Low Oxidation State Group 15 Chemistry:

New Synthetic Methods Towards Polymers

By

Erin L. Norton

A Thesis Submitted to the Faculty of Graduate Studies and Research through the Department of Chemistry and Biochemistry in Partial Fulfillment of the Requirements for the Degree of Master of Science at the University of Windsor

Windsor, Ontario, Canada

2008 Library and Bibliotheque et 1*1 Archives Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition

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

Your file Votre reference ISBN: 978-0-494-42260-1 Our file Notre reference ISBN: 978-0-494-42260-1

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

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

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

While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. •*• Canada © 2008 Erin L. Norton Abstract

During the past two decades significant advances have been made in the area of low valent phosphorus chemistry. Much of the more recent work has focused on developing controlled synthetic methods, as well as understanding reactivity of compounds bearing phosphorus atoms in unusually low oxidations states.

One of the most studied classes of stable compounds containing P(+l) centers are the triphosphenium salts that consist of a dicoordinate phosphorus cation ligated by two phosphines or a diphosphine ligand. New and improved syntheses of such salts containing halide counter anions were developed and are presented in this thesis.

Polyphosphazenes are an important class of phosphorus containing polymers.

Although methods exist to replace the phosphinyl fragments within the polymer backbone, methods to replace the nitrogen atoms were unknown. Methods for the substitution of these atoms with P(I) centers were developed and are anticipated to produce new groups of polymers with unique properties.

IV Co-Authorship Statement

Much of the material contained within this document has been previously published in peer-reviewed journals. In accordance with regulations defined by the Faculty of Graduate Studies, this dissertation is presented in manuscript format. I was the principal investigator for all publications and I had a significant role in the preparation of the manuscripts. I acknowledge my supervisor Dr. Charles L. B. Macdonald as a co­ author in this work as he also made significant contribution to the writing of manuscripts. Other listed authors on manuscripts contributed through raw data acquisition. Also, Jonathon Dube assisted in collection of raw data for both Chapters 3 and 4. The dissertation is based on the following publications:

Chapter 1 The writing of this chapter will go towards an invited Perspectives Article in the periodical Dalton Transactions of the Royal Society of Chemistry. It is expected that this work will be submitted shortly after the dissertation is defended.

Chapter 2 Norton, E.L., Szekely, K.S., Dube, J.W., Bomben, P.G., Macdonald, C.L.B., Inorg. Chem., 2008, 47, 1196-1203. DOI: 10.1021/ic701342u.

Chapters 3 & 4 This work has not been published.

v Dedicated to B. C. Norton

vi Acknowledgements

First and foremost I must thank my supervisor Dr. Chuck Macdonald for the opportunity he has given me, and all of the encouragement and support he provided me during my master's. No question was too stupid for Chuck and he was always encouraging when things weren't going well.

I would like to thanks all the Macdonald Group members especially Ben Cooper for his friendship and collaborative lab efforts; Bobby Ellis for showing me "the ropes" when I first started; the undergrads who have contributed to this work: Paolo Bomben,

Kara Szekely, and especially Jon Dube for his hard work and enthusiasm. Also thanks to the other "Macdonald's"-Erica Morasset, Chris Andrews, Rajoshree Roychowdhury,

Greg Farrar, Mike Stinchcombe and Tamara Milovic.

Special thanks to Drs. Sam Johnson, Jim Green, and Holger Eichorn for their help

in anything from Glove Box Maintenance to translating German journal articles. Also

thanks to the Johnson group members, Jill, Meghan, and Ritu for fun in the office.

I thank my family for all their support especially my mother who has been

through a lot in the past several years, but has taught me to persevere. A special thanks to

Aaron Rossini for encouragement and support.

Lastly, there is someone who will never read this thesis, but has probably

influenced it the most. My dad taught me to be respectful, to have integrity and to take

pride in everything I do. I know he wouldn't have understood an ounce of the chemistry

within, however he would have enthusiastically read this thesis from cover to cover.

vii Statement of Originality

I certify that this thesis, and the research to which it refers, are the product of my own work and that any ideas from the work of other people, published or otherwise, are fully acknowledged in accordance with the standard referencing practices of the discipline. I acknowledge the helpful guidance and support of my supervisor Dr. C.L.B. Macdonald.

viii Table of Contents Abstract iv Co-Authorship Statement v Dedication vi Acknowledgements vii Statement of Originality viii List of Tables x List of Figures and Schemes xi List of Abbreviations, Symbols, Nomenclature xiv Chapter 1 - Introduction into Low Oxidation State Main Group Chemistry 1 1.1 General Introduction 1 1.2 Oxidation States, Valence States and Our Model 2 1.3 Examples of Low Oxidation State Main Group Elements 5 1.3.1 Phosphine Ligands 6 1.3.2 iV-heterocyclic Carbene Ligands 17 1.3.3 Iniine-containing Ligands and Related Ligands 22 1.4 Dissertation Overview 27 1.5 References 29 Chapter 2 - A Convenient Preparative Method for Cyclic Triphosphenium Bromide and Chloride Salts 33 2.1 Introduction 33 2.2 Experimental 36 2.3 Results and Discussion 45 2.4 Conclusions 59 2.5 References 60 Chapter 3 - Small Molecular Targets as Monomers for Polymerization 63 3.1 Introduction 63 3.2 Experimental 68 3.3 Results and Discussion 75 3.4 Conclusions 87 3.5 References 88 Chapter 4 - Development of an Improved Synthesis of N-Heterocyclic Phosphenium Salts 90 4.1 Introduction 90 4.2 Experimental 92 4.3 Results and Discussion 96 4.4 Conclusions 102 4.5 References 104 Chapter 5 - Dissertation Summary and Future Considerations 106 5.1 Dissertation Summary 106 5.2 Future Considerations 107 5.3 References Ill Vita Auctoris 112

ix List of Tables

Table 2.1. P NMR shifts corresponding to the acyclic and cyclic intermediates as well as the product for the room temperature reaction of PCI3 + bis(diethylphosphino)propane (depp) 35 Table 2.2. Summary of X-ray crystallographic data for compounds 2.7[Br], 2.8[Br] and

2.8[HC12] 43 Table 2.3. Summary of X-ray crystallographic data for compounds [dppe(OH)2][Br] and [dppe(0)(OH)][Br] 44

Table 2.4. Selected Metrical Parameters for compounds 2.7[Br], 2.8[Br] and 2.8[HC12]. Distances are reported in angstroms and angles in degrees; for 2.7[Br], the corresponding values for each independent cation in the asymmetric unit are reported separately 52 Table 2.5. Selected Metrical Parameters for compounds [dppe(OH)2][Br]2 and [dppe(0)(OH)][Br3]. Distances are reported in angstroms and angles in degrees 57 Table 3.1. Summary of X-ray crystallographic data for compounds 3.7 74 Table 3.2. Selected Metrical Parameters for compound 3.7. Distances are reported in angstroms and angles in degrees; the corresponding values for each associated bond length are reported separately 80 Table 4.1. Summary of reaction attempts to form [Mes-DAB-P][Br] 94 Table 4.2. 31P{!H}-NMR experiment results for the reactions in Table 1. Most intense signal in each spectrum denoted by * 95

x List of Figures and Schemes

Figure 1.1. Some examples of our model of correlating Lewis Stucture to Oxidation State based on the number of "lone pairs" of electrons associated with the central atom. Phosphorus is used as an example 4 Figure 1.2. Ambiguity in Lewis-type chemical structure drawings of a single cation having the formula [P(PR3)2] 4 Scheme 1.1. First isolation of a cyclic triphosphenium salt 7 Scheme 1.2. Isolation of an acyclic triphosphenium salt 7 Scheme 1.3. First Synthesis of an diphosphine stabilized Arsenium salt 7 Scheme 1.4. An example of a reaction using a triphosphenium anion 8 Figure 1.3. Small neutral heterocycles containing triphosphenium moieties 8 Figure 1.4. Some examples of triphosphenium cations. The substituent labeled R is commonly phenyl, ethyl or cyclohexyl 9 Scheme 1.5. Synthesis of a triphosphenium iodide salt without the use of an external reducing agent 9 Scheme 1.6. Protonation and alkylation at the P(I) atom in various cyclic triphosphenium salts can be performed using the strong reagents triflic acid or methyl triflate respectively 12 Scheme 1.7. Complexation of a P(I) cation with a transition metal, in this case the common organometallic catalyst Schwartz's reagent Cp2ZrHCl 13 Figure 1.5. Lewis Structure depictions of triphosphenium cations and the analogous carbodiphosphiranes 14 Scheme 1.8. Examples of some reactions of carbodiphosphiranes with small inorganic molecules 15 Scheme 1.9. Reactions of carbodiphosphiranes with transition metals 15 Scheme 1.10. Classical syntheses of phosphamethinecyanine dyes 17 Scheme 1.11. Example of a triphosphenium inserting into an olefin bond between two carbene-like fragments 18 Scheme 1.12. Syntheses of phosphamethinecyanines directly from NHC and either PCI3 oraP(I)salt 19 Scheme 1.13. Treatment of elemental phosphorus P4 with CAAC C or an NHC results in the opening of the P4 tetrahedron 20 Scheme 1.14. Stabilization of Ge(II) by an NHC 21 Scheme 1.15. Generation of a Ge(II) dication through a non-RedOx ligation of two NHC substituents 21 Scheme 1.16. Reduction and coordination of a selenium atom by an NHC 22 Figure 1.6. Lewis structure depictions of an N-heterocyclic carbene and the phosphenium analogue 23 Scheme 1.17. Chlorophosphine produced through an electron transfer process with DAB andPCl3 24 Scheme 1.18. Synthesis of NHP's using methodology developed for the synthesis of P(I) salts 24 Figure 1.7. Two Lewis-type depictions of [NHSb]+ 25 Figure 1.8. An As(I) atom is stabilized by the tricoordinate DEV1PY ligand 26

xi Scheme 1.19. The difference in reactivity of NHSn and NHSi when treated with DAB. 27 Figure 1.9. Summary of Lewis depictions of Main Group Carbene Analogues as described by NBO calculation results 27 Figure 1.10. Structural models used throughout this dissertation for phosphorus (I)- containing compounds 28 Figure 2.1. Examples of relevant small main group molecules 33 Scheme 2.1. Dillon and Monks' mechanism of triphosphenium cation formation where X = CI or Br; the rate-determining step is the redox step involving the formal elimination of m X2 from the P -containing dication intermediate (Adapted from reference 23) 35 Figure 2.2. Structural depictions of the triphosphenium cations 2.7 and 2.8 36 Scheme 2.2. Some early reactions of Schmidpeter et al. that are consistent with the intermediate formation of triphosphenium chloride salts 45 Scheme 2.3. The differing behaviors observed for the reactions of phosphorus(III) halides with chelating bis(diphenylphosphino)alkane ligands. Only one of the potential halophosphonium products is included 46 Figure 2.3. 31P NMR spectra of the reactions of: (a) dppe and excess Br2; and PBr3 with dppp in the presence of (b) no cyclohexene, (c) one equivalent of cyclohexene and (d) 3 equivalents of cyclohexene. The signal corresponding to the dication 2.9 is indicated with an arrow. Relevant chemical shift assignments: ca. 60 ppm (s, [dppe-Br2]+2), 65 + + ppm (d, [(dppe)P] ), -230 ppm (t, [(dppe)P] ), ca. 34 ppm (s, dppe-02) 49 Figure 2.4. Thermal ellipsoid plots (30% probability surface) of the contents of the asymmetric units of the P1 bromide salts 2.7[Br] (a) and 2.8[Br] (b). Hydrogen atoms have been removed for clarity 50 Figure 2.5. Thermal ellipsoid plot (30% probability surface) of the contents of the asymmetric unit of [(dppp)P][HCl2], 2.8[HCi2]. Most of the hydrogen atoms have been removed for clarity 52 Figure 2.6. The molecular structure of [dppe(OH)2][Br]2 56 Figure 2.7. The molecular structure of [dppe(0)(OH)][Br3]. Only half of each Br3 anion depicted in the figure is present in each asymmetric unit 57 Scheme 2.4. Summary of the reactions of phosphorus(III) halides with chelating bis(diphenylphosphino)alkane ligands in the presence of excess cyclohexene 59 Scheme 3.1. General synthesis of some chlorinated phosphazenes. The major product is the six-membered ring 3.1 63 Scheme 3.2. General synthesis of a perchlorinated polyphosphazene from perchlorophosphazene 3.1. Both homo-substitued and mixed substituted polyphosphazenes are shown 64 Scheme 3.3. Phosphorus containing polymers made by the addition polymerization of phosphaalkenes which is analogous to one of the methods used for olefin polymerization. 65 Scheme 3.4. Proposed synthetic route to poly[bis(phosphino)benzeneP(I)] oligomers or polymers through ligand exchange with the P(I) salt 3.4 66 Figure 3.1. Solid-state structure of the dimer [(p-(Et02P)2C6H4)P]2[SnCl6] 67 Figure 3.2. Small molecular targets as possible monomers towards polymeric material. 75

xii Scheme 3.5. Frank's synthesis of a 4-membered all-phosphorus ring (3.16) with alternating 2-coordinate and 4-coordinate phosphorus atoms 77 Scheme 3.6. Synthesis of the 8-membered ring analogue (3.7) of the target molecule 3.14 by Schmidpeter 78 Figure 3.3. Solid State Structure of 8-membered ring 3.7. Hydrogen atoms have been removed for clarity 79 Scheme 3.7. The two syntheses developed to generate the eight-membered analogue of one of the four-membered ring target molecules 81 Figure 3.4. The all phosphorus target molecule 3.15 and the dimeric analogue 3.17.... 82 Scheme 3.8. Proposed synthetic route to 4-membered phosphorus ring 3.15 82 Scheme 3.9. Suggestion reported by Schmidpeter et. al, that the dimeric analogue of 3.15 is inaccessible by route of treating P4 with a phosphide 85 Figure 3.5. 31P{1H} NMR spectra of several reactions attempted to product a P-P-P type ligand. For spectra a) and b) the strong peak ~60ppm is indicative of the diphosphine CyP-PCy which is surprisingly absent from spectrum c) 86 Scheme 3.10. Alternative syntheses to achieve 4-membered phosphorus ring target 3.15 87 Figure 4.1. Typical examples of Fischer Carbene 4.1, Schrock Carbene 4.2, and N- Heterocyclic Carbene 4.3 90 Scheme 4.1. Reactions of NHP's with transition metals and other chemistry 92 Scheme 4.2. Previously reported methods for the synthesis of NHP's 96 Scheme 4.3. The relationship between the ionic phosphenium salts and the covalently bound phosphorus halide depending on the choice of solvent. (X = CI or Br) 97 Figure 4.2. Thermal ellipsoid plot (30% probability surface) of [Dipp-DAB-P][SnCl5] and [Dipp-DAB-P][l3]. Hydrogen atoms are omitted for clarity 98 Scheme 4.4. Synthesis of previously discussed triphosphenium bromide and chloride salts with the addition of cyclohexe to react with the X2 by-product 99 Figure 4.3. NMR spectra of three representative reactions into the synthesis of NHP's. Of note is the major signal for each spectrum is in the range of 180-200ppm, which agrees with literature reported values of similar NHP's 100 Scheme 5.1. Ligand exchange reactions with a) a chelating diphosphine (X=anionic heteroatom); b) a non-chelating diphosphine with an anionic linker 108 Scheme 5.2. Proposed synthesis for the formation of the 4-membered ring with N-aryl substitution 109 Figure 5.1. Recently reported compound by Gudat with a polarized P-P bond. R^Mes, tBu, R2=Ph, tetraethyl phospholide 109 Scheme 5.3. Insertion chemistry of a polarized P-P bond into common reagents 110 Scheme 5.4. General approach for cycloaddition reactions involving different main group elements 110

xiii List of Abbreviations, Symbols, Nomenclature

A Non-descript anion o A Angstrom Abs. coef. Absorption coefficient Ar Aryl

Bu Butyl, (CH2)3CH3 calc'd Calculated CSD Cambridge Structural Database Cp Cyclopentadienyl, C5H5 Cy Cyclohexyl Cy* Cyclohexene 8 Chemical Shift (NMR) d Doublet (NMR) DAB Diazabutadiene DFT Density functional theory depp l,3-bis(diethylphosphino)propane DIMPY a,oc' -diiminopyridine

Dipp 2,6-diisopropylphenyl, (CH(CH3)2)2C6H2 Dipp-DAB Af,Af'-bis-(2,6-diisopropylphenyl)butane-2,3-diimine DMSO Dimethylsulfoxide Dp Decomposition point

dppe 1,2-bis(diphenylphosphino)ethane, Ph2P(CH2)2PPh2

dppp 1,3-bis(diphenylphosphino)propane, Ph2P(CH2)3PPh2 E General Main Group Element

Et Ethyl, CH2CH3 HRMS High resolution mass spectrometry Hz Hertz

*Pr Iso-propyl, CH(CH3)2 !Pr-NHC l,3-diisopropyl-4,5-dimethylimidazol-2-ylidene

XIV JxY n-bond coupling constant between nuclei x and y L Neutral ligand M Metal Atom

Me Methyl, CH3

Mes Mesityl, 2,4,6-trimethylphenyl, Me3C6H2 mmol millimole Mp Melting point NacNac (3-diketiminate, [ArNC(Me)C(H)C(Me)NAr]" NBO Natural bond order NHC iV-heterocyclic carbene NHP TV-heterocyclic phosphenium NMR Nuclear Magnetic Resonance OTf Triflate, SO3CF3"

Ph Phenyl, C6H5 Pn Pnictogen, Group 15 element: N, P, As, Sb, Bi ppm parts per million q Quartet (NMR) R Any organic substituent RedOx reduction-oxidation s Singlet (NMR) t Triplet (NMR) l Bu tert-butyl, C(CH3)3 THF Tetrahydrofuran

TMS Trimethylsilyl, SiMe3 X Halogen atom xs excess

XV Chapter 1 - Introduction into Low Oxidation State Main Group Chemistry

1.1 General Introduction

Group 15 of the Periodic Table of the Elements is comprised of the elements nitrogen, phosphorus, arsenic, antimony and bismuth. A common name given to these elements is pnictogens that is derived from the Greek word pniktos, which means

"strangled".1 a symbol often used to denote pnictogen in a chemical structure is "Pn" and will be used throughout the thesis when an unspecified pnictogen atom is described. The elemental forms of the pnictogens span a wide range of forms including: a non-metal gas

(N2), non-metal solids (P-several allotropes), metalloids (As, Sb), and a metal (Bi).

The element to be discussed primarily in this dissertation is phosphorus, which can exist in a variety of solid-state structures. The most common allotrope is white or yellow phosphorus, which is a cubic solid array made up of tetrahedral P4 molecules.

This allotrope is very pyrophoric; however it is used extensively in inorganic synthesis.

Other allotropes (red and black) exist as polymeric materials with three coordinate phosphorus atoms.

The pnictogens all have five valence electrons and, at least for the heavier

elements in the group, are most typically found in compounds in either the +3 or +5

oxidation state. This thesis will primarily examine compounds with phosphorus atoms in

the +1 oxidation state (and compounds derived therefrom) along with structure and

reactivity of these compounds. Because of the unusual nature of this unusual oxidation

state for a lighter pnictogen atom, a discussion into the models used to assign and define

oxidation states is warranted.

1 1.2 Oxidation States, Valence States and Our Model

The concept of an element existing in a particular oxidation state is one of the oldest and most fundamental models used by chemists to rationalize the structural features and the reactivity of the compound in which that element is found.3 In its most simple form, the oxidation state of an atom provides a measure of the number of valence electrons associated with that atom therefore the oxidation state model can be used for the prediction or understanding of chemical behavior of the molecule.

In spite of their importance and ubiquity, the most common models used to assign oxidation states to atoms in molecules are far from perfect, especially for molecules that are bonded in a primarily covalent fashion.4"6 The most prevalent of such models involve the use of counting rules and a few assumptions (e.g., O is usually -2, H is usually +1, etc.) to assign oxidation states to each of the atoms in a molecule. While such models are useful for the balancing of Reduction-Oxidation (RedOx) reactions, they are not helpful for the rationalization of, for example, structural features, bonding models, or reactivity patterns. Numerous deficiencies of "counting rule" models have been presented by various authors, but a few prominent examples are worth highlighting for illustrative purposes. For group 14: the carbon atom in CH2CI2 is assigned an oxidation state of 0 on the basis or typical counting rules, however the structural features and reactivity patterns of that carbon atom are certainly more similar to those of the C(II) center in CHCI3, than to the C(0) centers in e.g., graphite or formaldehyde or the transient species chlorocarbene. In fact, the tetrahedral carbon atom in simple CR4 molecules (R = H,

alkyl, aryl, OH, halides, etc.) can be assigned oxidation states ranging all the way from -4

(e.g., CH4, which is not usually considered as a reducing agent) to +4 (e.g., CF4, which is

2 not usually considered as an oxidizing agent) in spite of the obvious structural similarities and the essentially identical bonding models used to describe the molecules.

In a related line of reasoning, such counting rules may suggest that the phosphorus atoms in L13P, PH3, and [PHJ"1" are all present in the -3 oxidation state in spite of obvious and significant differences in structural properties, bonding descriptions, and chemical reactivities between the phosphide, the phosphine and the phosphonium cation.

Clearly, the use of the oxidation states should be restricted to situations where they are appropriate, such as for the balancing of RedOx equations. One must not over-interpret the meanings of these assigned oxidation states.

Given the foregoing discussion, we prefer to employ an alternative model for the assignment of oxidation states of a type that are intended specifically to emphasize similarities in structural features, bonding descriptions and chemical behaviour. This model is more akin to the concept of valence states and is also related to the concept of the isolobal analogy.5 The protocol we use to assign the oxidation state of an element of interest is convenient and simple: the oxidation state of the element is determined by assessing the number of lone pairs of electrons on that atom, as illustrated in Figure 1.1.

Using phosphorus as an example, an atom with no lone pairs is considered to behave like

P(V), an atom with one lone pair is called P(III) and an atom with two lone pairs is treated as P(I), an atom with three lone pairs is P(-I) and one surrounded by four lone pairs is treated as P(-III). While even oxidation states are also possible, these must correspond to radicals using this convention and are much less common that the odd oxidation states.

3 Lewis ^-p p p p( .p. Structure y\ ^ \^ " ''' '••'

Oxidation State Pv P"1 P1 P-1 P-"1

Figure 1.1. Some examples of our model of correlating Lewis structure to oxidation state based on the number of lone pairs of electrons associated with the central atom. Phosphorus is used as an example.

The most obvious potential problem with the method we employ, which is also highlighted in Figure 1.2, is that several Lewis-type structures may appear to be reasonable for a given molecule. Often, a particular Lewis-type structure is implied to be more favorable on the basis of rules that have no relation to the actual distribution of electrons in a molecule (e.g., the minimization of formal charges). Again, it must be stressed that formal Lewis structures are simple models developed before the modern quantum mechanical descriptions of atomic structure and molecular bonding, that should not be taken too literally.5 More realistic assessments of the distribution of the electrons in molecules are provided by either experimental methods (e.g., structural features or chemical behaviour) or computational approaches.

I vPn -T© ,i»© J, V? ri.® 1 ® i -P-P-P— —P=P-P— — P = P = P —

— P*P-P•ffi — — P-P-P — ©

P(l) P(lll) P(V)

Figure 1.2. Ambiguity in Lewis-type chemical structure drawings of a single cation having the formula [P(PR3)2]. (arrows pointing towards the acceptor atom denote dative bonding)

4 It must also be noted that the use of the terms related to "valence" and "oxidation state" is potentially confusing and sometimes even misleading. For the purposes of this thesis and to conform to the language used in the literature, both terms will be used, however when the term "oxidation state" is used, it is used with the assumptions specified by the method outlined above.

For the remainder of the introductory chapter, some recent insights in regard to compounds containing main group elements from groups 13-16 in lower than normal oxidation or valence states will be examined. In light of the preceding description, it is clear that elements in lower than normal oxidation states are, by definition, more electron-rich than their more oxidized relatives. The consequences of such electron- richness are significant and sometimes dramatic: the presence of lone pairs of electrons typically results in a lower-than-usual coordination number (i.e. a coordinatively unsaturated center) and often provides for different reactivity than a pnictogen's more typical higher-oxidation state analogues. It is primarily because of the differences in reactivity that there has been a flurry of work on compounds containing elements in lower-than-usual oxidation states over the last decade or two, and much of this work has been recently reviewed.4

1.3 Examples of Low Oxidation State Main Group Elements

Main group p-block elements have traditionally been studied as elements that are a part of a larger ligand used for stabilizing transition metals (e.g., phosphines) or as part of organic molecules or biologically important compounds. In a somewhat different vein, in this introduction, the ligation of main group elements in unusually-low oxidation states

5 by common ligands will be discussed according to the type of ligand. While the view of such compounds as being "coordination complexes" of low oxidation state centers by various ligands is sometimes controversial, we do not necessarily wish to imply anything in regard to the nature of the bonding between the ligand and the element; this is simply a model one can use to classify and to illustrate the similarities between types of related compounds.

1.3.1 Phosphine Ligands

Phosphines have long been used as ligands in transition metal chemistry. In terms of the main group elements, the first example of a phosphorus (I) center stabilized by two phosphines was reported by Schmidpeter et al., in 1982.9 The initial motive behind the work was to isolate a phosphorus(I) halide by reducing PCI3 with SnCk (Scheme 1.1).

The reaction without a stabilizing ligand led to the formation of an orange precipitate, which is commonly reported as a by-product in reactions involving the reduction of halophosphines. The addition of a diphosphine to the reaction mixture resulted in a clear

solution, and the triphosphenium was isolated as the hexachlorostannate salt. The

methodology was then employed to generate various cyclic and acyclic triphosphenium

ions in the presence of an external reducing agent (SnC^ or phosphines with a halide

abstracting agent: AICI3).10"12 Of note is the apparent thermal instability of the acyclic

versions of the triphosphenium cations except for the case where the phosphine ligand

contains a hexa-alkyl bis-amino group (Scheme 1.2).11

6 2Sn 2 2 PCI, °' R P •* PR [SnCI6] + SnCI4 2 dppe 2 2

Scheme 1.1. First isolation of a cyclic triphosphenium salt.

2 AICI3 *p© 2PCI3 3 P(NMe ) / [AICI4] + [(Me2N)3PCI][AICI4] 2 3 (Me2N)3P P(NMe2)3

Scheme 1.2. Isolation of an acyclic triphosphenium salt.

The realization and isolation of these types of low valent pnictogen compounds

1 -a 1 o have motivated significant development in synthetic methodologies " and computational investigations into the most appropriate description of the bonding within triphosphenium and other low valent main group compounds. Computational studies have confirmed that conventional bonding descriptions are not appropriate for depicting these low valent pnictogens.17 The first arsenic(I) species was reported by Schmidbaur using a synthetic approach that is similar to one of those described above.

PPh, 2 SnCI2 PPh2 2ECI3 \e :E: [Sncy - SnCI4 / PPh, PPh2

E = P As

Scheme 1.3. First synthesis of an diphosphine stabilized arsenium salt.

Several anionic P(I) fragments have been reported by Fritz, with these compounds being used as ligands in transition metal chemistry 20-23

7 R2PX /PR2 R2P^p^PR2 Li® P© + p\\ | R,P Br—PR2 ^P\ 2r R2P PR2 R='Bu

Scheme 1.4. An example of a reaction using an anionic P(I) fragment.

One anionic diphosphine ligand was used by Karsch to form an overall neutral 4- membered ring species A.24 Both the arsenic and phosphorus variants were isolated and crystallized. In this example, an external reducing agent is not used and the ligand acted as the reducing agent. These types of compounds are the carbon analogues of some 8- membered phosphazene rings (B) reported by Schmidpeter in 1985.25'26

P*e Ph2P—R©

Ph2Rf >Ph 2 eW ©PPh2 \-./ l e ce Ph2Rr n :N:

SiMe3 ^P—PPh2

A B

Figure 1.3. Small neutral heterocycles containing triphosphenium moieties.

Since 2000, the work relating to the coordination of diphosphine-ligands to main

group elements has focused on the spontaneous yet controlled formation of these

complexes and understanding the mechanisms of the transformations and subsequent

reactivity of the products. In the case where the central atom is from Group 15 several

research groups including those of Dillon,14'27"29 Woollins18 and Macdonald13'15,17'30 have

developed facile clean syntheses with alternative more advantageous anions.

8 Dillon reported in 2000 that when PX3 (where X = CI, Br, or I) is treated with a diphosphine in the absence of a reducing agent, the desired cyclic triphosphenium salt is generated. In their report, it is suggested that excess diphosphine ligand can act as the reducing agent by reacting with X2 that is eliminated.27 These halo-oxidized diphosphine by-products were observed in situ only by 31P NMR and were not isolated. Dillon also reported the syntheses of numerous cyclic triphosphenium salts and some arsenic

31 analogues using SnCl2 as the reducing agent.

e e e e P P « P m P R p R R2R1. •PR ,^ 2 ,xR2,P. •PR .^2 --Z2- P--2z Phrn22P.r\ f PPi h' 2

\\ // \\ //

Figure 1.4. Some examples of triphosphenium cations. The substituent labeled R is commonly phenyl, ethyl or cyclohexyl.

The first report of the deliberate preparation and isolation of a triphosphenium species without the use of an external reducing agent came from the Macdonald group in

2003.13

P,3 _iEE^ PhJ'^fet* f>

Scheme 1.5. Synthesis of a triphosphenium iodide salt without the use of an external reducing agent.

9 This synthesis employs PI3 and generates the desired iodide salt in a clean and quantitative manner because the h that is produced is much less reactive compared to putative CI2 or Br2 by-products that would be produced using the lighter halophosphines.

One significant aspect of this result is that there is no observation by 31P NMR of an iodo-oxidized diphosphine when the reaction is conducted in a non-coordinating solvent such as dichloromethane. The synthesis of the triphosphenium iodide salts demonstrates that a mild synthetic route can be used to generate these types of compounds. Following the approach first used by Ellis and co-workers, in 2006 Woollins reported the use of P2I4 directly with l,8-bis(diphenylphosphino)naphthalene to generate the desired triphosphenium iodide salt.18

It is well known that in a solution containing a Lewis base, PI3 is in equilibrium with P2I4 and I2.32 While the mechanism underlying this RedOx process is not understood, the formation of the triphosphenium iodide salts can be understood in a purely formal sense by the further disproportionation of P2I4 into some form "P-I" that is chelated by the diphosphine ligand. In a decisive VT NMR study, Dillon and Monks demonstrated that the reduction of the phosphorus(III) center occurs after the chelation by the diphosphine ligand for a number of such species.29 The details of this mechanism will be discussed in the following chapter because of the importance of Dillon's findings in relation to our own.

In any event, the triphosphenium iodide salts are useful due to the convenient nature of metathesis chemistry for the exchange of the iodide anion for other types of

17 anions including [BPh4], [OTf], [PF6], [GaCl4] and others.

10 The triphosphenium cations are of interest because they are rare species that contain P(I) centers that are unusually stable. Furthermore, in spite of their overall positive charges there are two lone pairs of electrons on the dicoordinate phosphorus atom (in the most appropriate model of the electronic structure), which implies that the phosphorus center can still act as a Lewis Base. The models for illustrating these types of molecules (Figure 1.2) can either denote the ligands as having covalent bonds to the central phosphorus atom resulting in the central atom having a negative charge, or alternatively drawing a dative bond between the ligand and phosphorus center resulting in the central atom having a positive charge.4 Because of the lone pairs of electrons based on the central phosphorus, one can imagine its chemistry to be similar to that of a phosphide, and this does prove to be true in some cases; however, the reactivity is distinctive enough for one to consider the triphosphenium ions to be a class unto themselves.

Alkylation and protonation are readily observed with a large variety of alkyl groups with triphosphenium species (acyclic and cyclic).33,34 Such oxidation at the dicoordinate phosphorus atom is more facile in the acyclic cases, and mild alkylating reagents such as alkyl chlorides may be employed. For the cyclic versions, the much stronger reagents such as methyl triflate or triflic acid are required to methylate or protonate the P(I) center, respectively.33'35

11 1 9+ z+

/P Ph2P -PPh2 -^ Ph2P"">Ph2 ^^ Ph2P^ >Pn2

^^ OTfe OTfe

Scheme 1.6. Protonation and alkylation at the P(I) atom in various cyclic triphosphenium salts can be performed using the strong reagents triflic acid or methyl triflate, respectively.

In contrast to phosphides, depending on the nature of the oxidant, oxidation can occur at either the dicoordinate phosphorus atom or at the diphosphine ligand. When the oxidant has lone pairs of electrons (such as O, S, etc.), the oxidation occurs on the ligand phosphorus atoms because the LUMO is largely based on those atoms. Alternatively, when the oxidant does not have lone pairs of electrons (e.g., H+, CH3+, etc.), the oxidation occurs at the P(I) center because the HOMO is largely based there.

An example of an amazing reduction process with an unprecedented class of product was reported by Driess in which the phosphorus (I) center is re-coordinated by a well known organometallic reagent.36 Reaction of an excess of the Schwartz Reagent,

(Cp)2ZrHCl, with an acyclic triphosphenium ion results in a planar tetracoordinate phosphorus atom chelated by zirconium atoms. The P(I) center is formally reduced to P(-

o -I

III) and all of the phosphine ligands are lost in the process. The P NMR shows only 2 phosphorus containing products: the free ligand P(NMe2)3 at 123.9 ppm and the tetracoordinated phosphorus species at 254.2 ppm. Interestingly the signal for the P(I) center for the triphosphenium salt is observed at -194 ppm.

12 (Me3N)2Px.vP(NMe3)2 Cp2ZrHCI ^fx®/^*

••® *" H P. H BPh4

@BPh4 Cp2ZrC >K»2

Scheme 1.7. Complexation of a P(I) cation with a transition metal, in this case the common organometallic catalyst Schwartz reagent Cp2ZrHCl.

Although the chemistry of As(I) salts has not been studied in great detail to date, much of the observed reactivity is similar to that of the corresponding triphospheniums.

An interesting and informative result with the arsenic (I) species is that upon oxidation with atmospheric oxygen or other oxidizing reagents, the diphosphine ligand is oxidatively cleaved; however, the arsenic remains as an arsinidine-like "As-I" fragment that form (Asl)6 clusters which are capped by two iodides to give the dianions of [Asels]"

2 15

An important class of compounds analogous to the triphosphenium cations is the carbodiphosphiranes. These compounds can be rationalized as a carbon atom in the zero oxidation state (i.e., C(0)) stabilized by two phosphines ligands. The dicoordinate carbon has 2 lone pairs of electrons. This suggests that although the structure is similar to a carbene, the potential chemistry is very different due to the second lone pair of electrons associated with the carbodiphosphiranes. The first carbodiphosphirane compound,

'in hexaphenylcarbodiphosphorane, was first reported in 1961 by Ramirez. Since then much work has been done to synthesize these types of compounds as well as to coordinate them to transition metals and to explore other reactivity.38

13 -p—c—p— —p=c—P— —p=c=p-

*iv

Figure 1.5. Lewis structure depictions of triphosphenium cations and the analogous carbodiphosphiranes.

The preparation of these compounds usually begins with the displacement of the bromides from dibromomethane by two trialkylphosphines. The resultant bis- phosphonium salt can easily be deprotonated by a weak base followed by a second deprotonation of the ylide phosphonium salt with a much stronger base such as butyllithium or potassium metal. This general route is one of many ways to generate the carbodiphosphorane.

Carbodiphosphiranes are important because of their use as ligands in transition metal chemistry, as well as their Wittig-type applications in organic chemistry. The reactivity of these compounds is well established compared to the relatively unexplored

chemistry of the triphosphenium salts. Because of their similar electronic structures, the

results of reactivity studies of the triphosphenium cations should be compared to and

contrasted with the known chemistry of the carbodiphosphiranes.

14 e e Ph3pz=c—PPh3 Br

Br Ph3R^ N C—PPh3 0 I I Br2 II F,C—C—0 Ph3P=C—PPh2 (CF ) CO H 3 2 H90 ^ CF-, -CgHe

© .•CI © Ph3P^* PPh3

Ph2PCI Ph2C=C=0.

Ph3P^©^pph3 c CI-i' 0 Ph,C=C=C=PPh CO, ? PPh,

e Ph3P^ ^C02 ^C

©PPh3

Scheme 1.8. Examples of some reactions of carbodiphosphiranes with small inorganic molecules. (CO)5W=C=C=PPh3 Ph3Px ,C—Pt(cod) W C ( °)6 Ph3Px Ph2 \ 3 [(cod)Ptl2] ^ v-( Ni(CO)4^ \0—Ni(CO)3 K? toluene/^ ph/

•2 [HC(PPh3)2] © *&*© PhgP^'^PPhg

Ni(CO)4 AuCI(tht) THF Sv -tht -2CO N< Ph3Ps Ph3Px 1. CuCI ,C—Ni(CO)2 ,C—Au—CI 2. LiCp -LiCI Ph„P Ph3P

Cu

Ph3P - "- PPh3

Scheme 1.9. Reactions of carbodiphosphiranes with transition metals.

15 The carbodiphosphiranes are essentially the only examples of stable compounds containing a group 14 atom ligated by phosphines ligands other than a handful of compounds that are only stable in the coordination sphere of transition metals.39"41 The heavier group 14 analogues have yet to be reported. It is probable that the lack of effective synthetic methodologies has prevented the synthesis and study of further examples of these types of compounds.

In summary, phosphine ligands have been employed to stabilize a variety of low- valent group 14 and 15 species. In the case of the triphosphenium cations, it has been demonstrated that the phosphines ligands are very capable of reducing PX3 to generate

97

P(I) complexes spontaneously and cleanly. It is also noteworthy that, while phosphines ligands have not yet been used to stabilize isovalent reduced species containing group 16 elements, they are capable of facilitating a reduction of the EX4 (E=Se, Te) to EX2.

Recent work from the Ragogna group demonstrates that ECU is subsequently 49 reduced upon addition of varying stoichiometric amounts of triphenylphosphine. In particular, the addition of equimolar amounts of SeCLt and PPI13 generates the salt

[PPh3Cl][SeCl3] which indicates that the selenium was reduced from Se(IV) to Se(II) and

the phosphorus was oxidized from P(III) to P(V). This result clearly demonstrates that a

RedOx process is occurring in which the phosphine is facilitating the reduction of SeCLt-

Upon addition of a second equivalent of phosphine the selenium is completely

reduced to red selenium. With addition of a third equivalent of phosphine, the selenium

is even further reduced to Se(-II) with ligation of a phosphines to the selenium. Similar

results were reported for TeCU with equimolar amounts of triphenylphosphine with half

of the tellurium (Fv7) being completely reduced to tellurium metal and the phosphorus and 16 the other half remaining as Te(IV) in the dianion [TeCy ". Again, these results are indicative of the phosphine inducing spontaneous reduction of the chalcogen halide.

1.3.2 N-Heterocyclic Carbene Ligands

Phosphamethine cyanines are historically important compounds in main group chemistry because they were the first isolated group of compounds with (p-p)n: bonds between carbon and a main group element heavier than period II.43 These compounds were first synthesized by Dimroth through the reaction of tris(hydroxymethyl)phosphane or tris(trimethylsilyl)phosphines and an imidazolium salt.44 These dyes exhibit similar curve shapes and maximum absorptions in their UV/vis spectroscopy as compared to the corresponding methine and azamethine cyanines 43

A CI + P(CH2OH)3 •3CH20 -2HCI • HY

+ P(CH2OH)3 -3CH20 -2HCI -HY

X = S,NR1

Y = BF4, CI04 1 2 3 R = CH3, C2H5; R = H, Br, OCH3, C2H5, CH(CH3)2; R = H, Br, OCH3

Scheme 1.10. Classical syntheses of phosphamethinecyanine dyes.

17 Because the phosphorus reagents used in the classic syntheses of these dyes are either explosive or highly pyrophoric, other routes to their synthesis are desirable.

Schmidpeter et al. reported the reaction of a triphosphenium salt {P[P(NMe2)3]2}[BPh4] with carbene olefin dimers to generate the dicoordinate phosphorus cation analogous to the phosphamethine cyanine dyes through a unusual insertion into the C=C bond of the olefin.45

R R .-n R ••« R ,,x f W ^ + (Me2N)3P"--"P(NMe2)3 ^ / ssf f© ^N NT -2P(NMe2)3 V-NR RN e R R BPh4 0 BPh 4

Scheme 1.11. Example of a triphosphenium inserting into an olefin bond between two carbene-like fragments.

A recently discovered synthetic route, which makes use of the significant developments in the chemistry of stable N-heterocyclic carbenes (NHC's) since the early

1990's, is the direct reaction of PX3 (where X=C1, Br, I) with 2 equivalents of various

NHC's.16 This method requires an additional equivalent of NHC to react with the X2 that is formed during the reaction. The haloimidazolium halide salt that is formed is difficult to separate from the desired phosphenium salt because of similar solubilities. A more clean and efficient method for the generation of these carbene stabilized P(I) centers is by treating a cyclic triphosphenium salt with 2 equivalents of carbene. The ligand exchange from the diphosphine to the carbenes is expected because of the better donor abilities of the NHC's as well as the well-established ligand exchange chemistry of the triphosphenium species.4 Therefore, the only by-product of the reaction is the free diphosphine ligand which has very different solubility properties in comparison to the phosphenium salt and the diphosphine by-product can be easily washed away.16

18 R1 R1 CI \ n I R1 /\ R1 "^N N-" 3 \ / + PCI3

R2 R2 R2 e ci Clfc

R1 R1

e R1 /\ R1 p'e Nkp M 2 \ / + R2f\ PR2 -dppe NN, N N_//

H 2 * w R R R R2 R2

X = CI, BPh4 1 R = C2H5, CH(CH3)2, C(CH3)3, 1-Ad, Mes 2 R = H, CH3

Scheme 1.12. Syntheses of phosphamethinecyanines directly from NHC and either PCI3 or a P(I) salt.

These new methods of generating these historically important dicoordinate phosphenium centers are significant because of their use of stable, easily handled reagents and minimal, easily removed by-products. Furthermore, given the incredible diversity of NHC-like carbenes that have now been generated, the simplicity of this method is amenable to the production of a huge variety of dyes and related molecules.

Important new examples of low-valent phosphorus atoms stabilized by carbene ligands were prepared by Bertrand and co-workers. They found that the activation of P4,

(white phosphorus) by carbenes using either stable cyclic (alkyl)(amino)carbenes

(CAACs) or N-heterocyclic carbenes (NHCs) yields species such as D and E.47 The reaction of P4 with CAAC C yields a tetraphosphatriene which is formed in a 9:1 E:Z ratio with respect to central diphosphene moiety. When an NHC is treated with P4 initial

31P NMR results indicated the formation of the analogous (Z)- and (E)- tetraphosphatrienes. Over time however, these signals disappeared and a set of 10 signals

19 appeared in the ratio 1:1:1:1:1:1:1:3:1:1. These peaks were assigned to a P12 cluster containing two NHC groups, which was structurally characterized by single crystal X-ray diffraction.

1/2P4 P«~-P —P-n/wP

(E) & (Z)

Dipp Dipp Dipp N 1/2P4 ^ > -

N 1 1 I 1 1 Dipp Dipp Dipp

E

—*- Pi? cluster Scheme 1.13. Treatment of elemental phosphorus P4 with CAAC C or an NHC results in the opening of the P4 tetrahedron.

Other heavier main group elements have also been reportedly stabilized through carbene coordination. In Group 14, a carbene stabilized germanium (II) species has been reported along with some reactivity studies.48 An NHC was added to

tetramesityldigermene to generate the stabilized germylene. The germanium-carbene

bond is stable at room temperature, however upon treatment with 2,3-dimethylbutadiene

(DMB) at elevated temperatures in THF, there occurs displacement of the carbene ligand

and generation of the DMB-trapped germylene. This carbene-stabilized germanium

20 species can also act as a Lewis base through coordination to BH3 as confirmed by a single crystal X-ray diffraction study 48

'Pr jPr I -N Mes -N \ + „.Ge=Ge' Ge.., / V'Mes Mes Mes I I 'Pr jPr

Scheme 1.14. Stabilization of Ge(II) by an NHC.

A recent and remarkable report by Baines et al. involves the non-RedOx

2+ coordination of NHC-GeI2 to generate a germanium dication [Ge-NHC3] that remains in the +2 oxidation state throughout the reaction.49 This example is somewhat surprising considering the examples already discussed where PX3 is reduced from P(III) to P(I) upon ligation by two carbenes and subsequent loss of X2.

'Pr excess ''Dr-N/NN'''~'r 1 NHC= \_ +2 .Ge ,, >~G*..„, NHC NHC \ 21" NHC ipr

Scheme 1.15. Generation of a Ge(II) dication through a non-RedOx ligation of two NHC substituents.

An example with a group 16 element being stabilized by a carbene the reaction of a simple NHC with TeLt to generate the reduced carbene ligated Te(II) species.50 The crystal structure exhibited a T-shaped tellurium center, indicating two lone pairs on the

Te which is evidence of the tellurium being reduced in the reaction from Te(IV) to Te(II).

A final example is the coordination of carbene ligands to the Group 16 element selenium. The Ragogna group has reported the reaction of selenium tetrachloride with an

21 NHC to generate the neutral carbenium-selenide adduct which was characterized by single crystal X-ray diffraction.42 Because of the reactive nature of the CI2 (X=C1 or Br) produced, some of the carbene and selenium tetrachloride were consumed to form two chloroimidazolium salts with one hexachloroselenide. In the reaction with selenium tetrabromide, the desired adduct is formed with many side products that were not characterized.

'Pr 'Pr 'Pr I I CI -N € 3 SeCI4 + 4 St. ©) CI [SeCI6] > / -N CI -N I I 'Pr jPr 'Pr

Scheme 1.16. Reduction and coordination of a selenium atom by an NHC.

1.3.3 Imine-containing Ligands and Related Ligands

In terms of the differences in chemical behavior and observed structural features, perhaps the most interesting and diverse family of compounds that may contain low-valent main group elements are those bonded to various imine-containing ligands. The nature of compounds based on such ligands can be rationalized by the ligands' potentials for RedOx behavior.

Although saturated and unsaturated N-heterocyclic phosphenium (NHP) cations are isovalent with the well-known Arduengo-type ^-heterocyclic carbenes,51 their potential relationship to low oxidation state chemistry has only recently become apparent. In spite of their similarity to NHC's, NHP's are considerably more electrophilic, and have been employed as electron-withdrawing ligands for transition metal complexes and catalysts.52'53 More recently, the electrophilic nature of the NHP

22 moiety, which contains a P(III) center, has been employed by Gudat and co-workers to effect several types of bond activation and to generate unique highly-polarized diphosphines. 4

• • • • R-N/CNN-R R-N/FVR W W

Figure 1.6. Lewis structure depictions of an N-heterocyclic carbene and the phosphenium analogue.

Although many methods have been employed for their production, N- heterocyclic phosphenium cations are traditionally prepared by the treatment of a halogenated diaminophophine precursor with a halide abstracting agent.55 The precursor molecules are generally prepared by the treatment of PX3 with a diamido ligand. While this approach works for many ligands, some results of Denk,5 and later Gudat,56 with oc-diimines (a.k.a. diazabutadienes, DAB) suggested that low- valent phosphorus species might be involved in such syntheses. In particular, Denk found that the treatment of PCI3 with doubly-reduced DAB ligands resulted in precipitation of an orange, amorphous, polymeric material they described as "[PCl]n".

They pursued an alternative synthetic approach to the phosphenium salts. Gudat and co-workers found that the treatment of PCI3 with DAB ligands in the presence of base produced a suprising chlorophosphine in which a carbon of the DAB backbone has become chlorinated.56

23 N N- PCI3, Et3N R-N"XN-R

> CI

Scheme 1.17. Chlorophosphine produced through an electron transfer process with DAB and PC13.

The direct connection to low-valent phosphorus chemistry was elucidated in

2006. Cowley and co-workers found that when the synthetic methods used to produce

P(I) salts were performed in the presence of a-diimines (a.k.a. diazabutadienes, DAB) the resultant compounds are clearly salts containing phosphenium cations containing

P(III) centers.57'58 Identical reactivity is found when the analogous reactions are performed using arsenic trihalides. Our research group obtained similar results from a related system and it should be noted that our results have proven through the identities of the counter anions (e.g., [SnCls]", [I3]"), that a RedOx reaction had occured in the process, as illustrated in Scheme 1.18.59 Such atom to ligand electron transfer is apparently very favorable for P and As and renders this synthetic approach to phosphenium and arsenium cations one of the most efficient.

Ar— / \ Ar PCI3, SnCI2 [Sncy

Ar~-N N"Ar w //

Ar- / \ Ar PI3 Ar~"N N"Ar [l3]

Scheme 1.18. Synthesis of NHP's using methodology developed for the synthesis of P(I) salts, (methyl on products)

24 Interestingly, the cationic antimony analogues of NHCs that were reported by

Gudat and co-workers in 2004 display somewhat anomalous reactivity and exhibit metrical parameters (determined through single crystal X-ray diffraction) that are perhaps more suggestive of a diimine-coordinated Sb1 cation.60 Such a view is bolstered by the results of a comprehensive computational investigation of the isovalent analogues of NHCs from groups 13 to 16, which suggests that there are two

"lone pairs" of electrons on the Sb atom in the best description of the electronic structure of [NHSb]+. Unfortunately, no bismuth analogues of the five-membered

NHC species have yet been reported (although the series of four-membered ring variants reported by Veith and co-workers in the late 1980's contains a bismuth representative).61

s'b® of.e \JJ vs- W

Figure 1.7. Two Lewis-type depictions of [NHSb]+.

Perhaps the most direct support for the involvement of low-valent centers in imine-based complexes of the group 15 elements was provided by Cowley and co­ workers report of a cation containing As(I) ligated by a diiminopyridine (DIMPY) ligand, Figure 1.8.62 The metrical parameters (determined through single crystal X- ray diffraction) in this compound are most consistent with the ligand being present in its neutral (i.e. non-reduced) form, in clear contrast to the a-diimines such as DAB.

Furthermore, computational investigations by our group were in accordance with the assignment of a univalent arsenic center to the complex, suggest that one can rationalize the nature of such compounds on the basis of how easily the diimine may

25 be reduced: DAB is easily reduced and readily accepts two electrons from the low- valent Pn1 cation to generate a Pn(III) center, whereas DIMPY is much less susceptible to reduction and the low-valent center is retained.59 RN w.. R Figure 1.8. An As(I) atom is stabilized by the tricoordinate DIMPY ligand.

It should also be noted that [(DIMPY)As]+ is very reminiscent of the 10-Pn-3

(meaning that the Pn element is surrounded by a total of 10 electrons and is linked to only 3 other atoms) compounds first studied by Arduengo and Stewart in the 1980's and 90's,1'63 which are now being explored by Driess' research group.

The chemistry of NHCs has been explored and reviewed on numerous occasions and clearly is based on C(II); hence, it will not be explored in any detail herein.65"67 Similarly, the chemistry of the analogues containing silicon68'69 and germanium70 is best described and understood in terms of the presence of E(II) centers and not E(0) centers.

A group 14 example that validates the results Gudat obtained for antimony are

also from his group and based on Af-heterocyclic stannylenes (NHSn).71 They

observed that when NHSn was treated with more DAB ligand, the product is not the

spirocyclic RedOx product that is observed with the silyl analogue; rather, the Sn

atom exchanges between the two DAB ligands. This reactivity, in concert with the

observed metrical parameters of the compound, are consistent with the molecule

consisting of a Sn(0) atom coordinated by the neutral DAB ligand (Figure 1.8).

26 R" R R I 1 1 N-. N N Sn: + :sn N N I 1I lI R' R R

R R R I I I Nc -N N- sn Si N~ N N- I I I R R R

Scheme 1.19. The difference in reactivity of NHSn and NHSi when treated with DAB.

To conclude this discussion of the main group NHC analogues it is noteworthy to mention the recent thorough computational study of Tuonnonen et al..72 This report includes a detailed look into the electronic structure of the main group carbene analogues including calculated bond lengths and suggested Lewis depictions based on

NBO analysis.

© E E © © E © HN NH HN NH * * HN NH HN •* NH W \=J E = lrr E = B", Al-, Ga-, C, Si, E = Sb+, 02+, S2+, Ge, Sn, N+, P\ As* Se2+, Te2+

Figure 1.9. Summary of Lewis depictions of main group carbene analogues as described by NBO calculation results.

1.4 Dissertation Overview

The above section outlines some of the major advances made in the field of low oxidation state main group chemistry, and defines our model for assigning oxidation

27 states. The remainder of the dissertation describes recent work in low oxidation state phosphorus chemistry.

As stated previously, there is some ambiguity in the way phosphorus (I)- containing compounds are depicted. It should be noted that two different types of structures for these compounds will be used, however they are the equivalent, and one must remember that these depictions are only models and often times these models do not accurately convey the chemistry that the compound can undergo. The varying types of depictions are summarized in Figure 1.10.

-®,L-p4®- -p=p_fk- i •• i ii

Figure 1.10. Structural models used throughout this dissertation for phosphorus (I)- containing compounds.

The chapters in this thesis describe distinct projects. Chapter 2 presents a new and convenient synthesis for the synthesis of triphosphenium bromide and chloride salts.

This improved method is detailed along with a discussion into mechanistic studies performed. Chapter 3 outlines research into the synthesis of small target molecules containing a P(I) atom that could be used for ring opening polymerization. In Chapter 4, a preliminary study is presented on the synthesis of a /^-heterocyclic phosphenium halide

salt employing methods described in Chapter 2.

The dissertation is summarized in Chapter 5, along with some future considerations into the research discussed.

28 1.5 References

1. Arduengo, A. J.; Stewart, C. A.; Davidson, F.; Dixon, D. A.; Becker, J. Y.; Culley, S. A.; Mizen, M. B. J. Am. Chem. Soc. 1987,109, 627-647. 2. Corbridge, D. E. C. Phosphorus World, 2005. 3. Jensen, W. B. J. Chem. Educ. 2007, 84, 1418-1419. 4. Ellis, B. D.; Macdonald, C. L. B. Coord. Chem. Rev. 2007, 251, 936-973. 5. Parkin, G. J. Chem. Educ. 2006, 83, 791-799. 6. Smith, D. W. J. Chem. Educ. 2005, 82, 1202-1204. 7. McNaught, A. D.; Wilkinson, A. Compendium of Chemical Terminology, 2nd Edition, 1997. 8. Lewis, G. N. J. Am. Chem. Soc. 1916, 38, 762-785. 9. Schmidpeter, A.; Lochschmidt, S.; Sheldrick, W. S. Angew. Chem., Int. Ed. Engl. 1982, 21, 63-64. 10. Schmidpeter, A.; Lochschmidt, S.; Sheldrick, W. S. Angew. Chem., Int. Ed. Engl. 1985, 24, 226-227. 11. Schmidpeter, A.; Lochschmidt, S. Angew. Chem., Int. Ed. Engl. 1986,25, 253- 254. 12. Schmidpeter, A.; Lochschmidt, S.; Karaghiosoff, K.; Sheldrick, W. S. J. Chem. Soc, Chem. Commun. 1985, 1447-1448. 13. Ellis, B. D.; Carlesimo, M.; Macdonald, C. L. B. Chemical Communications 2003, 1946-1947. 14. Dillon, K. B.; Monks, P. K.; Olivey, R. J.; Karsch, H. H. Heteroatom Chem. 2004, 75,464-467. 15. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2004, 43, 5981-5986. 16. Ellis, B. D.; Dyker, C. A.; Decken, A.; Macdonald, C. L. B. Chem. Commun. 2005, 1965-1967. 17. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2006, 45, 6864-6874. 18. Kilian, P.; Slawin, A. M. Z.; Woollins, J. D. Dalton Trans. 2006, 2175-2183. 19. Gamper, S. F.; Schmidbaur, H. Chem. Ber.-Rec. 1993,126, 601-604. 20. Kovacs, I.; Fritz, G. Z Anorg. Allg. Chem. 1994, 620, 1364-1366. 21. Kovacs, I.; Fritz, G. Z Anorg. Allg. Chem. 1994, 620, 1367-1368.

29 22. Kovacs, I.; Fritz, G. Z Anorg. Allg. Chem. 1994, 620, 1-3. 23. Krautscheid, H.; Matern, E.; Fritz, G.; Pikies, J. Z. Anorg. Allg. Chem. 1998, 624, 1617-1621. 24. Karsch, H. H.; Witt, E. J. Organom. Chem. 1997, 529, 151-169. 25. Schmidpeter, A.; Burget, G. Angew. Chem., Int. Ed. Engl. 1985,24, 580-581. 26. Schmidpeter, A.; Steinmuller, F.; Sheldrick, W. S. Z Anorg. Allg. Chem. 1989, 579, 158-172. 27. Boon, J. A.; Byers, H. L.; Dillon, K. B.; Goeta, A. E.; Longbottom, D. A. Heteroatom Chem. 2000,11, 226-231. 28. Boyall, A. J.; Dillon, K. B.; Howard, J. A. K.; Monks, P. K.; Thompson, A. L. Dalton Trans. 2007, 1374-1376. 29. Dillon, K. B.; Monks, P. K. Dalton Trans. 2007,'1420-1424. 30. Ellis, B. D.; Macdonald, C. L. B. Phosphorus, Sulfur Silicon Relat. Elem. 2004, 779,775-778. 31. Barnham, R. J.; Deng, R. M. K.; Dillon, K. B.; Goeta, A. E.; Howard, J. A. K.; Puschmann, H. Heteroatom Chem. 2001,12, 501-510. 32. Kirsanov, A. V.; Gorbatenko, Z. K.; Feshchenko, N. G. Pure Appl. Chem. 1975, 44, 125-139. 33. Burton, J. D.; Deng, R. M. K.; Dillon, K. B.; Monks, P. K.; Olivey, R. J. Heteroatom Chem. 2005,16, 447-452. 34. Dillon, K. B.; Olivey, R. J. Heteroatom Chem. 2004,15, 150-154. 35. Boyall, A. J.; Dillon, K. B.; Monks, P. K.; Potts, J. C. Heteroatom Chem. 2007, 18, 609-612. 36. Driess, M.; Aust, J.; Merz, K.; van Wullen, C. Angew. Chem., Int. Ed. 1999,38, 3677-3680. 37. Ramirez, F.; Desai, N. B.; Hansen, B.; McKelvie, N. J. Am. Chem. Soc. 1961, 83, 3539-3540. 38. Kolodiazhnyi, O. I. In Phosphorus Ylides: Chemistry and Application in Organic Synthesis; Wiley-VCH: Weinheim, 1999, p 555 pp. 39. Ettel, F.; Huttner, G.; Imhof, W. J. Organomet. Chem. 1990,397, 299-307.

30 40. Ettel, F.; Huttner, G.; Zsolnai, L. Angew. Chem., Int. Ed. Engl. 1989, 28, 1496- 1498. 41. Ettel, F.; Schollenberger, M.; Schiemenz, B.; Huttner, G.; Zsolnai, L. J. Organomet. Chem. 1994, 476, 153-162. 42. Dutton, J. L.; Tabeshi, R.; Jennings, M. C; Lough, A. J.; Ragogna, P. J. Inorg. Chem. 2007,46, 8594-8602. 43. Jutzi, P. Angew. Chem., Int. Ed. Engl. 1975,14, 232-245. 44. Dimroth, K.; Hoffmann, P. Angew. Chem. 1964, 76, 433. 45. Schmidpeter, A.; Lochschmidt, S.; Willhalm, A Angew. Chem., Int. Ed. Engl. 1983, 22, 545-546. 46. Regitz, M.; Scherer, O. J.; Editors Multiple Bonds and Low Coordination in Phosphorus Chemistry, 1990. 47. Masuda, J. D.; Schoeller, W. W.; Donnadieu, B.; Bertrand, G. Angew. Chem., Int. Ed. 2007,46, 7052-7055. 48. Rupar, P. A.; Jennings, M. C; Ragogna, P. J.; Baines, K. M. Organometallics 2007,26,4109-4111. 49. Rupar, P. A.; Staroverov, V. N.; Ragogna, P. J.; Baines, K. M. J. Am. Chem. Soc. 2007,729,15138-15139. 50. Kuhn, N.; Abu-Rayyan, A.; Piludu, C; Steimann, M. Heteroatom Chem. 2005, 76,316-319. 51. Arduengo, A. J.; Harlow, R. L.; Kline, M. J. Am. Chem. Soc. 1991, 773, 361-363. 52. Abrams, M. B.; Scott, B. L.; Baker, R. T. Organometallics 2000, 79, 4944-4956. 53. Gudat, D.; Haghverdi, A.; Nieger, M. Angew. Chem., Int. Ed. 2000,39, 3084- 3086. 54. Burck, S.; Gudat, D.; Nieger, M. Angew. Chem., Int. Ed. 2004, 43, 4801-4804. 55. Denk, M. K.; Gupta, S.; Ramachandran, R. Tetrahedron Lett. 1996,37, 9025- 9028. 56. Gudat, D.; Haghverdi, A.; Hupfer, H.; Nieger, M. Chem. Eur. J. 2000, 6, 3414- 3425. 57. Reeske, G.; Hoberg, C. R.; Hill, N. J.; Cowley, A. H. J. Am. Chem. Soc. 2006, 128, 5300-5300.

31 58. Reeske, G.; Cowley, A. H. Inorg. Chem. 2007,46, 1426-1430. 59. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chim. Acta 2007, 360, 329-344. 60. Gudat, D.; Gans-Eichler, T.; Nieger, M. Chem. Commun. 2004, 2434-2435. 61. Caminade, A. M.; Veith, M.; Huch, V.; Malisch, W. Organometallics 1990, 9, 1798-1802. 62. Reeske, G.; Cowley, A. H. Chem. Commun. 2006, 1784-1786. 63. Arduengo, A. J., Ill; Stewart, C. A. Chem. Rev. 1994, 94, 1215-1237. 64. Driess, M.; Muresan, N.; Merz, K.; Paech, M. Angew. Chem., Int. Ed. 2005, 44, 6734-6737. 65. Crabtree, R. H. Coord. Chem. Rev. 2007,251, 595. 66. Hahn, F. E. Angew. Chem., Int. Ed. 2006, 45, 1348-1352. 67. Kuhl, O. Chem. Soc. Rev. 2007, 36, 592-607. 68. Baker, R. J.; Jones, C; Mills, D. P.; Pierce, G. A.; Waugh, M. Inorg. Chim. Acta 2008, 361, 427-435. 69. Driess, M.; Yao, S.; Brym, M.; Van Wuellen, C; Lentz, D. J. Am. Chem. Soc. 2006,128, 9628-9629. 70. Kuhl, O. Coord. Chem. Rev. 2004, 248, 411-427. 71. Gans-Eichler, T.; Gudat, D.; Naettinen, K.; Nieger, M. Chem. Eur. J. 2006,12, 1162-1173. 72. Tuononen, H. M.; Roesler, R.; Dutton, J. L.; Ragogna, P. J. Inorg. Chem. 2007, 46, 10693-10706.

32 Chapter 2 - A Convenient Preparative Method for Cyclic Triphosphenium Bromide and Chloride Salts

2.1 Introduction

The chemistry of compounds containing main group elements in unusually low oxidation states (or perhaps more appropriately valence states)1 has been an area of intense investigation over the last two decades.2 Such compounds are of both fundamental and practical interest in comparison to their "normal" oxidation state counterparts because of the differences in the structural features that are observed, the electronic structures that are adopted and, most importantly, the unique reactivity patterns engendered by the presence of the unusually electron rich centers present in such molecules. While much of the research into the lower oxidation state compounds containing the heavier elements of group 15 (pnictogens, Pn = P, As, Sb, Bi) has focused on the transient pnictinidines of the form R-Pn and their related precursor molecules, there has also been considerable interest in stable compounds containing Pn(I) centers. "

In particular, the seminal work in the 1980's of Schmidtpeter demonstrated that

"triphosphenium" cations 2.1, which contain P(I) centers, are stable compounds that can

Q be readily isolated in many instances. Triphosphenium cations contain an electron-rich, dicoordinate phosphorus center and are isovalent with bis(phosphine)iminium (PNP) cations 2.2 and neutral carbodiphosphoranes 2.3,9 which are topical molecules10"12 that are related to triphosphenium cations by the "diagonal relationship".13

R3P^^PR3' R3P=N=PR3 R3P=C=PR3

2.1 2.2 2.3

Figure 2.1. Examples of relevant small main group molecules.

33 More recently, the research groups of Dillon14"16 and Woollins,17 in addition to our group18, have investigated synthetic approaches to halide salts of triphosphenium cations19 (and some of the analogues containing dicoordinate arsenic atoms14'18'20"22).

While we were able to demonstrate that the direct reaction of chelating diphosphines with

PI3 in dichloromethane results cleanly in the formation of triphosphenium iodide salts with the concomitant elimination of I2,19 the corresponding reactions of PCI3 or PBr3 generate significant by-products in addition to the desired salts under similar conditions.15

An elegant 31P nuclear magnetic resonance (NMR) study by Dillon and Monks has recently elucidated the mechanism of the formation of triphosphenium cations under such conditions, as illustrated in Scheme 2.1. This proposed mechanism suggests two intermediate species, one acyclic (2.4) and one cyclic (2.5). When the study was conducted with bis(diethylphosphino)propane as the chelating diphosphine, the two intermediates were unambiguously assigned from the variable temperature and time spectra. Their diagnostic signals were significantly different than the triphosphenium product and an example of their NMR shifts is illustrated in Table 2.1. Of note is that the chelated phosphorus oxidation state is P(III) in both intermediates. This hypothesized mechanism also demonstrates that the rate-determining step of the reaction is the formal elimination of X2 (and consequent reduction of the P(III) center to a P(I) center), which rapidly oxidizes the diphosphine ligand, or other reducing agent present, to generate the undesirable by-products. The presence of such oxidized diphosphine by-products, which are difficult to separate from the desired products because of their similar solubilities, renders the isolation of pure products difficult or impossible and complicates the interpretation of reactivity studies on the triphosphenium halide salts.

34 Table 2.1. 31P NMR shifts corresponding to the acyclic and cyclic intermediates as well as the product for the room temperature reaction of PCI3 + bis(diethylphosphino)propane (depp).23 Species 5 (ppm) 2.4-acyclic intermediate 17.7 (s) -48.3 (d) 57.2 (d) 2.5-cyclic intermediate -102.6 (t) 43.0 (d) 2.6-triphosphenium product -254.1 (t) [deppP][Cl] 30.7 (d)

:X P-X: ..0 + R?P PR? ^ U •:. w R2P ©PR2 '*:'

2.4 I

:0 r%

+ *2P PR2 'I.' * „Y „ R2P PR2 . .V$

2.6 2.5

Scheme 2.1. Dillon and Monks' mechanism of triphosphenium cation formation where X = CI or Br; the rate-determining step is the RedOx step involving the formal elimination of X2 from the P(III) containing dication intermediate (Adapted from reference 23).

Presented herein, is a simple synthetic approach that allows for the preparation

and isolation of high-purity bromide and chloride salts of some cyclic triphosphenium

cations (2.7 and 2.8). Furthermore, we detail experimental evidence that provides some

insight into the nature of the process underlying the formation of the cations and by­

products using this new protocol.

35 Ph2R PPh2 Ph2P PPh2

2.7 2.8

Figure 2.2. Structural depictions of the triphosphenium cations 2.7 and 2.8.

2.2 Experimental

Reagents and General Procedures. All manipulations were carried out using standard inert atmosphere techniques. Phosphorus (III) bromide, phosphorus (III) chloride, 1,2- bis(diphenylphosphino)ethane (dppe), l,3-bis(diphenylphosphino)propane were purchased from Strem Chemicals Inc., and all other chemicals and reagents were obtained from Aldrich. Phosphorus (III) bromide and phosphorus (III) chloride were distilled before use, and all other reagents were used without further purification. All solvents were dried on a series of Grubbs'-type columns and were degassed prior to use. CDCI3 was dried over calcium hydride.

Instrumentation. NMR spectra were recorded at room temperature in CDCI3 solutions on a Bruker Avance 300 MHz spectrometer. Chemical shifts are reported in ppm,

! 13 31 relative to external standards (SiMe4 for H and C NMR, 85% H3PO4 for P NMR).

Coupling constant magnitudes, |/|, are given in Hz. Melting point (Mp) or decomposition points (Dp) were obtained on samples sealed in glass capillaries under dry

N2 using an Electrothermal melting point apparatus. High resolution mass spectrometry was performed at the McMaster Regional Centre for Mass Spectrometry. Elemental analyses were performed either at Guelph Chemical Laboratories or Atlantic Microlabs.

36 Preparation of [(dppe)P][Br] (2.7[Br]). To a colourless solution of PBr3 (531 mg,

1.961 mmol) in CH2CI2 (35 mL) was added cyclohexene (0.62 mL, 6.080 mmol) followed by addition of a colourless solution of dppe (1.641 g, 4.119 mmol) in CH2CI2.

The colourless solution was stirred at room temperature overnight, after which the white precipitate was filtered and volatile components were removed under reduced pressure.

THF (30 mL) was added to the resulting white foam and the heterogeneous mixture was sonicated at room temperature for 1 hr. The solid was then washed with THF (3x20 mL) and any remaining volatile components were removed under reduced pressure. The resultant white powder was dissolved in MeCN and after slow evaporation of the solvent a colourless crystalline material was obtained. Yield: 96% (957 mg, 1.883 mmol).

l 13 X ^Pf'H} NMR: -229.8 (t, VPP = 449, IP), 64.2 (d, Jw = 449, 2P). C{ H} NMR: 30.1

l 2 2 J (d, Jcp = 44), 126.2 (dd, VCP = 73, JCp = 9), 129.8 (t, JCp = 6), 132.4 (s), 133.5 (s). H

2 NMR: 3.70 (d, 7HP = 16, 4H), 7.56 (m, 12H) 7.84 (m, 8H). Dp: 203-206 °C. HRMS

+ (ESI) calc'd for C26H24P3 429.1091, found 429.1082 (-2.1 ppm). While we have been unable to obtain satisfactory microanalytical data for 2.7[Br], multinuclear NMR and powder X-ray diffraction experiments suggest that the product is pure.

Preparation of [(dppp)P][Br] (2.8[Br]). To a colourless solution of PBr3 (415 mg,

1.533 mmol) in CH2C12 (30 mL) was added cyclohexene (0.46 mL, 4.59 mmol) followed by addition of a colourless solution of dppp (905 mg, 2.195 mmol). The colourless solution was stirred at room temperature overnight and then filtered and volatile components were removed under reduced pressure. THF (35 mL) was added to the resulting white foam and the heterogeneous mixture was sonicated at room temperature

37 for 1 hr. The solid was then washed with THF (3x20 mL) and any remaining volatile components were removed under reduced pressure. The resultant white powder was dissolved in MeCN and after slow evaporation of the solvent a colourless crystalline material was obtained. Yield: 83% (666 mg, 1.273 mmol). 31P{lH] NMR: -209.3 (t, VPP

= 422, IP), 23.1 (d, %> = 422, 2P). 13C{1H} NMR: 19.4 (s), 25.1 (d, VCP = 47), 126.1

2 2 ! (dd, VCP = 75, 7Cp = 8), 129.7 (t, JCv = 6). H NMR: 2.74 (m, 2H), 3.34 (m, 4H), 7.50

+ (m, 12H), 7.75 (m, 8H). Dp : 192-194 °C. HRMS (ESI) calc'd for C27H26P3 443.1247, found 443.1269 (+4.9 ppm). Anal. Calc'd. For C27H24BrP3 (523.321): C 61.97, H 5.01.

Found: C 62.24, H 5.31%.

Preparation of [(dppe)P][Cl] (2.7[C1]). To a colourless solution of PC13 (500 mg, 3.64 mmol) in CH2CI2 (30 mL) was added cyclohexene (1.1 mL, 10.9 mmol) followed by addition of a colourless solution of dppe (1.89 g, 4.73 mmol). The colourless solution was stirred at room temperature overnight and then filtered and volatile components were removed under reduced pressure. THF (35 mL) was added to the resulting white foam and the heterogeneous mixture was sonicated at room temperature for 1 hr. The solid was then washed with THF (3x20 mL) and any remaining volatile components were removed under reduced pressure. The crude white powder was dissolved in a minimal amount of CH2CI2 and, after the precipitation of a yellow by-product, the solution was filtered and all remaining volatile components were removed under reduced pressure.

The resultant white powder was dissolved in MeCN and after slow evaporation of the

solvent a colourless crystalline material was obtained. Yield: 56% (948 mg, 2.04 mmol).

31 ! P{ H} NMR: -228.9 (t, VPP = 448, IP), 64.6 (d, VPP = 448, 2P). ^C^H} NMR: 29.9

38 X 2 2 (d, VCP = 44), 126.2 (dd, JCP = 75, JCP = 8), 129.7 (t, 7CP = 6), 132.3 (s), 133.4 (s). *H

2 NMR: 3.75 (d, JBP = 16, 4H), 7.57 (m, 12H), 7.86 (m, 8H). Dp : 90-91 °C. HRMS

+ (ESI) calc'd for C26H24P3 429.1091, found 429.1099 (+1.9 ppm). Anal. Calc'd. For

C26H24CIP3 (468.843): C 67.18, H 5.20. Found: C 67.07, H 4.84%.

Preparation of [(dppp)P][CI] (2.8[C1]). To a colourless solution of PC13 (210 mg, 1.53 mmol) in CH2CI2 (30 mL) was added cyclohexene (0.44 mL, 4.39 mmol) followed by addition of a colourless solution of dppp (780 mg, 1.89 mmol). The colourless solution was stirred at room temperature overnight and then filtered and volatile components were removed under reduced pressure. THF (35 mL) was added to the resulting white foam and the heterogeneous mixture was sonicated at room temperature for 1 hr. The solid was then washed with THF (3x20 mL) and any remaining volatile components were removed under reduced pressure. The crude white powder was dissolved in a minimal amount of CH2CI2 and, after the precipitation of a yellow by-product, the solution was filtered and all remaining volatile components were removed under reduced pressure.

The resultant white powder was dissolved in MeCN and after slow evaporation of the solvent a colourless crystalline material was obtained. Yield: 45% (345 mg, 0.72 mmol).

l l ^Pj'H} NMR: -209.4 (t, JPP = 422, IP), 23.1 (d, JP? = 422, 2P). "C^H} NMR: 19.3

2 2 (s), 25.1 (d, VCP = 47), 125.9 (dd, VCP = 75, /Cp = 10), 129.7 (t, /Cp = 6), 132.3 (s),

133.4 (s). *H NMR: 2.72 (m, 2H), 3.29 (m, 4H), 7.45 (m, 12H), 7.75 (m, 8H). Dp : 140-

+ 141 °C. HRMS (ESI) calc'd for C27H26P3 443.1247, found 429.1252 (+1.0 ppm). Anal.

Calc'd. For C27H26CIP3 (478.869): C 67.72, H 5.47. Found: C 67.54, H 5.74%.

39 Small scale preparation and 31P NMR spectroscopic identification of the cations in

[dppeBr2][A]2 (2.9[A]2) and [dpppBr2][A]2 (2.10[A]2); A = Br or Br3. An excess of liquid Br2 was added to dichloromethane solutions of either dppe or dppp and stirred for several hours. All volatile components were removed under reduced pressure and the

31 resultant solids were dissolved in CDC13. P{ *H} NMR: 2.9 60.1(s); 2.10 61.6(s).

Small scale preparation and 31P NMR spectroscopic identification of the cations in

[dppeCl2][A]2 (2.11[A]2) and [dpppCl2][A]2 (2.12[A]2); A = CI or Cl3. Pure Cl2 gas was bubbled through dichloromethane solutions of either dppe or dppp until the solutions retained the yellow-green colour. All volatile components were removed under reduced pressure and the resultant solids were dissolved in CDC13. ^P^H} NMR: 2.11 75.2(s);

2.12 75.8(s).

Preparation of [dppeBr2][Br3]2 (2.9[Br3]2). To a colourless solution of dppe (1.97 g,

4.94 mmol) in CH2C12 was added Br2 (0.51 mL, 9.89 mmol) dropwise until a bright yellow colour persisted. The solution was stirred at room temperature for 20min before a white precipitate formed. The reaction mixture was stirred overnight, after which the yellow solution was separation from the white solid. All volatile components remaining were removed and the white solid was collected. All attempts at recrystallization were unsuccessful because of the amorphous nature of the compound, as confirmed by powder

X-ray diffraction experiments. Yield: 55.5% (2.85 g, 2.74 mmol). ^P^H} NMR:

60.1(s). ^C^H} NMR: 25.2 (m), 117.7 (m), 130.8 (s), 134.0 (s), 136.9 (s). *H NMR:

4.64 (s, 4H), 7.80 (m, 12H), 8.35 (m, 8H). Mp: 128-132 °C.

40 Preparation of [dppeBr2][Br]2 (2.9[Br]2) from 2.9[Br3]2. To a yellow solution of

2.9[Br3]2 (200 mg, 0.21 mmol) in a 4:1 mixture of CH2C12 and CHC13 (25 mL) was added cyclohexene (110 |iL, 1.08 mmol, 5.15 equiv.), giving rise to a lighter yellow solution.

Following 3 hours of stirring, the reaction mixture the volatile components were removed under reduced pressure and a light grey powder was collected. Yield: 56% (77 mg, 0.12 mmol). ^P^H} NMR: 58.2(s). ). ^Q1!!} NMR: 21.7 (m), 122.1 (m), 130.5 (s), 132.0

(s), 135.3 (s). *H NMR: 3.56 (s, 4H), 7.67 (m, 12H), 8.05 (m, 8H). Mp: 187-190 °C.

We have been as yet unable to obtain satisfactory microanalysis for compounds 2.9[Br3]2

and 2.9[Br]2; please note that "2.9[Br]2" is a well-known organic reagent that is generally

prepared and used in situ.25 While the dication 2.9 is unambiguously identified as the

only observed component in both salts by the multinuclear NMR data presented above,

the identity of the anions is not provided by such experiments. To confirm the identity of

the anions for 2.9[Br]2 and 2.9[Br3], single crystals were isolated from NMR samples of

2.9[Br]2 and 2.9[Br3] in CDCI3. The solutions were exposed to ambient conditions and

the isolated cations in the crystal structure can be understood as cation 2.9 reacting with

ambient water.

X-ray Crystallography.

Each crystal was covered in Nujol and placed rapidly into the cold N2 stream of the Kryo-

Flex low temperature device. The data were collected using the SMART26 software on a

Bruker APEX CCD diffractometer using a graphite monochromator with Mo Koc

radiation (X = 0.71073 A). A hemisphere of data was collected using counting times of

10-30 seconds per frame. The data were collected at -100 °C. Details of crystal data,

41 data collection and structure refinement are listed in Table 2.2 and selected metrical parameters are compiled in Table 2. Data reductions were performed using the SAINT27 software and the data were corrected for absorption using SADABS. The structures were solved by direct methods using SIR9729 and refined by full-matrix least-squares on

F2 with anisotropic displacement parameters for the non-H atoms using SHELXL-9730 and the WinGX31 software package. Details of the final structure solutions were evaluated using PLATON and thermal ellipsoid plots were produced using SHELXTL.32

Powder X-ray diffraction experiments were performed with a Bruker D8 Discover diffractometer equipped with a Hi-Star area detector using Cu Ka radiation (A, = 1.54186

o A).

42 Table 2.2. Summary of X-ray crystallographic data for compounds 2.7[Br], 2.8[Br] and 2.8[HC12]. Compound [(dppe)P][Br] [(dppp)P][Br] [(dppp)P][HCI2] Empirical formula C26H24BrP3 C27H26BrP3 C27H27CI2P3 Formula weight 509.27 523.30 515.30 Temperature (K) 173(2) 173(2) 173(2) Wavelength (A) 0.71073 0.71073 0.71073 Habit, Colour Prism, Colourless Prism,Colourless Prism,Colourless Crystal system Triclinic Monoclinic Monoclinic Space group P-l P2i/c P2i/n Unit cell dimensions: a (A) 9.8923(17) 11.1018(11) 12.594(2) b(A) 13.499(2) 18.9235(19) 16.088(3) c(A) 19.672(4) 12.9372(13) 12.686(2) a(°) 71.401(2) 90 90 P(°) 79.610(2) 113.898(1) 91.516(2) 7(°) 69.134(2) 90 90 Volume (A3) 2319.9(7) 2484.9(4) 2569.3(8) 4 4 4 Density (calculated) 1.458 1.399 1.332 Absorption coefficient (mm"1) 1.990 1.860 0.454 F(000) 1040 1072 1072 0 range for data collection (°) 1.10 to 27.50 2.01 to 27.50 2.04 to 27.50 Limiting indices -122a(I)] R\ = 0.0348, Rl= 0.0471 /Jl = 0.0410 wR2 = 0.0812 wRl = 0.0886 wR2 = 0.0912 R indices (all data) Rl = 0.0496 Rl =0.0758 Rl = 0.0604 wR2 = 0.0943 wR2 = 0.0996 wR2 = 0.0997 Largest difference map Peak 0.654 and -0.299 0.591 and-0.301 0.434 and-0.313 and hole (e A"3) 2 2 *Rl(F)= E(|F0| - |FC|)/I|F0|} for reflections with F0 > 4(o(F0)). wR2(F ) = {£w(|F0| - m \FcffrLw{\F0ff} where w is the weight given each reflection.

43 Table 2.3. Summary of X-ray crystallographic data for compounds [dppe(OH)2][Br] and [dppe(Q)(OH)][Br].

Compound [dppe(OH)2][Br]2 [dppe(0)(OH)][Br3]

Empirical formula C26H26Br202P2 C26H25Br302P2 Formula weight 592.23 671.13 Temperature (K) 173(2) 173(2) Wavelength (A) 0.71073 0.71073 Habit, Colour Prism, Colourless Prism, Orange Crystal system Triclinic Monoclinic Space group P-l P-l Unit cell dimensions: a (A) 6.846(2) 8.6158(8) b(k) 9.187(3) 11.4881(11) c(A) 10.972(4) 14.2085(14) a(°) 109.032(4) 107.6360(10) P(°) 97.300(4) 98.1280(10) 7(°) 97.678(4) 97.8280(10) Volume (A3) 635.7(4) 1302.7(2)

1 1 2 Density (calculated) 1.547 1.711 Absorption coefficient (mm1) 3.335 4.790 F(000) 298 664 6 range for data collection (°) 2.00 to 27.49 1.53 to 27.50 Limiting indices -8 < h < 8, -102o(/)] Rl= 0.0830, Rl = 0.0259 wR2 = 0.1555 wR2 = 0.0774 R indices (all data) Rl =0.1378 Rl =0.0315 wR2 = 0.1738 wR2 = 0.0850 Largest difference map Peak 1.119 and-0.527 0.759 and -0.383 and hole (e A"3) a 2 2 Rl(F) = I(|F0| - |FC|)/Z|F0|} for reflections with F0 > 4(a(F0)). wR2(F ) = {Zw(|F0| - 2 2 2 m \Fc\ ) rLw(\F0\ f} where w is the weight given each reflection. 2.3 Results and Discussion

As early as 1985, the research group of Schmidpeter demonstrated that the direct reaction of phosphorus(III) chloride with some phosphine donors can result in the disproportionation of PCI3 and the generation of triphosphenium salts, as illustrated in

Scheme 2.2.3 '34 While the parent chloride salts were not isolated, chloride complexation

33 35 with AICI3 provided the salt [(Ph3P)2P][AlCl4] ' or salt metathesis with the transient

34 triphosphenium chloride salt and [Na][BPh4] to obtain the salt [{(Me2N)3P}2P][BPh4]; both of these results are consistent with the at least intermediate formation of triphosphenium chloride salts.

P I® D ~~1® (Me2N)3p- >(NMe2)3 Ph3P' >Ph3 ' Q + [BPh4l 3P(NMi2)3, Cl +3PPh3 0[A|C|4] + 2 [Na][BPh4] T +2 AICI3 -2[Na][CI] cr CI CI CI Me2N-P®-NMe2 ®[BPh4] Ph-P®Ph °[AICI4] NMe2 ph

Scheme 2.2. Some early reactions of Schmidpeter et al. that are consistent with the intermediate formation of triphosphenium chloride salts.

More recently, the research group of Dillon investigated the reaction of phosphorus(III) halides with a large variety of monodentate and bidentate phosphine donors using 31P NMR spectroscopy.14'15 In addition to the desired salts, (including

[dppeP][X] and [dpppP][X], i.e. 2.7[X] and 2.8[X]), the NMR spectra consistently indicated the formation of the oxidized (halogenated) phosphines as by-products in the reaction mixtures. Unfortunately, since the solubility characteristics of the by-products are very similar to those of the desired triphosphenium salts, the separation and purification of the target compounds are very difficult. However, we discovered that,

45 under certain conditions, it is possible to generate triphosphenium iodide salts through the

18 19 diphosphine-assisted disproportionation of PI3. ' When such reactions are performed in non-coordinating solvents, such as dichloromethane, the only phosphorus containing product is the desired triphosphenium iodide salt and the by-product is elemental iodine, as illustrated in Scheme 2.3.

I© I© /P\ ' /P\ ' e Q Ph2P PPh2 [X] Ph2P PPh2 [l] \^_y + 3/2 Ph2P PPh2 X + Ph2P PPh2 \^ y X = CI, Br XX X = l +

X X Q„.. \ I ' Q[X] 1/2 Ph2P© ©PPh2 e [X]

+2 1.3:[dppeBr2] +2 1.4: [dpppBr2] +2 1.5:[dppeCI2] +2 1.6:[dpppCI2]

Scheme 2.3. The differing behaviors observed for the reactions of phosphorus(III) halides with chelating bis(diphenylphosphino)alkane ligands. Only one of the potential halophosphonium products is included.

Unfortunately, when the analogous reactions of chelating diphosphines with PC13 or PBr3 are attempted in dichloromethane (also illustrated in Scheme 2.3), the unwanted and difficult to remove by-products are generated including: [dppe-Br2][Br]2, 2.9[Br]2;

[dppp-Br2][Br]2, 2.10[Br]2; [dppe-Cl2][Cl]2, 2.11[C1]2; and, [dppp-Cl2][Cl]2, 2.12[C1]2;.

The conventional explanation for the differences in the observed behavior is that the diphosphine reagents are acting as both ligands and reducing agents in the reactions involving the chloro- or bromophosphines, but only as ligands in the reaction with PI3 (at least when it is conducted in dichloromethane).15 We had postulated that an alternative but related interpretation for this observation is that the larger reduction potentials of intermediately-formed Cl2 and Br2 in comparison to that of I2 result in the rapid oxidation

46 of some of the diphosphine ligand and prevent their direct observation.19 Recently,

Dillon and Monks elucidated the mechanism of formation for cyclic triphosphenium cations (which is almost certainly also applicable to acyclic cations) and have suggested that the rate-determining step is the P(III) to P(I) RedOx process in which a formal equivalent of X2 (X = CI, Br) is removed from the intermediate salt and oxidizes the diphosphine. The reactions are observed to be more rapid for the bromophosphines than for the chlorophosphines; such behavior is consistent with both interpretations of the

RedOx process.

Regardless of the mechanistic details, the formation of by-products has rendered the isolation of chloride or bromide salts of triphosphenium cations difficult; thus, we sought a synthetic approach that would provide the desired salts with fewer or more- conveniently removable by-products. We reasoned that the ideal by-product for our purposes would be a neutral, relatively unreactive molecule that is either soluble in non- polar solvents or, more preferably, readily removed under reduced pressure so that it could be easily separated from the triphosphenium salts. Since we required a molecule that would have these properties and that would formally sequester one equivalent of X2

(X = CI, Br) from the reaction mixture, cyclohexene, which is a well-known reagent used for the removal of bromine used in organic syntheses,36 appeared to be a potentially reasonable candidate. Furthermore, preliminary 31P, !H and 13C NMR investigations revealed that cyclohexene does not react with any of the other reagents (PCI3, PBr3, dppe, dppp) so its suitability as a halogen-sequestering agent in this system was evaluated.

31P NMR spectra reveal that when a dichloromethane solution containing a

mixture of equimolar amounts of cyclohexene and PBr3 was mixed with a

47 dichloromethane solution containing one equivalent of either of the chelating diphosphines, the desired triphosphenium cation (2.7 or 2.8) is formed and the relative amount of oxidized product (2.9: 8 60 ppm or 2.10: 8 62 ppm) is reduced significantly

(ca. 40%) as assessed by integration of the corresponding signals in the spectra, as illustrated in Figure 2.3. Because some of the bromophosphonium by-product was still present, several syntheses were attempted with excess amounts of cyclohexene. As shown in Figure 2.3, when ca. 3 equivalents (or more) of cyclohexene are present in the reaction mixture, there are no longer any peaks attributable to the bromophosphonium by-product. lH NMR and GCMS experiments on sealed samples confirmed the formation of the expected trans- 1,2-dibromocyclohexane however, in practice, the by­ product may be conveniently removed with the other volatile components of the reaction mixture under reduced pressure.

48 (a) dppe + Br2

(b) 0 equiv.

JAA_

1.0 equiv. (c) 4 'X~*K*» •»i'Wwn»'i".nii"*inuWv»

3.0 equiv. (d)

50 -50 -100 -150 -200 ppm

31 Figure 2.3. P NMR spectra of the reactions of: (a) dppe and excess Br2; and PBr3 with dppp in the presence of (b) no cyclohexene, (c) one equivalent of cyclohexene and (d) 3 equivalents of cyclohexene. The signal corresponding to the dication 2.9 is indicated with an arrow. Relevant chemical shift assignments: ca. 60 ppm (s, [dppe-Br2]+2), 65 + + ppm (d, [(dppe)P] ), -230 ppm (t, [(dppe)P] ), ca. 34 ppm (s, dppe-02).

T1

While P NMR spectra suggest that the resultant solid is essentially pure, we have found that the product should be sonicated with THF prior to recrystallization in order to obtain the purest samples of the triphosphenium bromide salts. The slow evaporation of a solution of the cleaned solid in acetonitrile results in the production of crystalline salts characterized as the triphosphenium bromide salts 2.7[Br] and 2.8[Br].

The identities and purities of these materials have been confirmed in each case by multinuclear NMR experiments, HRMS experiments and microanalyses.

49 For the bromide salts 2.7[Br] and 2.8[Br], the recrystallization produced materials suitable for examination by single crystal X-ray diffraction. Details of the data collection and refinement for the structures are listed in Table 2.2 and the values of selected metrical parameters are collected in Table 2.3; the asymmetric unit of each salt is depicted in Figure 2.4. Perhaps not surprisingly, each of bromide salts is isostructural with the corresponding iodide salt19 and, consequently, the geometrical features of the cations in each structure fall in the ranges (P-P distances: 2.113(2) - 2.184(2) A) found in the Cambridge Structural Database37 for similar cyclic triphosphenium cations reported previously15'19'38"40 and do not mandate further discussion.

(a) (b)

Figure 2.4. Thermal ellipsoid plots (30% probability surface) of the contents of the asymmetric units of the P(I) bromide salts 2.7[Br] (a) and 2.8[Br] (b). Hydrogen atoms have been removed for clarity.

50 In contrast to the high-yield and clean synthesis observed for the bromide salts, we have found that the corresponding reactions involving PCI3 are somewhat more complicated. As in the case of the bromide analogues described above, 31P NMR experiments confirm the formation of the desired cation; however, the yields of the pure chloride salts are, in every instance, considerably lower than those of the bromide salts.

Furthermore, it must be noted that significant washing and fractional recrystallization are necessary to obtain chloride salts of sufficient purity for microanalysis and the presence of some P-containing by-products in the reaction mixtures suggested to us that a more in- depth examination of this system was mandated.

Our suspicions were confirmed upon examination of the crystal structure of a crystal obtained from the recrystallization of the crude reaction products. As illustrated in Figure 2.5, the salt that crystallized from the reaction mixture is not the anticipated chloride salt but is in fact the hydrogen dichloride salt [(dppp)P][HCi2], 2.8[HCi2]. The metrical parameters of the cation in the salt are indistinguishable from those found in other salts containing the cation and require no further comment. Similarly, the hydrogen dichloride anion, while relatively rare in terms of the number of examples (15) reported in the CSD, exhibits a Cl-Cl distance of 3.1161(11) A, Cl-H distances of 1.54(4) and

1.58(4) A, and a Cl-H-Cl angle of 178(3)°; all of these values are consistent with those that have been previously observed.

51 Figure 2.5. Thermal ellipsoid plot (30% probability surface) of the contents of the asymmetric unit of [(dppp)P][HCl2], 2.8[HC12]. The hydrogen atoms in the cation have been removed for clarity.

Table 2.4. Selected metrical parameters for compounds 2.7[Br], 2.8[Br] and 2.8[HCl2]. Distances are reported in Angstroms and angles in degrees; for 2.7[Br], the corresponding values for each independent cation in the asymmetric unit are reported separately.

[(dppe)P][Br] [(dppp)P][Br] [(dppp)P][HCl2] P1-P2 2.1231(9) [2.1216(9)] 2.1138(11) 2.1254(8) P2-P3 2.1308(9) [2.1270(9)] 2.1262(11) 2.1132(8) P1-C5 (or -C6) 1.818(2) [1.820(2)] 1.801(3) 1.811(2) P3-C4 1.817(2) [1.836(2)] 1.807(3) 1.814(2) C4-C5 1.525(3) [1.523(3)] 1.525(4) 1.531(3) C5-C6 1.538(4) 1.530(3) P1-P2-P3 88.37(3) [89.17(3)] 98.30(4) 96.70(3)

It should be mentioned that the hydrogen dichloride salt is obtained reproducibly even when after the thorough drying and redistillation of all of the reagents and solvents employed in the reaction. We also wish to note that whereas each of the pure halide salts appears to be stable indefinitely to atmospheric conditions in both the solid state and solution phases, dichloromethane or acetonitrile solutions containing mixtures of chloride

52 and hydrogen dichloride salts slowly decompose to generate an orange insoluble material.

Analytically pure samples of the chloride salts may be obtained by fractional crystallization of the crude products, however the need for such additional workup is less than ideal.

From the foregoing, it is clear that the nature of the products from the reactions outlined above appears to depend on the identity of the halogen used for the source of the

P(I) center. In an effort to understand the behavior observed, we undertook a series of further investigations. Analysis of the components present in the reaction mixtures using

GC-MS reveals that, in the cases where X = Br, the only bromine-containing product identified is trans- 1,2-dibromocyclohexane whereas when X = CI there are several chlorine-containing products including the expected trans-1,2-dichlorocyclohexane as well as 3-chlorocyclohexene. This latter product is instructive in that it is consistent with one of the products observed in the direct reaction of cyclohexene with CI2 41'42 and it produces HC1 as a by-product; this HC1 can presumably combine with chloride anions present in the reaction mixture to generate the observed [HC12]~ anions.

Because the presence of HC1 in the reaction mixture is the most likely source of the hydrogen dichloride anion, we reasoned that it might be possible to obtain the chloride salts in higher yield and purity through the use of a base to remove the HC1. To prove this hypothesis, a dilute solution of triethylamine in dichloromethane was added to the reaction mixture after 1 hour. After filtration, removal of volatile components, sonication and recrystallization, the resultant solid exhibited spectroscopic and physical properties identical to those observed for the analytically pure samples of the chloride salts 2.7[C1] and 2.8[C1], but the yield was not improved tremendously. Attempts at

53 neutralizing the HC1 with triethylamine during the reaction created additional P- containing by-products and were not pursued further.

Given that the halogenated products derived from the cyclohexene sequestering agent appeared to be consistent with those observed from the direct reaction of cyclohexene with elemental chlorine or bromine, we wished to ascertain whether the halogenated diphosphine by-products 2.9[A]2, 2.10[A]2, 2.11[A]2 and 2.12[A]2 (A = the appropriate halide or trihalide anions) also react with cyclohexene; i.e., these experiments were undertaken to assess whether the reaction may proceed through the initial oxidation of the diphosphine followed by reaction with cyclohexene or if such an intermediate is unnecessary. To this end, we prepared the halogenated compounds 2.9[A]2, 2.10[A]2,

2.11[A]2 and 2.12[A]2 through the reaction of either dppe or dppp with the elemental halogens. In the case of the brominated compounds, the materials appear to be obtained initially as the yellow-coloured tribromide salts (i.e. 2.9[Br3]2 and 2.10[Br3]2) when an excess of bromine is used; however, despite the possibility of their formation, we did not find any evidence of trichloride anions.

The treatment of the halogenated diphosphonium species 2.9[Br]2, 2.10[Br]2,

2.11 [Cl]2 and 2.12[C1]2 with cyclohexene does not change the appearance of the P

NMR spectra and suggests that these halophosphonium halide salts do not react with cyclohexene. In the case of the tribromide salts 2.9[Br3]2 and 2.10[Br3]2, we do observe the disappearance of the yellow colour upon addition of cyclohexene and we observe the formation of 1,2-dibromocyclohexane and obtain the corresponding bromide salts as the only phosphorus-containing product. Such behavior suggests that the halophosphonium salts are not effective halogenating agents for cyclohexene; however, the tribromide

54 anion is able to effect the halogenation (or alternatively, it suggests that cyclohexene may be a useful reagent for the conversion of tribromide salts to bromide salts).

Attempts to obtain satisfactory microanalysis for compounds 2.9[Br3]2 and

2.9[Br]2 have as of yet been unsuccessful. It should be noted "2.9[Br]2" is a well-known organic reagent, that is generally prepared and used in situ,25'43 for the conversion of alcohols into organo-bromides. The dication 2.9 is unambiguously identified by 31P, *H and C NMR experiments; however, the identities of the anions are less certain.

Therefore we undertook some experiments to corroborate our assignments. It should be noted that both of the compounds appear to decompose upon standing in inert atmosphere conditions either as solids or in solution and both salts react very rapidly with any ambient water. The addition of a large excess of water to the salts results in the

31 formation of Ph2P(0)CH2CH2P(0)Ph2 (dppe-02; P NMR 8 33 ppm). While 2.9[Br]2 is not crystalline, the exposure of NMR tubes containing freshly prepared and completely characterized solutions of the salt in CD2C12 to the moist ambient atmosphere results in the precipitation of colourless crystals suitable for analysis by X-ray crystallography.

The structure of the resultant salt, which has the composition [dppe(OH)2][Br]2 is shown in Figure 2.6 and is consistent with the formulation of the starting material as the bromide salt.

55 Figure 2.6. The molecular structure of [dppe(OH)2][Br]2.

In a similar vein, the exposure of NMR tubes containing freshly prepared and completely characterized solutions of 2.9[Br3]2 in CD2C12 to the moist ambient atmosphere produced orange crystalline material amenable to X-ray diffraction experiments. The resultant salt has the composition [dppe(0)(OH)][Br3] and the molecular structure is presented in Figure 2.7. Please note that two half-tribromide anions are found in the asymmetric unit (Br(l) and Br(3) are located on inversion centers).

56 Figure 2.7. The molecular structure of [dppe(0)(OH)][Br3]. Only half of each Br3 anion depicted in the figure is present in each asymmetric unit.

Table 2.5. Selected Metrical Parameters for compounds [dppe(OH)2][Br]2 and [dppe(0)(OH)][Br3]. Distances are reported in Angstroms and angles in degrees.

[dppe(OH)2][Br]2 [dppe(0)(OH)][Br3] P-OH 1.641(6) 1.5347(16)

P-0 1.4986(17)

These NMR scale experiments are reproducible, both of these products are consistent with the reported reactivity of the dication 2.9 and result from the formal elimination of HBr (2 equivalents for 2.9[Br]2 and 3 equivalents for 2.9[Br3]2). More importantly, both are consistent with the presence of the anions we proposed. While it remains possible that the as-prepared salt "2.9[Br3]2" might contain a mixture of [Br3]~

57 and [Br]" anions, it is clear that at least some tribromide anions are present. It should also be reemphasized that neither dppe nor Br2 reacts directly with water so the most plausible routes to the observed oxy-phosphonium salts is via 2.9[Br]2 and 2.9[Br3]2.

In light of the foregoing, it appears that dications such as 2.9, 2.10, 2.11 and 2.12 are not generated (even as intermediates) in the syntheses of the triphosphenium salts when sufficient cyclohexene is present. The appearance of some of the halogenated phosphonium species at lower concentrations of cyclohexene suggests that the diphosphine and the olefin are each oxidized readily and rapidly by either "free" X2 or, perhaps more likely, the corresponding [(dppe/p)PX] intermediate and that it is possible to preclude the formation of the P-containing by-product by simply decreasing the probability that a diphosphine vs. an olefin will be proximate to the oxidant upon its release of X2. In this light, the halogen sequestration strategy we have outlined above is best accomplished using an excess of cyclohexene.

For completeness, it should be noted that no reaction is observed when I2 is mixed with cyclohexene and, similarly, cyclohexene does not appear to sequester the I2 that is generated by the reaction of chelating diphosphines with PI3. Such behavior is as one would expect on the basis of the relative reduction potentials of I2 vs. Br2 vs. CI2 and/or on the basis of the relative strengths of the P-X bonds. The results of the reactions observed for the three different phosphorus(III) halides are summarized in Scheme 2.4.

58 © ~l Ph,P PPh, Ph2P PPh2 1© Ph P PPh U©[CIr ] 2 2 y ©r + xs o + XS o Ph2P PPh2 °[l] X=CI X' X

Ph2P PPh2 l» + CI CI X = Br + + XS o

+ 1© r / \- ' ew , * H-CI Ph2P' PPh2 [Br] Br Br

Scheme 2.4. Summary of the reactions of phosphorus(III) halides with chelating bis(diphenylphosphino)alkane ligands in the presence of excess cyclohexene.

2.4 Conclusions

A new convenient method for the synthesis of the bromide and chloride salts

2.7[X] and 2.8[X] has been developed. The new methodology appears to be general in that the addition of a halogen-scavenging reagent (in this case cyclohexene) facilitates the reduction of phosphorus trihalides by removal of the X2 by-product that is unavoidable.

Mechanistic studies indicate that an intermediate halo-oxidized diphosphine ligand is not being formed, because such compounds when made independently do not react with cyclohexene.

59 2.5 References

1. Parkin, G. J. Chem. Educ. 2006, 83, 791 -799. 2. Macdonald, C. L. B.; Ellis, B. D. In Encyclopedia of Inorganic Chemistry; 2nd ed.; King, R. B., Ed.; John Wiley and Sons, Inc.: Hoboken, N.J., 2005. 3. Ellis, B. D.; Macdonald, C. L. B. Coord. Chem. Rev. 2007, 251, 936-973. 4. Lammertsma, K. Top. Curr. Chem. 2003, 229, 95-119. 5. Lammertsma, K.; Vlaar, M. J. M. Eur. J. Org. Chem. 2002, 1127-1138. 6. Mathey, F. Angew. Chem., Int. Ed. Engl. 1987, 26, 275-286. 7. Mathey, F. Dalton Trans. 2007, 1861-1868. 8. Schmidpeter, A. Heteroatom Chem. 1999,10, 529-537. 9. Schmidbaur, H. Angew. Chem., Int. Ed. Engl. 1983, 22, 907-927. 10. Frenking, G.; Neumiiller, B.; Petz, W.; Tonner, R.; Oxler, F. Angew. Chem. Int. Ed. 2007, 46, 2986-2987. 11. Schmidbaur, H. Angew. Chem. Int. Ed. 2007, 46, 2984-2985. 12. Tonner, R.; Oxler, F.; Neumiiller, B.; Petz, W.; Frenking, G. Angew. Chem. Int. Ed. 2006, 45, 8038-8042. 13. Dillon, K. B.; Mathey, F.; Nixon, J. F. Phosphorus: The Carbon Copy; John Wiley & Sons: Chichester, 1998. 14. Barnham, R. J.; Deng, R. M. K.; Dillon, K. B.; Goeta, A. E.; Howard, J. A. K.; Puschmann, H. Heteroatom Chem. 2001,12, 501-510. 15. Boon, J. A.; Byers, H. L.; Dillon, K. B.; Goeta, A. E.; Longbottom, D. A. Heteroatom Chem. 2000,11, 226-231. 16. Dillon, K. B.; Monks, P. K.; Olivey, R. J.; Karsch, H. H. Heteroatom Chem. 2004, 75, 464-467. 17. Kilian, P.; Slawin, A. M. Z.; Woollins, J. D. Dalton Trans. 2006, 2175-2183. 18. Ellis, B. D.; Carlesimo, M.; Macdonald, C. L. B. Chem. Commun. 2003, 1946- 1947. 19. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2006, 45, 6864-6874. 20. Driess, M.; Ackermann, H.; Aust, J.; Merz, K.; Von Wullen, C. Angew. Chem., Int. Ed. 2002, 41,450-453. 21. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2004,43, 5981-5986.

60 22. Ellis, B. D.; Macdonald, C. L. B. Phosphorus, Sulfur Silicon Relat. Elem. 2004, 179, 775-778. 23. Dillon, K. B.; Monks, P. K. Dalton Trans. 2007, 1420-1424. 24. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996,15, 1518-1520. 25. Schmidt, S. P.; Brooks, D. W. Tetrahedron Lett. 1987, 28,161-16%. 26. ; Bruker AXS Inc.: Madison, WI, 2001. 27. ; Bruker AXS Inc.: Madison, WI, 2001. 28. ; Bruker AXS Inc.: Madison, WI, 2001. 29. Altomare, A.; Burla, M. C; Camalli, M.; Cascarano, G.; Giacovazzo, C; Guagliardi, A.; G., M. A. G.; Polidori, G.; Spagna, R.; CNR-IRMEC: Bari, 1997. 30. Sheldrick, G. M.; Universitat Gottingen: Gottingen, 1997. 31. Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837-838. 32. Sheldrick, G. M.; Bruker AXS Inc.: Madison, WI, 2001. 33. Schmidpeter, A.; Lochschmidt, S.; Sheldrick, W. S. Angew. Chem., Int. Ed. Engl. 1985, 24, 226-227. 34. Schmidpeter, A.; Lochschmidt, S. Angew. Chem., Int. Ed. Engl. 1986, 25, 253- 254. 35. Ellis, B. D.; Macdonald, C. L. B. Acta Crystallogr., Sect. E: Struct. Rep. Online 2006, 62, M1869-M1870. 36. March, J., Smith, M. B. March's Advanced Organic Chemistry: Reactions, Mechanisms, and Structure; 6th ed.; John Wiley & Sons, Inc.: Hoboken, New Jersey, 2007. 37. Allen, F. H. Acta Crystallogr., Sect. B: Struct. Sci. 2002, B58, 380-388. 38. Schmidpeter, A.; Lochschmidt, S.; Sheldrick, W. S. Angew. Chem., Int. Ed. Engl. 1982, 21, 63-64. 39. Ellis, B. D.; Macdonald, C. L. B. Acta Crystallogr., Sect. E: Struct. Rep. Online 2006, E62, ml235-ml236. 40. Deng, R. M. K.; Goeta, A. E.; Dillon, K. B.; Thompson, A. L. Acta Crystallogr., Sect. E: Struct. Rep. Online 2005, E61, m296-m298. 41. Poutsma, M. L. J. Am. Chem. Soc. 1963, 85, 3511-3512.

61 42. Poutsma, M. L. J. Am. Chem. Soc. 1965, 87, 4285-4293. 43. Schmidt, S. P.; Brooks, D. W. Tetrahedron Lett. 1987, 28, 767-768. Chapter 3 - Small Molecular Targets as Monomers for Polymerization

3.1 Introduction

Polymers containing inorganic elements have long been known to possess interesting properties, which have made them desirable to study from an academic perspective as well as using them industrially because of these exceptional attributes.

The major categories of inorganic polymers are a) polyphosphazenes, which are comprised of a phosphorus-nitrogen backbone; b) polysiloxanes, with a silicon-oxygen backbone; and, c) polysilanes, which are analogues of many organic polymers.1

The most prevalent group of phosphorus-containing polymers is the polyphosphazenes. Several classes of relatively small phosphazene rings (such as 3.1 and

3.2) can be thermally polymerized to generate polyphosphazenes, which are an industrially important class of polymers. As illustrated in Scheme 3.1, the formal monomeric unit of these trimers or tetramers is analogous to nitriles, which is the reason these molecules were first referred to as phosphonitriles.

CI CI

ClOp-N^p/_C| // \ CI CI N 3.2 \ /

C2H2CI4 or nPCI + nNH CI *• i| CI CI 5 4 Cl-P< ,P-CI C6H5CI I N I CI CI CI 3.1 I -P=N^ 3.3 I CI n

Scheme 3.1. General synthesis of some chlorinated phosphazenes. The major product is the six-membered ring 3.1.

63 The most common phosphazene precursors, (NPCl2)n, where n = 3 (3.1) or 4

(3.2), are generated through the reaction of PCI5 with NH4CI in a chlorinated solvent.

The chlorine atoms can be easily substituted with a variety of nucleophiles. This substitution can occur prior to polymerization or more often following polymerization and mixed-substituent polyphosphazenes may be obtained by careful control of the stoichiometry of the substitution reaction.

-P=N- CI CI I \ / CI R N^ N A I i| *- -P=N- Cl-P^. ^P-CI I INI CI 1. NaR CI CI ZNaR' R -2NaCI I 3.1 -P=N- I R'

Scheme 3.2. General synthesis of a perchlorinated polyphosphazene from perchlorophosphazene 3.1. Both homosubstituted and mixed substituted polyphosphazenes are shown.

Because of the tunability of the R-groups attached to the phosphorus centers, the polymers that are generated can have a wide variety of properties, which makes these polymers very important industrially. Several examples of the use of polyphosphazenes include inorganic rubber, flame retardant materials, as well as biomaterials for bone regeneration. The backbone of these polymers (P-N chain) provides a number of desirable attributes such as flexibility, fire resistance, and chemical and thermal stability.

The side groups of the polymers (attached to the phosphorus atoms) can control such properties as solubility, biological activities as well as electrical and optical properties.

64 Such tunability is achieved through very simple and mild reaction conditions, which is one of the reasons why these polymers are very well studied and exploited in industry.1

Gates and co-workers are currently developing other phosphorus-containing polymers that exploit the reactivity patterns of phosphorus, which are often similar to those of organic carbon-containing analogues.2 For example, they have attempted to mirror olefin polymerization by developing polymerization techniques using phosphaalkene monomers.3 They have successfully generated poly(methylenephosphine), which contains P(III) atoms (Scheme 3.3).

Ph Ph [hit] .. I /P~CN +P—C- Mes Ph 150°C I I Mes Ph n

Scheme 3.3. Phosphorus containing polymers made by the addition polymerization of phosphaalkenes, which is analogous to one of the methods used for olefin polymerization.

Manners and co-workers have shown that it is possible to generate phosphazenes and the corresponding polyphosphazenes in which one of the PR2 fragments has been

4 replaced with isolobal fragments such as CR, S, SO, S02, etc. In contrast, there have been almost no reports of analogous systems in which the N groups have been replaced and no reports of any general synthetic approaches that may be used to make either the precursors to polymers of such compounds. It must be noted that the N atoms in phosphazenes are isovalent with the P(I) centers we have investigated. Given that our previous work and that of others in P(I) chemistry suggested several possible approaches to this new class of compounds and related species, we decided to pursue this goal.

65 Because of the incredible diversity of physical properties already demonstrated by polyphosphazenes, we believe that polymers containing P(I) atoms would possess very interesting properties. Based on our previous work into the synthesis of very stable P(I) containing salts,5 we believe that this synthetic goal is feasible.

Since ligand substitution reactions are a well-established method for the exchange of the phosphine ligands supporting the P(I) centers in triphosphenium species 3.4,' our group's first attempts at generating oligomeric or polymeric materials focused on the use of bis(phosphinobenzene) ligands. When the phosphorus substituents are placed in the meta or para positions of the aromatic ring, the diphosphine cannot act as a chelating ligand; thus, each P(I) center must be ligated by two separate bis(phosphinobenzene) ligands after a ligand exchange reaction with the triphosphenium salts previously mentioned.

Ph2R PPh, \ / p® 3.4 0 R2 + -dppe m

PR P°2

PR2

Scheme 3.4. Proposed synthetic route to poly[bis(phosphino)benzeneP(I)] oligomers or polymers through ligand exchange with the P(I) salt 3.4.

When the diphosphinobenzenes were used as the linkers, this methodology was successful in liberating dppe during the ligand exchange reaction, however the obtained material was very difficult to dissolve in organic solvents making it difficult to

66 O 1 characterize completely. Numerous solution P NMR experiments revealed the formation of the anticipated P-P-P fragments (for a variety of non-chelating diphosphines) and solid-state P NMR spectra confirmed that such fragments were also present in the insoluble solid material. In the case of the ligand p-(Et20P)2C6H4 a soluble salt of the dimeric dication was obtained, and low-quality crystals suitable for X-ray diffraction were isolated to confirm the connectivity of the structure (shown in Figure

3.1).8

Figure 3.1. Solid-state structure of the dimer [(/KEtC^P^CeH^PMSnCy.

Because the insolubility of the materials containing bisphosphinobenezenes in organic solvents is likely related to the highly charged nature of the species, the next approach was to develop an anionic bis-phosphine ligand in order to generate an overall neutral polymeric material, which could theoretically be more soluble in common organic solvents. We reasoned that the candidates for anionic bis-phosphine ligands most similar to the benzene-linked diphosphines would contain aromatic anionic cyclopentadienyl rings with the two phosphine ligands arranged in a 1,3-substitution pattern.

67 Preliminary attempts of the reaction of these anionic ligands with [dppeP][I] provided positive and promising results. In particular, the 31P{ 'H} NMR spectra contain distinct doublet and triplet signals indicative of the formation of the desired triphosphenium fragments. The signals corresponding to the [dppeP][I] starting material are not visible. This suggests that synthesis of the desired oligomer is successful - further investigations are currently underway.9

The other major approach for the generation of such phosphorus-containing polymers is through the synthesis of small cyclic molecules that have the potential to polymerize under thermal or catalytic polymerization conditions. Herein is a discussion into the design and initial synthetic results for these small molecules.

3.2 Experimental

Reagents and General Procedures. All manipulations were carried out using standard inert atmosphere techniques. Phosphorus (III) bromide, phosphorus (III) chloride, diphenylphosphine, dicyclohexylphosphine, and dicyclohexylchlorophosphine were purchased from Strem Chemicals Inc., and all other chemicals and reagents were obtained from Aldrich. Phosphorus (III) bromide and phosphorus (III) chloride were distilled before use, and all other reagents were used without further purification. The compound [dppeP][Br] was synthesized according to procedures described in the previous chapter. All solvents were dried on a series of Grubbs'-type columns10 and were degassed prior to use. CDCI3 was dried over calcium hydride. C6D6 was dried over

Na/benzophenone. Dry THF used for 31P NMR experiments was degassed prior to use.

68 Instrumentation. NMR spectra were recorded at room temperature in THF, CDCI3 or

CeD6 solutions on a Bruker Avance 300 MHz spectrometer. Chemical shifts are reported

! 13 31 in ppm, relative to external standards (SiMe4 for H and C NMR, 85% H3PO4 for P

NMR). Coupling constant magnitudes, \j\, are given in Hz. Melting point (Mp) or decomposition points (Dp) were obtained on samples sealed in glass capillaries under dry

N2 using an Electrothermal melting point apparatus.

Preparation of NH(PPh2)2, 3.5. Hexamethyldisilazane (506 mg, 0.77 mmol) was added dropwise to a solution of chlorodiphenylphospine (1.37 g, 6.20 mmol) in Et20 (15 mL).

Upon addition, a significant amount of a white precipitate was formed, and therefore more Et20 (15 mL) was added to ensure adequate stirring. The solution was refluxed for

3 hours, after which the mixture was allowed to cool and the volatile components were removed under reduced pressure. The resulting white solid was collected. Yield: 94%

(1.13 g, 2.93 mmol). 3lP{lU} NMR: 43.21 (s). Mp: 143-146 °C (lit. 144.5-146.5°C).n

Preparation of NH(PCy2)2 3.6. Hexamethyldisilazane (320 |iL, 1.53 mmol) was added dropwise to a solution of chlorodiphenylphospine (700 mg, 3.0 mmol) in Et20 (30 mL).

The solution was refluxed for 2 days, after which the mixture was allowed to cool and the volatile components were removed under reduced pressure. The resulting white fluffy solid was collected. No yield was determined due to the always present impurities as ascertained by 31P NMR. 31P{!H} NMR: 63.4 (s).

69 Preparation of [N-PPh2-P-PPh2]2 3.7.

(A) A solution of 3.5 (380 mg, 0.99 mmol) and [dppeP][Br] (500 mg, 0.98 mmol) in THF

(30mL) was cooled to -78°C before dropwise addition of one equivalent of BuLi (0.61 mL, 1.6 M). The reaction mixture was allowed to warm to room temperature, during which time a bright yellow colour developed. After allowing the mixture to stir at room temperature for 2 hr, the white precipitate was filtered, followed by removal of all volatile components from the yellow filtrate under reduced pressure. No yield was determined due to the always present dppe impurity as ascertained by 31P NMR.

(B) Et3N (150 |j,L, 1.08 mmol) was added dropwise to a solution of 3.5 (320 mg, 0.79 mmol) and [dppeP][Br] (400 mg, 0.79 mmol) in CH2C12 (30 mL). The resulting yellow reaction mixture was allowed to stir at room temperature for 2 hr. The white precipitate was filtered, followed by removal of all volatile components from the yellow filtrate under reduced pressure. No yield was determined due to the always-present dppe impurity as ascertained by 31P NMR. ^P^H} NMR: -139.72 (t, VPP = 427, 2P), 36.49

(d, VpP = 427, 4P). Mp: 218-221 °C.

General Preparation of [K][PR2]. To a solution of KH (1 equiv.) at 0 °C in THF was added the appropriate dialkylphosphine (1 equiv.). The reaction mixture was allowed to warm to room temperature. When R=Ph (3.8) the reaction mixture was stirred for several hours before the volatile components were removed under reduced pressure. When

R=Cy (3.9) the reaction mixture was refluxed overnight, after which the solution was cooled to room temperature and the volatile components were removed under reduced pressure.

70 31 [K][PPh2] [K]3.8. P{'H} NMR: 21.6 (THF)

31 ! [K][PCy2] [K]3.9. P{ H} NMR: 27.2 (THF)

Reaction of [K][PCy2] with [dppeP][Br]. A solution of [K]3.9 (256 mg, 1.08 mmol) in

THF was added to a solution of [dppeP][Br] (275 mg, 0.54 mmol) in THF (total 30 mL) and an orange colour was observed. The reaction mixture was allowed to stir overnight, after which the volatile components were removed under reduced pressure. No yield was determined. 31P{*H} NMR: -27.21 (s), -12.0 (s), -10.0 (s), 29.2 (s).

Reaction of [K][PCy2] with PBr3. To a solution of PBr3 ((57.3 mg, 0.21 mmol) in THF was added a solution of [K]3.9 (150 mg, 0.64 mmol) in THF (20 mL total) at -78 °C. The reaction mixture was allowed to warm to room temperature and was left to stir overnight.

The yellow solution with brown precipitate was concentrated under reduced pressure, and the mixture was obtained as a viscous brown oil. No yield was determined. 31P{1H}

NMR: -137.4 (m), -27.1 (s), -20.8 (s), -14.2 (d, VPP = 203), -6.9 (s), 26.36 (m), 45.6 (s).

Reaction of [K][PCy2] with PBr3 and Cyclohexene. To a solution of PBr3 (114 mg,

0.42 mmol) and cyclohexene (129 |iL, 1.27 mmol) in THF was added a solution of

[K]3.9 (200 mg, 0.8 5mmol) in THF (25 mL total). The reaction mixture was a cloudy yellow solution and was allowed to stir at room temperature for several hours. The volatile components were removed under reduced pressure, and the resulting solid is orange in colour. No yield was determined. ^P^H} NMR: -25.5 (s), -11.6 (s), 1.3 (s),

6.6 (s), 8.2 (s), 60.0 (s), 65.6 (s).

71 Reaction of [dppeP][Br] with [Li][PlBu2]. To a slurry of [dppeP][Br] (530 mg, 1.04 mmol) in THF was added a solution of [Li][P'Bu2] (320 mg, 2.10 mmol) in THF (30 mL total). The reaction mixture immediately lost the yellow colour of the dissolved phosphide and proceeded to turn brown, then orange and a white precipitate was formed.

The reaction mixture was allowed to stir overnight at room temperature, after which the white precipitate was filtered and the orange filtrate was concentrated under reduced pressure to afford a thick orange oil. No yield was determined. 3lP{lH] NMR: 237.4 (t,

l VpP = 446Hz), -124.8 (m), -14.8, -12.0, 20.7, 35.8, 57.6 (d, JPP = 397Hz), 67.9 (d, VPP =

455Hz), 108.5 (d, VPP = 445Hz).

X-ray Crystallography.

Each crystal was covered in Nujol and placed rapidly into the cold N2 stream of the Kryo-

Flex low temperature device. The data were collected using the SMART12 software on a

Bruker APEX CCD diffractometer using a graphite monochromator with Mo Koc radiation (k = 0.71073 A). A hemisphere of data was collected using counting times of

10-30 seconds per frame. The data were collected at -100 °C. Details of crystal data, data collection and structure refinement are listed in Table 3.1 and selected metrical parameters are compiled in Table 3.2. Data reductions were performed using the

SAESfT13 software and the data were corrected for absorption using SADABS.1 The structures were solved by direct methods using SIR9715 and refined by full-matrix least- squares on F2 with anisotropic displacement parameters for the non-H atoms using

SHELXL-9716 and the WinGX17 software package. Details of the final structure

72 solutions were evaluated using PLATON and thermal ellipsoid plots were produced using

SHELXTL.18 3.1. Summary of X-ray crystallographic data for compounds 3.7.

Compound [N-PPh2-P-PPh2]2 Empirical formula C48H40N2P6 Formula weight 1830.46 Temperature (K) 173(2) Wavelength (A) 0.71073 Habit, Colour Prism, Colourless Crystal system Monoclinic Space group P2/c Unit cell dimensions: A (A) 12.153(2) B{k) 9.3574(16) C(A) 23.480(3) oc(°) 90 P(°) 121.056(6) 7(°) 90 Volume (A3) 2287.4(6) 2 Density (calculated) 1.329 Absorption coefficient (mm1) 0.389 F(000) 948 0 range for data collection (°) 1.96 to 27.50 Limiting indices -152a(I)] Rl = 0.0400, wR2 = 0.1052 R indices (all data) Rl = 0.0452 wR2 = 0.1138 Largest difference map Peak 0.556 and -0.302 and hole (e A"3) a 2 /?l(F) = Z(\F0\ - \FC\)/Z\F0\} for reflections with F0 > 4(a(F0)). wR2(F ) = {Zw(\F0f 2 2 2 2 1/2 |FC| ) /EM;(|FJ ) } where w is the weight given each reflection. 3.3 Results and Discussion

The conceptual design for the small molecules that can hopefully be used as monomers for polymeric material containing P(I) centers is multifaceted. Even though these molecules are relatively simple in appearance because of their small size and relatively uncomplicated structures, the synthetic approaches to making them can be difficult and complex. Each of the small cyclic molecules is designed with some degree of ring strain incorporated, to facilitate ring-opening polymerization. Another important feature of the desired precursor molecules is their overall charge; having small neutral molecules helps to increase solubility, processability and ease of characterization of the resulting polymeric material. Figure 3.2 is an inventory of the small cyclic target molecules.

p*G 'p'G p0 ©/' *\e ®/' "\© ©/' *\© R2R. ,PR2 RaP\ -PR2 R2^\ .PR2

Ce N P

R" R' R*

3.11 3.12 3.13

p© p© ©/" \© ©/' \© R2R. PR2 R2P\ PR2

N© Pe 3.14 3.15

Figure 3.2. Inventory of small molecular targets designed to be monomers for polymeric material.

In certain cases, some examples of the target molecules have been reported; however, none of these were prepared in a convenient, reproducible or more readily

75 generalized fashion. For example, one of the molecular inspirations (3.11) is based on the 4-membered triphosphenium containing species made by the Karsch group.19 Another important model for one of the target small molecules, (3.15) was reported by Frank and co-workers in 1996; this four-membered phosphorus ring is perhaps the most important of all the target molecules in that the resulting polymer would be a continuous chain of phosphorus atoms with alternating coordination numbers of two and four. These polymers could also be considered to be the all-phosphorus analogue of polyphosphazenes. Such a polymer would also be the valence isomer of the oligomeric polyphosphine species of the general form (PR)n, which have been studied extensively.21'22

It should be noted that the synthetic route to the four-membered phosphorus ring reported by Frank's group is actually quite unexpected and clearly serendipitous. By treating a cyclic bis(amino)chlorophosphine with lithium metal, a P-P linked dimer is formed along with the secondary "tetraphosphete" product 3.16. This four-membered phosphorus ring contains two "linked" triphosphenium fragments and has a neutral charge overall. The crystal structure of 3.16 exhibits P-P bonds that are slightly smaller than typical P-P single bonds and the angle at the triphosphenium P(I) center is 79.4°, which relates well to the other known triphosphenium four-membered ring examples.

This key example may provide structural insight into the design of our target molecules.

The large substituents on the chelating phosphines are certainly stabilizing the small four- membered phosphorus ring, and this information could aid in ligand design for our

synthetic approaches. Also, the electronic properties of the amido-substituents on the

four-coordinate phosphorus atoms may be important for the stability of the compound.

76 We have previously reported that calculations show that these bis-amido phosphorus ligands can have a profound effect on the reactivity of a triphosphenium species.

'Bu lBu 'Bu Me I 4Li, A Me. N N /Me \, \; 2 /Siv ^P-P\ Me N CI -4LiCI Me N ^N Me

'Bu 'Bu Bu

4Li, A t t -Li4[N(Bu)Si(Me)2N(Bu)]2

'Bu 'Bu 'Bu

dimerization Me. >i^ 'LPs. PC :SL Si >=P: X V Me N^ VP

3.16

Scheme 3.5. Frank's synthesis of a four-membered all-phosphorus ring (3.16) with alternating two-coordinate and four-coordinate phosphorus atoms.

The third example of inspiration is from Schmidpeter with the synthesis of an eight-membered ring that is essentially a dimeric version of the desired four-membered nitrogen-backed ring 3.14.24 The synthesis of this target makes use of P4, an elemental form of phosphorus that is extremely pyrophoric. By treating the elemental phosphorus with a P-N-P type ligand, the eight-membered ring was formed, recrystallized and characterized. 25

77 Ph P P 2 Li© Ph2P^PPh2 V= \ Ph © N' p-Ph + -LiPx-i ph-p N Ph P=P- x/2 P,. ' PDnh

3.7

Scheme 3.6. Synthesis of the eight-membered ring analogue (3.7) of the target molecule 3.14 by Schmidpeter.

The first of these potential target molecules to be investigated was the nitrogen- backed molecule 3.14 because of simple synthetic procedures to obtain the P-N-P backbone.11 The proposed route after the synthesis of the P-N-P backbone was to employ ligand exchange with the P(I) salts previously mentioned as previous work has suggested that such ligand exchange reactions are generally straightforward.

Attempts to isolate the Af-alkyl substituted four-membered ring (3.12) resulted in pale yellow crystals. The structure was revealed to be an eight-membered ring, which is

24 the same molecule that Schmidpeter has previously reported.

78 Figure 3.3. Solid-state structure of eight-membered ring 3.7. Hydrogen atoms have been removed for clarity.

The solid-state structure of 3.7 is interesting and worthy of discussion. The eight- membered ring has a boat-boat conformation26 with the dicoordinate phosphorus atoms pointing in the same direction. The angle at the dicoordinate phosphorus is quite small at

95.44(3)°, however this agrees well with previously reported triphosphenium P-P-P angles.25 This small angle consequently requires the angle at the nitrogen to be quite large at 130.94°(9). The bond lengths within the eight-membered ring are all intermediate between typical P-N and P-P single and double bonds, indicating some degree of electron derealization.

79 Table 3.2. Selected metrical parameters for compound 3.7. Distances are reported in o Angstroms and angles in degrees; the corresponding values for each associated bond length are reported separately. Angle Bond Length

P-P-P 95.44(3) P-P 2.1390(6) 2.1310(6) P-N-P 130.94(9) P-N 1.5955(14) 1.6023(15)

An attempt to synthesize the compound in a more controlled fashion was achieved by generating the lithium amide salt in situ followed by addition of the P(I) salt 3.4[Br].

After allowing the reaction to warm to room temperature and stirring for an adequate amount of time, lithium bromide was filtered off and the remaining mixture contained both the desired eight-membered ring, 3.7, as well as the liberated dppe, as determined by

31P NMR experiments. Removal of the dppe is somewhat difficult and has yet to be accomplished completely because of the similar solubilities of the two molecules.

An even milder synthesis was developed using NEt3 in which and no cooling is required during the addition of the base. Furthermore, while the solvent used for the reaction using BuLi was THF (in which the P(I) salts are only slightly soluble), the reaction using NE13 may be done in dichloromethane, which completely dissolves all of the reagents.

80 * \ I *• -dppe, -LiBr Ph Ph ' 3.4[Br] or v N P-Ph r ii ii Et3N, CH2CI2 Ph-Pv N Ph,P-. ^PPh, -dppe 2 N ' -[NHEt3][Br] Ph H J 3.7

Scheme 3.7. The two syntheses developed to generate the eight-membered analogue of one of the four-membered ring target molecules.

Syntheses of analogous molecules with alkyl-phospino substituents were attempted. Unfortunately, the desired P-N-P type ligands were much more difficult to isolate than the phenyl-phosphino substituted ligand. The only P-N-P type ligand to be synthesized with alkyl substituents was the cyclohexyl phosphine substituted bis- phosphino amine. The reaction between this white powder and 3.4[Br] was attempted

31 with either Et3N or BuLi as bases. The P NMR spectra of both reactions indicated that

31 dppe was indeed liberated from 3.4. For the reaction with Et3N, the P NMR spectrum of the reaction mixture also revealed a large number of phosphorus-containing products, none of which could be positively identified. For the reaction using BuLi as the base, the two other phosphorus-containing products observed (besides dppe) were oxidized dppe and the diphosphane Cy2P-PCy2. The lack of convenient and reliable synthetic routes to other P-N-P type ligands has hindered our development into a series of compounds related to 3.7.

As previously mentioned, the most desirable compound is actually the all- phosphorus containing four (3.15) or eight membered ring (3.17). Given that Frank's synthesis is somewhat unexpected, an improved synthetic methodology needed to be

81 developed. Based on the methods described above employed to generate the PN- containing rings, a P-P-P anionic fragment was envisioned to react with the P(I) salts to generate the analogous all-phosphorus containing rings.

R P=P. pe 2 \ ffi/r\e PR, R2PX /PR2 R R Pe 2 \ P=PR2

3.15 3.17

Figure 3.4. The all-phosphorus target molecule 3.15 and the dimeric analogue 3.17.

The first attempt at these anionic P-P-P type ligands resulted in a harsh reduction- type reaction of PBr3 with potassium phosphides. The phosphides were generated using dialkylphosphines (PHR2) and a variety of bases including KH, and KOlBu under standard procedures.

K [dppeP][Br] R.p.R dimerization ©, R R PR ~Pe -KBr, -dppe R2Pv 2 p P R=Ph 3.8 • • © 3.1S R=Cy 3.9 3.18

1 /2 PX3 + 3.4[Br] ( 0) X i 1) L1AIH4 R-p-p^p-R. I I 2) strong base i i R R R R 3.19 3.20 Scheme 3.8. Proposed synthetic route to 4-membered phosphorus ring 3.15.

The phosphides 3.8 and 3.9 were treated with the P(I) salts in an attempt to create a transient phosphinidene species 3.18, which was earlier proposed by Frank in their

82 four-membered phosphorus ring synthesis, and this transient species would hopefully undergo dimerization to form the target molecule 3.15. A P NMR experiment performed on the crude reaction mixture indicates that there was indeed liberation of dppe from the P(I) salt in addition to some of the starting phosphide. The only other signal from a phosphorus-containing product is observed at 29.2 ppm as a singlet; signals at such chemical shifts are often observed when the phosphine ligand has become oxidized. The 31P NMR spectrum suggests that the phosphides are capable of displacing the dppe ligand from the P(I) salt; however, the phosphides appear to be too reactive for the somewhat sensitive P(I) fragment, and some degree of decomposition is occurring.

The phosphides were also treated with PBr3, both with and without cyclohexene added to mop up any reactive X2 that would be produced during the reaction, with the goal of generating the bisphosphinophosphorus halide 3.19. The reaction was designed to convert the bisphophinophosphorus halide into the corresponding secondary phosphines, which would be followed by deprotonation with a strong base to generate the phosphide 3.20. This P-P-P ligand, analogous to the P-N-P ligands, would be mixed with

3.4[Br] to generate the four-membered (3.15), or the dimeric eight-membered ring (3.17)

(Scheme 3.8). It is anticipated that the P-P-P ligand would react in an analogous fashion to the P-N-P ligands.

The reaction of PBr3 and [K]3.9 with cyclohexene present, resulted in a mixture of phosphorus-containing products as discerned by the 31P{1H}-NMR spectrum of the crude reaction material. If the desired P-P-P fragment (3.19 or 3.20) were formed, we would expect to observe two signals; a doublet (attributed to the phosphide ligands) and a triplet (attributed to the dicoordinate phosphorus center). Unfortunately, the only signals

83 observed in the 31P NMR spectrum are singlets. The strongest signal has a chemical shift of 60 ppm, and is consistent with diphosphine compounds of the formula R2P-PR2. This assignment of the signal at 60 ppm was confirmed upon isolation of a single crystal of

CV2P-PCV2 from a reaction mixture with the lithium phosphide, [Li]3.9 with PBr3 and cyclohexene. The P NMR spectrum for this reaction also has a strong signal at 58.2 ppm, which is assigned to the diphosphine.

Schmidpeter has already commented on the plausibility of these phosphorus polymers featuring alternating dicoordinate and tetracoordinate phosphorus within the original report of the crystal structure of compound 3.7.26 He suggested that the putative oligomer of this type would decompose into a diphosphine (R2P-PR2) and elemental phosphorus (Px-i )• The basis of this suggestion can be attributed to their previous observation of the formation of the diphosphine (R2P-PR2) when a phosphide is treated with P4 and on the basis of overly simplistic bond energy considerations. However, it should be noted that when the stoichiometry is carefully controlled, the reaction produces a P-P-P type anion, which contains the desired triphosphenium moiety (Scheme 3.9). In these reports, the highly reactive and uncontrolled P4 is used as a source of (P+). With our proposed syntheses, the more reliable and controllable P(I) source would be used, eliminating potential problems associated with by-products.

84 •p=\ R P P—R || || «• 2Ph2P—PPh2 + P4 R-PN P

R

R R 2 ^P' + (n+1)/4P4 - R2P^ PR2 + © e ©

R R 2 ^p' + n/2 P4 - R2P-PR2 + 2P? e

Scheme 3.9. Hypothesis reported by Schmidpeter et al., that the dimeric analogue of 25 3.15 is inaccessible by route of treating P4 with a phosphide.

Interestingly, the reaction of PBr3 and [K]3.9 without cyclohexene provides more insightful results that suggest the target molecule is not entirely out of reach. The 31P

NMR spectrum of the crude reaction material has a large signal due to excess phosphide

3.9; however, several small signals can most likely be assigned as triplets. These possible triplets have chemicals shifts of -137.4 and 26.4 ppm. Also discernable is a doublet with a chemical shift of -14.2 ppm. Unfortunately, the coupling constants of the three signals cannot be assigned to any other signal, causing assignment of the peaks to be virtually impossible. Recrystallization of this crude reaction mixture was attempted; however, suitable crystals for X-ray diffraction studies were not grown.

85 a) LiPCy2 + PBr3 + cyclohexene

wm0^ ^ff^y^#WM^^M^yw^

b) KPCy2 + PBr3 + cyclohexene

1 1 V^F«w^^pn*nwflir\Kn^T fw W*llW*iWW>V»lM»»*WVJ JIWflywM i4W LiMA^.^, ^ *,>»., H^>Nw>.^v»*AVM^Mi»*iii^.<^it>i.»iMV

c) KPCy2 + PBr3

I i '—r "l I I I | I I I I | I I I I | I I I I | I 100 50 0 -50 -100 ppm

Figure 3.5. 31r.rP{ hH} NMR spectra of several reactions attempted to product a P-P-P type ligand. For spectra a) and b) the strong peak ~60ppm is indicative of the diphosphine Cy2P-PCy2 which is surprisingly absent from spectrum c).

Although some positive preliminary results were obtained from these phosphide reactions, no definitive conclusions can be drawn indicating that a different synthetic approach to our target molecules is necessary. Alternative synthetic strategies have been proposed as shown in Scheme 3.10.

86 %-R

SiMe-,

1 /2 PCI3 ^ R2P\p^PR2 1 LiAIH4 -SiMe3CI

CI 2. KH K© G

3.4[Br] ©/ \© R R. PR y -dppe 2 2 '0 Me SL ^SiMe 3 3 2R PCI R2P^ ^PR2 > 2 3.15 -LiCI ^> Li© G I -SiMe3CI SiMe3 , J BuLi

Me3Si. .SiMe3

SiMe3

Scheme 3.10. Alternative syntheses to achieve four-membered phosphorus ring target 3.15.

These approaches make use of milder reaction conditions, which are potentially favourable for the formation of the target molecule.

3.4 Conclusions

Small molecular targets have been designed based on several molecules within the literature. These targets are intended for polymerization into materials that contain

P(I) centers. The synthesis of 3.7, the dimeric analogue of the target 3.14, has been developed, which makes use of the P(I) salts that were discussed in the previous chapter.

Preliminary work into the all phosphorus target, 3.15, has provided some insight into the importance of ligand design as well as reactivity control in terms of the reagents. Mild conditions appear to be necessary in order to afford this most desirable target.

87 3.5 References

1. Mark, J. E., Allcock, H. R., West, R. Inorganic Polymers; Prentice Hall: Englewood Cliffs, NJ, 1992. 2. Dillon, K. B.; Mathey, F.; Nixon, J. F. Phosphorus: The Carbon Copy: From Organophosphorus to Phospha-organic Chemistry, 1998. 3. Tsang, C. W.; Yam, M.; Gates, D. P. J. Am. Chem. Soc. 2003,125, 1480-1481. 4. Gates, D. P.; Manners, I. J. Chem. Soc, Dalton Trans. 1997, 2525-&. 5. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2006, 45, 6864-6874. 6. Barnham, R. J.; Deng, R. M. K.; Dillon, K. B.; Goeta, A. E.; Howard, J. A. K.; Puschmann, H. Heteroatom Chem. 2001,12, 501-510. 7. Schmidpeter, A.; Lochschmidt, S.; Sheldrick, W. S. Angew. Chem., Int. Ed. Engl. 1985, 24, 226-227. 8. Gueorguieva, D. Honours, University of Windsor, 2004. 9. Farrar, G., Norton, E.L., Macdonald, C.L.B. unpublished results. 10. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996,15, 1518-1520. 11. Noth, H.; Meinel, L. Z. Anorg. Allg. Chem. 1967, 349, 225-&. 12. ; Bruker AXS Inc.: Madison, WI, 2001. 13. ; Bruker AXS Inc.: Madison, WI, 2001. 14. ; Bruker AXS Inc.: Madison, WI, 2001. 15. Altomare, A.; Burla, M. C; Camalli, M.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; G., M. A. G.; Polidori, G.; Spagna, R.; CNR-IRMEC: Bari, 1997. 16. Sheldrick, G. M.; Universitat Gottingen: Gottingen, 1997. 17. Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837-838. 18. Sheldrick, G. M.; Bruker AXS Inc.: Madison, WI, 2001. 19. Karsch, H. H.; Witt, E. J. Organomet. Chem. 1997, 529, 151-169. 20. Frank, W.; Petry, V.; Gerwalin, E.; Reiss, G. J. Angew. Chem., Int. Ed. Engl. 1996,55, 1512-1514. 21. Dyker, C. A.; Burford, N.; Lumsden, M. D.; Decken, A. J. Am. Chem. Soc. 2006, 128, 9632-9633.

88 22. Weigand, J. J.; Burford, N.; Lumsden, M. D.; Decken, A Angew. Chem., Int. Ed. 2006,45, 6733-6737. 23. Ellis, B. D.; Macdonald, C. L. B. Phosphorus, Sulfur Silicon Relat. Elem. 2004, 179, 775-778. 24. Schmidpeter, A.; Burget, G. Angew. Chem., Int. Ed. Engl. 1985, 24, 580-581. 25. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chem. 2006, 45, 6864-6874. 1. Schmidpeter, A.; Steinmuller, F.; Sheldrick, W. S. Z Anorg. Allg. Chem. 1989, 579, 158-172. 27. Eliel, E. L., Wilen, S.H. In Stereochemistry of Organic Compounds; John Wiley & Sons, Inc.: New York, 1994. 28. Schmidpeter, A.; Burget, G.; Sheldrick, W. S. Chem. Ber.-Rec. 1985,118, 3849- 3855.

89 Chapter 4 - Development of an Improved Synthesis of iV-Heterocyclic Phosphenium Salts

4.1 Introduction

Organometallic complexes and catalysts have seen significant advances within the last twenty years in the area of ligand design, and in particular the development of carbene ligands including Af-heterocyclic carbenes (NHC's). Carbenes in the coordination sphere of transition metals have been studied for many years, including

Fisher-type carbenes (electrophilic) and Schrock-type carbenes (nucleophilic).1 These

"classical carbenes" are not stable alone and only exist as organometallic complexes

(Figure 4.1). Arduengo reported the first stable, crystallographically-characterized N- heterocyclic carbene (NHC) compound in 1991 and this class of compounds quickly became important because of their enhanced ligand properties compared to phosphines.

OR H c R S R (CO)5\N=( Cp2(Me)Ta^ ^N' N" R H \=J

4.1 4.2 4.3

Figure 4.1. Typical examples of Fischer Carbene 4.1, Schrock Carbene 4.2, and N- Heterocyclic Carbene 4.3.

These NHC's are much more electron rich and more basic than traditional phosphine ligands, and because of their pure donor abilities they bind very tightly to metals.3 A large number of functionalized NHC's have been synthesized because of the ability to tune their many structural aspects for a variety of purposes. These NHC's have been recently reviewed, and continue to be a very prominent ligand-type still under development in many organic and inorganic research groups.3'4

90 The rarity of the stable C(II) center in the NHC's has led many researchers to synthesize analogous compounds with varying elements in place of the carbon. In general, lower oxidation states are more stable as you proceed down the periodic table, making the heavier NHC analogues very plausible. However, in reality, some of the heavier carbene analogues actually predate the first reports of stable NHC's.5"8 Synthetic methods for the main group element analogues can be complicated because a RedOx process is necessary for generating the target heterocyclic products. Reduction of many of the main group halide compounds, where the main group element in NHC analogues usually originates, can be difficult to control, and can also lead to a variety of undesired by-products. For group 14 the only stable halide complexes in the +2 oxidation state occur for tin and lead, although these elements tend to disproportionate when treated with other reagents. The halides of the lighter group members, silicon and germanium, must be reduced from a +4 oxidation state, which is synthetically challenging leads to unstable products.

There are many examples in the literature of the synthesis of the N-heterocyclic phospheniums (NHP's) as well as their use as ligands in transition metal chemistry

(Scheme 4.1).9"17 As indicated in Chapter 1, stable and isolable NHP's actually predate the reports of stable NHC's by almost two decades. In spite of the comparable nature of carbon and phosphorus-containing compounds, NHP's are actually much more electrophilic and thus give very different reactivity to the metal complexes than NHC's.

This superior property of the NHP's over the NHC's should make the NHP's even more worthy of study.

91 [M(CO)3(bipy)L]

M=Cr, Mo, W

Ph

CO /'

Scheme 4.1. Reactions of NHP's with transition metals or small inorganic and organic molecules.

Our focus with this research endeavour is to synthesize the phosphorus analogue of N-heterocyclic carbenes in a clean, spontaneous reaction with minimal, easily removed by-products, and easily manipulated counter anions.

4.2 Experimental

Reagents and General Procedures. All manipulations were carried out using standard inert atmosphere techniques. Phosphorus (III) bromide, phosphorus (III) chloride, diphenylphosphine, and dicyclohexylphosphine were purchased from Strem Chemicals

Inc., and all other chemicals and reagents were obtained from Aldrich. Phosphorus (III) bromide and phosphorus (III) chloride were distilled before use, and all other reagents were used without further purification. All solvents were dried on a series of Grubbs'- type columns18 and were degassed prior to use. CDCI3 was dried over calcium hydride.

C6D6 was dried over Na/benzophenone. Dry THF used for 31P NMR experiments was

92 degassed prior to use. The compound l,4-bis(2,4,6-trimethylphenyl)-2,3-dimethyl-l,4- diazabutadiene (Mes-DAB) was synthesized according to literature procedure.19

Instrumentation. NMR spectra were recorded at room temperature in THF, CDCI3 or

C6D6 solutions on a Bruker Avance 300-MHz spectrometer. Chemical shifts are reported

13 31 in ppm, relative to external standards (SiMe4 for *H and C NMR, 85% H3PO4 for P

NMR). Coupling constant magnitudes, \j I, are given in Hz.

General Preparation of [Mes-DAB-P][Br]. To a solution of PBr3 and cyclohexene in organic solvent (CH2CI2 or toluene), was added a solution of Mes-DAB in the same organic solvent (total 25 mL). Upon addition of the Mes-DAB the colour of the reaction mixture was bright red and gradually darkened to either dark brown or dark green.

Various methods as described below were used in an attempt to purify the compound, and each trial reaction was underwent recrystallization in CH2CI2 or toluene, however no single crystals suitable for X-ray diffraction studies were isolated in spite of numerous crystallization attempts and conditions.

93 Table 4.1. Summary of reaction attempts to form [Mes-DAB-P][Br].

mmol (equiv)

Reaction pBf3 MesDAB Cy* Solvent Notes

1 1.11(1) 1.11(1) 3.32(3) CH2CI2 Reaction performed in bridge system, crude product black foam 2 1.11(1) 1.11(1) 3.32(3) CH2CI2 Reaction mixture filtered, filtrate concentrated to black foam 3 1.11(1) 1.11(1) 3.32(3) CH2CI2 Crude reaction material washed with pentane, black foam 4 1.11(1) 1.11(1) 3.32(3) CH2CI2 Fresh PBr3 used, product dark red 5 1.11(1) 1.11(1) 3.32(3) CH2CI2 Crude reaction material sonicated in pentane, black solid 6 0.74(0.7) 1.11(1) 3.32(3) CH2CI2 Excess of Mes-DAB, sonicated in pentane, black solid 7 0.74(0.7) 1.11(1) 3.32(3) Toluene Excess of Mes-DAB, change solvent, sonication in pentane, red/black solid 8 0.74(0.8) 0.93(1) 3.32(3) Toluene Slight excess of Mes-DAB, crude reaction filtered, concentrated filtrate to black solid 9* 1.10(0.9)* 1.20(1) 1.20(1)* Toluene PC13 & SbPh3 used instead of PBr3 & Cy*, stir overnight, sonicate in pentane, filter orange precipitate, product white solid *alternative reagents used, see Notes in Table 4.1

94 Table 4.2. 31P{'H}-NMR experiment results for the reactions in Table 4.1. Most intense signal in each spectrum denoted by *. Reaction # Solvent 5 (ppm) 1 CDC13 -4.4 176.1* 192.1 2 CH2C12 -3.4 174.4 190.0* 3 CD2C12 175.0 189.3* 4 5 CD2C12 175.2 188.6* 6 CD2C12 -3.3 188.9* 7 C6D6 175.1* 187.1 8 C6D6 166.8* 170.5 9 C6D6 0.4 143.8*

Preparation of [Mes-DAB-P][PF6]. To a flask containing KPF6 (123 mg, 0.67 mmol) in an ice bath, was added a solution of [Mes-DAB-P][Br] (286 mg, 0.67 mmol) in CH2C12

(25 mL). The reaction mixture was allowed to stir at room temperature for 3 days before the volatile components were removed under reduced pressure. The black solid was collected and recrystallization was attempted in toluene, however no single crystals suitable for X-ray diffraction studies were isolated. Crude yield: 54% (176 mg, 0.35 mmol). 31P{XH} NMR: -143.95 (m, IP), -4.29, 170.25, 177.03, 198.77.

Preparation of [Mes-DAB-P][OTf]. To a solution of AgOTf (173 mg, 0.67 mmol) in toluene was added a solution of solution of [Mes-DAB-P][Br] (290 mg, 0.67 mmol) in

95 toluene (total 25 mL). The reaction mixture was allowed to stir at room temperature overnight before the white precipitate was filtered and the filtrate was concentrated to yield a black solid. Recrystallization of the black solid was attempted in toluene, however no single crystals suitable for X-ray diffraction studies were isolated. 31P{!H}

NMR:-3.28, 200.21.

4.3 Results and Discussion

There are several methods reported in the literature for the synthesis of the NHP salts. ' ' ° These methods (Scheme 4.1) include partial reduction of a diimine followed by treatment with PCI3; metathesis of a silicon-diamide, which was made from the reduction of SiCLt with the dilithiated diamido salt; and, treatment of the diamine compound with PCI3 and base to generate the C-C saturated NHP. The first and third examples required chloride abstraction to generate the phosphenium salts.

e ..eOTf

D2eq.Lit Ar^NH N^Ar 1)PCI3 , Ar^/^Ar

2)2HNEt3CI \ // 2) Me3SiOTf \_/

CLsrcl V©Cl p Ar-N N^Ar D2eq.Li , Ar-^'^Ar PCI3,A , Ar^ \N^Ar \\ // 2) SiCI4 \==J \=J

P l)NaBH4 ^ Ar^NH HN^Ar 1)PCi3,Et3N> Ar^N/ \N^Ar

2)H20 \ / 2) [MIA] \ I

Scheme 4.2. Previously reported methods for the synthesis of NHP's.

It is common in different organic solvents to observe these compounds as the covalently bound phosphorus halide and not as the phosphenium halide salt. In general

96 for the C-C unsaturated versions the 31P NMR signal for the phosphenium salt is shifted ca. 10 ppm downfield compared to the covalent phosphorus halide. For the saturated analogues the chemical shift difference is usually much larger at ca. 100 ppm downfield.

This suggests that the unsaturated versions exhibit partial ionic character even in their

"covalent" form.10'21

X® X

\_/ CH2CI2 \_/

Scheme 4.3. The relationship between the ionic phosphenium salts and the covalently bound phosphorus halide depending on the choice of solvent. (X = CI or Br)

Previously in our group, it has been shown that phosphenium salts can be

99 generated in high yield by the spontaneous reaction of PI3 or PCI3 with Dipp-DAB. In the case of the PCI3, a reducing agent (SnCli) was added because of the more reactive nature of the elemental chlorine that is produced transiently. These reactions produced the phosphenium salts cleanly and single crystals of [Dipp-DAB-P][l3] and [Dipp-DAB-

P][SnCl5] suitable for X-ray diffraction studies were isolated (Figure 4.2). These reactions are important in that they demonstrate that the RedOx process necessary for formation of the products is both spontaneous and high yielding. The drawback to these reactions however, is the nature of the anions: they are both large anions that are not easily manipulated using metathesis reactions, and are not desirable for further studies of the reactivity of the phosphenium cations.

97 Figure 4.2. Thermal ellipsoid plot (30% probability surface) of [Dipp-DAB-P][SnCl5] and [Dipp-DAB-P][I3]. Hydrogen atoms are omitted for clarity.

Because of the essential loss of X2 to drive the reaction, we assumed that the addition of a halogen scavenger that potentially accelerates the loss of X2 would be beneficial for the spontaneity of the reaction and would allow for the formation of the more chemically-useful halide salts. The obvious other benefit is the quenching of a potentially reactive by-product that could hinder further reaction. As demonstrated in a previous chapter, cyclohexene is a useful alkene that conveniently removes X2 as an easily removed by-product (Scheme 4.4). In this light, we wished to examine the use of cyclohexene for the preparation of phosphenium salts. It should be noted that preliminary studies confirmed that cyclohexene does not react with any of the starting materials and confirmed its suitability.

98 Pe 0 X Scheme 4.4. Synthesis of previously discussed triphosphenium bromide and chloride salts with the addition of cyclohexene to react with the X2 by-product.

These initial studies using either PBr3 or PCI3 with the addition of cyclohexene in order to generate the phosphenium halide salts has generated excellent results, as assessed by NMR spectroscopy. For each attempted reaction the 31P NMR indicates that there is no or minimal amounts of PX3 remaining in the crude reaction mixture, and the major peak in the spectra always appears between the chemical shift range of 165-193

•5-1 ppm. On the basis of comparison to P NMR chemical shifts of analogous compounds reported in the literature, the desired phosphenium halide salt and the covalently bound phosphorus halides were successfully generated. Thus, we are confident to state that we are generating the desired product, even though the product is not entirely pure.

These phosphorus atom of the phosphenium salts are formally considered to be in the +3 oxidation state. For our work, the implication is that during the reactions the phosphorus is spontaneously reduced to P(I) by a loss of X2, and then subsequently oxidized concomitantly with the 2 electron reduction of the diimine ligand. The nature of the ions observed in some of these cases confirms that such RedOx processes are occurring.

99 a) PBr3 + MesDAB + cyclohexene

*m**i*m'imiM»n*k**mm*0mmmimm*iim1 tmrmi^tmmmitmmmtm'mm*

b) [MesDAB-P][Br] + [K][PF6]

c) [MesDAB-P][Br] + [Ag][OTf]

lbJ|UILlUUk||||gb

I I I I I I I I I I I I I I I I | I I I I | I I I I | 200 -200 ppm

Figure 4.3. NMR spectra of three representative reactions into the synthesis of NHP's. Of note is the major signal for each spectrum is in the range of 180-200 ppm, which agrees with literature reported values of similar NHP's.10'22

All of the crude materials from these reaction attempts were subject to standard recrystallization techniques in order to perform X-ray diffraction studies to further confirm the identity of the products. Several methods were employed for recrystallization including slow evaporation, vapour diffusion, and vacuum slow

100 evaporation. Various solvents were also attempted; however, to date we have been unable to isolate single crystals suitable for X-ray diffraction studies. Careful, thorough washing of the crude reaction material with non-polar solvents or thorough removal of all volatile components could aid in isolating crystals for X-ray diffraction studies.

We also postulate that the although the addition of cyclohexene has proven useful in other reactions to remove the X2 that is produced, perhaps there is an interaction between the cyclohexene or the 1,2-dihalocyclohexane with the target phosphenium product which is hindering removal of the excess cyclohexene or the 1,2- dihalocyclohexane. This could provide insight into the reason that crystalline material has not yet been isolated.

Because of this inability to grow single crystals from all of the crude reaction materials where it is known by 31P NMR that the desired phosphenium is the major product, we decided to try an alternative halogen scavenger. It is well known that triphenyl antimony can act as a halogen scavenger and can either adopt a covalent or ionic structure.23

Although the phosphenium salt is ionic, Gudat has shown that when a non-polar solvent such as toluene is used, the product prefers to be in the covalently bound phosphorus halide.24 This manipulation could allow for the separation of the desired product from the potential antimonate salt. Our initial reaction resulted in an orange precipitate, which was filtered, and the resultant filtrate was concentrated to afford a white solid. The 31P{1H} NMR spectrum displayed the major signal at 143.8 ppm, which is somewhat upfield compared to the spectra of the reaction mixtures where the reaction was performed with cyclohexene. We are currently investigating the use of triphenyl

101 antimony and other reagents as halogen scavengers, and are confident that variation of certain reaction conditions such as temperature or equivalents of reagents, will lead us to achieve a clean facile synthesis of the phosphenium halide.

Another avenue to pursue is the in situ metathesis reaction to generate a phosphenium salt with an anion, which is comparable in size to the cation, crystallization.

Attempts to generate the hexafluorophosphate and triflate salt appear to be successful based on the little to no change in the 31P NMR chemical shift of the NHP, which one would expect if the cation is not changing during the metathesis reaction. Further investigation with these two anions, as well as the BARF

((perfluorinated)tetraphenylborate) anion, which is known to be a non-coordinating anion, is ongoing.

Perhaps the next vista and interesting target molecules will be those bearing ligands that are intermediate between diimines and dipyridine. The recent synthesis of a carbene in such an environment suggests that such compounds should be accessible for perhaps all of the group 14 and 15 elements. In light of unusual properties that appear to be almost intermediate between those of the lower and higher oxidation state alternatives, such compounds should exhibit some very interesting chemistry.3,25

4.4 Conclusions

The synthesis of a phosphenium halide salt has been achieved, with confirmation by 31P NMR spectroscopy. Optimization of the synthesis is ongoing with an attempt to isolate crystalline material suitable for X-ray diffraction studies, as well as producing the phosphenium salts in high yield. Studies into the use of an alternative halogen scavenger

102 to cyclohexene have begun and are being continued. Preliminary attempts to isolate the product as either the triflate, hexafluorophosphate through metathesis reactions with the crude phosphenium bromide material have been successful according to 31P NMR and work is ongoing into metathesis reactions with alternative non-coordinating anions.

103 4.5 References

1. Crabtree, R. H. The Organometallic Chemistry of the Transition Metals; 4th ed.; John Wiley & Sons, Inc.: Hoboken, New Jersey, 2005. 2. Arduengo, A. J.; Harlow, R. L.; Kline, M. J. Am. Chem. Soc. 1991,113, 361-363. 3. Kuhl, O. Chem. Soc. Rev. 2007, 36, 592-607. 4. Crabtree, R. H. Coord. Chem. Rev. 2007, 251, 595. 5. Brooker, S.; Buijink, J. K.; Edelmann, F. T. Organometallics 1991,10, 25-26. 6. Gehrhus, B.; Hitchcock, P. B.; Lappert, M. F. Angew. Chem., Int. Ed. Engl. 1997, 56,2514-2516. 7. Neumann, W. P. Chem. Rev. 1991, 91, 311-334. 8. Tsumuraya, T.; Batcheller, S. A.; Masamune, S. Angew. Chem.. 1991,103, 916- 944. 9. Abrams, M. B.; Scott, B. L.; Baker, R. T. Organometallics 2000,19, 4944-4956. 10. Denk, M. K.; Gupta, S.; Ramachandran, R. Tetrahedron Lett. 1996,37, 9025- 9028. 11. Gudat, D.; Haghverdi, A.; Hupfer, H.; Nieger, M. Chem. Eur. J. 2000,6,3414- 3425. 12. Reeske, G.; Cowley, A. H. Inorg. Chem. 2007,46,1426-1430. 13. Reeske, G.; Hoberg, C. R.; Hill, N. J.; Cowley, A. H. J. Am. Chem. Soc. 2006, 128, 5300-5300. 14. Burck, S.; Daniels, A.; Gans-Eichler, T.; Gudat, D.; Nattinen, K.; Nieger, M. Z. Anorg. Allg. Chem. 2005, 631, 1403-1412. 15. Burck, S.; Gudat, D.; Nieger, M.; Three, J. Dalton Trans. 2007, 1891-1897. 16. Hardman, N. J.; Abrams, M. B.; Pribisko, M. A.; Gilbert, T. M.; Martin, R. L.; Kubas, G. J.; Baker, R. T. Angew. Chem., Int. Ed. 2004,43, 1955-1958. 17. Nakazawa, H.; Miyoshi, Y.; Katayama, T.; Mizuta, T.; Miyoshi, K.; Tsuchida, N.; Ono, A.; Takano, K. Organometallics 2006, 25, 5913-5921. 18. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996,15, 1518-1520. 19. Tom Dieck, H., Svoboda, M., Greiser, T., Z. Naturforsch., B: Anorg. Chem., Org. Chem., 1981, 36B, 823-832.

104 20. Burck, S.; Gudat, D.; Naettinen, K.; Nieger, M.; Niemeyer, M.; Schmid, D. Eur. J. Inorg. Chem. 2007, 5112-5119. 21. Burford, N.; Losier, P.; Macdonald, C; Kyrimis, V.; Bakshi, P. K.; Cameron, T. S. Inorg. Chem. 1994, 33, 1434-1439. 22. Ellis, B. D.; Macdonald, C. L. B. Inorg. Chim. Acta 2007, 360, 329-344. 23. Brusso, J. L.; Clements, O. P.; Haddon, R. C; Itkis, M. E.; Leitch, A. A.; Oakley, R. T.; Reed, R. W.; Richardson, J. F. J. Am. Chem. Soc. 2004,126, 8256-8265. 24. Burck, S.; Gudat, D.; Nieger, M. Angew. Chem., Int. Ed. 2004,43,4801-4804. 25. Tuononen, H. M.; Roesler, R.; Dutton, J. L.; Ragogna, P. J. Inorg. Chem. 2007, 46,10693-10706.

105 Chapter 5 - Dissertation Summary and Future Considerations

5.1 Dissertation Summary

This dissertation has outlined and discussed the findings of research into low oxidation state group 15 compounds, and in particular, low oxidation state phosphorus compounds. This area of work has seen great advances in the past two decades and continues to grow as an active area within main group chemistry.1

An improved synthesis for triphosphenium halide salts was developed, and in particular, the triphosphenium bromide and chloride salts were obtained for the first time in pure forms. The use of the halogen-scavenging agent cyclohexene was employed to remove the dihalogen produced during the reaction. Previously, the dihalogen that was formed in these reactions could react with the diphosphine ligand to generate the oxidized diphosphine products that are very difficult to separate from the triphosphenium salt.2

The new synthesis also afforded excellent yields for the bromide salt (96%), which is well above the reported yield for the iodide salt (52%).

The design of small molecular targets has initiated a large research area within the

Macdonald group focusing on the synthesis of oligomers and/or polymers that contain low oxidation state phosphorus moieties. An eight-membered ring that was previously reported by Schmidpeter et al.,3 was synthesized in a more controlled fashion, through the ligand exchange of the P(I) salt and a P-N-P type ligand. Based on the success of this reaction, an analogous P-P-P type ligand is envisioned in order to create the all- phosphorus analogue of polyphosphazenes. The preliminary synthetic results indicate that a more mild reaction pathway is necessary, and highlight the need for special attention into the design of the P-P-P ligand.

106 Lastly, preliminary results into the synthesis of iV-heterocyclic phosphenium halide salts proved successful on the basis of 31P NMR experiments. The syntheses employed the halogen scavenger approach that was used in Chapter 2. Alternative halogen scavengers are being investigated as well as isolating the product with a different anion through metathesis reactions.

5.2 Future Considerations

The research endeavours outlined in this dissertation are far from over and if anything have actually inspired many more ideas and projects in order to better understand these low oxidation state main group systems.

The new synthesis developed for the triphosphenium salts will hopefully prove to be general for other difficult chemical transformations. Given the low cost of many main group halides, it is not surprising that they are a common starting materials for many syntheses. These reagents often release X2 into the reaction system, generating a potentially harmful by-product. If the removal of the X2 can be controlled by a halogen- scavenging agent such as cyclohexene, this could aid in the synthesis of compounds which are otherwise unattainable.

Another area of interest is the use of other halogen-scavenging agents, depending on desired reaction conditions and other reagents present. We have already attempted to use SbPli3 in our synthesis of the phosphorus analogues of NHC's; however, the reaction requires optimization and investigation is ongoing. Other alkenes or low coordinate inorganic compounds should also be considered.

107 With the synthetic approaches to the triphosphenium salts being vastly improved, more fundamental work into the corresponding inorganic polymers can be conducted more conveniently. Since the triphosphenium salts are quite stable and easily handled, it is our goal to pursue ligand exchange reactions with small chelating diphosphine ligands to produce small molecules, or non-chelating diphosphine ligands to produce oligomers or polymers.

•p0 ©/••\© R P PR 2 2 R2P PR2 -dppe

Ph ®/.p^® Ph e Ph' VJ s Ph -dppe R2 P e \--/© R2 R?P PR^ e.P e n

Scheme 5.1. Ligand exchange reactions with a) a chelating diphosphine (X=anionic heteroatom); b) a non-chelating diphosphine with an anionic linker.

The small target molecules described in Chapter 3 are still being pursued because of their uniqueness and because of the desire to make polymers that contain a low- coordinate P(I) center in the monomeric unit. We are determined to generate these small molecules in a controlled, methodical fashion, which will also allow us to regulate by­ product formation as well as to understand the mechanisms behind the transformations.

Once the syntheses of some of these targets are achieved, studies into the plausibility of polymerizing the molecules through ring-opening polymerization will be conducted.

Recently, the synthesis of N-aryl substituted P-N-P ligands have been reported.4

These ligands will be useful in the synthesis of the N-substituted small molecular target described in Chapter 3. With the steric bulk of the aryl group to buttress the PR2

108 fragments, it is anticipated that the four-membered ring will be more favourable than the dimeric form, rendering these small molecules as viable synthetic targets.

r, 1© A Ph2p pph2 ••© ^ ®/\© *- R2PV PR2 + -dppe \ / N R2P\ /PR2 ' N Ar I Ar

Scheme 5.2. Proposed synthesis for the formation of the four-membered ring with N- aryl substitution.

An additional future project involves the NHC phosphorus analogues or NHP's.

Gudat et al. recently reported the synthesis of a diphosphine that is polarized along the P-

P bond (Figure 5.1).5

R1 d

Figure 5.1. Recently reported compound by Gudat with a polarized P-P bond. R^Mes, 4Bu, R2=Ph, tetraethyl phospholide.

Using the developed synthesis for the NHP's, we aim to synthesize the NHP's and subsequently treat them with a dialkylphosphide to generate the same types of compounds polarized diphosphines. In their report they highlight the insertion reactions of which these polarized diphosphine compounds are capable (Scheme 5.3).5 We are very

interested in pursuing this type of insertion chemistry with small inorganic molecules

such as NH3 or even H2.

109 R R

H90 P—PPh2 K + HPPh2 N R R

R HO'Pr P-OPr + HPPh2

-Me3SnCI +Me3SnCI N R

R R L/H2OI2 /-ci + \_PPh2 P-CI + Ph2P-SnMe3 -N R R R

-PPh, R

Scheme 5.3. Insertion chemistry of a polarized P-P bond into common reagents.

More generally, we intend to investigate the versatility of the RedOx cycloaddition approach to the preparation of other heterocycles containing fragments derived from P(I) and other low-valent main group elements. Such an approach could potentially provide for the simple and convenient synthesis of many classes of compounds.

cyclic or acyclic

1 1 2 EXn + E E E L, I R2 M R1 R2 E = Group 13-16 E1&E2 = Group 14-16

Scheme 5.4. General approach for cycloaddition reactions involving different main group elements.

110 5.3 References

1. Ellis, B. D.; Macdonald, C. L. B. Coord. Chem. Rev. 2007, 251, 936-973. 2. Dillon, K. B.; Monks, P. K. Dalton Trans. 2007, 1420-1424. 3. Schmidpeter, A.; Burget, G. Angew. Chem., Int. Ed. Engl. 1985, 24, 580-581. 4. Biricik, N., Durap, F., Kayan, C, Gumgum, B. Heteroatom Chem. 2007,18, 613- 616. 5. Burck, S.; Gudat, D.; Nieger, M. Angew. Chem., Int. Ed. 2004, 43, 4801-4804.

111 Vita Auctoris

Name: Erin Lee Norton

Place of Birth: St. Catharines, Ontario

Year of Birth: 1981

Education: Laura Secord Secondary School, St. Catharines, Ontario 1995-2000 University of Waterloo, Waterloo, Ontario 2000-2005 B.Sc.

University of Windsor, Windsor, Ontario 2005-2008 M.Sc.

112