallium oxide thin films versus films with separated indium oxide/gallium oxide areas based on the gallium-to-indium ratio in the clusters used. We also hope that the preparation of these clusters will enable the isolation of additional heterometallic clusters. We are currently looking into the incorporation of metals other than those in Group 13, which if achieved may also hold 82 potential for the fabrication of additional thin film metal oxides. With the knowledge that heterometallic clusters are achievable, other researchers may wish to reexamine previous syntheses for the possibility of adding additional heterometallic structures to the examples presented in this chapter. Bridge to Chapter V I have described the synthetic preparation and uses of a series of homometallic and heterometallic tridecameric hydroxide clusters in Chapters II, III, and IV. In an effort to further derivatize these compounds and expand their applications, we next sought ways to achieve post-synthetic modification of the tridecameric structures. 83 CHAPTER V POST-SYNTHETIC MODIFICATION AND POTENTIAL CHEMICAL APPLICATIONS OF CLUSTER COMPOUNDS Introduction I contributed to the work herein by developing and perfOlming the procedures and composing the communication detailing the results. Matthew M. Baseman contributed to the results through the determination of yields for the metal exchange reactions and exploration of other possible isolable clusters. Laura A. Thompson also contributed to the findings through additional exploration ofpotential metal exchange reactions. Dr. Lev N. Zakharov perfonned all crystallographic studies. Prof. Darren W. Johnson was the principal investigator for this work and provided editorial assistance. 84 Throughout this dissertation I have described a series of homo- and heterometallic tridecameric Group 13 metal hydroxide clusters. 1-4 The first of these isolated was [Ga13(,u3-0H)6(Ll-OH)lS(H20)24](N03)IS (Ga13).4 Group 13 metal hydroxide clusters such as Ga13 represent an area of broad scientific interest. In addition to their importance to basic scientific research, they have many practical applications and real-world implications. Many aluminum hydroxide species have either been proven or proposed to exist in natural aqueous solutions.s-7 Some such species are used to flocculate and remove heavy metal ions from solution, while others have been shown to transport heavy metals out of mining areas contributing to heavy metal pollution. Additionally, some of the compounds are directly phytotoxic.7.8 The traditional syntheses of these clusters can often be problematic, with extensive reaction/crystallization times and low isolated yields. We have previously demonstrated an improved synthetic procedure for making the GaB and Ga7+xln6-x compounds that has proven to be high-yielding and scalable (Chapter III).3 Using this, we have shown that Ga13 and [Ga7In6Vl3-0H)6(,u- OH)lS(H20)24](N03)IS (Ga7In6) hydroxide clusters can be used as single-source solution precursors for thin film metal oxide semiconductors.3 Following the successful isolation of a series oftridecameric Group 13 hydroxide clusters (as described in Chapters II, III, and IV) and demonstration of some of their applications (i.e. thin film precursors, Chapter III) I sought to further expand their usefulness. 1-3 Post-synthetic modification represents a potential method, giving several options to control the properties of the cluster compounds. Using structures prepared by our developed synthetic route, I proposed structural modification through two potential 85 routes: introduction of new metals by metal exchange reactions or functionalization of the clusters through ligand exchange reactions. There is plior literature precedent in related cluster compounds for both metal exchange and ligand substitution. 9.1 0 In both situations, solution stability ofthe cluster compounds is a paramount concern in achieving the desired transfonnation. In metal exchange reactions, the outer metal centers must be labile enough to be replaced by competing metal ions. In the case ofligand exchange, the clusters must persist in solution and not fully disassociate, which would leave no exterior metal ions to bind to ligands. There are three potential degrees of solution stability. The clusters may be fully stable in solution, they may partially dissociate, or they may fully dissociate. We can dissolve the clusters in, and recrystallize them from, protic solvents (H20 and MeOH) as well as successfully implement them as single-source precursors for metal oxide thin films, which gives evidence for some degree of solution stability/aggregation.3 In order to test the solution stability of our clusters, we sought to find out how and to what extent our clusters participate in metal and/or ligand exchange reactions by investigating three pertinent research questions: 1) Are the six peripheral metal centers labile in solution, 2) Do homogeneous conditions facilitate exchange of water ligands by organic ligands, and 3) Would heterogeneous or biphasic reaction conditions facilitate ligand exchange? 86 Summary of Experimental Approach First off, I chose to investigate whether the six outer ring metal ions in Gal3 are labile in solution. If Gal3 clusters partially or fully dissociate in solution, the presence of an excess of In(N03)3 might promote the fonnation ofheterometallic clusters upon solvent evaporation, as in our previously established synthesis of Ga7In6. No evidence for an Inl3 cluster has been observed, so any indium incorporation should be due to dissociation and replacement of outer ring gallium ions. These studies were perfonned with and without DBNA additive and over different time periods. The second question concerned organic ligand exchange with the H20 ligands of GaB in solution. Protons on the clusters (from bridging hydroxide and H20 ligands) rapidly exchange in protic solvents, and as a result lack an appropriate handle for 1H NMR characterization. Substitution of water ligands for ligands with non-exchangeable protons could provide such a handle for NMR characterization. Free ligands should exhibit chemical shifts different from those ofligands bound to clusters, which in turn may also exhibit different chemical shifts from ligands bound to monomeric gallium. In experiments investigating this issue, Gal3 was dissolved in either H20 or MeOH and combined with various ligands, sometimes with pyridine to help deprotonate and solubilize the ligand. Examples of suitable ligand classes examined include benzene carboxylates, bipyridines, and catechols. My research in this area examined the extent of solution stability and feasibility ofligand exchange and characterization. The last research question described in this chapter concerns whether or not organic-soluble ligands can be bound to Gao under heterogeneous or biphasic reaction 87 conditions. Organic-soluble ligands were dissolved in p-xylene and other high-boiling organic solvents containing a suspension of solid Ga\3. The solutions were then heated above 100 DC to drive off coordinating water ligands, promoting the exchange of ligands. X-ray crystallography and NMR were then used to analyze the success and degree of ligand exchange.] J Biphasic layering experiments were also conducted, in which ligands were dissolved in organic solvents and the Gal3 clusters in H20 or MeOH. To address the questions outlined above, experimentation was undertaken to examine the overarching question of solution stability. All experiments utilized single crystalline Ga 13, chosen because of synthetic ease, structural simplicity (only gallium atoms), and apparent solution stability. Some of the first studies looking at the solution stability of Gal3 were performed using 69Ga NMR. Samples were prepared as aqueous solutions of Ga(N03)3 and Ga\3 and spectra were recorded on a 500 MHz NMR spectrometer. The chemical shift of aqueous Ga(N03)3 was recorded, set to 0 ppm (a value characteristic of octahedral gallium), and used as a reference peak. In solution, the Gal3 cluster showed a similar peak centered around 0 ppm, perhaps slightly broader in nature. As a result, it is unlikely that GaB is fOinling tetrahedral species in solution, as tetrahedral species would exhibit a different chemical shift. However, these results were inconclusive in terms of showing whether GaB was stable in solution. Rather, they showed that gallium was present in octahedral coordination only and gallium NMR was ruled out as an effective technique to evaluate solution speciation. 88 Metal Exchange Background Our foray into this field was initially quite serendipitous, discovering an interesting synthesis ofthe tridecameric hydroxide compounds featuring the addition of organic molecules with nitroso functional groups to methanolic solutions of metal nitrate salts. Expanding on this, the synthesis of a series ofheterometallic Galin tridecameric clusters was achieved. These compounds have indium atoms replacing gallium atoms in the outer ring of metal atoms. Indium composition can range from one to six atoms in these positions, as detailed in Chapters III and IV. The heterometallic clusters are synthesized from a mixture of Ga(N03)3 and In(N03)3 starting materials. I streamlined this technique using di-n-butylnitrosamine to give 80+ % yields of the Ga13 compound and 90+ % yields of the heterometallic GalIn clusters.3 One significant drawback to this synthesis is the large excess of In(N03h required to isolate clusters with greater indium substitution. For example, Ga7In6 requires a 1:12 ratio of Ga(N03h to In(N03h. We suspect this may be due to the solubility of In(N03h in di-n-butylnitrosamine; it is soluble while Ga(N03h and the tridecamelic clusters are both insoluble. I examined ways to address this excess requirement, in addition to examining the solution stability of these clusters. I discovered that the same heterometallic clusters can be produced stm1ing with Ga13 by performing a metal exchange reaction with In(N03h to yield [Ga7+xln6-x(u3- OH)6(u-OH)18(H20h4](N03)lS, which requires significantly less In(N03h than our previously repOlied procedure. Metal exchange reactions have been reported in Keggin and Anderson clusters (Figure 5.1), which are similar to our tridecameric M 13 compounds. In Keggin-AII3' the 89 Figure 5.1. Rep.-esentation of Keggill-Al n, Anderson, and M n clusters. In both the Keggin and Anderson clusters, the central metal can be exchanged; in our M I3 clusters, the six peripheral metal atoms can be exchanged. I~ central tetrahedral metal atom can be replaced by Ga or Ge. - In Anderson clusters, a variety of di- and trivalent metal atoms have been substituted in the central position of the M070 2/- clusters. 13 . '4 Based on these well-established phenomena in these related cluster types, I explored the possibility for metal exchange in our clusters. In this section, the results of metal exchange studies are presented wherein methanolic solutions of Gan and In(N03)3 are generated, which upon crystallization produce heterometallic Ga7+xIn6-x clusters. These clusters are structurally identical to those previously isolated. The metal exchange-where octahedral, non-central metal atoms are substituted-represents the first example I have observed of such behavior in Group 13 metal hydroxides. In the case of Keggin-Al 13 metal exchange, only the single central metal which is bound by oxide jigands is replaced, whereas up to tive metal atoms in the exterior position are exchangeable using the method described in this chapter.~·15 This method reduces the excess of indium equivalents that are required to form the heterometallic clusters with 90 higher indium atom substitution, which in turn reduces the amount of di-n- butylnitrosamine required, as this amount is proportional to that of the metal nitrate.3 Metal Exchange Procedure My decision to make GalIn heterometallic structures by the addition of In(N03)3 to a solution containing GaB was chosen because neither In13, nor any other oligomeric indium hydroxide cluster has been observed. As a result, it was assumed that the presence of indium in the final structure would occur through incorporation of indium atoms into the GaB structure by a metal exchange process. GaB (0.10 g, 0.037 mmol), In(N03)J-nH20 (0.27 g, 0.897 mmol), and di-n-butylnitrosamine (0.20 g, 1.27 mmol) were dissolved in 10 mL of methanol. After the solution was evaporated, crystals formed which were detennined to be [GasIn5(,u3-0H)6(,u-OH)IS(H20)z4](N03)15 by single-crystal X-ray diffraction. Such metal exchange can also be achieved in water or MeOH; GaB and In(N03)3'nH20 are dissolved in 10 mL of H20 or MeOH with no nitrosamine present. These reactions give the same products as those isolated with the nitrosamine present. The use of H20 lengthens evaporation times, but both procedures (MeOH or H20 only) eliminate the use of additional di-n-butylnitrosamine, a desirable outcome due to the expense and hazards associated with this compound. The reverse reaction can also be achieved. If Ga7In6 is dissolved in MeOH with Ga(N03)J-nH20, gallium centers can be introduced into the structures upon recrystallization. While interesting, this result is 91 less synthetically useful than the procedure to synthesize the indium-inclusive structures from Gau (Figure 5.2). I also examined the effect of time on structural composition and it appems th- /, .\ I;""'; c Figure 5.3. Preliminary EXAFS data. This figure (generated with the Altemis software package) depicts three different sets of EXAFS data in R space. The blue trace (merge) represents aqueous Gal:!. The red trace (fit I) is a fit generated f[-om single- crystal X-ray diffraction data of solid phase Gal:!. The green trace (merge) is aqueous data for Ga(N03h possibilities include many metals that have not successfully been incorporated into MI3 clusters, offering a chance to isolate new heterometallic structurcs. To this end, other metal nitrate salts have been examined for incorporation into the M 13 fi·amework. Experiments to investigate this possibility were set lip similarly to the indium exchange reaction. Metal nitrate salts (or salts with other counteranions as availability necessitated) were dissolved in solutions of GaD, followed by evaporation of solvent. Initially, metals most similar to indium in hydrolysis constant (pKiI ) were examined, as they should be hydrolyzed under similar conditions to In(llI) ions. Such metals include H '+ A! H sHe 1+ d TJ 4+ Th I . f' . . d ..g~, " c', t, an 1. ese meta centers encompass a vanety 0 JOl1IC ra II, so it might be possible to discern the importance of that factor in isolation of the 94 heterometallic clusters, as we feel this also explains the inability to isolate In13. Additional metal ions such as Ru3+ and Fe3+ were tested based on our interest in the potential heterometallic structures. In initial studies, anywhere from six to thitty equivalents (versus Ga13) of metal salts were dissolved in MeOH solutions with GaB and di-n-butylnitrosamine. With Ru3+, both the chloride and nitrosyl nitrate salts were tested. Both afforded viscous brown/red liquids. In the case of Fe(N03)3/Ga13 solutions, the effect of pH on reactions was tested through added Hel and NaOH. In all cases, the only crystalline product isolated was recrystallized GaB, although in some cases no crystalline material was isolated at all. Looking to the future, the addition of base may be necessary with metals of higher pKa values in order to initiate formation of the hydroxide btidges necessary for synthesis and crystallization of the clusters. We remain optimistic that this method will provide access to a host of new heterometallic structures. DFT studies could also prove useful in evaluting the relative stabilities ofheterometallic structures in order to detennine which metal ions might prove fruitful. Ligand Exchange Background Another method I examined to modify the metal hydroxide clusters is through ligand exchange reactions. In the tridecameric clusters, each of the outer six metal ions is coordinated to four H20 ligands. In solution, these H20 ligands are expected to be in rapid exchange with solvent molecules (either H20 or MeOH), giving the opportunity to replace them. Ample precedent for ligand-supported clusters exists, such as the heidi- 95 bound versions of the Gal3 and AIl3 clusters that were described in Chapter I, which have a triply-deprotonated heidi ligand that replaces three of the H20 ligands on each of the six peripheral gallium or aluminum ions, as well as one of the hydroxide bridges. 10,16,17 Also discussed in Chapter I is the general lack of larger nuclearity metal hydroxide clusters isolated without stabilization by organic ligands. It is known that in the case of Keggin- AI 13, water ligands bound to the octahedral aluminum atoms can be substituted by various phenol-based ligands, which eventually causes decomposition of the cluster.9 This is due to the presence of only one water ligand per aluminum atom in Keggin-AI13' A bidentate ligand replaces the water ligand upon fonnation of the initial bond, and then causes breakage of a hydroxide bridge bond with fonnation of the second phenolate-aluminum bond. Decomposition of the entire cluster occurs rapidly thereafter. An even number of water ligands per metal center in our M13 clusters should decrease the likelihood of this occurrence. Ligand substitution is a worthwhile area of exploration for several reasons. First, the choice of peripheral ligands is usually dictated by which ligands facilitate cluster fonnation. In our case, the all-inorganic clusters with only H20 ligands on the outer ring metal ions are readily available, giving the opportunity to tailor ligand choice according to which ligands interest us, not simply which nucleate or template cluster growth. The ability to choose supporting ligands gives control over the properties and functionality exhibited by the clusters. The appropriate ligand could impose different solubility properties by altering the charge of the cluster or adding aliphatic content. Reactivity could also be imparted, such as a case involving substitution by a bidentate ligand 96 containing two olefin groups, which could undergo olefin metathesis with neighboring ligands to generate a large macrocycle surrounding the M 13 core (Figure 5.4). L (~~\LXfL L~'L@..; Ligand Exchange.. M > ~. ( M ')1 ., I Coupling " ,.I L L L\ i / L '--L-/ Macrocycle Figure 5.4. Functionalization of M 13 compound with organic ligands. In this idealized situation, organic ligands would bind to a M 13 core. The ligands would then be coupled together using additional chemistry, followed by hydrolysis or another process to remove the metal core, affording a large macrocycle. Functionalization ofM 13 clusters with appropriate ligands may facilitate self- assembly onto solid supports as well. For example, one could imagine assembling a monolayer of single-molecule magnets on a surface of interest by functionalizing them with appropriate ligands to bind to the surface. Additionally, a multidentate ligand may facilitate the synthesis and isolation of higher order structures. With aluminum and non- Group 13 metals such as iron and manganese, clusters with 15 and 19 metal ions are known in ligand-supported forms. In these cases, ligands with a high coordination number are often used, such as H3heidi or H5hpdta.lO.18 The addition of a ligand such as these to our tridecameric clusters may help isolate larger species. Ligand exchange reactions could prove useful in the research area of mineral milnics as ,vell. .i\lulninuln oxo/hydroxo clusters have been touted as natural mineral mimics, but there is limited research supporting this claim due to the general synthetic inaccessibility of such clusters.5.6 Mimics are useful because precise characterization of 97 bulk minerals can prove difficult. This can be likened to protein research, where small- molecule active site mimics are often synthesized to gain better understanding of the native protein. In this example, M13 clusters are the potential active site mimic, and natural minerals constitute the native protein. In one example of mineral ligand exchange, Barron et al. show that the mineral boehmite (AIO(OH)) can be functionalized with molecules such as benzoic acid. I I Gallium and aluminum have been shown to exhibit parallel hydrolytic behavior, so the observation of a similar functionalization of GaB or Al 13 with organic ligands such as benzoic acid would provide experimental evidence ofM13 mineral mimicry. It is conceivable that the M 13 clusters could act like a large single metal ion, a bulk mineral, or something different. Heterogeneous Ligand Exchange My first attempt at ligand exchange utilized a heterogeneous approach. Finely ground crystals of GaB were suspended in an organic solvent (typically an aromatic hydrocarbon) in which potential ligands were dissolved. The first solvent system explored was p-xylene, due to its high boiling point. A number of organic ligands including benzoic acid. toluic acid, malonic acid, and 2,2' -bipyridine were added to the suspension containing solid GaB. Control reactions with no ligand present were also set up. In p-xylene, reactions were generally heated for 24 hours at 120°C. Upon cooling, it was noted that the solutions had developed a yellow color and contained a powdery solid. There was also the distinct smell of an aromatic aldehyde after the heating period. 98 Because a single-crystalline product was not obtained, NMR analysis was used to analyze the resultant reaction mixtures. NMR spectra of the solid and liquid products recovered from the p-xylene solvent showed a complicated selies of peaks (Figure 5.5). It was noted that the spectra showed peaks possibly conesponding to a carboxylic acid, an aldehyde, and up to 23 peaks of varying intensity between 4 and 5.5 ppm. A number of control reactions run without ligands present in various solvents (p-xylene, toluene, and hexadecane) showed similar oxidation behavior. Control reactions run with Ga(N03)3, In(N03)3, and KN03 in the absence ofligands resulted in similar NMR spectra exhibiting peaks conesponding to oxidation products, though with different peak intensities (Ga(N03)3 is shown in Figure 5.5). Using gas chromatography-mass spectroscopy (GC/MS) it was determined the solution originally containing GaB and p-xylene still contained mostly p-xylene (55 %), but also p-tolualdehyde (15 %) and p-toluic acid (5 %), in addition to other lesser products. From these data, it was hypothesized that p- xylene was being oxidized (presumably by molecular oxygen) into these various products. Degassed reactions run under nitrogen exhibited much smaller peaks conesponding to the oxidation products, supporting the hypothesis that molecular oxygen plays a role in the observed oxidations, with activation by the GaB compound. In hopes of avoiding this oxidation side reaction, a variety of solvents were examined in place ofp-xylene, as it appeared too susceptible to oxidation. Solvents such as benzene, toluene, 1,1 ,2,2-tetrachloroethane, nitrobenzene, hexadecane, and liquid CO2 failed to yield concrete, characterizable results. Neat reactions were run, either by heating benzoic acid to above its melting point with Gal3 present, or using benzyl alcohol 99 as a solvent/ligand combination. Similar to p-xylene, NMR spectra suggested the oxidation of toluene and hexadecane. The other solvents tested failed to yield positive results as well. A wide variety of potential ligands were tested, including nitrilotriacetic acid disodium, toluic acid, malonic acid. and 2.2'-bipyridine. When 22' -bipyridine was used, new peaks shifted downfield appeared, but these were shown to correspond to protonated ligand, not a gallium-bipy complex. Figure 5.5. NMR spectra from ligand exchange reactions. The top spectrum corresponds to GaD heated with p-xylene, open to air. Note appearance of peaks in aldehyde and carboxylic acid range. The middle spectrum represents Ga(N03h with p- xylene. Similar peaks appear in this spectl"Llm as with GaD. The bottom spectrum shows Ga 13 with p-xylene under reduced oxygen conditions. The reaction was degassed and kept under nitrogen atmosphere. The size of the apparent aldehyde and carboxylic acid peaks are much reduced, suggesting molecular oxygen plays a role in the reaction. 100 Reflecting on these results following a helpful discussion with Dr. Lev Zakharov, it was detennined that due to tight crystal packing there is most likely not enough space between GaB clusters in the solid state to allow ligand substitution. It is possible that ligands on the surface of the crystals were substituted, but these represent only a small portion of the total clusters present and leave the rest with HzO ligands or decomposed due to heating. Powder X-ray diffraction ofthe powder isolated after the reactions showed no diffraction peaks, but it is suspected to be a dehydrated fonn of gallium, such as GaZ03, GaO(OH), or Ga(OH)3. Important conclusions from these studies include the notion that solid phase ligand exchange is unlikely to yield successful ligand exchange results. However, the potential catalytic behavior ofGal3 toward the oxidation ofp- xylene and other solvents with molecular oxygen might be an interesting follow-up study, especially if this activation can be perfonned selectively. Homogeneous Ligand Exchange Experiments utilizing homogenous ligand exchange (Gal3 and potential ligand dissolved in the same solution) were perfom1ed using a variety of different ligands. Gal3 is known to dissolve in MeOH, HzO, ethylene glycol, and propylene glycol (and perhaps to a small degree in diethylene glycol, though not in butylene glycol or any of the other highly polar organic solvents tested). This solvent variety provides a starting point for solution phase ligand substitution. Propylene and ethylene glycol are themselves potential ligands; if either of these molecules replace water ligands, it may drastically 101 alter the solubility of Ga 13 by changing the charge of the molecule and adding aliphatic content. Dissolution in either ethylene or propylene glycol was followed by recrystallization attempts through vapor diffusion of various solvents. With ethylene glycol solutions, isopropyl alcohol, acetone, tetrahydrofuran, and 1,4-dioxane were used as diffusion solvents; isopropyl alcohol was used as a diffusion solvent with propylene glycol solutions. None of these recrystallization attempts yielded crystalline product. Similar attempts were made with diethylene glycol, butylene glycol, propanedithiol, and ethylenediamine. No appreciable amount ofGa13 dissolved in any of the solvents; rigorous stin'ing only resulted in additional breakup of the Ga13 crystals. In addition to the ligand exchange reactions where the solvent was a potential ligand itself, other homogeneous GaB/ligand substitutions were screened. Ligands used for these potential transfonnations included 2,2' -bipyridine, 4-aminobenzoic acid, catechol, H3heidi, nitrilotriacetic acid, and thiosalicylic acid. Generally, twelve equivalents of the ligand were added for each equivalent of Ga13, though in the case of H3heidi, only six equivalents were added and in the case of 4-aminobenzoic acid 50 equivalents were added. Only six equivalents of H3heidi were deemed necessary based on the stoichiometries of Ga13heidi6 and Alnheidi6compounds previously reported in the literature. I0,17,l9 H3heidi was dissolved in MeOH with GaB and a small amount of pyridine to deprotonate the ligand, although such a base was not added during the screening of other potential ligands. lO In the case of H3heidi ligand, this procedure afforded an insoluble precipitate, which could not be redissolved. With catechol, the first attempt at ligand substitution was not perfonned under air-free conditions. This was later 102 thought to lead to decomposition of the ligand, so additional attempts utilized air-free techniques and were run under argon.9 Multiple recrystallization attempts with these different potential1igands did not yield a crystalline product. NMR characterization did not suggest coordination of new ligands either. Biphasic Ligand Exchange Beyond homogeneous and heterogeneous ligand exchange methods, a biphasic approach was examined. In biphasic ligand exchange, the Gal3 clusters and potential ligands were dissolved in various solvents, which were then vigorously mixed in order to facilitate transfer between the different solvent layers. Biphasic reactions are of interest because they eliminate the need for ligand and Gal3 co-solubility, which expands ligand choice. Consequently, a number of conditions were screened using a biphasic ligand exchange approach. N,N-dimethyl-4-nitrosoaniline was chosen due to its highly colored nature and known affinity for Ga(III). As a result, any ligand transfer to the Ga13layer should be readily evident. It was found that some color transfer occurred even in control reactions, and no crystalline product was isolated. Reactions run with terephthalic acid yielded similar results, with no crystalline product. 103 Conclusion Previous reports (such as those summarized in Chapter I) have shown that metal hydroxide nanoclusters are often isolated when stabilized by exterior ligands, so d .c I' d d 1 . 1 . 10 17-22 Add' . 11 b"prece ence lor 19an -supporte c usters certam y eXIsts. ' lhona y, su stltutlOn of H20 ligands has been shown in Keggin-AI 13 clusters.9 It seems reasonable that the correct conditions must simply be found to substitute and re-isolate ligand-substituted M 13 clusters such as Gal3Lx or Al 13 Lx (L = any potential ligand bound to the tridecameric clusters). Initial attempts to utilize a heterogeneous method proved unsuccessful, most likely due to lack of access to the proposed exchange sites because of tight crystal packing. Attempts at homogeneous and biphasic ligand exchange have also proven less than fruitful. Given this finding, there are several future research questions that apply to the area ofligand exchange. Are H20 ligands exchangeable under homogenous conditions? Based on the heterogeneous ligand exchange attempts, it appears that in order for all H20 ligands to be replaced (or at least some on all clusters) the clusters must be dissolved in order to be accessible to potential ligands. Ga13 is soluble in H20, MeOH, ethylene glycol, and propylene glycol. Each of these solvents is a potential ligand candidate itself, with the latter two being the most interesting due to their potential bidentate chelating nature. Bidentate ligands are preferable to monodentate to slow the rate ofligand exchange on the gallium ions. Therefore, future exploration in this area would likely focus on bidentate, polar molecules, such as catechol and salicylic acid. If a bidentate ligand can 104 be shown to bind to GaB, it could be exploited for macrocycle applications described in the ligand exchange introduction (Figure 5.4). Will biphasic conditions facilitate H20 ligand exchange? The potential ligands are currently limited to those soluble in solvents that dissolve GaB, so a biphasic exchange might prove useful. This way, any organic soluble ligand could be used. GaB could be dissolved in a suitable solvent (most likely H20 or MeOH) and layered with an organic solution containing the appropriate ligand followed by vigorous stilTing to mix the layers. In this case, fluorescent or colored ligands could be used, enabling visual confirmation ofligand substitution by tracking transfer of color from one solvent to the next. A color shift or fluorescence quenching may be observed upon binding to GaB, providing ease of characterization. Can Gan or other M 13 act as a metal templatefor large macrocycles? Ligand substitution might open the door for future metal-templated reactions. By exchanging H20 ligands for a predesigned ligand containing two reactive groups in close proximity to each other, then reacting the bi-functionalligand with another on an adjacent gallium ion, a large, continuous ligand shell encircling the cluster could be formed. By etching the cluster or protonating the ligand, a large macrocycle could subsequently be generated. One potential class ofligands to use in such experiments would be a bi-functional weak- strong binding ligand with two olefin moieties. Once bound to adjacent gallium or indium ions, olefin metathesis could be perfOlTIled to link the ligands and form a macrocycle. This macrocycle could itse1fpossess interesting properties (similar to cucurbituril or cyclodextrin), and could be used to template synthesis of other metal 105 clusters (Figure 5.4). A macrocycle made in this manner would be expected to be high- yielding and monodisperse, as the number of ligands being coupled to form the macrocycle would be predetermined by the number of ligands bound to the cluster. There are plenty of potential experiments to run with regard to metal exchange as well. As mentioned, computational experiments might be beneficial to determine what metal ions would favorably replace gallium atoms. A rigorous test of different pH conditions might prove fruitful with some of the different metal ions, though a balance must be found where the clusters remain sufficiently intact but hydroxide bridges can be formed to bridge additional metal ions. Bridge to Chapter VI We have demonstrated the successful synthesis ofa variety of tridecameric hydroxide clusters, from homometallic GaB and AlB to heterometallic GalIn and AllIn clusters. We have isolated ten compounds in total so far, and have attempted to isolate other compounds through metal and ligand exchange reactions. In the process of trying to isolate additional compounds such as In 13 and heterometallic compounds containing non- Group 13 metals, we have isolated several other examples of compounds containing Group 13 metal ions. Four of these compounds will be discussed in Chapter VI; two containing gallium, one with indium, and one with aluminum. 106 CHAPTER VI ISOLATION OF TWO ADDITIONAL CLASSES OF GROUP 13-CONTAINING MOLECULES Introduction During the course of the projects described in the preceding chapters, several more Group 13 containing structures were isolated, including Ga- and In-acetate 1-D chains and Anderson molybdate clusters with the central metal substituted by gallium or aluminum. The chain structures were produced by a similar technique to the tridecameric clusters featured in Chapters II-V, with ethyl acetate hydrolysis supplying the acetate groups. Two unusual complexes of formula [M(u-OH)(u-02CCH3hJn (M = In or Ga) were crystallized by diffusion of ethyl acetate into methanolic solutions of the M(N03)3 salts and N-nitrosopyrrolidine (Scheme 6.1). These represent the first examples of such simple l-D chains comprising Group 13 elements, and their fonnation results from the surprising hydrolysis of ethyl acetate to provide the bridging acetate ligands. These structures were synthesized and isolated by myself with single-crystal X-ray characterization by Dr. Lev Zakharov. The principal investigator for this research was Prof. Darren W. Johnson, who also provided editorial assistance. The results of these 107 studies were published in volume 48 of Inorganic Chemist'T, a publication of the American Chemical Society, in March 2009.1 The second set of results detailed will be the synthesis of gallium- and aluminum- centered Anderson-type molybdate clusters. In this case, the synthesis and isolation were again performed by myself, with single-crystal X-ray characterization perfonned by Dr. Lev Zakharov. The principal investigator was Prof. Darren W. Johnson, who provided editorial assistance. The results of this study were published in Acta Crystallographica Section E, a publication of the International Union of Crystallography.2 Interest in Infinite Group 13 Networks Interest in the area of infinite Group 13 carboxylate compounds originates from several areas, stimulated largely by work on metal-organic frameworks (MOFs) and zeolites.3-7 Numerous examples of 1-0,2-0, and 3-D indium (and fewer of gallium) carboxylate extended structures exist, often taking advantage of benzene di- and tricarboxylates as bridging ligands between metal centers.8-17 In these cases, ligand geometry seems to have a dominant effect on the final extended structure. 12 However, the steric bulk of the bridging ligands also plays a role, with t-butyl or other bulky ligands often forming rings instead of chains. 18.I9 One structure type was notably absent from this literature-a truly one-dimensional chain of gallium or indium with bridging acetate and hydroxide-a surprising omission given the interest in chain structures comprising Group 13 elements.2o-22 108 In addition to the multi-dimensional extended indium structures mentioned above, there are examples of non-linear or extended discrete structures of Group 13 ions, such as an elegant "gallic wheel'" produced by Christou and coworkers. I') The metal ions in these gallic wheel and indium carboxylate structures possess similar coordination environments to the 1-0 chain structures presented in this chapter, but form more complicated extended structures. 14 •15 Several additional repOlis have detailed a related manganese 1-0 chain structure, which is the only other example of this infinite linear chain topology with acetate and hydroxide I have encountered.23 -26 A recent report describes a structure similar to the indium chain, but with iodate bridging ligands in place of acetate. 27 M = Ga or In Scheme 6.1. General synthetic I'oute to I-D acetate chain compounds. Gallium or indium nitrate are combined in methanolic solution with N-nitrosopyrrolidine, which is allowed to evaporate, then ethyl acetate is added via vapor diffusion. After several months, single-crystalline product of In-chain and Ga-chaill are isolated. In th is chapter two new infinite chain structures of form ul a [In(;i -OH)(,u- chain, Figure 6.2) are described. To the best of my knowledge, the Tn-chain 11110 G~- chain structures are the first examples of such acetate-hydroxy bridged chain compounds 109 comprising Group 13 metals, and also showcase an interesting reaction wherein ethyl acetate is hydrolyzed, then incorporated as acetate in the resulting structures. Synthetic Procedure for I-D Chains These compounds were discovered during investigation of the tridecameric Group 13 hydroxo/aquo nanoclusters such as GaB, AlB, and Ga7In6, which I described in Chapters II, III, and IV.28-30 During my efforts to isolate an analogous Inl3 structure (which has remained elusive), I instead isolated the indium chain compound In-chain (Figure 6.1). The same methodology was successfully applied to isolate the gallium chain structure Ga-chain (Figure 6.2). M(N03)3 salts (M = Ga or In, 1 equiv.) were dissolved in methanol, then N-nitrosopyrrolidine (1.85 equiv.) was added, affording a clear yellow tinted solution. HPLC grade methanol was obtained from J.T. Baker and used without further purification (water content < 0.1 %). No attempt was made to keep the methanol anhydrous. Hydrated metal nitrate salts were also used, so all the water necessary for incorporation as hydroxide bridges was available from these salts, water in the solvent, and other adventitious water. The methanolic solutions were evaporated open to air at room temperature, giving a viscous yellow solution. The vials were then sealed in jars containing ethyl acetate (~20 mL) to allow for vapor diffusion into the yellow solution. After several months, crystals suitable for single-crystal X-ray diffraction formed, revealing the In-chain and Ga-chain structures, respectively. In- chain was also isolated with di-n-butylnitrosamine replacing N-nitrosopyrrolidine. 110 Ho\vever, using di-n-butylnitrosamine with Ga(N03)3 produces GaLl instead. 30 Interestingly, when the same conditions are applied to Fe(N03h, an iron trimer (well- known in the literature) is isolated instead ofa chain structure. 31 .32 Figmoe 6.1. Ortep visualization (50 % ptOobability) of In-acetate I-D chain compoHnd. The structural details present in the solid state can be seen in the above figure. Indium atoms are octahedral and bound to four acetate bridges and two hydroxide bridges. This structural morphology continues to form an infinite one-dimensional chain compound. Structural Description of Chain Molecules In-chain and Ga-chain consist of infinite linear chains of bridged metal ions (Figure 6.1). Each trivalent octahedral metal ion is bound to two ,LI-OH groups and four ,LI- I ,3-0Ac ligands. These are shared with neighboring metal ions, giving a formula of 111 [M(,u-OH)(02CCH3)2]1I, which results in an overall neutral charge. Interestingly, no interstitial molecules are present in In-chain and neighboring chain molecules show little intermolecular interaction. Many examples of Group 13-carboxylate structures contain interstitial molecules, so it is unusual that chain structure In-chain lacks such solvent molecules.x.lo.I'.'3-'5.27 1~ I :) I d Io I T o ~ I Figure 6.2. Ortep representation (50 % probability) of the crystal structure of the gallium 1-D chain compound. This crystal structure features the presence of interstitial molecules of water and acetic acid. These interstitial molecules hydrogen bond to each other and to the chain molecules. 112 Unlike the indium chain, the gallium chain contains acetic acid and water in the interstitial space between chains (Figure 6.2). As with the /1-1 ,3-0Ac bridging ligands, acetic acid solvent is also derived from the hydrolysis of ethyl acetate. Acetic acid and water form a network of interchain hydrogen bonds throughout Ga-chain (Figure 6.2). Acetic acid hydrogen bonds to the hydroxyl bridges and water molecules, while the water molecules hydrogen bond with the interstitial acetic acid and the oxygen atoms of the acetate bridges. The acetic acid and water molecules in Ga-chain are disordered over two positions related by a mirror plane. A similar network of hydrogen bonds, but without the disorder, is found in the structure of the previously reported manganese chain.25 The presence of water and acetic acid in Ga-chain, but not in In-chain can be rationalized in two ways. Gallium has a higher Lewis acidity than indium, and is therefore more likely to hydrolyze ethyl acetate, giving acetate or acetic acid. This may produce a higher acetate concentration in the crystallization solution. Gallium also has a smaller atomic radius than indium, which may leave more space in the crystal structure for incorporation of interstitial water and acetic acid. Hydrolysis of Ethyl Acetate As mentioned, the origin of the acetate bridges, derived from the hydrolysis of ethyl acetate, is also of interest in these structures. Ethyl acetate is hydrolyzed and incorporated into the chain structures in two different ways: as bridging acetate ligands 113 and as acetic acid in the interstitial spaces. Traditionally, extended structures bearing carboxylate ligands introduce the carboxylate moiety directly, e.g., as benzene carboxylates or sodium acetate. However, in a related example, Borovik and co-workers have described a case of mild ethyl acetate hydrolysis and acetonitrile hydration followed by structural incorporation of acetate and acetamidate ligands.33 In this structure, the acetate moiety occupies a 1,3-bridging position between two cobalt(II) ions while hydrogen bonding with urea groups, similar to the binding motif seen in the gallium chain structure where acetate oxygen atoms hydrogen bond to water molecules. Borovik was also able to isolate the same structure by direct treatment with an acetate source, whereas we have yet to isolate the chain structures by such means or in the absence of the nitroso compound. Conclusion to Group 13 Chain Compounds The application of indium and gallium in organic transformations has been widely studied, with a range of transformation types examined, but reviews of the topic do not discuss hydrolysis reactions. 34-37 Nevertheless, these observations suggest that indium and gallium hydroxide complexes may be worth exploring as mild reagents for hydrolysis activation. For instance, in this report up to three acetates are formed per metal ion based on the stoichiometry found in the crystals, and the crystallization vials also smell of acetic acid, providing further support of ester hydrolysis using Group 13 metals.37 114 Furthennore, it may be possible to use such an approach in the hydration of other molecules such as acetonitrile, as demonstrated by Borovik and co-workers. 33 Summary of Crystallographic Data Crystal data for C4H 7InO.dGa-chain}: M r = 249.92, colorless block, 0.34 x 0.18 x 0.09 mm, orthorhombic, space group Cmcm, T= 173 K, MoKx (0.71703 A), a = 14.570(5) A, b = 6.831(2) A, c = 7.269(2) A, V= 723.5(4) A3, Z= 4,/Jcaled = 2.294 g/cm3, f1 = 3.23 mm-1, F(OOO) = 480, 2Bmax = 56.5°, 1962 reflections collected, 481 unique [Rint = 0.011]. R indices [I> 20"(1)]: Rl = 0.0167, wR2 = 0.0475, GOF = 1.131. OJlStal data for CrJl13Ga08 (In-chain): M r = 282.88, colorless block, 0.28 x 0.08 x 0.05 mm, monoclinic, space group P2(1)/m, T= 173 K, MoKu (0.71703 A), a = 7.9516(7) A, b = 6.7651 (6) A, c = 10.6719(10) A, f3 = 106.1500(10)°, V = 551.42(9) A3, Z = 2, pealed = 1.704 g/cm3, f1 = 2.52 mm- I, F(OOO) = 288, 2Bmax = 54.0°, 6194 reflections collected, 1302 unique [Rint = 0.024]. R indices [I> 20"(1)]: Rl ::= 0.0304, wR2 = 0.0782, GOF = 1.106. Description of Anderson Clusters Anderson first described a planar polyanion in 1937.38 Anderson-type clusters (as they are now known) are well established and many papers detailing their preparation and applications have been published.39.40 Compounds containing Anderson-type clusters 115 have been explored for applications as stmctural aesthetics, biologically active compounds, and catalysts.39 They have also been shown to function as building blocks for larger molecular assemblies, where they can be linked to form extended networks with pores and cavities.41.42 The majority of Anderson clusters are based on M070 246- or W70246- frameworks, and many structures have been synthesized with substitution of the central octahedron or variable bridging ligands. A similar planar arrangement of seven metals is observed at the core ofthe recently reported stmcture of [Gal3(u3-0H)6(u2- OH)18(H20h4](N03)IS.28 Attempts to synthesize derivatives ofthis compound led to isolation of (NH4h[Ga(u3-0H)6M06018]-7H20 (GaMo6) and (NH4h[Al(u3- OH)6M06018]-nH20 (AIMo6) compounds that are described in the following pages.2 Extensive literature reports have covered the different stmctural variants and chemistry ofhexamolybdoaluminate(III) polyanions. 39 The gallium-substituted B-type Anderson compound has been synthesized previously, by adding a solution of gallium metal in concentrated HN03 to a solution ofMo03 in aqueous NaOH.43 ,44 The mixture was heated overnight and rinsed several times with acetone, and vacuum dried overnight, affording product as the sodium salt. However, a crystal stmcture determination of GaMo6 had not been reported.44 Synthesis by the metal exchange method repOlied in this chapter represents a far more benign method than that previously reported.44 The synthesis reported herein also represents an alternative preparation of the AI-substituted stmcture, with no acid addition required (as is usually the case). There are several literature repOlis of this stmcture, so the focus of this section is on GaMo6. 116 Synthetic Preparation of Substituted Anderson Clusters Due to the similarity of the structure of the M070 246- complex to the inner planar core of the reported Ga13 metal hydroxide cluster [Gal3(u3-0HM,u2- might be obtained by reaction of (NH4)6Mo7024 with six or more equivalents of Ga(N03)3.28 Setting up the reaction in a similar manner to the synthesis of GaB (dissolving starting materials in a MeOH:H20 mixture and adding di-n-butylnitrosamine), provided single crystals of the gallium-substituted B-type Anderson cluster, characterized by X-ray crystallography (Figure 6.3)? Figure 6.3. ChemDraw representation of the GaMo6 anion. This structure has a molecular formula of (NH4)3[Ga(u3-0H)6Mo6018]-7H20. The [M07024] starting material has a -6 charge which decreases to a -3 charge upon protonation of the six ,u3-oxide bridges, as well as the substitution of Ga(III) for Mo(VI). 117 Commercial products (NH4)6M07024 (Baker and Adamson) and Ga(N03)3'nH20 (Strem) were used to obtain GaMo6. (NH4)6M07024 (0.25 g, 0.2 mmol) and Ga(N03)3 (0.7 g, 1.92 mmol) were dissolved in a 1: 1 H20/MeOH mixture (10 mL) in a 20 mL scintillation vial. The mixture was heated slightly (while open to atmosphere in a fume hood), with some cloudiness remaining in the mixture. Di-n-butylnitrosamine (0.45 g, 0.5 mL, 2.8 mmol) was added, and was not initially miscible. Additional MeOH (~2 mL) brought most into solution, and the mixture was then filtered. The remaining solution was evaporated, and after 9 days clear, colorless crystals with block-like habit had formed around the outside edge of the vial. The crystals were isolated in 90 % crude yield. Fewer equivalents of Ga(N03h (relative to (NH4)6M07024) and no organic additive (H20 as solvent only) were also used successfully to produce crystals of GaMo6•2 AIMo6 is synthesized in a similar manner, simply substituting Al(N03k9H20 for Ga(N03h·nH20. Crystallographic Details The planar structure consisting of seven metals observed in the parent Anderson compound is also present in the structure of GaMo6, with average Ga-O bond lengths of 1.97 (1) A. The crystal structure of (NH4h[GaM06(OH)601S]-7H20, contains two centrosymmetric GaMo6 B-type Anderson cluster units consisting of central Ga06 octahedra surrounded by a hexagonal assembly ofMo06 edge-sharing octahedra (Figures 6.4 and 6.5). Like other B-type Anderson clusters where the central Mo atom is 118 substituted with a di- or trivalent metal ion, the central six JI:1-oxido bridges are protonated.:1 2a(!), R[F2 > 2a(F2)] = 0.040, wR(Y) = 0.110, S= 1.17. Conclusion to Ga- and AI-substituted Anderson Molybdates Two different Anderson molybdate structures were isolated with substitution of the central molybdenum atom by either gallium or aluminum. In both cases, these structures had been described previously, but in the case of the gallium-substituted compound, no crystal structure had been isolated to date. Upon substitution of the central molybdenum atom, the six j.13-0XO ligands are protonated, which changes the overall charge of the structure from -6 to -3. The synthetic preparation used to isolate these 121 compounds represents a much-simplified route as well compared to previous isolations.4~.44 Figure 6.6. PyMOL representation of the crystal structure of GaMo(j. This figure depicts the protonated nature of the ,LJ3-bridges in the structure. Depth can also be welJ- visualized here. The remaining bridges and capping ligands are all oxo. Gallium is blue, molybdenum is purple, oxygen is red, and hydrogen is white. Bridge to Chapter VII This chapter concludes the research results portion of this dissertation. Chapter VII presents concluding thoughts and a summary of the Chapters [-VI. It also contains a number of ideas for future work on this project. 122 CHAPTER VII CONCLUSIONS AND FUTURE DIRECTIONS Introduction As a graduate education draws to a close, it is only natural to look back on the accomplishments one has been able to achieve. Throughout the ups and (much more numerous) downs, it is necessary to focus both on your successes, to draw inspiration from them, but also on what can be learned from your failures. Because synthetic chemistry is a field of experimentation, success can be in frustratingly short supply. However, at the end of the day, it is both the successes and failures that contribute to your education, and learning from both is absolutely necessary. In fact, it is sometimes even more important to look at what dreams and goals have not been realized. Any good research question leads to further experiments. Not simply to keep ourselves employed, although that is certainly a benefit, but because good science should always lead us to additional questions. Each successful experiment should result in many more questions than answers. For we can never hope to come close to understanding all the intricacies of the natural world, we can simply keep asking ourselves, "Why?". In this spirit, this chapter concludes this dissertation by summarizing the accomplishments described in the first six chapters and proposing future areas of investigation for this project. 123 Research Summary The results described throughout this dissertation have had two things in common: 1) the molecules contain Group13 metals and 2) those metals are bridged by hydroxide ligands. The initial review chapter describes the examples of this structural combination in the literature. The molecules covered included both the purely inorganic and those with organic ligands bound to the clusters. Next, Chapter II describes a tridecameric aluminum hydroxide. This compound, [AI13(;13-0H)6(;1-0H)18(H20h4]15+ (AIl3), was originally isolated by another research group, but only in extremely low yield. 1 Our synthetic procedure afforded the first intentional synthesis of the compound while increasing the amounts that could be isolated.2 Applying this synthetic procedure to mixtures ofGa(N03)3 and In(N03)3 hydrates resulted in the isolation of the first ever heterometallic Group 13 hydroxide compounds, as described in Chapter II1.3 The [Ga7In6(;13-0HM,u-OH)18(H20)24] 15+ compound described in Chapter III was used successfully as a single-source precursor for solution processed amorphous metal oxide thin films. Building off the initial heterometallic result, Chapter IV describes a whole series of these compounds, encompassing six different GalIn structures and two different AllIn structures. This represents the largest collection of metal hydroxide compounds reported, as most generally detail only one or two clusters. Chapter V discusses the various attempts to post-synthetically modify these tridecameric clusters. The most successful method so far is a metal exchange procedure where crystals of [Ga13(;13- OH)6(;1-0H)18(H20)24] 15+ (Gal3) can be redissolved with MeOH or H20 with In(N03)3 to afford the heterometallic GalIn compounds. Ligand exchange was also examined, but a 124 variety of approaches have yet to yield concrete results. Finally, Chapter VI describes four additional Group13 hydroxide structures: two infinite one-dimensional Group13 hydroxy-aceate chains and two Group 13 substituted Anderson clusters.4 ,5 The chain compounds consist of gallium or indium metal centers bridged together by two acetate ligands and one hydroxide ligand.s The Anderson clusters were originally M070 246- anions, but the central molybdenum atom has been exchanged for gallium or aluminum atoms and in six 1'3-0 ligands protonated to fonn six 1'3-0H bridges.4 The results presented in this dissertation represent significant advances in the field, adding a host of new structures, while shedding light on their fonnation and synthesis. Future Work As stated in the introduction, good scientific research should always result in more questions. This research project is certainly no exception. With that in mind, the following sections will describe several areas of research that I think would be particularly interesting to investigate. Some have preliminary results that have been presented already in this dissertation, while others have not yet been discussed. Oxidation Chemistly One area of particular interest, one that could potentially be described as the proverbial "low-hanging fruit", would be investigation of the oxidation chemistry that 125 occurred when Gal3 was heated in various solvents, such as toluene, xylene, and hexadecane (Chapter V). In the presence of Gal3, about 15 % of the p-xylene solvent was oxidized to p-tolualdehyde, with an additional 5 % ofp-toluic acid generated. 55 % of the p-xylene remained, in addition to various other byproducts. These percentages were determined by GC/MS analysis. Ga(N03h also showed some oxidation activity. Determining the differences in reaction products (e.g. product yields, reaction selectivities, and degree of catalysis) could lead to a useful synthetic method for producing these oxidation products from readily available starting materials in a benign method, using molecular oxygen and the gallium compounds. Metal Exchange Metal exchange reactions have been examined as a method to access new heterometallic clusters, and this approach has been successful in the case of Anderson and Keggin clusters.4,6-8 To date this approach has not successfully generated novel structures with the tridecameric cluster framework. However, based on the results with indium substitution in Gal3, I believe that ifnew heterometallic compounds are isolable, the potential is high that this method represents a route to their isolation. While a variety of metals have been examined for metal exchange potential, less has been done in the area of pH control (Chapter V). Only iron was subjected to a variety of pH conditions when attempting to perfonll metal exchange with Fe(N03h and Gal3. pH variation might be necessary in the case of some metals to facilitate formation of the hydroxide bridges, 126 as we suspect to be the case in the synthesis of AI13 (Chapter II).2 In addition to pH variation, looking at metal salts with other counteranions might be useful. Because we generate the clusters through the nitrate salts of gallium, aluminum, and indium, nitrate salts of potential metal exchange candidates were examined. However, many other counteranions are available, and may be more successful. Anion Exchange Anion exchange is another potential way to introduce new metals into the M13 framework. There are many known anionic clusters, from the polyoxometallates to copper sulfide clusters. 8 The exchange ofN03' anions for these anions would introduce different metal ions into the clusters, and in the case of copper sulfide anions, would potentially generate precursors for materials such as copper-indium-gallium-sulfide thin films which are of interest for solar energy applications.9.10 To date, anion exchange columns have been used for anion exchange. Running solutions of Ga13 over cr anion exchange columns resulted in the isolation of powders. These were dissolved with various silver salts, attempting to drive the anion exchange through precipitation of Agel. Only amorphous powders were isolated from the evaporated solutions. More careful evaporation or vapor diffusion methods might be effective in generating single-crystalline product. Several students have already begun to explore this area more thoroughly, using both anion exchange resins and gel crystallization methods. 127 Nanocomposite Formation In the same vein of anion exchange, investigating the interactions of the polycationic M 13 clusters with negatively charged compounds such as polyoxometallates or Anderson clusters could be a potential area of research. Other research groups have been able to isolate hybrid structures ofKeggin-AI13 with Anderson clusters, as well as gels fonned from vanadium polyoxometallates and Keggin_AI13.11-14 Initial experiments in this area mixing Ga13 and M070 246- have led to the isolation of amorphous precipitates which could not be redissolved and recrystallized. Layering experiments to slowly mix the solutions of anion and cation also afforded amorphous materials. However, fonnation of these powders is highly suggestive of an interaction between the M 13 clusters and the anionic clusters. The two components are likely coming together too rapidly, not allowing organization into crystalline fonn. Some of the same methods discussed for anion exchange could be of use here, as the reaction is essentially an anion exchange reaction as well. By further slowing the rate that the solutions mix, it may be possible to fonn single crystals instead of powders. Gel crystallization might be a way to accomplish this, by trapping one component in the gel while the other diffuses into it. Ligand Exchange Of the areas in this future work section, ligand exchange has seen the most prior research, as described in Chapter V. A host of potential ligands and solvents have been tested under various conditions such as heterogeneous, homogeneous, and biphasic. In 128 all cases tested so far, no single crystalline materials were isolated. However, the ligand- supported structures detailed in Chapter I certainly give credence to the possibility of ligand exchange in these clusters. 15.16 In these syntheses, the ligands have been introduced during the cluster synthesis. It is not known however if the metal hydroxide cluster forms first, and the ligands bind afterward, or if the ligands coordinate first and initiate formation of the cluster element. Experiments utilizing pre-made clusters might help answer this question and shed light on cluster formation-a point of debate. 15,!7 As with metal exchange, pH control might be important to achieve ligand exchange. pH control was explored with exchange reactions of the H3heidi ligand, where PYIidine was added as in the literature preparation of Ga2heidi2, Gasheidi4, and Gal3heidi6.15 Use ofa wider range of ligands would also be an area to explore, particularly in the case of biphasic reactions. Additional EXAFS Studies Based on the results of full data analysis of previous EXAFS experiments, additional studies could prove useful. Particularly, I think EXAFS might be useful in attempting to answer the question of "regioisomers" addressed in Chapter IV. In the GalO1n3 compounds for instance, we do not know the relative positions of the indium atoms; whether they are all separated from each other by gallium atoms, two are neighboring and one is separated by a gallium atom, or all three are neighboring each 129 other. If the clusters are stable in solution, EXAFS studies might shed light on this by identifying whether the indium atoms have nearby neighbors that are also indium atoms. Interactions ofM I3 Clusters ·with Nucleic Acids Several research groups have previously examined the interactions of negatively charged polyoxometallate clusters (POMs) with DNA, RNA, and proteins. These clusters have been shown to selectively precipitate prions,18 hydrolytically cleave RNA model compounds,19 and inhibit replication of viruses such as HIV.2o Nanoscience is becoming increasingly interested in using the unique properties of nucleic acid polymers for detection, labeling, and templated synthesis, however molecular scale insight into the interactions between nucleic acids and inorganic clusters is relatively rare. There seems to be a particular lack of knowledge regarding positively charged compounds of this type interacting with DNA, RNA, and proteins. In order to study and understand the fundamental interactions between positively charged tridecameric Group 13 hydroxide compounds and DNA/RNA, studies to examine the interaction of clusters such as Gal3 with short DNAs and/or RNAs are proposed. Understanding the interactions between these two types of molecules has implications ranging from mineral catalysis of RNA synthesis to modern pursuits at the bio-nano interface. The similar size regimes present in the two molecule classes (~18 A for a M 13 cluster, ~3.5 Arise/base) should present interesting possibilities to examine different binding interactions. Beyond examining the basic chemical interactions between these two types of molecules, several additional factors could be examined, 130 including the effects of oligonucleotide length, oligonucleotide sequence, and different metals in the clusters on binding. Aside from binding interactions, it is known that POMs can cleave the phosphodiester bond of the RNA model compound 2-hydroxypropyl-4-nitrophenyl phosphate (HPNP).19 Based on previous work showing that Ga(N03h solutions hydrolyze EtOAc (Chapter VI), Ga(N03h itself or Gal3 may also hydrolyze the phosphodiester bond in HPNP, which if successful could be expanded to look at actual RNA phosphodiester bonds and whether hydrolysis can be achieved in a specific manner. Because of the relatively small sizes of both nanopmiicles and potential nucleic acid chelators a number of characterization methods can be used to understand M 13-nucleic acid interactions. 131 REFERENCES Chapter I (1) Baes, C. F.; Mesmer, R. E. The HydroZvsis ofMetal Cations; Wiley: New York, 1976. (2) Casey, W. H. "Large Aqueous Aluminum Hydroxide Molecules" Chemical Reviews 2006, 106, 1-16. (3) Michot, L. J.; Montarges-Pelletier, E.; Lartiges, B. S.; de la Caillerie, l B. D.; Briois, V. "Formation Mechanism of the Ga13 Keggin Ion: A Combined EXAFS and NMR Study" Journal o/the American Chemical Society 2000,122,6048-6056. (4) Bi, S. P.; Wang, C. Y.; Cao, Q.; Zhang, C. H. "Studies on the Mechanism of Hydrolysis and Polymerization of Aluminum Salts in Aqueous Solution: Correlations Between the "Core-links" Model and "Cage-like" Keggin-AI 13 Model" Coordination Chemistry Reviews 2004,248,441-455. (5) Jordan, P. A; Clayden, N. l; Heath, S. L.; Moore, G. R.; Powell, A K.; Tapparo, A. "Defining Speciation Profiles of A13+ Complexed with Small Organic Ligands: The AI3+-heidi System" Coordination Chemistry Reviews 1996, 149, 281-309. (6) de 1a Caillerie, l B. D.; Man, P. P.; Vicente, M. A.; Lambert, J. F. ,,27AI MQ- MAS NMR as a Tool for Structure Determination in Nanocomposite Materials: The Nature of Al Pillars in "Alu-heidi" Pillared Clays" Journal ofPhysical Chemistry B 2002,10~4133-4138. (7) Schmitt, W.; Jordan, P. A.; Henderson, R. K.; Moore, G. R.; Anson, C. E.; Powell. A. K. "Synthesis. Structures, and Properties of Hydrolytic A1(III) Aggregates and Fe(IlI) Analogues Formed with Iminodiacetate-based Chelating Ligands" Coordination Chemistry Reviews 2002,228,115-126. 132 (8) Goodwin, J. C.; Teat, S. J.; Heath, S. 1. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(I1I) Clusters" Angewandte Chemie International Edition 2004, 43, 4037-4041. (9) Heath, S. 1.; Jordan, P. A.; Johnson, I. D.; Moore, G. R.; Powell, A. K.; Helliwell, M. "Comparative X-ray and 27Al-NMR Spectroscopic Studies of the Speciation of Aluminum in Aqueous Systems - Al(I1I) Complexes ofN(CH2C02H)2(CH2CH20H)" Journal ofInorganic Biochemistry 1995, 59, 785-794. (10) Tam, S. c.; Williams, R. J. P. "One Problem of Acid Rain - Aluminum" Journal ofInorganic Biochemistry 1986, 26, 35-44. (11) Valle, G. c.; Bombi, G. G.; Corain, B.; Favarato, M.; Zatta, P. "Crystal and Molecular Structures of Diaqua(nitrilotriacetato)aluminum(I1I) and di-Il-hydroxo- bis(nitrilotriacetato) Dialuminate(III) Dianion" Journal ofthe Chemical Society-Dalton Transactions 1989,1513-1517. (12) Furrer, G.; Phillips, B. 1.; Ulrich, K. u.; Pothig, R.; Casey, W. H. "The Origin of Aluminum Floes in Polluted Streams" Science 2002, 297, 2245-2247. (13) Stewart, T. A.; Trudell, D. E.; Alam, T. M.; Ohlin, C. A.; Lawler, C.; Casey, W. H.; Jett, S.; Nyman, M. "Enhanced Water Purification: A Single Atom Makes a Difference" Environmental Science & Technology 2009,43,5416-5422. (14) Schweitzer, G. K.; Pesterfield, 1. 1. The Aqueous Chemistry ofthe Elements; 1st ed.; Oxford University Press: New York, 2010. (15) Johansson, G. "The Crystal Structure ofa Basic Aluminum Selenate" Arkivfoer Kemi 1963, 20, 305-319. (16) Parker, W.O., Jr.; Millini, R.; Kiricsi, I. "Metal Substitution in Keggin-Type Tridecameric Aluminum-Oxo-Hydroxy Clusters" Inorganic Chemistly 1997, 36,571. (17) Lee, A. P.; Phillips, B. 1.; Olmstead, M. M.; Casey, W. H. "Synthesis and Characterization ofthe Ge04AldOHh4(OH2)128+ Polyoxocation" Inorganic Chemistry 2001, 40, 4485. (18) Lee, A. P.; Furrer, G.; Casey, W. H. "On the Acid-Base Chemistry of the Keggin Polymers: GaAl l2 and GeAI 12" Journal ofColloid and Intel:!rlce Science 2002,250,269. (19) Panthani, M. G.; Akhavan, V.; Goodfellow, B.; Schmidtke, J. P.; Dunn, 1.; Dodabalapur, A.; Barbara, P. F.; Korgel, B. A. "Synthesis ofCuInS2, CuInSe2, and Cu(lnxGal-x)Se2 (CIGS) Nanocrystal 'Inks' for Printable Photovoltaics" Journal ofthe American Chemical Society 2008,130,16770-16777. 133 (20) Kim, H. S.; Byrne, P. D.; Facchetti, A; Marks, T. J. "High Performance Solution- Processed Indium Oxide Thin-Film Transistors" Journal ofthe American Chemical Society 2008,130, 12580-12581. (21) Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. (22) Johansson, G. "The Crystal Structures of [AbCOHh(H20)s](S04)2'2H20 and [AbCOH)2CH20)s] (Se04)2'2H20" Acta Chemica Scandinavica 1962, 16, 403-420. (23) Seichter, W.; Mogel, H.-J.; Brand, P.; Salah, D. "Crystal Structure and Formation of the Aluminum Hydroxide Clm'ide [AI13(OH)24(H20)24]CIls·13H20" European Journal ofInorganic Chemistly 1998, 795-797. (24) Casey, W. H.; Olmstead, M. M.; Phillips, B. L. "A New Aluminum Hydroxide octamer, [AIs(OH)14(H20)IS](S04)S·16H20" Inorganic Chemistry 2005,44,4888. (25) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric AI13 Nanocluster [AI!3C,1l3-0H)6(ur OH)l S(H20)24] (N03) IS" Inorganic Chemistry 2008,47,1267-1269. (26) Breuil, H. "Sur les Chlorures et Bromures Basiques Daluminium" Annales De Chimie France 1965, 10,467-493. (27) Rather, E.; Gatlin, 1. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, V.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization of the Inorganic Nanocluster [Ga13(u3-0H)6(urOH)ls(H20h4](N03)IS" Journal ofthe American Chemical Society 2005,127,3242-3243. (28) Wulfsberg, G. Inorganic Chemistry; University Science Books: Sausalito, 2000. (29) Gerasko, O. A; Mainicheva, E. A; Naumov, D. Y.; Kuratieva, N. V.; Sokolov, M. N.; Fedin, V. P. "Synthesis and Crystal Structure of Unprecedented Oxo/Hydroxo- Bridged Polynuclear Gallium(III) Aqua Complexes" Inorganic Chemistly 2005, 44, 4133. (30) Powell, A K.; Heath, S. L. "X-ray Structural Analysis of Biologically Relevant Aluminium(III) Complexes" Coordination Chemistry Revievvs 1996, 149, 59-80. 134 (31) Petrosyants, S. P.; Malyarik, M. A.; Ilyukhin, A. B. "Complexation of Aluminum and Gallium with Aminoacetic Acid in Aqueous Solutions - Crystal Structure of Diaqua- di-,u-hydroxo-bis(iminodiacetato)dialuminum(III) and Bis(iminodiacetato)gallate(III) of Potassium" Zhurnal Neorganicheskoi Khimii 1995, 40, 769-775. (32) Yokoyama, T.; Abe, H.; Kurisaki, T.; Wakita, H. ,, 13C and 27AI NMR Study on the Interaction Between Aluminium Ion and Iminodiacetic Acid in Acidic Aqueous Solutions" Analytical Sciences 1999, 15, 393-395. (33) Sobota, P.; Mustafa, M. 0.; Utko, J.; Lis, T. "Preparation and Crystal Structure of Di-,u-acetato-,u-hydroxo-hexa(ethyl acetate)dialuminium Tris(tetrachloroaluminate)" Journal ofthe Chemical Society-Dalton Transactions 1990, 1809-1812. (34) Hoveyda, H. R.; Rettig, S. J.; Orvig, C. "Coordination Chemistry of2-(2'- hydroxyphenyl)-2-benzoxazole with Gallium(III) and Aluminum(III): Two Uncommon Group 13 Metal Environments Stabilized by a Biologically Relevant Binding Group" Inorganic Chemistry 1993,32,4909-4913. (35) Mainicheva, E. A.; Gerasko, O. A.; Sheludyakova, L. A.; Naumov, D. Y.; Naumova, M. I.; Fedin, V. P. "Synthesis and Crystal Structures of Supramolecular Compounds of Polynuclear Aluminum(III) Aqua Hydroxo Complexes with Cucurbit[6]uril" Russian Chemical Bulletin, International Edition 2006,55,267-275. (36) Feng, T. L.; Gurian, P. L.; Healy, M. D.; Barron, A. R. "Aluminum Citrate: Isolation and Structural Characterization of a Stable Trinuclear Complex" Inorganic Chemistry 1990, 29, 408-411. (37) Malone, S. A.; Cooper, P.; Heath, S. L. "Synthesis and Structure ofa New Aluminium Citrate Trimer from Aqueous Solution at Very Low pH" Dalton Transactions 2003,4572-4573. (38) Vicente, M. A.; Lambert, J. F. "AI-pillared Saponites - Part 4. Pillaring with a New AI13 Oligomer Containing Organic Ligands" Physical Chemistl)! Chemical Physics 1999,1,1633-1639. (39) Wieghardt, K.; Kleineboymann, M.; Nuber, B.; Weiss, J. "Macrocyclic Complexes ofIndium(III) - Novel ,u-hydroxo-bridged and ,u-oxo-bridged Complexes - Crystal Structures of [L4In4(u-OH)6] (S20 6h-4H20 and [L2In2(CH3C02)4(u-0)'2NaCI04 (L = 1,4,7-triazacyclononane)" Inorganic Chemistry 1986,25,1654-1659. (40) Mensinger, Z. L.; Betterton, S. A.; Kamunde-Devonish, M. K.; Zakharov, L. N.; Gatlin, J. T.; Keszler, D. A.; Johnson, D. W. "Preparation ofa Series of Novel Heterometallic Tridecameric Group 13 Hydroxide Clusters" in preparation. 135 (41) Schmitt, W.; Baissa, E.; Mandel, A.; Anson, C. E.; Powell, A. K. "[AI 1S(u3- 0)4(u3-0H)6(u-OH)14(hpdta)3t - A New Abs Aggergate Which fonns a Supnnolecular Zeotype" Angewandte Chemie International Edition 2001, 40, 3578-3581. (42) Casey, W. H., Personal Communication. (43) Mensinger, Z. L.; Zakharov, L. N.; Johnson, O. W. "Synthesis and Crystallization of Infinite Indium and Gallium Acetate 10 Chain Structures and Concomitant Ethyl Acetate Hydrolysis" Inorganic Chemistly 2009, 48, 3505-3507. (44) Anokhina, E. V.; Vougo-Zanda, M.; Wang, X. Q.; Jacobson, A. J. "In(OH)BOC·0.75H2BOC (BDC = benzenedicarboxylate), a Hybrid Inorganic-organic Vernier Structure" Journal ofthe American Chemical Society 2005,127,15000-15001. (45) Gomez-Lor, B.; GutielTez-Puebla, E.; Iglesias, M.; Monge, M. A.; Ruiz-Valero, C.; Snejko, N. "In2(OH)3(BDCh,s (BDC=1 ,4-benzendicarboxylate); An In(III) supramolecular 3D framework with Catalytic Activity" Inorganic Chemistry 2002,41, 2429-2432. (46) Lin, Z.-Z.; Jiang, F. L.; Chen, L.; Yuan, D. Q.; Hong, M. C. "New 3-D chiral framework of indium with 1,3,5-benzenetricarboxylate" Inorganic Chemistry 2005, 44, 73-76. (47) Volkringer, c.; Loiseau, T. "A new Indium Metal-organic 3D Framework with 1,3,5-benzenetricarboxylate, MIL-96 (In), Containing ,u3-oxo-centered Trinuc1ear Units and a Hexagonal 18-ring Network" Materials Research Bulletin 2006, 41, 948-954. (48) Vougo-Zanda, M.; Wang, X. Q.; Jacobson, A. J. "Influence of Ligand Geometry on the Formation ofIn-O Chains in Metal-oxide Organic Frameworks (MOOFs)" Inorganic Chemistl~Y 2007, 46, 8819-8824. (49) Ako, A. M.; Waldmann, 0.; Mereacre, V.; Klower, F.; Hewitt, 1. .1.; Anson, C. E.; Gudel, H. U.; Powell, A. K. "Odd-Numbered Fe(III) Complexes: Synthesis, Molecular Structure, Reactivity, and Magnetic Properties" Inorganic Chemistly 2007,46, 756-766. (50) Stamatatos, T. C.; Foguet-Albiol, D.; Poole, K. M.; Wernsdorfer, W.; Abboud, K. A.; O'Brien, T. A.; Christou, G. "Spin Maximization from S = 11 to S = 16 in Mn? Oisk- Like Clusters: Spin Frustration Effects and Their Computational Rationalization" Inorganic Chemistry 2009, 48, 9831-9845. 136 Chapter II (1) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric Al l3 Nanocluster [All3(u3-0H)6(u2- OH)lS(H20)24](N03)lS" Inorganic Chemistry 2008,47, 1267-1269. (2) Casey, W. H. "Large Aqueous Aluminum Hydroxide Molecules" Chemical Reviews 2006, 106, 1-16. (3) Casey, W. H.; Phillips, B. L.; Furrer, G. "Aqueous Aluminum Polynuclear Complexes and Nanoclusters: A review" In Nanoparticles and the Environment 2001; Vol. 44,167-190. (4) Goodwin, J. C.; Teat, S. J.; Heath, S. L. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(III) Clusters" Angewandte Chemie International Edition 2004,43,4037-4041. (5) Heath, S. L.; Jordan, P. A.; Johnson, 1. D.; Moore, G. R.; Powell, A. K.; Helliwell, M. "Comparative X-ray and 27AI-NMR Spectroscopic Studies of the Speciation of Aluminum in Aqueous Systems - AI(III) Complexes ofN(CH2C02Hh(CH2CH20H)" Journal ofInorganic Biochemistry 1995, 59, 785-794. (6) Mainicheva, E. A.; Gerasko, O. A.; Sheludyakova, L. A.; Naumov, D. Y.; Naumova, M. 1.; Fedin, V. P. "Synthesis and Crystal Structures of Supramolecular Compounds of Polynuclear Aluminum(III) Aqua Hydroxo Complexes with Cucurbit[6]uril" Russian Chemical Bulletin, International Edition 2006,55,267-275. (7) Schmitt, W.; Baissa, E.; Mandel, A.; Anson, C. E.; Powell, A. K. "[AI1s(u3- O)4(u3-0H)6(u-OH)14(hpdta)3t - A New Ahs Aggergate Which forms a Suprmolecular Zeotype" Angewandte Chemie International Edition 2001, 40, 3578-3581. 137 (8) Schmitt, W.; Jordan, P. A; Henderson, R. K.; Moore, G. R.; Anson, C. E.; Powell, A K. "Synthesis, Structures, and Properties of Hydrolytic AI(III) Aggregates and Fe(III) Analogues Fonned with Iminodiacetate-based Chelating Ligands" Coordination Chemistry Reviews 2002, 228, 115-126. (9) Seichter, W.; Mogel, H.-J.; Brand, P.; Salah, D. "Crystal Structure and Fonnation of the Aluminum Hydroxide Cloride [AldOHh4(H20)24]Cl1s·13 H20" European Journal ofInorganic Chemistly 1998, 795-797. (10) Allouche, L.; Gerardin, C.; Loiseau, T.; Ferey, G.; Taulelle, F. "A13o : Giant Aluminum Polycation" Angewandte Chemie International Edition 2000, 39, 511-514. (11) Rowsell, J.; Nazar, L. F. "Speciation and Thennal Transfonnation in Alumina Sols: Structures of the Polyhydroxoaluminum Cluster [Aho08(OH)s6(H20b] 18+ and its 0- Keggin Moiete" Journal C!fthe American Chemical Society 2000, 122, 3777-3778. (12) Casey, W. H.; Olmstead, M. M.; Phillips, B. L. "A New Aluminum Hydroxide Octamer, [A1g(OH)14(H20)18](S04)s'16H20" Inorganic Chemistly 2005,44,4888. (13) Lee, A. P.; Phillips, B. L.; Olmstead, M. M.; Casey, W. H. "Synthesis and Characterization of the Ge04AI12(OHh4(OH2) 128+Polyoxocation" Inorganic Chemistry 2001, 40, 4485. (14) Casey, W. H.; Phillips, B. L. "Kinetics of Oxygen Exchange Between Sites in the Ga04AldOHh4(H20)127+(aq) Molecule and Aqueous Solution" Geochimica et Cosn1ochimica Acta 2001,65, 705. (15) Johansson, G. "The Crystal Structure of Some Basic Aluminum Salts" Acta Chemica Scandinavica 1960, 14, 771-773. (16) Rather, E.; Gatlin, J. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, V.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization ofthe Inorganic Nanocluster [Ga13(fJ3-0HMfJ2-0H)18(H20h4](N03)IS" Journal C!fthe American Chemical Society 2005, 127, 3242-3243. (17) Gerasko, O. A.; Mainicheva, E. A.; Naumov, D. Y.; Kuratieva, N. V.; Sokolov, M. N.; Fedin, V. P. "Synthesis and Crystal Structure of Unprecedented Oxo/Hydroxo- Bridged Polynuclear Gallium(III) Aqua Complexes" Inorganic Chemistl:V 2005, 44, 4133. 138 (18) Wulfsberg, G. Inorganic Chemistry; University Science Books: Sausalito, 2000. (19) Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. (20) Anderson, J. S. "Constitution of the Poly-acids" Nature (London) 1937, 140, 850. (21) Evans, H. T., Jr "The Crystal Structures of Ammoium and Potassium Molydotellurates" Journal ofthe American Chemical Society 1948, 70, 1291. (22) Van der Sluis, P.; Spek, A. L. "BYPASS: an Effective Method for the Refinement of Crystal Structures Containing Disordered Solvent Regions" Acta Crystallographica 1990, A46, 194-201. Chapter III (1) Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. (2) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric AI 13 Nanocluster [Alu(P3-0H)6(u2- OH)18(H20)24](N03)lS" Inorganic Chemistry 2008, 47, 1267-1269. (3) Rather, E.; Gatlin, J. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, Y.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization of the Inorganic Nanocluster [Ga13(u·3-0H)6(u2-0H)18(HzOh4](N03)lS" Journal ofthe American Chemical Society 2005,127,3242-3243. (4) Casey, W. H. "Large Aqueous Aluminum Hydroxide Molecules" Chemical Reviews 2006, 106, 1-16. 139 (5) Casey, W. H.; Olmstead, M. M.; Phillips, B. L. "A New Aluminum Hydroxide Octamer, [Al8(OH)14(H20)18](S04)S·16H20" Inorganic Chemistry 2005,44,4888. (6) Goodwin, J. C.; Teat, S. J.; Heath, S. L. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(III) Clusters" Angewandte Chemie International Edition 2004, 43,4037-4041. (7) Seichter, W.; Mogel, H.-J.; Brand, P.; Salah, D. "Crystal Structure and Formation of the Aluminum Hydroxide Cloride [Al 13(OHh4(H20h4]Cl 1s '13 H20" European Journal ofInorganic Chemistly 1998, 795-797. (8) Schmitt, W.; Baissa, E.; Mandel, A.; Anson, C. E.; Powell, A. K. "[AltS(u3- 0)4(,u3-0H)6Cu-OH)14(hpdtaht - A New AIls Aggergate Which f0l111s a Suprmolecular Zeotype" Angewandte Chemie International Edition 2001, 40, 3578-3581. (9) Bradley, S. M.; Kydd, R. A.; Yamdagni, R. "Detection of a new Polymeric species formed through the Hydrolysis of Gallium(III) Salt Solutions" Journal ofthe Chemical Society-Dalton Transactions 1990,413-417. (10) Gerasko, O. A.; Mainicheva, E. A.; Naumov, D. Y.; Kuratieva, N. V.; Sokolov, M. N.; Fedin, V. P. "Synthesis and Crystal Structure of Unprecedented Oxo/Hydroxo- Bridged Polynuclear Gallium(III) Aqua Complexes" Inorganic Chemistry 2005, 44, 4133. (11) Kudynska, J.; Buckmaster, H. A.; Kawano, K.; Bradley, S. M.; Kydd, R. A. "A 9 GHz cw-electron-paramagnetic Resonance Study of the Sulphate Salts of Tridecameric [MnxAI l3.x0 4(OH)24(H20) !2J(7.x)" Journal o.fChemical Physics 1993, 99, 3329. (12) Lee, A. P.; Phillips, B. L.; Olmstead, M. M.; Casey, W. H. "Synthesis and Characterization of the Ge04Alt2(OH)24(OH2) 128+ Polyoxocation" Inorganic Chemistry 2001, 40,4485. (13) Oszk6, A.; Kiss, J.; Kiricsi, I. "XPS investigations on the feasibility of isomorphous substitution of octahedral A13+ for Fe3+ in Keggin ion salts" Physical Chemistry Chemica! Physics 1999, 1, 2565. (14) Parker, W.O., Jr.; Millini, R.; Kiricsi, I. "Metal Substitution in Keggin-Type Tridecameric Aluminum-Oxo-Hydroxy Clusters" Inorganic Chemistly 1997, 36, 571. 140 (15) Chiang, H. Q.; Hong, D.; Hung, C. M.; Presley, R. E.; Wager, J. F.; Park, C.-H.; Keszler, D. A.; Hennan, G. S. "Thin-film Transistors with Amorphous Indium-gallium Oxide Channel Layers" Journal ofVacuum Science and Technology 2006,24,2702. (16) Wager, J. F.; Keszler, D. A; Presley, R. E. Transparent Electronics; 1st ed.; Springer, 2008. (17) Fujdala, K. L.; Tilley, T. D. "An Efficient, Single-source Molecular Precursor to Silicoaluminophosphates" Journal ofthe American Chemical Society 2001, 123, 10133- 10134. (18) Kim, H. S.; Byrne, P. D.; Facchetti, A.; Marks, T. J. "High Perfonnance Solution- Processed Indium Oxide Thin-Film Transistors" Journal ofthe American Chemical Society 2008,130,12580-12581. (19) Murugavel, R.; Walawalkar, M. G.; Dan, M.; Roesky, H. W.; Rao, C. N. R. "Transfonnations of Molecules and Secondary Building Units to Materials: A Bottom-up Approach" Accounts o.fChemical Research 2004, 37, 763-774. (20) Sekar, P.; Greyson, E. c.; Barton, J. E.; Odom, T. W. "Synthesis ofNanoscale NbSez Materials from Molecular Precursors" Journal (~fthe American Chemical Society 2005, 127,2054-2055. (21) Nomura, K.; Ohta, H.; Takagi, A.; Kamiya, To; Hirano, M.; Hosono, H. "Room- temperature Fablication of Transparent Flexible Thin-film Transistors Using Amorphous Oxide Semiconductors" Nature 2004,432,488-492. (22) Presley, Ro E.; Honga, Do; Chianga, H. Q.; Hunga, C. M.; Hoffmanb, R. L.; Wager, J. F. "Transparent Ring Oscillator Based on Indium-gallium Oxide Thin-film Transistors" Soild State Electronics 2006, 50, 500. (23) Anderson, J. T.; Munsee, C. L.; Hung, C. M.; Phung, T. M.; Hennan, G. S.; Johnson, D. c.; Wager, J. F.; Keszler, D. A "Solution-Processed HafSOx and ZircSOx Inorganic Thin-Film Dielectrics and Nanolaminates" Advanced Functional Materials 2007, 77,2117-2124. (24) Meyers, S. T.; Anderson, J. T.; Hong, D.; Hung, C. M.; Wager, J. F.; Keszler, D. A "Solution-Processed Aluminum Oxide Phosphate Thin-Film Dielectrics" Chemistry o.f Materials 2007, 19,4023-4029. 141 (25) Ako, A M.; Waldmann, 0.; Mereacre, V.; K1ower, F.; Hewitt, I. J.; Anson, C. E.; Gude1, H. D.; Powell, A K. "Odd-Numbered Fe(III) Complexes: Synthesis, Molecular Structure, Reactivity, and Magnetic Properties" Inorganic Chemistry 2007,46, 756-766. (26) Van der Sluis, P.; Spek, A. L. "BYPASS: an Effective Method for the Refinement of Crystal Structures Containing Disordered Solvent Regions" Acta Crystallographica 1990, A46, 194-201. (27) Chen, Z.; Zhou, Y.; Weng, L.; Zhang, H.; Zhao, D. "Hydrothermal Synthesis of Two Layered Indium Oxalates with 12-membered Apertures" Journal ofSolid State Chemistry 2003,173,435. Chapter IV (1) Baes, C. F.; Mesmer, R. E. The Hydrolysis OfMetal Cations; Wiley: New York, 1976. (2) Seichter, W.; Mogel, H.-J.; Brand, P.; Salah, D. "Crystal Structure and Fomlation of the Aluminum Hydroxide Cloride [AI 13(OH)z4(HzO)z4] C1J s' 13HzO" European Journal ofInorganic Chemistry 1998, 795-797. (3) Schmitt, W.; Jordan, P. A; Henderson, R. K.; Moore, G. R.; Anson, C. E.; Powell, A. K. " Synthesis, Structures, and Properties of Hydrolytic Al(lll) Aggregates and Fe(lll) Analogues Formed with Iminodiacetate-based Che1ating Ligands" Coordination Chemistry Reviews 2002, 228, 115-126. (4) Goodwin, J. C.; Teat, S. J.; Heath, S. L. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(III) Clusters" Angewandte Chemie International Edition 2004, 43, 4037-4041. (5) Casey, W. H.; Olmstead, M. M.; Phillips, B. L. "A New Aluminum Hydroxide Octamer, [Als(OH)14(HzO)18](S04)S'16HzO" Inorganic Chemistry 2005,44,4888. 142 (6) Gerasko, O. A.; Mainicheva, E. A.; Naumov, D. Y.; Kuratieva, N. V.; Sokolov, M. N.; Fedin, V. P. "Synthesis and Crystal Structure of Unprecedented Oxo/Hydroxo- Bridged Polynuclear Gallium(III) Aqua Complexes" Inorganic Chemistry 2005, 44, 4133. (7) Casey, W. H. "Large Aqueous Aluminum Hydroxide Molecules" Chemical Reviev.'s 2006, 106, 1-16. (8) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric Al 13 Nanocluster [AI13(u3-0H)6(u2- OH)IS(H20h4](N03)lS" Inorganic Chemistry 2008,47, 1267-1269. (9) Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008,47,9484-9486. (10) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Synthesis and Crystallization ofInfinite Indium and Gallium Acetate 1D Chain Structures and Concomitant Ethyl Acetate Hydrolysis" Inorganic Chemistly 2009, 48, 3505-3507. (11) Powell, A. K.; Heath, S. L. "X-ray Structural Analysis of Biologically Relevant Aluminium(III) Complexes" Coordination Chemist'~y Reviews 1996, 149, 59-80. (12) Schmitt, W.; Baissa, E.; Mandel, A.; Anson, C. E.; Powell, A. K. "[AI lS(u3- 0)4(u3-0H)6(u-OH)14(hpdtaht - A New AIls Aggergate Which fonns a Supnnolecular Zeotype" Angewandte Chemie International Edition 2001, 40, 3578-3581. (13) Furrer, G.; Phillips, B. L.; Ulrich, K. D.; Pothig, R.; Casey, W. H. "The Origin of Aluminum Floes in Polluted Streams" Science 2002, 297, 2245-2247. (14) Swaddle, T. W.; Rosenqvist, J.; Yu, P.; Bylaska, E.; Philiips, B. L.; Casey, W. H. "Kinetic Evidence for Five-coordination in AI(OHh+(aq) ion" Science 2005, 308, 1450- 1453. (15) Rather, E.; Gatlin, J. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, V.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization of the Inorganic Nanocluster [Ga13(u3-0H)6(urOH)ls(H20h4](N03) IS" Journal ofthe American Chemical Society 2005, 127, 3242-3243. (16) Parker, W.O., Jr.; Millini. R.; Kiricsi, 1. "Metal Substitution in Keggin-Type Tridecameric Aluminum-Oxo-Hydroxy Clusters" Inorganic Chemistry 1997, 36, 571. (17) Lee, A. P.; Furrer, G.; Casey, W. H. "On the Acid-Base Chemistry of the Keggin Polymers: GaAI 12 and GeA1 12" Journal ofColloid and Inte11ace Science 2002,250,269. 143 (18) Johansson, G. "The Crystal Structures of [Ab(OHh(H20)g](S04h·2H20 and [Ab(OH)2(H20)g](Se04)2'2H20" Acta Chemica Scandinavica 1962, 16,403-420. (19) Malone, S. A; Cooper, P.; Heath, S. L. "Synthesis and Structure of a New Aluminium Citrate Trimer from Aqueous Solution at Very Low pH" Dalton Transactions 2003, 4572-4573. (20) Jordan, P. A; Clayden, N. J.; Heath, S. L.; Moore, G. R.; Powell, A K.; Tapparo, A. "Defining Speciation Profiles of A13+ Complexed with Small Organic Ligands: The Ae+-heidi System" Coordination Chemistry Reviews 1996, 149,281-309. (21) Heath, S. L.; Jordan, P. A; Johnson, 1. D.; Moore, G. R.; Powell, A. K.; Helliwell, M. "Comparative X-ray and 27AI_NMR Spectroscopic Studies of the Speciation of Aluminum in Aqueous Systems - AI(III) Complexes ofN(CH2C02Hh(CH2CH20H)" Journal ofInorganic Biochemistry 1995,59, 785-794. (22) de la Caillerie, J. B. D.; Man, P. P.; Vicente, M. A; Lambeli, J. F. ,,27AI MQ- MAS NMR as a Tool for Structure Determination in Nanocomposite Materials: The Nature of Al Pillars in "Alwheidi" Pillared Clays" Journal ofPhysical Chemistry B 2002, 10~4133-4138. (23) Van der Sluis, P.; Spek, A L. "BYPASS: an Effective Method for the Refinement of Crystal Structures Containing Disordered Solvent Regions" Acta C,ystallographica 1990, A46, 194-201. (24) Cooper, G. J. T.; Newton, G. N.; Long, D.-L.; Kogerler, P.; Rosnes, M. H.; Keller, M.; Cronin, L. "Exploring a Series oflsostructural Dodecanuc1ear Mixed Ni:Co Clusters: Toward the Control of Elemental Composition Using pH and Stoichiometry" Inorganic Chemistry 2009, 48, 1097-1104. (25) Chen, Z.; Zhou, Y.; Weng, L.; Zhang, H.; Zhao, D. "Hydrothermal Synthesis of Two Layered Indium Oxalates with 12-membered Apertures" Journal o.fSolid State Chemistry 2003,173,435. (26) Wulfsberg, G. Inorganic ChemistlY; University Science Books: Sausalito, 2000. (27) Edwards, D. D.; Folkins, P. E.; Mason, T. O. "Phase Equilibria in the Ga203- In203 System" Journal o.fthe American Ceramic Society 1997, 80, 253-257. (2R) Shannon. R. n.: Prewltt. C T. "Svnthesls :lnn Stmctllre ofPh:lses in Tn"O,-G:l"O, , / -:J - - ;> - - ~ - - -;> _. -. - J -- ~-- - - -- ---- -- - -- -- - - --- - - - - ----- - - -- ---L -.J - -~L - .J System" Journal o.fInorganic and Nuclear Chemistry 1968, 30, 1389-&. (29) Casey, W. H., Personal Communication. 144 Chapter V (1) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric Al 13 Nanoc1uster [AI13(u3-0H)6(ur OH)18(H20)24](N03)ls" Inorganic Chemistry 2008,47,1267-1269. (2) Mensinger, Z. L.; Betterton, S. A.; Kamunde-Devonish, M. K.; Zakharov, L. N.; Gatlin, J. T.; Keszler, D. A.; Johnson, D. W. "Preparation of a Series of Novel Heterometallic Tridecameric Group 13 Hydroxide Clusters" in preparation. (3) Mensinger, Z. L.; Gatlin, J. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. (4) Rather, E.; Gatlin, 1. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, V.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization of the Inorganic Nanoc1uster [Ga13(u3-0H)6(u2-0H)18(H20)24](N03)lS" Journal o/the American Chemical Society 2005, 127,3242-3243. (5) Casey, W. H. "Large Aqueous Aluminum Hydroxide Molecules" Chemical Reviel+'S 2006, 106, 1-16. (6) Casey, W. H.; Phillips, B. L.; Furrer, G. "Aqueous Aluminum Polynuclear Complexes and Nanoclusters: A Review" In Nanoparticles and the Environment 2001; Vol. 44,167-190. (7) Furrer, G.; Phillips, B. L.; Ulrich, K. u.; Pothig, R.; Casey, W. H. "The Origin of Aluminum Floes in Polluted Streams" Science 2002, 297, 2245-2247. (8) Stewart, T. A.; Trudell, D. E.; Alam, T. M.; Ohlin, C. A.; Lawler, c.; Casey, W. H.; Jett, S.; Nyman, M. "Enhanced Water Purification: A Single Atom Makes a Difference" Environmental Science & Technology 2009,43,5416-5422. 145 (9) Forde, S.; Hynes, M. J. "Kinetics and Mechanism of the Reactions of the All) Keggin Oligomer, [AI04AldOHb(H20)12f+, with a Series of Phenolic Ligands" New Journal ofChemistry 2002, 26, 1029-1034. (10) Goodwin, J. c.; Teat, S. J.; Heath, S. L. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(III) Clusters" Angewandte Chemie International Edition 2004,43,4037-4041. (11) Landry, C. c.; Pappe, N.; Mason, M. R.; Apblett, A. W.; Tyler, A. N.; MacInnes, A. N.; Barron, A. R. "From Minerals to Materials: Synthesis of Alumoxanes from the Reaction of Boehmite with Carboxylic Acids" Journal ofMaterial" Chemistly 1995, 5, 331-341. (12) Parker, W.O., Jr.; Millini, R.; Kiricsi, I. "Metal Substitution in Keggin-Type Tridecameric Aluminum-Oxo-Hydroxy Clusters" Inorganic Chemistry 1997, 36,571. (13) Lorenzo-Luis, P. A.; Gili, P. "Polyoxometalates with an Anderson-Evans Structure" Recent Research Developments in Inorganic Chemistry 2000, 2, 185-196. (14) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Triammonium Hexahydroxidooctadecaoxidohexamolybdogallate(III) Heptahydrate" Acta Crystallographica Section E 2008, 64, i8-i9. (15) Lee, A. P.; Phillips, B. L.; Olmstead, M. M.; Casey, W. H. "Synthesis and Characterization ofthe Ge04AI12(OH)z4(OH2)128+ Polyoxocation" Inorganic Chemistry 2001, 40, 4485. (16) de la Caillerie, J. B. D.; Man, P. P.; Vicente, M. A.; Lambert, J. F. ,,27AI MQ- MAS NMR as a Tool for Structure Determination in Nanocomposite Materials: The Nature of Al Pillars in "Aln-heidi" Pillared Clays" Journal ofPhysical Chemistly B 2002,106,4133-4138. (17) Heath, S. L.; Jordan, P. A.; Johnson, I. D.; Moore, G. R.; Powell, A. K.; Helliwell, M. "Comparative X-ray and 27AI-NMR Spectroscopic Studies of the Speciation of Aluminum in Aoueous SYstems - AUnn l:omnlexes ofNrl:H,l:O,H),rl:H,l:H,OH)" ----.-------.---------.1-----.- -.1------- ---, - / -----.1------ --- ,---..... - -4.--/ .... ,---4.---4.---/ Journal ofInorganic BiochemistlY 1995, 59, 785-794. 146 (18) Schmitt, W.; Baissa, E.; Mandel, A.; Anson, C. E.; Powell, A K. "[AI 15(,u3- O)4(,u3-0H)6(,u-OH)14(hpdtaht - A New Al 15 Aggergate Which forms a Suprmolecular Zeotype" Angewandte Chemie International Edition 2001,40,3578-3581. (19) Vicente, M. A; Lambert, J. F. "AI-pillared Saponites - Part 4. Pillaring with a New Al 13 Oligomer Containing Organic Ligands" Physical Chemistry Chemical Physics 1999,1,1633-1639. (20) Feng, T. L.; Gurian, P. L.; Healy, M. D.; Barron, A R. "Aluminum Citrate: Isolation and Structural Characterization of a Stable Trinuclear Complex" Inorganic ChemistJy 1990, 29, 408-411. (21) Petrosyants, S. P.; Malyarik, M. A.; Ilyukhin, A B. "Complexation of Aluminum and Gallium with Aminoacetic Acid in Aqueous Solutions - Crystal Structure of Diaqua- di-I-l-hydroxo-bis(iminodiacetato)dialuminum(III) and Bis(iminodiacetato)gallate(III) of Potassium" Zhurnal Neorganicheskoi Khimii 1995, 40, 769-775. (22) Powell, A. K.; Heath, S. L. "X-ray Structural Analysis of Biologically Relevant Aluminium(III) Complexes" Coordination Chemistry Reviews 1996, 149, 59-80. Chapter VI (1) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Synthesis and Crystallization ofInfinite Indium and Gallium Acetate 1D Chain Structures and Concomitant Ethyl Acetate Hydrolysis" Inorganic ChemistJy 2009,48,3505-3507. (2) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Triammonium Hexahydroxidooctadecaoxidohexamolybdogallate(III) Heptahydrate" Acta Crystallographica Section E 2008, 64, i8-i9. 147 (3) Eddaoudi, M.; Moler, D. B.; Li, H. L.; Chen, B. L.; Reineke, T. M.; O'Keeffe, M.; Yaghi, O. M. "Modular Chemistry: Secondary Building Units as a Basis for the Design of Highly Porous and Robust Metal-organic Carboxylate Frameworks" Accounts of Chemical Research 2001, 34, 319-330. (4) Ferey, G. "Hybrid Porous Solids: Past, Present, Future" Chemical Society Reviews 2008,37, 191-214. (5) Kitagawa, S.; Kitaura, R.; Noro, S. "Functional Porous Coordination Polymers" Angewandte Chemie International Edition 2004,43,2334-2375. (6) Moulton, B.; Zaworotko, M. J. "From Molecules to Crystal Engineering: Supramolecular Isomerism and Polymorphism in Network Solids" Chemical Reviews 2001,101,1629-1658. (7) Yaghi, O. M.; O'Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J. "Reticular Synthesis and the Design of New Materials" Nature 2003,423, 705-714. (8) Gomez-Lor, B.; Gutierrez-Puebla, E.; Iglesias, M.; Monge, M. A.; Ruiz-Valero, c.; Snejko, N. "Novel 2D and 3D Indium Metal-organic Frameworks: Topology and Catalytic Properties" Chemistry ofMaterials 2005, 17, 2568-2573. (9) Hong-Ling, G.; Bin, D.; Yi, L.; Peng, c.; Dai-Zheng, L.; Shi-Ping, Y; Zong-Hui, J. "Syntheses and Structures of 1D and 2D Coordination Polymers Derived from Pyridine-2,4,6-tricarboxylic Acid" Inorganic Chemistly Communications 2005, 8, 151- 154. (10) Lin, Z.-Z.; Chen, L.; Jiang, F. L.; Hong, M. C. "The Indium-carboxylate Chain Structure with the Rectangular Tunnels" Inorganic Chemistry Communications 2005,8, 199-201. (11) Lin, Z.-Z.; Jiang, F. L.; Chen, L.; Yuan, D. Q.; Hong, M. C. "New 3-D Chiral Framework of Indium with 1,3,5-benzenetricarboxylate" Inorganic Chemistly 2005, 44, 73-76. (12) Vougo-Zanda, M.; Wang, X. Q.; Jacobson, A. J. "Influence of Ligand Geometry on the Fonnation ofIn-O Chains in Metal-oxide Organic Frameworks (MOOFs)" Inorganic Chemistry 2007,46,8819-8824. 148 (13) Liu, Y. L.; Eubank, J. F.; Cairns, A. J.; Eckert, J.; Kravtsov, V. c.; Luebke, R.; Eddaoudi, M. "Assembly of Metal-organic Frameworks (MOFs) Based on Indium-trimer Building Blocks: A Porous MOF with soc Topology and High Hydrogen Storage" Angewandte Chemie International Edition 2007, 46, 3278-3283. (14) Volkringer, C.; Loiseau, T. "A New Indium Metal-organic 3D Framework with 1,3,5-benzenetricarboxylate, MIL-96 (In), Containing }i3-oxo-centered Trinuc1ear Units and a Hexagonal 18-ring Network" Materials Research Bulletin 2006, 41, 948-954. (15) Anokhina, E. V.; Vougo-Zanda, M.; Wang, X. Q.; Jacobson, A. J. "In(OH)BDC·0.75H2BDC (BDC = benzenedicarboxylate), a Hybrid Inorganic-organic Vernier Structure" Journal ofthe American Chemical Society 2005, 127, 15000-15001. (16) Gomez-Lor, B.; Gutierrez-Puebla, E.; Iglesias, M.; Monge, M. A.; Ruiz-Valero, C.; Snejko, N. "In2(OH)3(BDC)1.5 (BDC=1,4-benzendicarboxylate): An In(III) Supramolecular 3D Framework with Catalytic Activity" Inorganic Chemistry 2002,41, 2429-2432. (17) Volkringer, c.; Loiseau, T.; Ferey, G.; Morais, C. M.; Taulelle, F.; Montouillout, V.; Massiot, D. "Synthesis, Crystal Structure, and 71 Ga Solid-state NMR of a MOF-type Gallium Trimesate (MIL-96) with }i3-oxo-bridged Trinuc1ear Units and a Hexagonal 18- ring Network" Microporous and Mesoporous Materials 2007,105, 111-117. (18) Christian, P.; Rajaraman, G.; Harrison, A.; McDouall, J. J. W.; Raftery, J. T.; Winpenny, R. E. P. "Structural, Magnetic, and DFT Studies of a Hydroxide-bridged Crs Wheel" Dalton Transactions 2004, 1511-1512. (19) King, P.; Stamatatos, T. C.; Abboud, K. A.; Christou, G. "Reversible Size Modification ofIron and Gallium Molecular Wheels: A GalO "Gallic Wheel" and Large GalS and FelS Wheels" Angewandte Chemie International Edition 2006, 45, 7379-7383. (20) Feng, M. L.; Li, X. L.; Mao, J. G. "New Organically Templated Gallium and Indium Selenites or Selenates with One-, Two-, and Three-Dimensional Structures" Crystal Growth and Design 2007, 7, 770-777. (21) Su, J.; Wang, Y.; Yang, S.; Li, G.; Liao, F.; Lin, J. "New Series ofIndium Formates: Hydrothermal Synthesis, Structure, and Coordination Modes" Inorganic Chemistry 2007, 46, 8403-8409. 149 (22) Zhou, J.; Bian, G. Q.; Zhang, Y.; Zhu, Q. Y.; Li, C. Y.; Dai, J. "One-Dimensional Indium Sulfides with Transition Metal Complexes of Polyamines" Inorganic Chemistry 2007,46,6347-6352. (23) Mishra, A.; Tasiopoulos, A J.; Wemsdorfer, W.; Moushi, E. E.; Moulton, B.; Zaworotko, M. J.; Abboud, K. A; Christou, G. "Single-Molecule Magnets: A Family of Mn(III)/Ce(IV) Complexes with a [MngCeOg]J2+ Core" Inorganic Chemistry 2008,47, 4832-4843. (24) Price, D. J.; Batten, S. R.; Moubaraki, B.; MUlTay, K. S. "Synthesis, Structure, and Magnetism of {[Mn(u-OH)(,u-OAc2]-HOAc-H20h and the Facilitation of Long- range Magnetic Order Through Hydrogen Bonding" Polyhedron 2003, 22, 2161-2167. (25) Tasiopoulos, A. J.; Harden, N. C.; Abboud, K. A; Christou, G. "Preparation and Crystal Structures of Mn(H), Mixed-valent Mn(II)/Mn(III), and Mn(III) Polymeric Compounds" Polyhedron 2003, 22, 133-143. (26) Tasiopoulos, A. J.; Wemsdorfer, W.; Abboud, K. A.; Christou, G. "Two Isomeric [Mn120dOMeh(02CPh)16(H20)2f Single-molecule Magnets and a Mn(III) Polymer Prepared by a Reductive Aggregation Synthetic Route" Polyhedron 2005, 24, 2505-2512. (27) Liu, X. M.; Li, G. H.; Hu, Y. W.; Yang, M.; Kong, X. G.; Shi, Z.; Feng, S. H. "Hydrothemlal Synthesis and Crystal Structure of Polar and Nonpolar Compounds in Indium Iodate Family" Crystal Grmvth and Design 2008,8,2453-2457. (28) Rather, E.; Gatlin, J. T.; Nixon, P. G.; Tsukamoto, T.; Kravtsov, V.; Johnson, D. W. "A Simple Organic Reaction Mediates the Crystallization of the Inorganic Nanocluster [Ga13(,uJ-OH)6(,urOH)18(H20)24](NOJ)lS" Journal ofthe American Chemical Society 2005,127,3242-3243. (29) Gatlin, J. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric AI 13 Nanocluster [AII3(JlJ-OH)6(Jl2- OH)lg(H20)24](NOJ)lS" Inorganic Chemistry 2008,47, 1267-1269. (30) Mensinger, Z. L:, Gatlin, .f. T.; Meyers, S. T.; ZakhaTOv, L. N.; Keszler, D. A; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoc1usters and Inks for OX1de Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. 150 (31) Reisner, E.; Telser, J.; Lippard, S. J. "A Planar Carboxylate-Rich Tetrairon(II) Complex and Its Conversion to Linear Triiron(II) and Paddlewheel Diiron(II) Complexes" Inorganic Chemistly 2007,46,10754-10770. (32) Weinland, R.; Gussmann, E. "Uber Ferriacetate und tiber die Essigsiiurereaktion mit Eisenchlorid, sowie tiber die 'Basische Fiillung' (About Fenic Acetate and Acetic Acid Reaction with Fenic Chloride, and about the 'Basic Precipitation')" Zeitschriftfur Anorganische und Allgemeine Chemie 1910, 66, 157. (33) Zinn, P. J.; Sorrell, T. N.; Powell, D. R.; Day, V. W.; Borovik, A. S. "Acetonitrile Hydration and Ethyl acetate Hydrolysis by Pyrazolate-bridged Cobalt(II) Dimers Containing Hydrogen-bond Donors" Inorganic Chemistry 2007, 46, 10120-10132. (34) Amemiya, R.; Yamaguchi, M. "GaCh in Organic Synthesis" European Journal of Organic Chemistly 2005,5145-5150. (35) Fringuelli, F.; Piermatti, 0.; Pizzo, F.; Vaccaro, L. "Indium Salt-promoted Organic Reactions" Current Organic Chemistry 2003,7,1661-1689. (36) Frost, C. G.; Hartley, J. P. "New Applications ofIndium Catalysts in Organic Synthesis"Mini-Reviews in Organic Chemistly 2004,1,1-7. (37) Ranu, B. c.; Dutta, P.; Sarkar, A. "An Efficient and General Method for Ester Hydrolysis on the Surface of Silica Gel Catalyzed by Indium Triiodide Dnder Microwave Irradiation" Synthetic Communications 2000,30,4167-4171. (38) Anderson, J. S. "Constitution ofthe Poly-acids" Nature (London) 1937, 140, 850. (39) Lorenzo-Luis, P. A.; Gili, P. "Polyoxometalates with an Anderson-Evans Structure" Recent Research Developments in Inorganic Chemistry 2000, 2, 185-196. (40) Lee, D.; Joo, H. c.; Kwon, J. S.; Cho, M. A. "Tripotassium Hexahydrogen Hexamolybdocobaltate(III) Heptahydrate, K3[H6CoM06024]·7H20" Acta Crystallographica Section E-Structure Reports Online 2001,57,1112-1114. (41) Son, .I.-H.; Choi, H.; Kwon, Y.-D. "Porous Crystal Formation from Polyoxometalate Building Blocks: Single-Crystal Structure of [A104AldOH)dH20h4][Al(OH)6M0601sh(OH)·29.5H20" Journal ofthe American Chemical Society 2000, 122, 7432-7433. 151 (42) Himeno, S.; Murata, S.; Eda, K. "A Route to a Keggin-Type a- [(XlII04)Mo1203s(OH)t Anion Through an Anderson-type [XIII(OH)6M06018t anion: X = Ga" Dalton Transactions 2009,6114-6119. (43) Kitazumi, I.; Nakashima, Y.; Himeno, S. "Capillary Electrophoretic Determination of Ga(I1I) Based on the Formation of a Heteropolyoxomolybdate Complex" Journal ofChromatography A 2003,993,211-215. (44) Rollins, O. W.; Earley, J. E. "Hexamo1ybdogallate(III): The Free Acid Solution, the Ammonium and the Aquopentamminecobalt(III) Salt" Journal o(the American Chemical Society 1959, 8i, 5571-5572. Chapter VII (1) Seichter, W.; Mogel, IL-J.; Brand, P.; Salah, D. "Crystal Structure and Formation of the Aluminum Hydroxide Cloride [Al/3(OH)24(H20)24]Cl 1s·13H20" European Journal ofinorganic Chemistry 1998, 795-797. (2) Gatlin, 1. T.; Mensinger, Z. L.; Zakharov, L. N.; MacInnes, D.; Johnson, D. W. "Facile Synthesis of the Tridecameric Al/3 Nanocluster [Al/3(,u3-0HMJl2- OH)I8(H20)24](N03)ls" Inorganic Chemistry 2008,47, 1267-1269. (3) Mensinger, Z. L.; Gatlin, 1. T.; Meyers, S. T.; Zakharov, L. N.; Keszler, D. A.; Johnson, D. W. "Synthesis of Heterometallic Group 13 Nanoclusters and Inks for Oxide Thin-Film Transistors" Angewandte Chemie International Edition 2008, 47, 9484-9486. (4) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Triammonium Hexahydroxidooctadecaoxidohexamolybdogallate(III) Heptahydrate" Acta Crystallographica Section E 2008, 64, i8-i9. (5) Mensinger, Z. L.; Zakharov, L. N.; Johnson, D. W. "Synthesis and Crystallization ofInfinite Indium and Gallium Acetate 1D Chain Structures and Concomitant Ethyl Acetate Hydrolysis" Inorganic Chemistly 2009,48,3505-3507. 152 (6) Lee, A. P.; Furrer, G.; Casey, W. H. "On the Acid-Base Chemistry of the Keggin Polymers: GaAl12 and GeA1 12" Journal o.lColloid and Intel/ace Science 2002, 250, 269. (7) Lee, A. P.; Phillips, B. L.; Olmstead, M. M.; Casey, W. H. "Synthesis and Characterization of the Ge04AI dOHh4(OH2)128+ Polyoxocation" Inorganic Chemistry 2001,40,4485. (8) Parker, W.O., Jr.; Millini, R.; Kiricsi, 1. "Metal Substitution in Keggin-Type Tridecameric Aluminum-Oxo-Hydroxy Clusters" Inorganic ChemistlY 1997, 36, 571. (9) Nose, K.; Soma, Y.; Ornata, T.; Otsuka-Yao-Matsuo, S. "Synthesis of Ternary CulnS2 Nanocrystals; Phase Determination by Complex Ligand Species" Chemistly o.f Materials 2009, 21, 2607-2613. (10) Panthani, M. G.; Akhavan, V.; Goodfellow, B.; Schmidtke, J. P.; Dunn, L.; Dodabalapur, A.; Barbara, P. F.; Korgel, B. A. "Synthesis ofCulnS2, CulnSe2, and Cu(InxGal-JSe2 (CIGS) Nanocrystal 'Inks' for Printable Photovoltaics" Journal o/the American Chemical Society 2008,130,16770-16777. (11) Choi, H.; Kwon, Y.-u.; Han, O. H. "Nanocomposite Gels between [V 100 28t and [AI04AldOHh4(H20)12]7+ Polyoxometalate Clusters" Chemistly o.lMaterials 1999, 11, 1641-1643. (12) Son, J.-H.; Choi, H.; Kwon, Y.-u. "Porous Crystal Formation from Polyoxometalate Building Blocks: Single-Crystal Structure of [AI04AI dOH) dH20h4][AI(OH)6Mo6018h(OH)·29.5H20" Journal o.fthe American Chemical Society 2000, 122, 7432-7433. (13) Son, J. H.; Kwon, Y.-V.; Han, O. H. "New Ionic Crystals of Oppositely Charged Cluster Ions and Their Characterization" Inorganic Chemistry 2003,42,4153-4159. (14) Ibrahim, M.; Dickman, M. H.; Suchopar, A.; Kortz, U. "Large Cation-Anion Materials Based on Trinuclear Ruthenium(III) Salts of Keggin and Wells-Dawson Anions Having Water-Filled Channels" Inorganic Chemistry 2009,48, 1649-1654. (15) Goodwin, J. C.; Teat, S. J.; Heath, S. L. "How Do Clusters Grow? The Synthesis and Structure of Polynuclear Hydroxide Gallium(III) Clusters" Angewandte Chemie International Editon 2004, 43, 4037-4041. (16) Vicente, M. A.; Lambert, J. F. "AI-pillared Saponites - Part 4. Pillaring with a New AI l3 Oligomer Containing Organic Ligands" Physical Chemistry Chemical Physics 1999,1,1633-1639. 153 (17) Casey, W. H.; Olmstead, M. M.; Phillips, B. L. "A New Aluminum Hydroxide Octamer, [AIR(OH)14(H20)lR](S04)s·16H20" Inorganic Chemistry 2005,44,4888. (18) Lee, 1. S.; Long, J. R.; Prusiner, S. B.; Safar, J. G. "Selective Precipitation of Prions by Polyoxometalate Complexes" Journal ofthe American Chemical Society 2005, 127,13802-13803. (19) Absillis, G.; Cartuyvels, E.; Van Deun, R.; Parac-Vogt, T. N. "Hydrolytic Cleavage of an RNA-Model Phosphodiester Catalyzed by a Highly Negatively Charged Polyoxomolybdate [Mo70 24t Cluster" Journal ofthe American Chemical Society 2008, 130, 17400-17408. (20) Judd, D. A.; Nettles, J. H.; Nevins, N.; Snyder, J. P.; Liotta, D. C.; Tang, J.; Ermolieff, J.; Schinazi, R. F.; Hill, C. L. "Polyoxometallate HIV-1 Protease Inhibitors. A New Mode ofProtease Inhibition" Journal afthe American Chemical Society 2001,123, 886-897.