Oxyanion Reduction in a Non-Heme Iron System By

OXYANION REDUCTION IN A NON-HEME IRON SYSTEM

COURTNEY L. FORD

DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemistry in the Graduate College of the University of Illinois at Urbana-Champaign, 2018

Urbana, Illinois

Doctoral Committee:

Assistant Professor Alison R. Fout, Chair Professor Thomas B. Rauchfuss Professor Yi Lu Professor Eric Oldfield

ABSTRACT

The secondary coordination sphere of metalloenzymes is implicated in controlling nuclearity, enhancing substrate selectivity, and stabilizing reactive intermediates, and is thus essential in promoting the reactivity of the metal center in the active site. Inspired by the small molecule activation performed by biological systems and the work of others in the area of bioinorganic chemistry creating biomimetic systems, our group designed a tripodal ligand platform featuring a secondary coordination sphere. The ligand is tautomerizable, which allows for versatility in both the primary and secondary coordination spheres. Anionic coordination positions a hydrogen-bond acceptor in the secondary sphere, whereas dative coordination positions a hydrogen-bond donor in the secondary sphere. Previous reports detailed the metallation routes to access both tautomeric forms of the ligand in iron(II) complexes. Moreover, the ability of each arm of the ligand to tautomerize independently was observed in an iron(II)-hydroxo complex that included both tautomeric forms. The secondary coordination sphere also displayed a propensity to engage in hydrogen-bonding interactions with ancillary ligands and allowed for the isolation of a terminal iron(III)-oxo complex. Our lab has previously demonstrated the accessibility of the iron(III)-oxo species from the reduction of nitrite. Herein, the reduction of other oxyanions, the halogen and selenium oxyanions and chromate, to furnish the iron(III)-oxo complex was explored. In Chapter 2, the reduction of the halogen oxyanions to form the iron(III)-oxo complex and an iron(II)-halide complex is detailed. The series of iron(II)-halide complexes was independently synthesized to confirm the formulation of the oxyanion reduction products. The reduction of perchlorate was made catalytic, albeit with low turnover number. The synthesis of a zinc-perchlorate complex gave insight into the binding of the oxyanions to a metal center and the role of the secondary coordination sphere. In the reduction of the selenium oxyanions in Chapter 3, elemental selenium was formed along with the iron(III)-oxo complexes. The identification of the elemental selenium was confirmed through the formation of triphenylphosphine selenide. 31P NMR spectroscopy also enabled the quantification of the elemental selenium upon formation of the phosphine selenide. The reduction of the iron(III)-oxo complex by 1,2-diphenylhydrazine, a two proton/two electron donor, was investigated in Chapter 4 and led to the isolation of a new iron(II)-hydroxo complex with three hydrogen-bond donors in the secondary coordination sphere. The reactivity of

ii this complex was briefly explored, as was an alternative route to access the iron(III)-oxo complex. Addition of acid and water to an octahedral iron(III) species cleanly furnished the iron(III)-oxo complex. A new tripodal ligand platform, which featured only two arms capable of hydrogen- bonding interactions, was synthesized. Its metallation to furnish iron(II) and cobalt(II) complexes was investigated in Chapter 5. The impact of reducing the number of hydrogen-bonding interactions in the secondary coordination sphere was studied. The structures of the metal(II) complexes were as expected, with anionic or dative coordination of the ligand accessible and hydrogen-bonding interactions to an axial ligand present. However, upon oxidation of the iron(II) complexes, formation of a µ-oxo iron(III) dimer was observed, illustrating a limitation of this system. Overall, this work demonstrated the importance of the secondary coordination sphere in achieving small molecule activation in a biomimetic system. The ability of our tripodal ligand system to support the terminal iron(III)-oxo complex allowed for the deoxygenation of many oxyanion substrates, most notably perchlorate. Reducing the number of hydrogen-bonding interactions in the secondary coordination sphere resulted in the formation of a dimeric complex upon oxidation.

iii

In loving memory of Austin Joseph Kearns June 28, 1990 – February 6, 2017

ACKNOWLEDGEMENTS

I cannot thank my family enough for their unending support throughout my time in graduate school, even though they thought I was crazy for moving away from Kentucky. They believed in me and encouraged me to persevere when I wanted to give up, and kept me grounded through the highs and lows of the past five years. I can only hope to give back a fraction of what they have given to me. I particularly want to give thanks to my brother, Andrew. His quick wit and uproarious stories during our long phone conversations provided the much-needed comedic relief to lift my spirits (and many of my colleagues, with whom I shared a number of his tales). I am also indebted to my advisor, Professor Alison Fout. She was a constant source of guidance, support, and patience as I struggled to become a better student and scientist. She taught me to how to approach and solve problems, both in chemistry and in life. She also indulged, and shared, my love of college basketball by making the entire lab fill out brackets for March Madness. I am also grateful for the support of my committee members, Professor Yi Lu, Professor Eric Oldfield, and Professor Tom Rauchfuss, and Professor Cathy Murphy. I must also acknowledge my past and present lab mates, or co-conspirators as Alison must have thought on many occasions, for constantly pushing me to be a better scientist and colleague. I consider Zack Gordon, Yun Ji Park, Abdulrahman Ibrahim, Gabe Espinosa Martinez, Kenan Tokmic, Bailey Jackson, Jack Killion, Michael Drummond, Tabitha Miller, Joe Nugent, Safiyah Muhammad, and Daniel Najera to be my second family. They have given me great advice on all matters, from science to life, and I am privileged to have worked with them. Dr. Ellen Matson was also an important mentor during my graduate career, especially in the early days. I am so appreciative of everything she taught me, from the technical skills of working in the glovebox to a love for the best reality TV franchise, The Bachelor. I also had the great joy of training two talented undergraduate students: Niknaz Iranmanesh and Nick Capra. Finally, I would like to thank my fiancé, Ben Ryan. Thank you for being an incredible and supportive partner. Thank you for telling the best jokes and for cooking dinner every night. Most of all, thank you for motivating me to be my best self. I can’t wait to see where life takes us.

TABLE OF CONTENTS

CHAPTER 1: THE ROLE OF THE SECONDARY COORDINATION SPHERE IN BIOLOGICAL AND SYNTHETIC SYSTEMS ...... 1

CHAPTER 2: HALOGEN OXYANION REDUCTION BY A NON-HEME IRON SYSTEM ..10

CHAPTER 3: DIANIONIC OXYANION REDUCTION BY A NON-HEME IRON SYSTEM ...... 49

CHAPTER 4: INVESTIGATIONS INTO THE SYNTHESIS AND REDUCTION OF A NON-HEME IRON(III)-OXO ...... 79

CHAPTER 5: EFFECTS OF VARYING SECONDARY SPHERE INTERACTIONS: SYNTHESIS OF COBALT AND IRON COMPLEXES OF A TRIPODAL LIGAND FEATURING TWO HYDROGEN-BOND DONORS OR ACCEPTORS ...... 96

CHAPTER 1: THE ROLE OF THE SECONDARY COORDINATION SPHERE IN BIOLOGICAL AND SYNTHETIC SYSTEMS

1.1 Role of the secondary coordination sphere in biological systems Nature is able to activate a multitude of small molecules under ambient conditions in metalloenzymes, many of which contain a first-row transition metal in the active site.1-3 These active sites are also defined by extensive non-covalent secondary sphere interactions with the surrounding amino acid residues of the protein superstructure.3,4 While the primary coordination sphere of the transition metal center, composed of the covalent interactions between the metal ion and its ligands, controls its electronic structure and Lewis acidity, the secondary sphere also plays a significant role in the reactivity of the metalloenzyme.5,6 Of the of secondary sphere interactions, hydrogen-bonding interactions are the most prevalent in biological systems, where polar and protic residues are positioned in proximity to a metal center.3 Remarkably, these weak interactions, between 5 and 15 kcal/mol, control nuclearity, stabilize reactive intermediates, and enhance substrate selectivity.3,5 The importance of hydrogen-bonding interactions within biological systems is illustrated by the reversible binding of dioxygen for transport within a respiring organism in hemoglobin. The dioxygen binds trans to the proximal histidine of the heme active site, which is conserved over a broad range of species. In contrast, free iron(II)-porphyrins are readily oxidized in the presence of dioxygen, either through the formation of a µ-oxo iron(III)-porphyrin dimer or the generation of an iron(III)-superoxide.7 Hydrogen-bonding interactions with the proximal histidine tune the strength of the Fe–O bond, influencing the affinity of the heme center for dioxygen.7 Moreover, hydrogen-bonding interactions in the distal pocket are proposed to aid in the stabilization of the bound dioxygen unit. In the case of human hemoglobin, a distal histidine acts as a hydrogen-bond donor (Figure 1.1). When this residue is mutated, the binding affinity of dioxygen is significantly decreased.5 The secondary sphere also aids in the reduction of oxyanions. In the enzyme chlorite dismutase, the positioning of an arginine residue within the active site is proposed to foster rebound of hypochlorite and an iron(IV)-oxo upon cleavage of a Cl–O bond in chlorite (Figure 1.1).8 Peroxyhypochlorite is formed and subsequently decomposes to chloride and dioxygen. Mutation of this residue drastically reduces the catalytic efficiency of the enzyme and increases the release

1 of hypochlorite from the active site.8 The enzyme (per)chlorate reductase, which is expressed in bacteria residing in contaminated water supplies to reduce (per)chlorate, relies on the presence of aromatic residues in the tunnel leading from the exterior of the enzyme to the active site for high catalytic efficiency and selectivity even in low (per)chlorate concentrations.9 Coates and co- workers propose that monodentate or bidentate binding of an aspartate residue in the active site positions the aromatic tunnel residues to be in either an open or closed conformation, controlling the movement of the chlorine oxyanion substrate into and the release of water from the active site.9 Arg His

N N NH H H N O CO2H 2 NH HO C Cl O 2 O O N N N N FeII FeIV N N N N N N HN HO2C CO2H His HN His Figure 1.1 Examples of secondary sphere interactions in biological systems: stabilization of dioxygen via a hydrogen bond from the distal histidine in human hemoglobin (left) and positioning of an arginine residue to promote rebound in chlorite dismutase (right).

Thus, both the primary and secondary coordination spheres contribute to the metal- mediated catalysis of metalloenzymes. While the tuning of the primary coordination sphere has been extensively studied in synthetic systems, incorporating and modulating a secondary coordination sphere in model complexes remains challenging. The impact of such work, however, has great impact; metalloenzymes catalyze difficult processes with high efficiency and are unmatched by their synthetic analogues. Investigating the role of the secondary sphere in synthetic systems will further the understanding of biological small molecule activation and aid in the design of improved catalysts.

1.2 Incorporation of a secondary coordination sphere in non-heme biomimetic complexes Many investigations have explored the structure-function relationship of the secondary sphere of metalloenzymes through the study of synthetic models. Early contributions to this body

2 of work involved appending a secondary sphere to porphyrin platforms.10 In particular, the reversible binding of dioxygen was modeled by incorporating bulky substituents on at least one face of a heme. The increased steric bulk prevented oxidation of the iron(II)-porphyrin centers through µ-peroxo pathways and allowed for the spectroscopic characterization of synthetic heme dioxygen adducts.10 However, the syntheses of these modified porphyrins were lengthy and low- yielding. By comparison, the synthesis of non-heme ligand scaffolds is much easier and the resulting model complexes still provide insight into important biological reactions. The geometry imposed by many non-heme ligands is C3 symmetric, as the d-orbital splitting suggests high-valent multiply bonded metal-oxygen or metal-nitrogen species, reminiscent of proposed intermediates in biological small molecule activation, can be stabilized.11 Early work by Masuda using a tripodal pyridylamine ligand demonstrated the utility of a C3 symmetric platform featuring a secondary coordination sphere to stabilize a thermally unstable copper(II)-hydroperoxo species (Figure 1.2).12 Furthering these investigations, Borovik and co-workers synthesized a tripodal urea-based ligand, which allowed for the isolation of the first terminal iron(III)-oxo complex.13 Characterization of this complex and its manganese analogue contributed to the mechanism of dioxygen activation at non-heme metal centers.13-15 The structure of the urea-based ligand, however, necessitated trianionic coordination to a metal center and was unable to support a combination of hydrogen-bond donors and acceptors in the secondary sphere. tBu + tBu 2 - tBu tBu HN HO tBu O O NH NH O O O HN HN O N N O N Cu N Fe N N N N

Figure 1.2 Secondary sphere interactions in synthetic non-heme systems: Masuda’s copper(II)-hydroperoxo complex (left) and Borovik’s terminal iron(III)-oxo complex (right).

Love and co-workers designed a dipodal ligand scaffold that was capable of tautomerization to tune coordination of the ligand to a metal center and to modulate the hydrogen- bonding interactions of the secondary sphere.16 Tautomerization between a pyrrole-imine and an azafulvene-amine was proposed to allow for anionic or dative coordination to a metal center and

3 to position hydrogen-bond acceptors or donors in the secondary sphere. Tautomerization of each arm to the neutral azafulvene-amine tautomer upon the addition of a metal(II) salts to the free ligand was achieved (Figure 1.3).16 However, in the analogous tripodal system, the desired coordination mode was not accessible, as the metal center coordinated to the pyrrole and imine nitrogens of two arms or to the pyrrole nitrogens of two separate ligands.17,18 Our group sought to expand the use of the pyrrole-imine/azafulvene-amine tautomeric pair in a C3 symmetric ligand platform by installation of an apical nitrogen donor to enforce the desired binding mode of a metal center. tBu tBu tBu tBu N N NH NH X X MX2 M NH HN N N M = Fe, Co, Zn X = Cl, Br

Figure 1.3 Tautomerization from the pyrrole-imine to the azafulvene-amine upon the addition of metal(II) salts to Love’s dipodal ligand platform.

1.3 Synthesis and reactivity of a coordinatively versatile tripodal ligand platform featuring a secondary coordination sphere Motivated by the work of Masuda, Borovik, and Love, our group designed a tripodal non- Cy heme ligand platform, tris(5-cyclohexylimino-pyrrol-2-ylmethyl)amine (N(pi )3), that could tautomerize to include hydrogen-bond donors or acceptors in the secondary coordination sphere (Figure 1.4).19 Moreover, tautomerization would allow for versatility in the coordination mode of the ligand to a metal center. Deprotonation of the pyrrole-imine tautomer generates an anionic nitrogen donor with a hydrogen-bond acceptor in the secondary coordination sphere. In contrast,

Cy Cy Cy Cy Cy Cy N HN N N NH HN tautomerization NH HNHN N N N

N N

Cy Cy N(pi )3 N(afa )3 Cy Figure 1.4 Tautomerization of the N(pi )3 ligand platform.

4 coordination of the azafulvene-amine tautomer to a metal center is dative and positions a hydrogen- bond donor in the secondary coordination sphere. Combining the triply deprotonated ligand salt and an iron(II) salt generated the potassium Cy II 19 salt of the iron(II)-aquo complex, K[N(pi )3Fe(OH2)] (Fe -OH2; Figure 1.5). Hydrogen- bonding interactions were observed between two imine moieties and the two hydrogen atoms of the aquo ligand. Protonation of this species via the addition of acid yielded an iron(II)-hydroxo Cy Cy II 19 complex, N(pi )(afa )2FeOH (Fe -OH; Figure 1.5). Notably, the arms of the ligand in this complex were in two different tautomeric forms. One arm was coordinated to the iron(II) center anionically and its imine moiety was engaged in hydrogen-bonding with the hydrogen atom of the hydroxo ligand. The remaining arms were bound datively to the metal center, with the amine moieties of the secondary sphere acting as hydrogen-bond donors to the oxygen atom of the axial Cy II hydroxo functionality. The iron(II)-triflate species, N(pi )3Fe(OTf)2 (Fe -OTf; Figure 1.5), in which the three arms of the ligand platform were datively coordinated to the metal center, was furnished from the addition of the metal(II) salt to the free ligand, akin to the synthesis of the azafulvene-amine complexes reported by Love and co-workers.16,19 This series of iron(II) complexes demonstrated the ability of the arms of the ligand platform to tautomerize independently, according to the needs of the axially bound ligand. Moreover, the ability of the ligand to traverse three charge states, neutral, monoanionic, or trianionic, was also observed.

Cy Cy K Cy Cy Cy Cy Cy Cy Cy N N HN N N HN N N N H H O H O 3.1 KH HCl Fe NH HNHN FeCl2 N Fe N N N N N - 3 H2 - KCl N - 2 KCl N N Cy II II N(pi )3 Fe -OH2 Fe -OH

OTf Cy Cy O CF3 Cy S NH HN NH O O Fe(OTf) 2 N Fe N N

N FeII-OTf Cy Figure 1.5 Synthesis of a series of iron(II) complexes of the N(pi )3 ligand.

5 Cy Hydrogen-bonding interactions in the secondary coordination sphere of the N(pi )3 system were further explored through the installation of ancillary pseudohalide ligands, which can act as surrogates for more reactive anionic ligands, such as superoxide. Extensive solid-state structural analysis of a series of metal(II)-pseudohalide complexes revealed variability in the hydrogen- bonding environment of the secondary sphere.20 In each of the pseudohalide structures, one hydrogen-bonding interaction was observed between one arm of the ligand and the ancillary anion. In the case of azide and isocyanate, a second arm of the ligand exhibited positional disorder. Either the amine moiety was pointed toward the metal center, forming a second hydrogen-bonding interaction between the amine of the ligand platform and the pseudohalide, or the amine moiety pointed away from the metal center (Figure 1.6).20 This positional disorder demonstrated the Cy flexibility of the secondary sphere in the N(pi )3 system.

O Cy OTf O Cy OTf Cy Cy Cy Cy C HN C HN HN HN NH HN N N Fe Fe N N N N N N

N N

Figure 1.6 Positional disorder in the solid-state structure of an iron(II)-isocyanate complex.

Small molecule activation by the iron(II)-triflate species FeII-OTf was examined. The addition of 1 equiv nitrite to FeII-OTf furnished the triflate salt of the terminal iron(III)-oxo Cy III 21 complex [N(afa )3FeO]OTf (Fe -O; Figure 1.7). When the amount of nitrite was reduced to 0.5 equiv, two iron species were generated, FeIII-O and an iron(II)-nitrosyl complex, indicating the production nitric oxide, akin to the reduction of nitrite in biological systems.21 Cy OTf Cy OTf Cy Cy Cy O CF3 Cy HN S NH NH HN HN NH O O O NO - Fe N Fe N N 2 N N N - NO N N FeII-OTf FeIII-O Figure 1.7 Reduction of nitrite by FeII-OTf to form FeIII-O and nitric oxide.

Upon the reduction of FeIII-O, the iron(II) hydroxo complex FeII-OH was formed, a result of the movement of a proton from the secondary sphere to a bound substrate.22 This reaction further Cy corroborated the flexibility and versatility of the N(pi )3 system. The ligand platform could

6 coordinate to a metal center neutrally or as a monoanion or a trianion. The hydrogen-bonding interactions between the ligand platform and an ancillary ligand also exhibited variety, as evidenced by the positional disorder in the solid-state structure of the iron(II)-isocyanate complex. Moreover, the ligand was able to shuttle protons from the secondary sphere to a bound substrate upon the reduction of FeIII-O, a process of paramount importance in biological systems. The ability of the iron(II)-triflate species FeII-OTf to model nitrite reduction was also observed. The utilization of the iron(II)-triflate species FeII-OTf to reduce oxyanion substrates is explored herein. Through the incorporation of hydrogen-bonding interactions, inert anions, such as perchlorate, selenite, and selenate, were reduced through deoxygenation, resulting in the formation of the terminal iron(III)-oxo complex FeIII-O. The reactivity of FeIII-O toward reductants was investigated and led to the isolation of a new iron(II)-hydroxo complex with all three arms of the ligand platform were in the azafulvene-amine tautomeric form. The synthesis and metallation of a ligand scaffold featuring two arms capable of engaging in hydrogen-bonding with ancillary ligands afforded insight into the role of the secondary sphere in our system.

1.4 References (1) Lin, Y.-W. Rational Design of Metalloenzymes: From Single to Multiple Active Sites. Coord. Chem. Rev. 2017, 336, 1-27.

(2) Dawson, J. H. Probing Structure-Function relations in Heme- Containing Oxygenases and Peroxidases. Science 1988, 240, 433- 439.

(3) Borovik, A. S. Bioinspired Hydrogen Bond Motifs in Ligand Design: The Role of Noncovalent Interactions in Metal Ion Mediated Activation of Dioxygen. Acc. Chem. Res. 2005, 38, 54-61.

(4) Ozaki, S.-I.; Roach, M. P.; Matsui, T.; Watanabe, Y. Investigations of the Roles of the Distal Heme Environment and the Proximal Histidine Iron Ligand in Peroxide Activation by Heme Enzymes via Molecular Engineering of Myoglobin. Acc. Chem. Res. 2001, 34, 818-825.

(5) Shook, R. L.; Borovik, A. S. Role of the Secondary Coordination Sphere in Metal- Mediated Dioxygen Activation. Inorg. Chem. 2010, 49, 3646-3660.

(6) Cook, S. A.; Borovik, A. S. Molecular Designs for Controlling the Local Environments around Metal Ions. Acc. Chem. Res. 2015, 48, 2407-2414.

7 (7) Gell, D. A. Structure and Function of Haemoglobins. Blood Cells Mol. Dis. [Online early access]. DOI: 10.1016/j.bcmd.2017.10.006. Published Online Oct 31, 2017. https://www.sciencedirect.com/science/article/pii/S1079979617302085 (accessed Mar 17, 2018).

(8) Schaffner, I.; Hofbauer, S.; Krutzler, M.; Pirker, K. F.; Furtmüller, P. G.; Obinger, C. Mechanism of Chlorite Degradation to Chloride and Dioxygen by the Enzyme Chlorite Dismutase. Arch. Biochem. Biophys. 2015, 574, 18-26.

(9) Youngblut, M. D.; Tsai, C.-L.; Clark, I. C.; Carlson, H. K.; Maglaqui, A. P.; Gau-Pan, P. S.; Redford, S. A.; Wong, A.; Tainer, J. A.; Coates, J. D. Perchlorate Reductase is Distinguished by Active Site Aromatic Gate Residues. J. Biol. Chem. 2016, 291, 9190- 9202.

(10) Momenteau, M.; Reed, C. A. Synthetic Heme-Dioxygen Complexes. Chem. Rev. 1994, 94, 659-698.

(11) Saouma, C. T.; Peters, J. C. M≡E and M=E Complexes of Iron and Cobalt that Emphasize Three-Fold Symmetry (E = O, N, NR). Coord. Chem. Rev. 2011, 255, 920-937.

(12) Wada, A.; Harata, M.; Hasegawa, K.; Jitsukawa, K.; Masuda, H.; Mukai, M.; Kitagawa, T.; Einaga, H. Structural and Spectroscopic Characterization of a Mononuclear Hydroperoxo- Copper(II) Complex with Tripodal Pyridylamine Ligands. Angew. Chem. Int. Ed. 1998, 37, 798–799.

(13) MacBeth, C. E.; Golombek, A. P.; Young, V. G.; Yang, C.; Kuczera, K.; Hendrick, M. P.; Borovik, A. S. O2 Activation by Nonheme Iron Complexes: A Monomeric Fe(III)-Oxo Complex Derived from O2. Science 2000, 289, 938-941.

(14) Shirin, Z.; Hammes, B. S.; Young, V. G., Jr.; Borovik, A. S. Hydrogen Bonding in Metal Oxo Complexes: Synthesis and Structure of a Monomeric Manganese(III)-Oxo Complex and Its Hydroxo Analogue. J. Am. Chem. Soc. 2000, 122, 1836-1837.

(15) MacBeth, C. E.; Gupta, R.; Mitchell-Koch, K. R.; Young, V. G., Jr.; Lushington, G. H.; Thompson, W. H.; Hendrich, M. P.; Borovik, A. S. Utilization of Hydrogen Bonds To Stabilize M-O(H) Units: Synthesis and Properties of Monomeric Iron and Manganese Complexes with Terminal Oxo and Hydroxo Ligands. J. Am. Chem. Soc. 2004, 126, 2556- 2567.

(16) Reid, S. D.; Wilson, C.; Blake, A. J.; Love, J. B. Tautomerisation and Hydrogen-Bonding Interactions in Four-Coordinate Metal Halide and Azide Complexes of N-Donor-Extended Dipyrromethanes. Dalton Trans. 2009, 39, 418-425.

(17) Hart, J. S.; White, F. J.; Love, J. B. Donor-Extended Tripodal Pyrroles: Encapsulation, Metallation, and H-Bonded Tautomers. Chem. Commun. 2011, 47, 5711-5713.

8 (18) Hart, J. S.; Nichol, G. S.; Love, J. B. Directed Secondary Interactions in Transition Metal Complexes of Tripodal Pyrrole Imine and Amide Ligands. Dalton Trans. 2012, 41, 5785- 5788.

(19) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of Iron(II) Aqua and Hydroxyl Complexes Featuring a Tripodal H-Bond Donor and Acceptor Ligand. Inorg. Chem. 2014, 53, 4450-4458.

(20) Matson, E. M.; Park, Y. J.; Bertke, J. A.; Fout, A. R. Synthesis and Characterization of - M(II) (M = Mn, Fe, and Co) Azafulvene Complexes and Their X3 Derivatives. Dalton Trans. 2015, 44, 10377-10384.

(21) Matson, E. M.; Park, Y. J.; Fout, A. R. Facile Nitrite Reduction in a Non-Heme Iron System: Formation of an Iron(III)-Oxo. J. Am. Chem. Soc. 2014, 136, 17398-17401.

(22) Gordon, Z.; Drummond, M. J.; Matson, E. M.; Bogard, J. A.; Schelter, E. J.; Lord, R. L.; Fout. A. R. Tuning the Fe(II/III) Redox Potential in Nonheme Fe(II)-Hydroxo Complexes Through Primary and Secondary Sphere Modifications. Inorg. Chem. 2017, 56, 4852-4863.

9 CHAPTER 2: HALOGEN OXYANION REDUCTION BY A NON-HEME IRON SYSTEM†

2.1 Introduction The uses of halogen-containing oxyanions are both extensive and varied. Chlorine oxyanions have notable applications in bleaching agents, propellants, and explosives.1-4 The bromine oxyanion bromate is a flour additive that strengthens bread dough and is a side product of ozonation or chlorination of bromide-containing water.5-7 Iodine oxyanions are often components of iodized salt for human and animal consumption and find frequent use as oxidants in synthetic transformations, such as the oxidations of 1,2-diols, 1,2-diketones, and sulfides.8-10 Because the halogen oxyanions are highly water-soluble and strong oxidants, they have become pervasive contaminants in many sources of drinking water and pose a significant threat to human health.1-4,6 Removal of these pollutants using traditional water purification techniques, such as ion exchange resins, can be effective; however, these techniques are not selective for halogen oxyanions.6,11 Thus in water sources containing high concentrations of other oxyanion salts, particularly nitrate and sulfate, the halogen oxyanions may not be removed below regulatory standards.11 Furthermore, ion exchange resins must be regenerated or replaced frequently, generating concentrated waste streams. Chemical remediation techniques offer an alternative, as they would release benign products, such as a halide salt and water, from the reduction of the halogen oxyanions. However, few reagents are capable of reducing halogen oxyanions, particularly perchlorate, at an appreciable rate because of their low binding affinity to transition metal centers.12,13 These inorganic oxyanions have long been touted for their weak complexation, poor nucleophilicity, and, consequently, their kinetic inertness toward oxidation and reduction.13 Therefore, harsh reaction conditions (such as low pH, high temperature, photolysis, and/or long reaction times) are often required to facilitate oxyanion reduction.13-16 One of the few homogenous systems reported to catalyze oxyanion + reduction was the cationic oxorhenium(V) complex [Re(O)(hoz)2] (hoz = 2-(2′-hydroxyphenyl)- 2-oxazoline), developed by Abu-Omar and co-workers.13,14,17 In this system, perchlorate was

† Portions of this chapter are reproduced from the following publication with permission from the authors: Ford, C. L.*; Park, Y. J.*; Matson, E. M.; Gordon, Z.; Fout, A. R. Science 2016, 354, 741-743. III The solid-state structural analysis of [Fe -O]ClO4 was completed by Dr. Ellen M. Matson.

10 proposed to bind to the rhenium(V) center, followed by oxygen atom transfer to generate chlorate. The resulting dioxorhenium(VII) complex was reduced to regenerate the starting rhenium(V) through oxo transfer to a thioether (Figure 2.1). The chlorate formed was further deoxygenated by the catalyst to form chloride, which began to inhibit catalysis after 100 turnovers. +

O O O N Re N O O

+ - - SOR2 [Re(O)(hoz)2] + ClO4

+ SR O 2 O O O O OClO3 Re Re N N N N O O O O O O

+ - + SR - ClO3 2 O O O Re N N O O O

+ Figure 2.1 Proposed mechanism of perchlorate reduction by [Re(O)(hoz)2] .

+ Werth and co-workers further improved the [Re(O)(hoz)2] system by adhering the catalyst and palladium nanoparticles onto a porous carbon support under a hydrogen atmosphere.15,16 The heterogeneous Re(hoz)2-Pd/C system allowed for perchlorate to be reduced to chloride in aqueous solution, whereas Abu-Omar and co-workers had performed their perchlorate reduction studies in acetonitrile. Furthermore, the hydrogen atmosphere provided the reducing equivalents necessary to regenerate the oxorhenium(V) catalyst with concomitant release of water. Although deleterious hydrogenation reactions that deactivated the catalyst and catalyst leaching were observed in the

Re(hoz)2-Pd/C system, the substitution of one oxazoline ligand for a thiazoline ligand and the replacement of palladium nanoparticles with rhodium nanoparticles significantly mitigated these problems.16

11 Despite many efforts, with some successes, towards chemically reducing the halogen oxyanions using synthetic systems, the most efficient reduction of the halogen oxyanions is observed in microbial metalloenzymes during anaerobic respiration. The biological reduction of the chlorine oxyanions has been extensively studied and involves the metalloenzymes (per)chlorate reductase and chlorite dismutase (Figure 2.2).

- - - - ClO4 / ClO3 ClO2 Cl + O2

(per)chlorate reductase chlorite dismutase

Asp N N FeIII N N O O MoIV N S S CO2H HO2C S S HN His

O MoIV S N O H NH N S HN N NH O O O N 2 O O O H2N N N O P P H O O HO OH

Molybdenum-Bound Molybdopterin Guanine Dinucleotide Cofactor Figure 2.2 The active sites of (per)chlorate reductase and chlorite dismutase (top) and the structure of the molybdopterin guanine dinucleotide cofactor present in (per)chlorate reductase (bottom).

(Per)chlorate reductase, which catalyzes the reduction of (per)chlorate to chlorite, contains a bis(molybdopterin guanine dinucleotide)molybdenum cofactor in the active site.18 The primary coordination sphere of the molybdenum(IV) center is completed with a bidentate aspartate. During catalysis, (per)chlorate binds to the molybdenum center, shifting the coordination mode of the aspartate from bidentate to monodentate. Subsequent oxygen atom transfer from the bound (per)chlorate furnishes a molybdenum(VI)-oxo and either chlorate or chlorite. Two sequential proton-coupled electron transfer steps regenerate the molybdenum(IV) catalyst as water is lost from the active site (Figure 2.3).

12 Asp

O O + - + 2 H + 2 e S MoIV S - H O 2 - - S S + ClO4 - ClO2

O Asp Asp O O Cl O O Cl O O O O O O IV S MoVI S S Mo S S S S S

O Asp Asp O O Cl O Cl O O O O O O O IV S Mo S S MoVI S S S S S

O O Asp Cl + - O O O + 2 H + 2 e - H2O S MoIV S S S

Figure 2.3 Proposed mechanism of (per)chlorate reductase.

Chlorite dismutase is responsible for the dismutation of chlorite to dioxygen and chloride. The active site of the enzyme contains a ferric heme center, which binds chlorite to initiate catalysis. The chlorite is deoxygenated, forming an iron(IV)-oxo porphyrin cation radical complex and hypochlorite.19 Recombination of the hypochlorite and the iron(IV)-oxo generates a transient iron(III)-peroxyhypochlorite that then releases chloride and dioxygen (Figure 2.4).

13 Cl O O

N N FeIII N N

- N + ClO2 CO2H HO C 2 HN - O2 His - Cl-

O O Cl Cl O O N N N N IV FeIII Fe N N N N

N N CO H CO2H 2 HO2C HO2C HN HN His His

O Cl O N N FeIII N N

N CO2H HO C 2 HN His Figure 2.4 Proposed mechanism of chlorite dismutase.

While the mechanisms (per)chlorate and chlorite reduction have been extensively studied, the mechanism of the biological reductions of bromate and iodate are unclear. Enzymes specific to the reduction of these anions have yet to be identified. Instead, it is hypothesized that promiscuous oxyanion reducing enzymes, particularly nitrate reductase, account for the reduction of bromate and iodate in microbes.20,21 The active sites of many of these enzymes contain molybdopterin cofactors, allowing for the reduction of non-native oxyanion substrates. Reductase enzymes feature extensive hydrogen-bonding networks, which facilitate the movement of protons and water about the active site and stabilize reactive intermediates.18 Additionally, aromatic residues near the active site of (per)chlorate reductase aid in the binding of perchlorate to the molybdenum center.18 Thus, the non-covalent interactions found within oxyanion-reducing metalloenzymes play an important role in facilitating reactivity. Inspired by the active sites of (per)chlorate reductase and chlorite dismutase, we developed a non-heme platform with a secondary coordination sphere. Most transition metal systems are not

14 capable of these reductions13-16, but we hypothesized that incorporating secondary sphere interactions into transition metal complexes would aid in oxyanion deoxygenation. The ability of the ligand to undergo tautomerization may be a key feature during multi-electron deoxygenation reactions because it could facilitate proton and electron transfer between a substrate and a metal center. Furthermore, the secondary coordination sphere helped orient substrates binding to a metal center, as demonstrated by nitrite reduction in our system, in which a single hydrogen bond stabilized a key metal-nitrito intermediate.22 This chapter describes the reduction of the chlorine, bromine, and iodine oxyanions to form an iron(III)-oxo complex and an iron(II)-halide complex. The catalytic reduction of perchlorate to generate water and chloride was also investigated. Synthesis of a zinc-perchlorate complex provided insight into role of the secondary coordination sphere in these transformations.

Cy 2.2 Synthesis of [N(afa )3FeO]ClO4 Initial attempts to synthesize an iron(II)-perchlorate complex yielded an unexpected result, Cy the reduction of perchlorate to generate an iron(III)-oxo complex. Ligand, N(pi )3, was added to a solution of Fe(ClO4)2 generated from the in situ salt metathesis of FeCl2 and 2 equiv AgClO4, immediately turning the solution deep red-brown. After stirring for 1 h at room temperature, the crude reaction mixture was analyzed by 1H NMR spectroscopy, which revealed three products Cy III Cy III (Figure 2.5): [N(afa )3FeO]ClO4 ([Fe -O]ClO4), [N(afa )3FeOH](ClO4)2 ([Fe -OH](ClO4)2), Cy and protodemetallated ligand (N(pi )3 • 3 HClO4). Previous work from our lab detailed the Cy III synthesis of the triflate analogues of these compounds ([N(afa )3FeO]OTf (Fe -O), Cy III Cy [N(afa )3FeOH](OTf)2 (Fe -OH), and (N(pi )3 • 3 HOTf). The perchlorate compounds were readily assigned because the resonances in their 1H NMR spectra were indistinguishable from III Cy those of the triflate compounds. Furthermore, Fe -OH and N(pi )3 • 3 HOTf were previously III Cy generated via the addition of 1 equiv HOTf to Fe -O and the addition of 3 equiv HOTf to N(pi )3, III respectively. These results led us to two hypotheses: (1) [Fe -O]ClO4 was generated from the reduction of perchlorate and (2) acid was formed during the course of the reaction, resulting in the III Cy formation of [Fe -OH](ClO4)2 and N(pi )3 • 3 HClO4.

15 Cy Cy Cy 2 ClO Cy 3 ClO4 Cy Cy Cy Cy ClO4 Cy Cy 4 Cy Cy N HN HN HN N N NH HN HN NH NH HN O HO Fe(ClO4)2 Fe + Fe + NH HNHN N N N N N N NH HNHN

N N N N

Cy III III Cy N(pi )3 [Fe -O]ClO4 [Fe -OH](ClO4)2 N(pi )3 • HClO4 Cy Figure 2.5 The addition of Fe(ClO4)2 to N(pi )3

When the reaction was repeated in the presence of triethylamine and washed with diethyl ether upon completion, a different mixture of products was observed by 1H NMR spectroscopy III (Figure 2.6): [Fe -O]ClO4 and the perchlorate salt iron(II)-chloride complex Cy II II 1 [N(afa )3FeCl]ClO4 ([Fe -Cl]ClO4). [Fe -Cl]ClO4 was identified by comparison of its H NMR Cy II spectrum with the triflate analogue, [N(afa )3FeCl]OTf (Fe -Cl). Moreover, a triethylammonium salt was present in the diethyl ether wash. These results supported the hypotheses stated above; (1) III perchlorate was fully reduced to chloride, accounting for the formation of both [Fe -O]ClO4 and II [Fe -Cl]ClO4 and (2) the added triethylamine was protonated by acid formed during the course of the reaction, forming the triethylammonium salt and precluding the formation of [FeIII- Cy OH](ClO4)2 and N(pi )3 • 3 HClO4.

Cy Cy ClO Cy ClO4 Cy Cy Cy Cy 4 Cy Cy N HN NH N N NH HN Fe(ClO ) HN NH 4 2 O Cl NEt3 Fe + Fe NH HNHN N N N N N N

N N N

Cy III II N(pi )3 [Fe -O]ClO4 [Fe -Cl]ClO4 Cy Figure 2.6 The addition of Fe(ClO4)2 to N(pi )3 in the presence of NEt3.

III [Fe -O]ClO4 was further characterized by solid-state structural analysis. Refinement of the data revealed an iron(III)-oxo in a trigonal bipyramidal geometry with an outer-sphere III perchlorate anion (Figure 2.7). The bond lengths of [Fe -O]ClO4 were similar to those previously reported for FeIII-O, with an Fe–O distance of 1.8055(11) Å, compared to 1.8079(9) Å for FeIII- O.22 All three amino moieties of the secondary coordination sphere were pointing inward to engage III in hydrogen-bonding with the oxo moiety. When crystals of [Fe -O]ClO4 were analyzed by IR spectroscopy, a stretch corresponding to the C=N bond was observed at 1664 cm-1, matching that of FeIII-O (1667 cm-1) and confirming all three arms of the ligand were in the azafulvene-amine

16 tautomeric form.22,23 The solution magnetic moment of the iron center calculated by the Evan’s

Method (µeff = 6.12(8) µB) was consistent with a high-spin, S = 5/2 iron(III) formulation.

H7 Cl1 H6 H7 H5 H5 O1 H6 O1 Zn1 Fe1

III Zn-ClO [Fe -O]ClO4 4

III Figure 2.7 Molecular structures of [Fe -O]ClO4 and Zn-ClO4 shown with 50% probability ellipsoids. Select hydrogen atoms, outer-sphere anions, and solvent have been omitted for clarity.

Cy 2.3 Synthesis of [N(afa )3ZnClO4]ClO4 We sought to understand the role of the secondary coordination sphere in the reduction of Cy perchlorate. Because perchlorate was readily reduced in the presence of N(pi )3 and an iron(II) III center to form [Fe -O]ClO4, we were unable to crystallize any iron(II)-perchlorate species.

Instead, we employed zinc, a redox innocent metal, as a surrogate for iron. Zn(ClO4)2 was generated in situ from the salt metathesis of ZnCl2 and 2 equiv AgClO4 and added to 1 equiv Cy Cy N(pi )3, furnishing the zinc-perchlorate species [N(afa )3ZnClO4]ClO4 (Zn-ClO4; Figure 2.8).

Cy Cy Cy O Cy ClO4 Cy O O Cy N Cl NH N N HN NH O Zn(ClO4)2 Zn NH HNHN N N N

N N

Cy N(pi )3 Zn-ClO4

Figure 2.8 Synthesis of Zn-ClO4.

Crystals suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a saturated solution of the target molecule in dichloromethane. Refinement of the data revealed a trigonal bipyramidal zinc center with one perchlorate anion bound to the metal center through one oxygen atom and an outer-sphere perchlorate anion (Figure 2.7). There were no hydrogen-

17 bonding interactions between the amino moieties of the secondary coordination sphere and the bound perchlorate anion; however, one amino moiety was engaged in hydrogen-bonding with the outer-sphere perchlorate anion. Zn-ClO4 was robust and did not display any perchlorate deoxygenation. As a result, we hypothesized that the role of the secondary coordination sphere was not in facilitating coordination of the perchlorate anion to the metal center, but was instead critical in stabilizing the iron(III)-oxo formed upon perchlorate deoxygenation. This stabilization was evidenced by the three hydrogen-bonding interactions between the amino moieties of the III ligand and the oxo moiety in the solid-state structure of [Fe -O]ClO4.

2.4 Chlorine, bromine, and iodine oxyanion reduction Cy Following the reduction of perchlorate upon the addition of N(pi )3 to Fe(ClO4)2, we sought to further investigate the ability of our non-heme system to reduce oxyanions. Because the Cy II iron(II)-triflate species, N(afa )3Fe(OTf)2 (Fe -OTf), had previously demonstrated the one- electron reduction of nitrite to nitric oxide with concomitant formation of FeIII-O,22 we probed its reactivity toward the family of halogen oxyanions. Ultimately, we observed the deoxygenation of - - - - - perchlorate (ClO4 ), chlorate (ClO3 ), chlorite (ClO2 ), hypochlorite (ClO ), bromate (BrO3 ), iodate - - III (IO3 ), and periodate (IO4 ) to generate Fe -O and the corresponding iron(II)-halide complex, Cy II Cy II Cy II [N(afa )3FeCl]OTf (Fe -Cl), [N(afa )3FeBr]OTf (Fe -Br), or [N(afa )3FeI]OTf (Fe -I). II The addition of [NBu4][ClO4] to a yellow solution of Fe -OTf resulted in an immediate color change to deep red-brown. Analysis of the reaction mixture by 1H NMR spectroscopy Cy revealed a mixture of products, akin to those observed upon the addition of N(pi )3 to Fe(ClO4)2: III III Cy Fe -O, Fe -OH, and N(pi )3 • 3 HOTf. Consistent with previous results, when the reaction was repeated in the presence of triethylamine, the 1H NMR spectrum of the reaction mixture revealed a mixture of FeIII-O and FeII-Cl (Figure 2.9). In addition, triethylammonium triflate was formed, suggesting that acid was generated during the course of the reduction of [NBu4][ClO4].

18 Cy Cy 2 OTf Cy 3 OTf Cy Cy OTf Cy Cy Cy Cy HN HN HN NH HN HN NH NH HN O HO Fe + Fe + N N N N N N NH HNHN ] O 4 OTf Cl N N N Cy ][ O CF 4 Cy 3 Cy [NBu III III Cy S NH Fe -O Fe -OH N(pi )3 • HOTf HN NH O O

N Fe N N [NBu N 4 ][Cl NEt O Cy OTf Cy OTf 4 Cy Cy Cy 3 ] Cy FeII-OTf HN NH NH HN HN NH O Cl Fe + Fe N N N N N N

N N

FeIII-O FeII-Cl

II Figure 2.9 Addition of [NBu4][ClO4] to Fe -OTf in the absence (top) and presence (bottom) of NEt3.

We hypothesized that the reduction of perchlorate proceeded by stepwise deoxygenation from perchlorate to chlorate, chlorite, hypochlorite, and finally chloride. To support this hypothesis, we added each chlorine oxyanion to a solution of FeII-OTf in the presence of triethylamine. In each case, we observed the same mixture of products, FeIII-O and FeII-Cl (Figure 2.10). The ratio of these products was further investigated to affirm that both the oxo and the chloride originated from the chlorine oxyanion. Poor solubility of FeII-Cl precluded quantification of the complexes by integration of the 1H NMR spectrum of the reaction mixture. Thus, we relied on the reaction of FeIII-O with 1,2-diphenylhydrazine (DPH) to generate azobenzene to quantify the two products. Because FeII-Cl did not react with DPH, the addition of DPH (in equimolar amounts based on FeII-OTf) to the chlorine oxyanion reduction mixture would yield a mixture of azobenzene and unreacted DPH whose ratio would match that of the oxygen and chlorine present in the starting oxyanion, assuming complete reduction. This method successfully demonstrated the ratios of oxygen and chlorine present in perchlorate, chlorate, and hypochlorite; the ratio of azobenzene to DPH quantified with 1H NMR spectroscopy were 4.36:1, 2.54:1, and 0.67:1, respectively. Although chlorite is reduced to FeIII-O and FeII-Cl by FeII-OTf, the ratio of these products was not quantified because the highest commercially available technical grade of chlorite is 80% and chlorite is not readily purified from other chlorine oxyanion impurities.

19 OTf Cy Cy Cy Cy Cy OTf Cy OTf O CF Cy Cy 3 Cy HN NH S NH NH HN NH HN NH ClO - HN O O x O Cl NEt (x + 1) 3 Fe x Fe + 1 Fe N N N x = 1, 3, or 4 N N N N N N

N N N FeII-OTf FeIII-O FeII-Cl Figure 2.10 Chlorine oxyanion reduction by FeII-OTf to yield FeIII-O and FeII-Cl.

Following the reduction of the chlorine oxyanions to generate FeIII-O and chloride, we explored the deoxygenation of the bromine and iodine oxyanions to form FeIII-O and an iron(II)- halide. Due to the instability and lack of commercial availability of many of the bromine and iodine oxyanions, the anions studied were bromate, iodate, and periodate. The addition of bromate to FeII-OTf in the presence of triethylamine resulted in the formation of FeIII-O and FeII-Br (Figure 2.11). The ratio of these two products was quantified using the reaction of DPH with FeIII-O, as was used for the chlorine oxyanions. The integration of azobenzene versus DPH in the reaction mixture when DPH was added (equimolar to FeII-OTf) was 3.30:1, consistent with the complete reduction of bromate in our non-heme iron system.

Cy Cy OTf Cy Cy OTf Cy Cy HN NH NH HN NH NH O Br x Fe + 1 Fe N N N N N N

- N OTf N Cy O x O CF Br Cy 3 Cy III II S NH x = 3 Fe -O Fe -Br HN NH O O

(x + 1) N Fe N N

N IO Cy Cy Cy OTf Cy OTf x = 3 orx 4- Cy Cy FeII-OTf HN NH NH HN HN NH O I x Fe + 1 Fe N N N N N N

N N

FeIII-O FeII-I Figure 2.11 Bromine and iodine oxyanion reduction by FeII-OTf to yield FeIII-O and an iron(II)-halide.

The reductions of iodate and periodate by FeII-OTf proceeded similarly in the presence of triethylamine to furnish FeIII-O and FeII-I (Figure 2.11); however, the reduction of iodate required the addition of 18-crown-6 to improve the solubility of the oxyanion in acetonitrile. Quantification

20 of the FeIII-O and FeII-I produced could not be completed by the addition of DPH to the reaction because both iodate and periodate react readily with DPH. Instead, DPH (equimolar to FeII-OTf) was added after the reaction mixture was worked up to remove any residual oxyanion salt and triethylammonium triflate. Integration of the 1H NMR spectrum allowed for the ratio of azobenzene to unreacted DPH to be determined. In the case of iodate, the ratio of azobenzene to DPH indicated 2.61 equiv FeIII-O (3.00 equiv expected) had been generated; in the case of periodate, 4.33 equiv FeIII-O (4.00 equiv expected) had been generated.

2.5 Catalytic perchlorate reduction Because FeII-OTf could be regenerated from FeIII-O using DPH, we explored the possibility that perchlorate reduction could be made catalytic. FeII-OTf was stirred with

[NBu4][ClO4] and 4 equiv DPH overnight (Figure 2.12). We proposed that the complete conversion of DPH to azobenzene in this reaction would correspond to a turnover number of 4 because 4 deoxygenation steps are required for the release of chloride from perchlorate and the catalyst must be regenerated after each deoxygenation. Analysis of the crude reaction mixture by 1H NMR spectroscopy revealed 3.4 equiv of DPH had been converted to azobenzene and the formation of water from this conversion was confirmed by Karl Fischer titration. In the subsequent workup, FeII-Cl was crystallized in 75% yield. The yield of of FeII-Cl and the conversion of DPH to azobenzene corresponded to a turnover number of 3. Although meager, the catalytic reduction of perchlorate is a difficult reaction and few synthetic systems are competent for this transformation. Attempts to achieve higher turnover numbers were thwarted by the formation of FeII-Cl. II When testing the stoichiometric reactivity of Fe -Cl with [NBu4][ClO4], we observed incomplete conversion of the iron(II) complex to FeIII-O, with FeII-Cl remaining even after stirring the reaction overnight. In an effort to circumvent this problem, we tested several chloride abstraction reagents to remove the axial chloride ligand from FeII-Cl and generate an iron(II) center that would be more reactive toward perchlorate. AgOTf was added to FeII-Cl, but chloride abstraction to form AgCl was incomplete, and we were concerned about possible off-pathway oxidation of any iron(II) species in solution by silver. TlPF6 was also tested to relieve the possibility of oxidative side reactions, but did not fully abstract the chloride from FeII-Cl. Trimethylsilyl trifluoromethanesulfonate (TMSOTf) succeeded in complete chloride abstraction from FeII-Cl to

21 generate FeII-OTf and TMSCl, but was incompatible under the catalytic reaction conditions. The use of NaClO4 and triphenylsilyl perchlorate (Ph3SiOClO3) as the source of perchlorate were also tested. We proposed they may form NaCl or Ph3SiCl as byproducts and thereby increase turnover of the catalyst; however, neither perchlorate reagent improved the catalysis. NaClO4 was much II less soluble in acetonitrile than [NBu4][ClO4], and Ph3SiOClO3 did not prevent formation of Fe - Cl.

- + [NBu4][ClO4] + e

- [NBu4][OTf]

Cy Cy OTf OTf Cy Cy OTf Cy Cy Cy O CF HN NH Cy 3 Cy NH HN NH S NH HN HN NH O Cl O O Fe + Fe N N N N N N N Fe N N

N N N FeII-OTf FeIII-O FeII-Cl

+ e- + 2 H+

- H2O

2 H+ + 2 e- = DPH Figure 2.12 Catalytic perchlorate reduction by FeII-OTf.

Furthermore, while we were unable to determine whether the reaction proceeds by sequential two-electron reductions, from perchlorate to chlorate to chlorite to hypochlorite to chloride, we hypothesize that the reaction proceeds through four steps because four oxygen atoms must be transferred from perchlorate to generate chloride. Moreover, the stoichiometric reaction of each chlorine oxyanion yielded the same mixture of products, FeIII-O and FeII-Cl, in the expected ratios, which further asserts a four step process to generate chloride from perchlorate.

Cy Cy Cy 2.6 Synthesis of [N(afa )3FeCl]OTf, [N(afa )3FeBr]OTf, and [N(afa )3FeI]OTf The series of iron(II)-halide complexes was independently synthesized and structurally characterized to confirm the formulation of the iron(II) products of halogen oxyanion reduction. Cy II The triflate salt of the iron(II)-chloride complex [N(afa )3FeCl]OTf (Fe -Cl) was furnished

22 Cy through two routes: (1) addition of a dichloromethane solution of N(pi )3 to an acetonitrile solution of FeClOTf generated in situ from the salt metathesis of 1 equiv FeCl2 and 1 equiv AgOTf Cy and (2) combining 1 equiv N(pi )3, 0.5 equiv FeCl2, and 0.5 equiv Fe(OTf)2(MeCN)2 in tetrahydrofuran (Figure 2.13). Crystals suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a concentrated solution of the target molecule in dichloromethane. Refinement of the data revealed a trigonal bipyramidal iron(II) center with an axial chloride ligand (Figure 2.14).

Cy Cy OTf Cy Cy Cy Cy N NH N N HN NH 0.5 FeCl2 Cl 0.5 Fe(OTf) (MeCN) 2 2 Fe NH HNHN N N N

N N 0.5 Fe(OTf) 0.5 Fe Cy II N(pi )3 Br Fe -Cl 2 2 (MeCN) 0.5 FeI2 0.5 Fe(OTf) (MeCN) 2 2 2 Cy OTf Cy OTf Cy Cy Cy Cy NH NH NH NH NH HN Br I Fe Fe N N N N N N

N N

FeII-I FeII-Br Figure 2.13 Synthesis of the iron(II)-halide complexes: FeII-Cl, FeII-Br, and FeII-I.

The solid-state structure of FeII-Cl exhibited positional disorder in two arms of the ligand platform, where the amine moiety of the secondary coordination sphere was either rotated inward to hydrogen-bond with the chloride ligand or rotated outward, pointing away from the chloride

Br1 I1 H5 Cl1 H7 H5 H7 H5 H6 H7 H6 Fe1 Fe1 H6 Fe1

FeII-Cl FeII-Br FeII-I

Figure 2.14 Molecular structures of FeII-Cl, FeII-Br, and FeII-I shown with 50% probability ellipsoids. Select hydrogen atoms and outer-sphere anions have been omitted for clarity.

23 ligand. One arm had equal occupancy (50% each) in each position, while the second disordered arm had 18% occupancy of the ligand arm rotated inward and 82% occupancy of the arm rotated outward. In the latter case, the outwardly rotated arm allowed for hydrogen-bonding between the secondary sphere amine and the outer-sphere triflate anion. The rotation of the arms of the ligand was reflected in two C=N stretching frequencies observed by IR spectroscopy at 1649 and 1632 cm-1. These values also confirmed that each arm was in the azafulvene-amine binding mode. The II solution magnetic moment of Fe -Cl was determined to be 5.41(12) µB using the Evan’s Method, consistent with a high-spin iron(II) center.

Table 2.1 Selected structural parameters of FeII-Cl, FeII-Br, and FeII-I. FeII-Cl FeII-Br FeII-I Fe–X (Å) 2.3868(4) 2.5351(6) 2.918(3) Fe–N (Å) 2.2668(14) 2.250(3) 2.249(6) 2.0834(15) 2.133(3) 2.097(5)

Fe–Nafa (Å) 2.1144(15) 2.121(3) 2.098(6) 2.1142(14) 2.083(3) 2.110(5)

N–H stretch (cm-1) 3292, 3229 3288, 3225 –

C=N stretch (cm-1) 1649, 1632 1643, 1634 –

µeff (µB) 5.41(12) – –

Cy II The triflate salt of the iron(II)-bromide complex [N(afa )3FeBr]OTf (Fe -Br) was Cy synthesized from the addition of 1 equiv N(pi )3 to 0.5 equiv FeBr2 and 0.5 equiv

Fe(OTf)2(MeCN)2 in tetrahydrofuran (Figure 2.13). Crystals suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a concentrated solution of the target molecule in acetonitrile and dichloromethane (1:1). Refinement of the data revealed an iron(II) center in a trigonal bipyramidal geometry with bromide bound axially to the iron center and an outer-sphere triflate anion (Figure 2.14). Positional disorder in the ligand platform was also observed in this structure. One arm of the ligand had 64% occupancy when rotated inward to hydrogen-bond with the bromide ligand and 36% occupancy when rotated outward to hydrogen-bond with the outer- sphere triflate anion. The IR spectrum of the complex contained two C=N stretching frequencies

24 at 1643 and 1634 cm-1, which were consistent with the azafulvene-amine binding mode of the ligand. Determination of the solution magnetic moment of FeII-Br was precluded by its poor solubility. Cy II Similarly, the triflate salt of the iron(II)-iodide complex [N(afa )3FeI]OTf (Fe -I) was Cy synthesized from the addition of 1 equiv N(pi )3 to 0.5 equiv FeI2 and 0.5 equiv Fe(OTf)2(MeCN)2 in tetrahydrofuran (Figure 2.13). Crystals grown from the vapor diffusion of diethyl ether into a saturated solution of the target molecule in acetonitrile were structurally characterized. Refinement of the data revealed an iron(II) center in a trigonal bipyramidal geometry with substitutional disorder in the axial ligand, which was occupied by either a triflate anion or an iodide (Figure 2.14). The iodide had a partial occupancy of 46%, while the triflate anion was disordered over two positions, which had occupancies of 29% and 25%. Because of the substitutional disorder, this complex was not further characterized. However, the 1H NMR spectrum of this complex was collected, and the chemical shifts of the observed resonances matched those of FeII-OTf, but were significantly broadened (Figure 2.15).

II Fe -OTf

II Fe -I

1 II II Figure 2.15 H NMR spectra of Fe -OTf and Fe -I in CD3CN at room temperature.

2.7 Conclusions The reduction of the chlorine, bromine, and iodine oxyanions by FeII-OTf was investigated. Each halogen oxyanion was fully reduced to FeIII-O and the corresponding iron(II)- halide complex. We hypothesized that reductions occurred through sequential deoxygenations of

25 the substrate until the halide anion was released and subsequently performed anion metathesis with the FeII-OTf remaining in solution. The series of iron(II)-halide complexes was independently synthesized and characterized by solid-state structural analysis to confirm the formulation of the iron(II) products from the reduction of the oxyanions. In the solid-state structures of FeII-Cl and FeII-Br, hydrogen-bonding interactions were observed between the secondary sphere amine moieties and the bound halide ligands. In order to investigate the role of the secondary coordination sphere in the reduction of the halogen oxyanions, Zn-ClO4 was synthesized and crystallographically characterized. The solid- state structure revealed binding of perchlorate to the metal center, but no hydrogen-bonding interactions between the bound substrate and the secondary sphere. Thus, we concluded the secondary sphere was critical in stabilizing the iron(III)-oxo formed from reduction of the substrate, as the oxo moiety engages in hydrogen-bonding to all three secondary sphere amine groups, and was not needed to facilitate substrate binding to the metal center. The catalytic reduction of perchlorate was achieved, but attempts to improve the catalysis were unsuccessful because of the difficulty of abstracting the chloride from FeII-Cl. As a result, the catalyst was gradually inactivated as the concentration of chloride in solution increased. Although the turnover number recorded was meager, this work demonstrated the ability of our system to facilitate multi-electron processes relevant to biological systems and to reduce perchlorate to the benign products chloride and water.

2.8 Experimental General considerations. To avoid contact with oxygen and water, all manipulations were carried out under an atmosphere of nitrogen in an MBraun inert atmosphere drybox or using standard Schlenk techniques. Solvents for air- and moisture-sensitive manipulations were dried and deoxygenated using a Glass Contour System and stored over 4 Å molecular sieves prior to use. Celite 545 was heated to 150 °C under dynamic vacuum for 24 h prior to use in the drybox. All reagents were purchased from commercial sources and used as received unless otherwise noted. Tetrabutylammonium perchlorate was recrystallized from a saturated solution in tetrahydrofuran layered with hexane and stored at -35 °C. Potassium bromate was recrystallized from a saturated aqueous solution stored at 0 °C, filtered, washed with methanol, and dried under vacuum at 150 °C. 1,2-Diphenylhydrazine was recrystallized from a saturated solution in diethyl ether layered

26 with hexane and stored at -35 °C. Triethylamine was distilled and stored over 4 Å molecular sieves 24 25 Cy 23 at -35 °C prior to use. Ferrous trifluoromethanesulfonate, triphenylsilyl perchlorate, N(pi )3, Cy 23 and N(afa )3Fe(OTf)2 were synthesized according to literature procedure. NMR solvent

(acetonitrile-d3) was degassed and stored over 4 Å molecular sieves prior to use. NMR spectra were recorded at ambient temperature on a Varian spectrometer operating at 500 MHz (1H NMR), 126 MHz (13C NMR), and 470 MHz (19F NMR) and referenced to the peak of the residual solvent (δ parts per million and J in Hz). Solid-state infrared spectra were measured using a PerkinElmer Frontier FT-IR spectrophotometer equipped with a KRS5 thallium bromide/iodide universal attenuated total reflectance accessory. UV-vis spectra were recorded using a HP 8453 UV-vis spectrophotometer. Elemental analyses were performed by the University of Illinois at Urbana- Champaign School of Chemical Sciences Microanalysis Laboratory in Urbana, IL. The quantification of water was performed by an Aquatest CMA Karl Fischer Coulometric Titrator from Photovolt Instruments with HYDRANAL from Fluka Analytical. Solid-state structural characterization was done using a Bruker D8 Venture Duo or Bruker X8ApexII diffractometer at the George L. Clark X-Ray Facility and 3M Material Laboratory at the University of Illinois at Urbana-Champaign. WARNING: Perchlorate salts are explosive and should be handled with care in small quantities using a plastic spatula.

Cy III Preparation of [N(afa )3FeO]ClO4 ([Fe -O]ClO4). A 20 mL scintillation vial was charged with

FeCl2 (0.0100 g, 0.0789 mmol) and 5 mL of acetonitrile. In the dark, AgClO4 (0.0327 g, 0.158 mmol) was added, leading to the formation of a white precipitate. The slurry was stirred for 30 Cy min before being filtered over Celite into a 2 mL dichloromethane solution of N(pi )3 (0.0459 g, 0.0789 mmol). After 10 min, 1 drop of triethylamine was added (when triethylamine was not Cy Cy added, the crude reaction mixture contained [N(afa )3FeO]ClO4, [N(afa )3FeOH](ClO4)2, and Cy N(pi )3 • 3 HClO4; Figure 2.16). Stirring was continued for 1 h at room temperature during which time the solution turned dark brown. Volatiles were removed under reduced pressure and the resulting brown powder was washed with diethyl ether. The 1H NMR spectrum of the crude Cy Cy reaction mixture revealed a mixture of [N(afa )3FeO]ClO4 and [N(afa )3FeCl]ClO4. Cy [N(afa )3FeO]ClO4 crystallized from the vapor diffusion of diethyl ether into a concentrated solution of acetonitrile (0.0352 g, 0.0467 mmol, 59%). Crystals suitable for X-ray diffraction were

27 grown by layering a concentrated solution of acetonitrile and dichloromethane (1:1) with diethyl ether. Analysis for C36H51N7FeClO5•½CH2Cl2: Calcd. C, 55.10; H, 6.59; N, 12.32. Found C, -1 55.10; H, 6.46; N, 12.31%. IR = 1664 cm (C=N). µeff = 6.12(8) µB.

●

Cy Fe(ClO ) + N(pi ) * 4 2 3 ▲ ▲ * ▲ *

Cy Fe(ClO ) + N(pi ) + NEt 4 2 3 3 ◼ ◼ * * *

[FeIII-O]ClO 4 * * *

Cy ClO Cy 3 ClO4 Cy 2 ClO4 Cy ClO4 Cy Cy 4 Cy Cy Cy Cy Cy Cy HN HN HN NH NH HN NH HN HN NH HN NH O HO Cl

Fe Fe Fe N N N N NH HNHN N N N N N

N N N N * ● ▲ ◼

1 III Figure 2.16 H NMR spectra of the reactions forming [Fe -O]ClO4 without the addition of NEt3 (top) and with the addition of NEt3 (middle) and a recrystallized sample of III [Fe -O]ClO4 (bottom) (CD3CN, 21 °C).

Cy Preparation of [N(afa )3ZnClO4]ClO4 (Zn-ClO4). A 20 mL scintillation vial was charged with

ZnCl2 (0.0100 g, 0.0734 mmol) and 5 mL of acetonitrile. AgClO4 was added (0.0380 g, 0.183 mmol), resulting in the formation of a white precipitate. After stirring for 30 min, the suspension Cy was filtered over Celite. To the colorless solution, N(pi )3 (0.0426 g, 0.0734 mmol), dissolved in 2 mL dichloromethane, was added. The solution was stirred for 3 h and volatiles were removed

28 under reduced pressure, yielding the product as an off-white solid. Crystals suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a saturated dichloromethane solution (0.0402 g, 0.0476 mmol, 65% crystalline yield). Analysis for C36H51N7ZnO8•CH2Cl2:

11.0 10.5 10.0 9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 ppm 0 6 7 6 6 6 4 7 4 9 2 2 3 4 5 4 3 ...... 5 0 6 5 4 6 0 3 7 0 4 7 6 2 0 4 8 6 . . . . . 7 6 4 2 1 9 7 5 5 3 2 1 1 1 3 3 5 5 ...... 1 1 1 1 1 1 6 5 3 2 2 1 1 1 1 1 0 0

170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 ppm 1 13 Figure 2.17 H (top) and C (bottom) NMR spectra of Zn-ClO4 (CD3CN, 21 °C).

29 Calcd. C, 47.73; H, 5.74; N, 10.53. Found C, 48.15; H, 5.51; N, 10.76%. IR = 3301 and 3252 cm- 1 (N-H), 1639 cm-1 (C=N).

General procedure for stoichiometric chlorine oxyanion reductions. A 20 mL scintillation vial Cy was charged with N(afa )3Fe(OTf)2 (0.0491 g, 0.0525 mmol) and 5 mL of acetonitrile. The - chlorine oxyanion salt was added as a solid [ClOx = Ca(OCl)2 (0.0018 g, 0.0125 mmol); NaClO2

(0.0015g, 0.0167 mmol); NaClO3 (0.0013 g, 0.0125 mmol); [NBu4][ClO4] (0.0034 g, 0.0100 mmol)], followed by triethylamine (0.0051 g, 0.0500 mmol) dissolved in 2 mL acetonitrile. The resulting solution was stirred for 16 h. Volatiles were removed under reduced pressure and the resulting orange-brown powder was washed with diethyl ether before being analyzed by 1H NMR spectroscopy. The 1H NMR spectrum of the reduction of each oxyanion contained a mixture of Cy Cy [N(afa )3FeO]OTf and [N(afa )3FeCl]OTf. When the reaction was completed without the

II 4 Fe -OTf + [NBu ][ClO ] + NEt 4 4 3 * * ◼ ◼ *

●

II 4 Fe -OTf + [NBu ][ClO ] 4 4 ◼ ▲ * ▲ * ◼ ▲ *

Cy OTf Cy 3 OTf Cy 2 OTf Cy OTf Cy Cy Cy Cy Cy Cy Cy Cy HN HN HN NH NH HN NH HN HN NH HN NH O HO Cl

Fe Fe Fe N N N N NH HNHN N N N N N

N N N N * ● ▲ ◼

Figure 2.18 A representative example of chlorine oxyanion reduction: the addition of 1 eq II [NBu4][ClO4] to 4 eq Fe -OTf with NEt3 (top) or without NEt3 (bottom) (CD3CN, 21 °C).

30 Cy Cy addition of triethylamine, a mixture of [N(afa )3FeO]OTf, [N(afa )3FeOH](OTf)2, Cy Cy 1 [N(afa )3FeCl]OTf, and (N(pi )3 • 3 HOTf were observed by H NMR spectroscopy.

General procedure for the quantification of stoichiometric chlorine oxyanion reduction Cy products. A 20 mL scintillation vial was charged with N(afa )3Fe(OTf)2 (0.0491 g, 0.0525 mmol), 1,2-diphenylhydrazine (0.0092 mg, 0.0500 mmol), 20 mg MgSO4, the chlorine oxyanion salt [Ca(OCl)2 (0.0018 g, 0.0125 mmol); NaClO3 (0.0013 g, 0.0125 mmol); [NBu4][ClO4] (0.0034 g, 0.0100 mmol)], and 6 mL of acetonitrile. The solutions were stirred for 16 h before volatiles were removed under reduced pressure. The yellow-orange solid obtained was washed several times with diethyl ether to remove the unreacted 1,2-diphenylhydrazine and azobenzene. The remaining

OTf H Cy N Cy O CF3 Cy (x + 1) N S NH H HN NH O O ClO - (x + 1) x N Fe N N N - x N N x = 1, 3, and 4

H N N N N H - ClO 7 0 6 0 . . 0 1

- ClO 3 6 0 9 0 . . 1 1

- ClO 4 6 0 3 0 . . 4 1 8.4 8.3 8.2 8.1 8.0 7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 7.1 7.0 6.9 6.8 6.7 6.6 6.5 6.4 6.3 6.2 6.1 6.0 ppm Figure 2.19 Quantification of stoichiometric chlorine oxyanion reduction by 1H NMR spectroscopy demonstrating the ratio of azobenzene:1,2-diphenylhydrazine matches the III II expected ratio of Fe -O:Fe -Cl (CD3CN, 21 °C).

31 orange solid was eluted with 2 mL of acetonitrile and 2 mL dichloromethane. The ratio of azobenzene to unreacted 1,2-diphenylhydrazine in the diethyl ether washes were quantified by 1H NMR spectroscopy using ferrocene as an internal standard (0.0047 g, 0.0250 mmol).

Procedure for Karl Fischer titration of stoichiometric perchlorate reduction. A 20 mL Cy scintillation vial was charged with N(afa )3Fe(OTf)2 (0.0294 g, 0.0315 mmol), [NBu4][ClO4] (0.0021 g, 0.0100 mmol), 1,2-diphenylhydrazine (0.0055 g, 0.0300 mmol), and 4 mL of acetonitrile. The vial was sealed with tape and the solution was stirred for 16 h. Three aliquots of 0.4 mL were removed from the vial using a gas-tight syringe and analyzed immediately by Karl Fischer titration. The background water content was analyzed by repeating the reaction without Cy N(afa )3Fe(OTf)2 (Table 2.2).

II Table 2.2 Quantification of H2O produced from the reaction of 4 eq Fe -OTf, 1 eq [NBu4][ClO4], and 4 eq 1,2-diphenylhydrazine by Karl Fischer titration (in µg H2O).

II 4 Fe -OTf + [NBu4][ClO4] + 4 DPH 10.3 Background 8.7 (1.0 mL aliquot; µg H2O) 8.4 Average: 9.1 ± 1.0 27.4 Reaction Mixture 33.1 ( 0.4 mL aliquot; µg H2O) 28.3 Average: 29.6 ± 3.1 Total: 26.0 ± 3.3

Equiv H2O Detected (Expected): 2.56 (4.00)

Stoichiometric reduction of bromate and quantification of bromate reduction products. A 20 Cy mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), KBrO3

(0.0025 g, 0.0150 mmol), NEt3 (0.0061 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was stirred 16 h, during which time the color changed from yellow to red-brown. Volatiles were removed under reduced pressure and the resulting brown powder was triturated with diethyl ether 1 Cy and analyzed by H NMR spectroscopy, which showed a mixture of [N(afa )3FeO]OTf and Cy [N(afa )3FeBr]OTf.

32 Cy Cy OTf Cy Cy OTf Cy Cy HN NH NH HN NH NH O Br Fe Fe N N N N N N

N N * ◼

* * * ◼ ◼ ◼

1 Figure 2.20 H NMR spectrum of the products of bromate reduction (CD3CN, 21 °C).

The quantification of bromate reduction products was performed as follows; a 20 mL scintillation Cy vial was charged with N(afa )3Fe(OTf)2 (0.0561 g, 0.0600 mmol), KBrO3 (0.0025 g, 0.0150 mmol), 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol), MgSO4 (approximately 15 mg), and 4 mL acetonitrile. The mixture was stirred 16 h, with the solution changing from yellow to pale orange. Volatiles were removed under reduced pressure and the isolated solid was washed with 3 mL diethyl ether to remove azobenzene and 1,2-diphenylhydrazine. The orange diethyl ether solution was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol).

33 OTf H Cy N Cy O CF3 Cy 4 N S NH H HN NH O O KBrO3 4 N Fe N N N - 3 N N

H N N N N H

Figure 2.21 Quantification of bromate reduction by 1H NMR spectroscopy revealed the ratio III II of azobenzene:1,2-diphenylhydrazine matches the expected ratio of Fe -O:Fe -Br (CD3CN, 21 °C).

Stoichiometric iodate reduction and the quantification of iodate reduction products. A 20 Cy mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), KIO3 (0.0032 g, 0.0150 mmol), 18-crown-6 (0.0040 g, 0.0150 mmol), NEt3 (0.0061 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was stirred 16 h, during which time the color changed from yellow to red-brown. Volatiles were removed under reduced pressure and the resulting brown powder was triturated with diethyl ether and analyzed by 1H NMR spectroscopy, which showed a mixture of Cy Cy [N(afa )3FeO]OTf and [N(afa )3FeI]OTf.

34 Cy OTf Cy OTf Cy Cy Cy Cy HN NH NH HN HN NH O I Fe Fe N N N N N N

N N * ◼

◼ ◼ * * *

1 Figure 2.22 H NMR spectrum of the products of iodate reduction (CD3CN, 21 °C).

The quantification of iodate reduction products was performed as follows; a 20 mL scintillation Cy vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), KIO3 (0.0032 g, 0.0150 mmol),

18-crown-6 (0.0040 g, 0.0150 mmol), NEt3 (0.0061 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was stirred 16 h, during which time the color changed from yellow to red-brown. Volatiles were removed under reduced pressure and the resulting brown powder was washed with diethyl ether. The remaining brown powder was filtered and collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. The brown solution was evaporated to dryness and combined with 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol) in 4 mL acetonitrile. The solution was stirred 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2- diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a ratio of 1.21 azobenzene to 2.49 1,2-diphenylhydrazine. This corresponds to 33% conversion of Cy 1,2-diphenylhydrazine, which is equivalent to the reduction of 2.61 eq [N(afa )3FeO]OTf.

35 Fc

H N N H N N

Figure 2.23 Quantification of iodate reduction products by 1H NMR spectroscopy demonstrated the formation of 2.61 eq FeIII-O from iodate reduction based on the ratio of azobenzene:1,2-diphenylhydrazine (CD3CN, 21 °C).

Stoichiometric periodate reduction and the quantification of periodate reduction products. Cy A 20 mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), NaIO4

Cy OTf Cy OTf Cy Cy Cy Cy HN NH NH HN HN NH O I Fe Fe N N N N N N

N N * ◼

* * * ◼ ◼

1 Figure 2.24 H NMR spectrum of the products of periodate reduction (CD3CN, 21°C).

36 Cy A 20 mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), NaIO4

(0.0026 g, 0.0120 mmol), NEt3 (0.0061 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was stirred 16 h, during which time the color changed from yellow to red-brown. Volatiles were removed under reduced pressure and the resulting brown powder was washed with diethyl ether. The remaining brown powder was filtered and collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. The brown solution was evaporated to dryness and combined with 1,2- diphenylhydrazine (0.0111 g, 0.0600 mmol) and 4 mL acetonitrile. The solution was stirred 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2- diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a ratio of 1.65 azobenzene to 2.17 1,2-diphenylhydrazine. This corresponds to 43% conversion of Cy 1,2-diphenylhydrazine, which is equivalent to the reduction of 4.32 eq [N(afa )3FeO]OTf.

H N N N N H

Figure 2.25 Quantification of periodate reduction products by 1H NMR spectroscopy demonstrated the formation of 4.32 eq FeIII-O from periodate reduction based on the ratio of azobenzene:1,2-diphenylhydrazine (CD3CN, 21 °C).

37 Cy Catalytic perchlorate reduction. A 20 mL scintillation vial was charged with N(afa )3Fe(OTf)2

(0.0281 g, 0.0300 mmol), [NBu4][ClO4] (0.0103 g, 0.0300 mmol), 1,2-diphenylhydrazine (0.0221 g, 0.120 mmol), and 15 mg MgSO4. Following the addition of 5 mL acetonitrile, the solution was stirred 16 h. The resulting yellow-orange solution was filtered over Celite and volatiles were removed under reduced pressure. 1H NMR spectroscopy, using ferrocene (0.0112 mg, 0.0600 mmol) as an internal standard, was used to determine the conversion 1,2-diphenylhydrazine to azobenzene in the crude reaction mixture, which was 85%. The remaining orange powder was dissolved in 2 mL tetrahydrofuran and the concentrated solution was layered with 8 mL diethyl Cy ether. The vial was stored at -35 °C for 16 h, during which time [N(afa )3FeCl]OTf crystallized Cy from the solution. [N(afa )3FeCl]OTf was isolated by filtration, washed with 4 mL diethyl ether, eluted with 5 mL dichloromethane, and dried under reduced pressure (0.0186 g, 0.0226 mmol, 75%).

N N

H N N H

Figure 2.26 Quantification of the conversion of 1,2-diphenylhydrazine to azobenzene during 1 catalytic perchlorate reduction via H NMR spectroscopy (CD3CN, 21 °C).

38 Figure 2.27 1H NMR spectrum of FeII-Cl isolated following catalytic perchlorate reduction (CD3CN, 21 °C).

Cy Chloride abstraction from [N(afa )3FeCl]OTf with AgOTf. A 20 mL scintillation vial was Cy charged with [N(afa )3FeCl]OTf (0.0206 g, 0.0250 mmol), AgOTf (0.0064 g, 0.0250 mmol), and 5 mL acetonitrile. The vial was taped to perform the reaction in the dark and prevent oxidative side-reactions. The solution was stirred for 16 h and filtered to remove AgCl before volatiles were removed under reduced pressure. The orange-yellow residue obtained was then analyzed by 1H Cy NMR spectroscopy and still contained [N(afa )3FeCl]OTf.

Cy Chloride abstraction from [N(afa )3FeCl]OTf with TlPF6. A 20 mL scintillation vial was Cy charged with [N(afa )3FeCl]OTf (0.0206 g, 0.0250 mmol), TlPF6 (0.0087 g, 0.0250 mmol), and 5 mL acetonitrile. The solution was stirred for 16 h and filtered to remove TlCl before volatiles were removed under reduced pressure. The orange-yellow residue obtained was then analyzed by 1 Cy H NMR spectroscopy and still contained [N(afa )3FeCl]OTf.

Cy Chloride abstraction from [N(afa )3FeCl]OTf with TMSOTf. A 20 mL scintillation vial was Cy charged with [N(afa )3FeCl]OTf (0.0206 g, 0.0250 mmol), TMSOTf (0.0056 g, 0.0250 mmol), and 5 mL acetonitrile. The solution was stirred for 16 h before volatiles were removed under reduced pressure. The orange-yellow residue obtained was then analyzed by 1H NMR Cy spectroscopy and contained no [N(afa )3FeCl]OTf.

39 II Fe -Cl

II Fe -Cl + AgOTf

II Fe -Cl + TlPF 6

II Fe -Cl + TMSOTf

Figure 2.28 1H NMR spectra of FeII-Cl (top) and chloride abstraction reactions with FeII-Cl demonstrating complete abstraction was only achieved with TMSOTf (CD3CN, 21 °C).

Cy Preparation of [N(afa )3FeCl]OTf. A 20 mL scintillation vial was charged with FeCl2 (0.0100 g, 0.0789 mmol) and 5 mL of acetonitrile. AgOTf was added (0.0203 g, 0.0789 mmol) to the vial in the dark and a white precipitate formed immediately. The suspension was stirred for 30 min Cy before being filtered over Celite. To the clear solution, N(pi )3 (0.04590 g, 0.0789 mmol) dissolved in 2 mL of dichloromethane was added dropwise and the color changed to orange. After stirring 1 h, the volatiles were removed under reduced pressure and the resulting pale orange solid was washed with diethyl ether to give the final product (0.0535 g, 0.0650 mmol, 82.4%). Crystals suitable for X-ray diffraction were grown from a concentrated solution of dichloromethane layered

40 with diethyl ether. Analysis for C37H51N7FeClSO3F3•½CH2Cl2: Calcd. C, 52.09; H, 6.06; N, 11.34. Found C, 52.10; H, 5.93; N, 11.36%. IR = 3292 and 3229 cm-1 (N-H), 1649 and 1632 cm-1 (C=N).

µeff = 5.41(12) µB.

Cy Alternative preparation of [N(afa )3FeCl]OTf. A 20 mL scintillation vial was charged with Cy FeCl2 (0.0050 g, 0.040 mmol), Fe(OTf)2(MeCN)2 (0.0162 g, 0.040 mmol), N(pi )3 (0.0458 g, 0.079 mmol), and 5 mL of tetrahydrofuran. The solution was stirred for 2 h, producing an orange suspension. Volatiles were removed under reduced pressure, leaving in a pale orange solid. The product was recrystallized from a saturated solution of dichloromethane layered with diethyl ether (0.0561 g, 0.068 mmol, 87% crystalline yield).

140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 -10 ppm 1 II Figure 2.29 H NMR spectrum of Fe -Cl (CD3CN, 21 °C).

Figure 2.30 Electronic absorption spectrum of FeII-Cl (0.01 mM in MeCN, 21 °C).

41 Cy Preparation of [N(afa )3FeBr]OTf. A 20 mL scintillation vial was charged with FeBr2 (0.0100 Cy g, 0.0464 mmol), Fe(OTf)2(MeCN)2 (0.0202 g, 0.0464 mmol), and 3 mL acetonitrile. N(pi )3 (0.0539 g, 0.0927 mmol) was dissolved in 2 mL dichloromethane and added dropwise to the acetonitrile solution. The solution was stirred 2 h at room temperature, during which time the color changed from pale yellow to orange. Volatiles were removed under reduced pressure. The resulting orange powder crystallized from the vapor diffusion of diethyl ether into a concentrated dichloromethane (2 mL) and acetonitrile (2 mL) solution to give the final product (0.0396 g, 0.0457 mmol, 49%). Crystals suitable for X-ray diffraction were grown using the same conditions. The poor solubility of the product precluded the determination of its solution magnetic moment by the

Evan’s Method. Analysis for C37H51N7FeBrSO3F3: Calcd. C, 51.28; H, 5.93; N, 11.31. Found C, 51.35; H, 5.80; N, 11.53%. IR = 3288 and 3225 cm-1 (N-H), 1643 and 1634 cm-1 (C=N).

1 II Figure 2.31 H NMR spectrum of Fe -Br (CD3CN, 21 °C).

Cy Preparation of [N(afa )3FeI]OTf. A 20 mL scintillation vial was charged with FeI2 (0.0200 g,

0.0646 mmol), Fe(OTf)2(MeCN)2 (0.0282 g, 0.0646 mmol), and 3 mL acetonitrile. The solution Cy was stirred and quickly turned deep purple. N(pi )3 (0.0752 g, 0.129 mmol) was dissolved in 2 mL dichloromethane in a separate vial and added dropwise to the acetonitrile solution. The reaction was continued for 2 h at room temperature, during which time the solution turned pale brown. Volatiles were removed under reduced pressure. The resulting powder was dissolved in 4 mL tetrahydrofuran, dried, and washed with 4 mL dichloromethane. Crystals suitable for X-ray diffraction were grown from a concentrated solution of acetonitrile layered with diethyl ether. Cy Refinement of the data revealed that the crystals were a mixture of N(afa )3Fe(OTf)2 and Cy [N(afa )3FeI]OTf, which precluded further characterization.

42 1 II Figure 2.32 H NMR spectrum of Fe -I (CD3CN, 21 °C).

43 III Table 2.3 Crystallographic Parameters of [Fe -O]ClO4 and Zn-ClO4. Cy Cy [N(afa )3FeO]ClO4 [N(afa )3ZnClO4]ClO4

Empirical Formula C36H51ClFeN7O5 C36H51Cl2N7O8Zn Formula Weight 753.14 846.10 Temperature 100 K 100 K Wavelength 0.71073 0.71073 Crystal System Monoclinic Orthorhombic Space Group P 2(1)/n P bca a = 12.355(2) Å a = 18.0429(5) Å Unit Cell Dimensions b = 23.342(4) Å b = 21.6526(6) Å

c = 14.136(3) Å c = 21.6723(5) Å

α = 90° α = 90°

β = 104.367(6)° β = 90°

γ = 90° γ = 90° Volume 3949.1(11) 8466.8(4) Z 4 8 Reflections Collected 118889 82242 Independent 8190 7775 Reflections Goodness-of-Fit on F2 1.052 1.102 Final R Indices R1 = 0.0338 R1 = 0.0515 [I>2sigma(I)] wR2 = 0.0804 wR2 = 0.1499 R1 = 0.0422 R1 = 0.0669 R Indices wR2 = 0.0848 wR2 = 0.1583

44 Table 2.4 Crystallographic Parameters of FeII-Cl, FeII-Br, and FeII-I.

Cy Cy Cy [N(afa )3FeCl]OTf [N(afa )3FeBr]OTf [N(afa )3FeI]OTf

Empirical Formula C37H51ClF3FeN7O3S C37H51BrF3FeN7O3S C37H51IF3FeN7O3S Formula Weight 822.21 857.66 904.66 Temperature 100 K 100 K 100 K Wavelength 0.71073 0.71073 0.71073 Crystal System Monoclinic Monoclinic Orthorhombic Space Group P 2(1)/c P 2(1)/c P bca a = 12.5299(5) Å a = 12.8435(4) Å a = 22.3195(12) Å b = 24.7035(9) Å b = 24.2569(6) Å b = 18.3777(10) Å Unit Cell Dimensions c = 13.2003(5) Å c = 13.2346(4) Å c = 22.4140(11) Å

α = 90° α = 90° α = 90°

β = 96.0698(14)° β = 98.2510(10)° β = 90° γ = 90° γ = 90° γ = 90° Volume 4063.0(3) 4080.5(2) 9193.8(8) Z 1 1 8 Reflections Collected 109519 34833 126468 Independent 8993 7518 8453 Reflections Goodness-of-Fit on F2 1.082 1.034 1.110

Final R Indices R1 = 0.0357 R1 = 0.0488 R1 = 0.0960 [I>2sigma(I)] wR2 = 0.0889 wR2 = 0.1134 wR2 = 0.2453 R1 = 0.0428 R1 = 0.0626 R1 = 0.1113 R Indices wR2 = 0.0935 wR2 = 0.1209 wR2 = 0.2575

45 2.9 References (1) Urbansky, E. T.; Schrock, M. R. Issues in Managing the Risks Associated with Perchlorate in Drinking Water. J. Environ. Manage. 1999, 56, 79-95.

(2) Srinivasan, R.; Sorial, G. A. Treatment of Perchlorate in Drinking Water: A Critical Review. Separ. Purif. Tech. 2009, 69, 7-21.

(3) Ye, L.; You, H.; Yao, J.; Su, H. Water Treatment Technologies for Perchlorate: A Review. Desalination 2012, 298, 1-12.

(4) Bardiya, N.; Bae, J.-H. Dissimilatory Perchlorate Reduction: A Review. Microbiol. Res. 2011, 166, 237-254.

(5) Kurokawa, Y.; Maekawa, A.; Takahashi, M.; Hayashi, Y. Toxicity and Carcinogenicity of Potassium Bromate: A New Renal Carcinogen. Environ Health Perspect. 1990, 87, 309- 335.

(6) Matos, C. T.; Velizarov, S.; Reis, M. A. M.; Crespo, J. G. Removal of Bromate from Drinking Water Using the Ion Exchange Membrane Bioreactor Concept. Environ. Sci. Technol. 2008, 42, 7702-7708.

(7) Chen, H.; Xu, Z.; Wan, H.; Zheng, J.; Yin, D.; Zheng, S. Aqueous Bromate Reduction by

Catalytic Hydrogenation over Pd/Al2O3 Catalysts. Appl. Catal., B 2010, 96, 307–313.

(8) Lyday, P. A.; Kaiho, T. Iodine and Iodine Compounds. In Ullmann's Encyclopedia of Industrial Chemistry, 7th ed.; Wiley-VCH: Weinheim, Germany, 2015; pp 1–13.

(9) Buist, G. J. The Oxidation of Organic Compounds by Non-Metallic Anions. In Comprehensive Chemical Kinetics; Bamford, C. H.; Tipper, C. F. H., Eds. Reactions of Non-Metallic Inorganic Compounds, Vol. VI; Elsevier: New York, 1972, pp 435-392.

(10) Leonard, N. J.; Johnson, C. R. Periodate Oxidation of Sulfides to Sulfoxides. Scope of the Reaction. J. Org. Chem. 1962, 27, 282-284.

(11) Choe, J. K.; Mehnert, M. H.; Guest, J. S.; Strathmann, T. J.; Werth, C. J. Comparative Assessment of the Environmental Sustainability of Existing and Emerging Perchlorate Technologies for Drinking Water. Environ. Sci. Technol. 2013, 47, 4644–4652.

(12) Brown, G. M.; Gu, B. The Chemistry of Perchlorate in the Environment. In Perchlorate; Gu, B.; Coates, J. D., Eds.; Springer: Boston, 2006; pp. 17–47.

(13) Abu-Omar, M. M. Effective and Catalytic Reduction of Perchlorate by Atom Transfer: Reaction Kinetics and Mechanism. Comm. Inorg. Chem. 2003, 24, 15-37.

(14) Abu-Omar, M. M.; McPherson, L. D.; Arias, J.; Béreau, V. M. Clean and Efficient Catalytic Reduction of Perchlorate. Angew. Chem. Int. Ed. 2000, 39, 4310-4313.

(15) Liu, J.; Han, M.; Wu, D.; Chen, X.; Choe, J. K.; Werth, C. J.; Strathmann, T. J. A New Bioinspired Perchlorate Reduction Catalyst with Significantly Enhanced Stability via Rational Tuning of Rhenium Coordination Chemistry and Heterogeneous Reaction Pathway. Environ. Sci. Technol. 2016, 50, 5874-5881.

(16) Liu, J.; Choe, J. K.; Wang, Y.; Shapley, J. R.; Werth, C. J.; Strathmann, T. J. Bioinspired Complex-Nanoparticle Hybrid Catalyst System for Aqueous Perchlorate Reduction: Rhenium Speciation and Its Influence on Catalyst Activity. ACS Catal., 2015, 5, 511-522.

(17) McPherson, L. D.; Drees, M.; Khan, S. I.; Strassner, T.; Abu-Omar, M. M. Multielectron Atom Transfer Reactions of Perchlorate and Other Substrates Catalyzed by Rhenium Oxazoline and Thiazoline Complexes: Reaction Kinetics, Mechanisms, and Density Functional Theory Calculations. Inorg. Chem. 2004, 43, 4036-4050.

(18) Youngblut, M. D.; Tsai, C.-L.; Clark, I. C.; Carlson, H. K.; Maglaqui, A. P.; Gau-Pan, P. S.; Redford, S. A.; Wong, A.; Tainer, J. A.; Coates, J. D. Perchlorate Reductase is Distinguished by Active Site Aromatic Gate Residues. J. Biol. Chem. 2016, 291, 9190- 9202.

(19) Schaffner, I.; Hofbauer, S.; Krutzler, M.; Pirker, K. F.; Furtmüller, P. G.; Obinger, C. Mechanism of Chlorite Degradation to Chloride and Dioxygen by the Enzyme Chlorite Dismutase. Arch. Biochem. Biophys. 2015, 574, 18-26.

(20) Hijnen, W. A. M.; Voogt, R.; Veenendaal, H. R.; Van Der Jagt, H.; Van Der Kooij, D. Bromate Reduction by Denitrifying Bacteria. Appl. Environ. Microbiol. 1995, 61, 239-244.

(21) Amachi, S.; Kawaguchi, N.; Muramatsu, Y.; Tsuchiya, S.; Watanabe, Y.; Shinoyama, H.; Fijii, T. Dissimilatory Iodate Reduction by Marine Pseudomonas sp. Strain SCT. Appl. Environ. Microbiol. 2007, 73, 5725-5730.

(22) Matson, E. M.; Park, Y. J.; Fout, A. R. Facile Nitrite Reduction in a Non-Heme Iron System: Formation of an Iron(III)-Oxo. J. Am. Chem. Soc. 2014, 136, 17398-17401.

(23) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of Iron(II) Aqua and Hydroxyl Complexes Featuring a Tripodal H-Bond Donor and Acceptor Ligand. Inorg Chem 2014, 53, 4450–4458.

(24) Hagen, K. S. Iron(II) Triflate Salts as Convenient Substitutes for Perchlorate Salts: Crystal Structures of [Fe(H2O)6](CF3SO3)2 and Fe(MeCN)4(CF3SO3)2. Inorg. Chem. 2000, 39, 5867-5869.

(25) Prakash, G. K. S.; Keyaniyan, S.; Aniszfeld, R.; Heiliger, L.; Olah, G. A.; Stevens, R. C.; Choi, H. K.; Bau, R. Triphenylsilyl Perchlorate Revisited: 29Si and 35Cl NMR Spectroscopy and X-ray Crystallography Showing Covalent Nature in Both Solution and the Solid State.

47 Difficulties in Observing Long-Lived Silyl Cations in the Condensed State. J. Am. Chem. Soc. 1987, 109, 5123-5126.

48 CHAPTER 3: DIANIONIC OXYANION REDUCTION BY A NON-HEME IRON SYSTEM

3.1 Introduction Selenium is an essential element for all forms of life because of its incorporation into selenocysteine and selenoenzymes;1,2 however, the difference between the daily recommended dose and a toxic dose is only one order of magnitude. The common forms of selenium found in 2- the environment are selenide, elemental selenium and the selenium oxyanions: selenate (SeO4 ) 2- and selenite (SeO3 ). Of these selenium species, only the oxyanions are soluble in water and thus exhibit higher toxicity.3 These oxyanions originate from both natural sources, such as volcanoes and seleniferous soils, and anthropogenic sources, including waste from mining, the refinement of metal ores, and other industrial processes,3 but the speciation of selenium in the environment is largely dependent upon the activity of microbes.1 In selenium-rich environments, many species of bacteria and archaea have developed mechanisms to transform the water-soluble selenium oxyanions to a less soluble and less bioavailable form of selenium: red allotropes of selenium (Figure 3.1).3

Se(0) 2- 2- Se(0) SeO4 SeO3 Se(0) various selenite selenate reductase reduction processes

S MoIV S S S

Figure 3.1 Biological selenium oxyanion reduction to elemental selenium showing the proposed active site of selenate reductase.

Selenate reduction to selenite is a dissimilatory process catalyzed by the molybdenum- containing metalloenzyme selenate reductase in the periplasm.3 Although the active site of the enzyme has not been crystallographically characterized, data from X-ray absorption spectroscopy suggests the active site of the oxidized form contains four Mo–S interactions, from two molybdopterin cofactors, one Mo=O bond, and one Mo–O(H) bond.4 The reduced form also contains four Mo–S interactions and either one or two Mo–O(H) bonds.4 Based on the structural similarities of selenate reductase and other molybdenum metalloenzymes, a molybdenum(IV)/(VI)

49 catalytic cycle is likely operative. Oxygen atom transfer from selenate to the molybdenum(IV) center results in the formation of selenite and a molybdenum(VI)-oxo (Figure 3.2). Subsequent reduction and release of water regenerates the molybdenum(IV) center. OH OH 2- O SeO4 VI S MoIV S S Mo S 2- - SeO3 S S S S

- - OR O OR SeO 2- 4 VI S MoIV S S Mo S 2- - SeO3 S S S S R = Me, Ph Figure 3.2 Proposed oxygen atom transfer from selenate to form selenite in selenate reductase 1- (top) and in the model system [Mo(OR)(S2C2Me2)2] (R = Me, Ph; bottom) of Holm and co- workers.5

The reduction of selenate to selenite has been modeled by Holm and co-workers in a 1- 5 molybdenum(IV)-(bis)dithiolene system, [Mo(OR)(S2C2Me2)2] (R = Me, Ph; Figure 3.2). Addition of selenate to the molybdenum(IV)-methoxide or -phenoxide complex resulted in formation of a molybdenum(VI)-oxo via oxygen atom transfer. This species, however, was not stable to isolation and was detected by transient electronic absorption spectroscopy. The generation of selenite was confirmed by 77Se NMR spectroscopy. The selenite produced from selenate reductase can be further reduced to amorphous red selenium through multiple mechanisms. In contrast with the large number of species capable of dissimilatory selenate reduction, few selenite-respiring species have been identified.3 Alternative methods of selenite reduction include glutathione-mediated reduction via formation of selenodiglutathione (GS-Se-GS) or reduction by promiscuous respiratory reductases, such as nitrite or sulfite reductase.3 Once elemental selenium is formed, it aggregates to form nanoparticles and is then extruded from the cell into the extracellular matrix through processes that are not well- understood. The remediation of selenate and selenite from sources of contaminated water is of particular importance because selenium bioaccumulates, adversely impacting aquatic life and human health.6 Removal of selenium oxyanions by traditional ion exchange resins is hindered by lower concentrations of selenium oxyanions compared to competing anion contaminants.

50 Moreover, concentrated waste streams are produced when the resins are regenerated. Chemical reduction of the selenium oxyanions is also possible, but is generally more effective at removing selenite than selenate.6 Bioreactors have demonstrated their potential to remediate contaminated water sources, but they do not perform well when other oxyanions, such as nitrate and sulfate, are present in solution. Ultimately, the most effective remediation techniques will form insoluble elemental selenium as it can be removed from the water supply by filtration or gravity settling.6 The remediation of the hexavalent chromium oxyanion chromate from drinking water sources is also of particular importance. Although there are naturally occurring sources of chromate, in the form of ores and minerals,7 the appearance of chromate in water supplies has been linked to the improper disposal of industrial waste.8 The chromate dianion is structurally similar to sulfate and is readily taken up by cells through the sulfate transport system. Once inside the cell, the chromium(VI) center is thought to be reduced to chromium(V) with concomitant formation of reactive oxygen species.9 This deleterious reactivity gives rise to the extreme toxicity of chromate, which is described as one of the greatest threats to human health by the United States Environmental Protection Agency.10 Because chromium(III) species are generally less soluble and thus less bioavailable than chromium(VI) species, the reduction of chromate to chromium(III) is an attractive remediation route and has been utilized in wastewater treatment.8 However, many species of bacteria in chromate-contaminated environments have developed mechanisms to mitigate chromate toxicity, including reducing cellular uptake of the anion and expressing enzymes that reduce chromate, making water treatment by bioremediation a possibility.8 In the cases where chromate is reduced enzymatically, Cr(OH)3 is formed and rapidly mineralizes on the surface of the chromate-reducing bacteria. Many enzymes that reduce chromate have been discovered, but most have not yet been structurally characterized. In addition, the mechanism of the reduction of chromate to chromium(III) is not well-understood. Cy II Following the success of the iron(II)-triflate species N(afa )3Fe(OTf)2 (Fe -OTf) in reducing monoanionic halogen oxyanions, including perchlorate, we wished to expand the scope of oxyanions that could be reduced by FeII-OTf. Because the biological reduction of the dianionic oxyanions selenite and selenate by molybdenum-containing enzymes is similar to that of the halogen oxyanions, we wished to explore the reactivity of selenite and selenate in our system. The reduction of chromate was also of interest to us, as it is a toxic water contaminant that can be

51 reduced to more benign species in biological systems. We proposed that the secondary sphere interactions imparted by our tripodal ligand platform would aid in deoxygenation of the dianionic Cy III oxyanion substrates and facilitate formation of the iron(III)-oxo species [N(afa )3FeO]OTf (Fe - O) . This chapter describes the reduction of selenite, selenate, and chromate in our non-heme iron system.

3.2 Selenite reduction We began studying the reduction of the selenium oxyanions with the reaction of excess II 1 Na2SeO3 and Fe -OTf in acetonitrile. After stirring the mixture 16 h, it was analyzed by H NMR spectroscopy and only unreacted FeII-OTf was present. However, if a drop of methanol was added at the beginning of the reaction, formation of FeIII-O was observed. We hypothesized deoxygenation of selenite by FeII-OTf had occurred and were eager to explore whether the selenium oxyanion could be reduced to elemental selenium. In order to investigate the reduction of selenite to elemental selenium, the stoichiometric II reaction of 3 equiv Fe -OTf with 1 equiv Na2SeO3 was pursued (Figure 3.3). Because Na2SeO3 is largely insoluble in acetonitrile, as evidenced in the previous reaction by the formation of FeIII-O only after the addition of a drop of methanol, a solubilizing agent was required to observe reactivity. Initial efforts to improve selenite solubility were via encapsulation of the sodium cations with benzo-15-crown-5, but the reaction failed to go to completion. When methanol was added to the acetonitrile solution in a 1:3 ratio by volume in place of benzo-15-crown-5, unreacted FeII- OTf persisted even after stirring the reaction mixture 16 h. However, the amount of FeIII-O produced was increased, leading us to use methanol, as opposed to benzo-15-crown-5, in future experiments.

Cy Cy OTf Cy Cy OTf Cy O CF3 Cy HN S NH NH HN HN NH O O O Na2SeO3 3 3 Fe N Fe N N - Se(0) N N N

N N FeII-OTf FeIII-O Figure 3.3 Proposed reduction of selenite to form elemental selenium by FeII-OTf.

II During the course of the reaction of 3 equiv Fe -OTf with 1 equiv Na2SeO3 the formation of a red precipitate was noted. Because biological selenite reduction furnishes red amorphous

52 elemental selenium,8 we proposed that the red precipitate in our reaction was also elemental selenium. Quantitative isolation of the red material for characterization proved difficult as it was colloidal. Thus, we turned to alternative methods to identify and quantify the amount of the red precipitate that was formed. The crude reaction mixture was dried, washed thoroughly with water to remove any residual selenite, and analyzed by inductively coupled plasma optical emission spectroscopy (ICP-OES). While this technique confirmed the presence of selenium in the sample, it could not differentiate between the oxidation states of selenium and thus did not provide definitive evidence that the red precipitate was elemental selenium. We also explored 77Se NMR spectroscopy to characterize the selenium-containing products II from the crude reaction of Fe -OTf and Na2SeO3 because of literature precedent in using the technique to identify selenium allotropes and oxyselenium species.11,12 Elemental selenium is insoluble in most solvents, but exhibits minimal solubility in carbon disulfide.13 Thus, we prepared samples for solution 77Se NMR spectroscopy in carbon disulfide. We first independently synthesized a sample of red amorphous selenium and observed a single resonance in the 77Se NMR spectrum at 613 ppm (Figure 3.4), which was consistent with that of other characterized selenium 11,14 II allotropes. However, the red material formed from the reaction of Fe -OTf and Na2SeO3 was not soluble enough in carbon disulfide to collect its 77Se NMR spectrum. As a result, we were unable to confirm the identity of the red precipitate using this method.

77 Figure 3.4 Se NMR spectrum of independently prepared red selenium (CS2, 21 °C).

Instead, we relied on the reaction of elemental selenium with a phosphine to yield a phosphine selenide. This transformation is well-established, as formation of phosphine selenides is commonly used to investigate the σ-donor ability of phosphines.15 Coupling between the phosphorus nucleus and an NMR-active selenium-77 nucleus (7.6% natural abundance) gives 31 1 characteristic satellites in the P NMR spectrum and the value of the resulting JP,Se coupling constant lends insight into the electronic structure of the phosphorus center. The reaction of the II red solid obtained from the reduction of selenite by Fe -OTf with triphenylphosphine (PPh3) was pursued because PPh3 is a crystalline solid that is readily available and easily handled.

53 15 Furthermore, triphenylphosphine selenide (Se=PPh3) was previously characterized. Its quantitative synthesis was achieved by the addition of selenium to PPh3 in a refluxing solution of chloroform or toluene. Because selenite reduction by FeII-OTf was conducted in acetonitrile,

Se=PPh3 was independently synthesized in acetonitrile solution. The product was obtained in 31 1 quantitative yield and its P NMR spectrum matched the published spectrum (δ = 35.9 ppm; JP,Se = 732 Hz; Figure 3.5).15 When a portion of the red precipitate was collected and allowed to stir 31 with PPh3 in acetonitrile, Se=PPh3 was observed by P NMR spectroscopy and confirmed the formation of elemental red selenium from the reduction of selenite.

31 Figure 3.5 P NMR spectrum of independently prepared Se=PPh3 (CH2Cl2, 21 °C).

The quantitative isolation of the elemental selenium precipitate from the crude reaction mixture was not possible; isolation of the precipitate and subsequent addition of PPh3 gave a II maximum yield of 63% Se=PPh3. Alternatively, PPh3 was added with Na2SeO3 to Fe -OTf to generate Se=PPh3 in situ. Using this method, triphenylphosphine oxide (O=PPh3) was formed and necessitated the addition of excess PPh3 to the selenite reduction reaction. The control reaction of

Na2SeO3 and PPh3 under identical conditions yielded neither Se=PPh3 nor O=PPh3 and unreacted 31 PPh3 was the only phosphorus-containing product observed by P NMR spectroscopy. Side III reactivity between Fe -O and PPh3 was proposed to be the source of O=PPh3, as evidenced by Cy the formation of the deoxygenated iron(III) species N(pi )3Fe. Its characteristic C=N stretching frequency at 1581 cm-1 was observed in the IR spectrum of the control reaction of FeIII-O and

PPh3. Additionally, the control reaction generated O=PPh3, albeit in much lower yield. When the II reduction of selenite was repeated with 4 equiv PPh3, 1 equiv Na2SeO3, and 3 equiv Fe -OTf and 31 analyzed by P NMR spectroscopy the yield of Se=PPh3 was greatly improved (0.93 equiv, which corresponds to a 93% yield of elemental selenium). O=PPh3 (0.88 equiv) and unreacted PPh3 (2.15

54 equiv) were the remaining phosphorus-containing products that were present in the 31P NMR spectrum. The amount of FeIII-O formed from the reduction of selenite was examined using 1,2- diphenylhydrazine (DPH), analogous to the method used for the quantification of FeIII-O produced from iodate and periodate (Chapter 2). DPH was added (equimolar to FeII-OTf) following II completion of the reaction of 3 equiv Fe -OTf with 1 equiv Na2SeO3 and filtration of the solution to remove as much of the colloidal selenium as possible. Integration of the 1H NMR spectrum allowed for the ratio of azobenzene to DPH to be determined, providing an indirect measure of the amount of FeIII-O that was present in solution prior to the addition of DPH. This procedure III demonstrated the formation of 2.94 equiv Fe -OTf from the reduction of 1 equiv Na2SeO3 (3.00 III equiv would indicate quantitative reduction of Na2SeO3). When the DPH quantification of Fe - III O was performed on the selenite reduction reaction with 4 equiv PPh3 added, the amount of Fe - O calculated to have been present was lower (1.81 equiv), likely from the deoxygenation of FeIII-

O by PPh3. Based on the indirect measurements of the amount of elemental selenium and FeIII-O II formed from the reduction of 1 equiv Na2SeO3 by 3 equiv Fe -OTf, we proposed selenite reduction to form elemental selenium was nearly quantitative in our system. Although this reduction is not unprecedented and is readily achieved in biological systems, our system represents a rare example of a synthetic system that facilitates the process. As with the reduction of the halogen oxyanions in Chapter 2, we hypothesize that the secondary sphere plays a role in stabilizing the iron(III)-oxo formed from the deoxygenation of the selenium-containing substrate. The formation of an iron(III)-selenide species was also investigated as an alternative to trapping elemental selenium with PPh3. Although there is literature precedent for the formation of an iron(III)-selenide complex from the addition of elemental selenium to an iron(II) center coordinated to a tripodal ligand scaffold featuring a secondary coordination sphere,16 we were unable to furnish an analogous species in our system. FeII-OTf did not react with elemental selenium or diphenyldiselenide, nor we did not observe formation of any iron(III)-selenide species during selenite reduction by FeII-OTf.

55 3.3 Selenate reduction The more difficult reduction of selenate by FeII-OTf was also explored to determine whether selenate could be reduced to elemental selenium, akin to the reduction of selenite (Figure 3.6). The solubility of selenate salts is much lower than that of selenite salts, even in water. II Therefore, it was not surprising that the addition of Na2SeO4 to Fe -OTf in acetonitrile did not yield FeIII-O. As with selenite reduction, the poor solubility of the selenium oxyanion salt necessitated the addition of a solubilizing agent to observe reactivity between the oxyanion substrate and FeII-OTf. The use of benzo-15-crown-5 did not result in the desired deoxygenation reactivity. Cy Cy OTf Cy Cy OTf Cy O CF3 Cy HN S NH NH HN HN NH O O 2- O SeO4 4 4 Fe N Fe N N - Se(0) N N N

N N FeII-OTf FeIII-O Figure 3.6 Idealized reduction of selenate by FeII-OTf.

However, upon the addition of methanol to the acetonitrile solution, in a ratio of 1 to 3 by volume, FeIII-O was generated. Analysis of the stoichiometric reaction of 4 equiv FeII-OTf with 1 1 equiv Na2SeO4 in methanolic solution by H NMR spectroscopy revealed a small amount of FeIII-O and a significant amount of unreacted FeII-OTf. When the reaction was repeated with the addition of 1 equiv PPh3, 0.14 equiv Se=PPh3 was formed, providing an indirect measurement of the amount of elemental selenium formed from the complete deoxygenation of selenate. The quantification of FeIII-O using DPH demonstrated that 1.49 equiv had been formed from the reduction of 1 equiv Na2SeO4 (4.00 equiv would indicate quantitative reduction of Na2SeO4; Table II 3.1). We proposed the poor yield of elemental selenium from the reaction of Fe -OTf and Na2SeO4 compared to that of Na2SeO3 could have been the result of the poor solubility of the selenate salt, the unfavorable reduction potential of selenate, a different mechanism of deoxygenation, or a combination of the three.

To more thoroughly understand the role of solubility in the reduction of Na2SeO4, we tested II a variety of selenate sources. The addition of Ag2SeO4 to Fe -OTf did not greatly improve the formation of FeIII-O and the 1H NMR spectrum of the crude reaction mixture revealed FeIII-O, unreacted FeII-OTf, and a new paramagnetic product with a characteristic resonance at 48 ppm.

56 This product was later identified as the complex obtained from the reaction of silver salts with the Cy iron(II)-hydroxo complex [N(afa )3FeOH]OTf (Chapter 4). The formation of side products and the potential for deleterious reactivity with silver salts in solution led us to abandon the use of

Ag2SeO4. II We next investigated the reactivity of Fe -OTf toward K2SeO4 in the presence of crypt- II 222. The reaction of 4 equiv Fe -OTf, 1 equiv K2SeO4, 2 equiv crypt-222, and 5 equiv PPh3 yielded 0.26 equiv Se=PPh3. While this was an improvement from the yield of Se=PPh3 obtained from the reduction of Na2SeO4, we repeated the reaction in methanolic acetonitrile solution (1 to 3 ratio by volume) to further improve this yield. The addition of methanol to the reaction mixture 31 increased the yield of Se=PPh3 to 0.40 equiv. The P NMR spectrum also showed O=PPh3 (0.19 III equiv) and unreacted PPh3 (4.43 equiv). Quantification of Fe -O using DPH revealed 2.46 equiv had been formed from the reduction of 1 equiv K2SeO4 (Table 3.1).

III Table 3.1 Comparison of the optimized yields of Se=PPh3 and Fe -O from the reduction of II Na2SeO4 and K2SeO4 by Fe -OTf. III Se=PPh3 (equiv) Fe -O (equiv)

II 4 Fe -OTf + Na2SeO4 0.14 1.49

II 4 Fe -OTf + K2SeO4 + 2 crypt-222 0.40 2.46

III The improved yields of Se=PPh3 and Fe -O from the reduction of K2SeO4 compared to

Na2SeO4 suggested solubility played a role in the reduction; however, it remained unclear if the unfavorable reduction potential of selenate or if a different deoxygenation mechanism were also contributing to the low yields of elemental selenium. If the initial deoxygenation step of selenate were a two electron oxygen atom transfer reaction the resulting selenium-containing species would be selenite, as demonstrated by Holm and co-workers.5 Thus, reduction of selenate by FeII-OTf may not proceed through selenite because selenite reduction to elemental selenium was quantitative.

3.4 Selenium dioxide reduction If the deoxygenation of selenate were to occur through one electron steps, the neutral oxyselenium species SeO2 could be formed. To further examine this possibility, we explored the

57 II II reactivity of Fe -OTf with SeO2. Upon the addition of 1 equiv SeO2 to 2 equiv Fe -OTf, we observed no formation of FeIII-O. Instead, the only products observed by 1H NMR spectroscopy Cy II were the protodemetallated ligand N(pi )3 • 3 HOTf and unreacted Fe -OTf. When the reaction 1 was repeated in the presence of 1 equiv NEt3, a red precipitate was observed and the H NMR spectrum of the reaction revealed the formation of FeIII-O. The addition of base suppressed the Cy II 1 formation of N(pi )3 • 3 HOTf, although Fe -OTf was still observed in the H NMR spectrum of the crude reaction mixture, an indication that the reaction did not go to completion.

Confirming the identity of the red precipitate was more difficult in this case, as SeO2 reacts cleanly with 3 equiv PPh3 to generate 1 equiv Se=PPh3 and 2 equiv O=PPh3. As a result, PPh3 could not be added to the reaction mixture and the red precipitate had to be collected after the reaction had concluded. Because the reaction did not go to completion, it was possible that unreacted SeO2 would be collected along with the red precipitate and would react with the PPh3 that was later added. However, the reaction of SeO2 and PPh3 would be the only source of O=PPh3, 31 which could be used to determine the amount of Se=PPh3 formed from SeO2. Based on the P III NMR spectrum, the yield of Se=PPh3 was 0.26 equiv. DPH was utilized to quantify the Fe -O formed from the reduction of SeO2 (0.78 equiv; 2.00 equiv would indicate quantitative reduction of SeO2). III While the low yields of elemental selenium and Fe -O from the reduction of SeO2 by II Cy Fe -OTf seemed to be consistent with selenate reduction proceeding through SeO2, N(pi )3 • 3 HOTf was never detected during the course of selenate reduction, and no base was present in solution to preclude its formation. We were unable to determine if selenate reduction produces II SeO2 in our system. It is also possible that the diminished reactivity of Fe -OTf toward SeO2, as compared to other oxyanion substrates, stems from SeO2 being a neutral species. There is likely a greater driving force for anionic substrates to bind to the dicationic iron(II) center of FeII-OTf.

3.5 Chromate reduction The reduction of the dianionic chromate anion by FeII-OTf was also investigated. The II reactivity of Fe -OTf toward K2CrO4 was initially tested by stirring an equimolar mixture of the two in acetonitrile for 16 h. Formation of FeIII-O was observed by 1H NMR spectroscopy, but unreacted FeII-OTf was also present. Analogous to the improved reactivity upon addition of crypt-

222 and methanol to K2SeO4, the addition of the two solubilizing agents improved the reduction

58 Cy of K2CrO4. The formation of N(pi )3 • 3 HOTf was noted and prompted the addition of NEt3 to the reaction mixture. We were unable to identify the chromium-containing species produced from the reduction of chromate. The amount of FeIII-O was, however, quantified to provide insight into the number of deoxygenation steps that taken place. Based on the ratio of azobenzene to DPH following the addition of DPH to the iron-containing products, 2.47 equiv FeIII-O were formed

(4.00 equiv would indicate complete deoxygenation of 1 equiv K2CrO4). We propose this corresponds to two deoxygenation reactions per chromate anion, although this cannot be confirmed without identifying the resulting chromium species.

3.6 Investigating the scope of oxyanion reduction: unsuccessful substrates Throughout the course of our oxyanion reduction studies, we encountered many oxyanion II substrates that were not reduced by Fe -OTf. Among the unsuccessful substrates were Na2SO3,

NaHSO3, NaAsO2, K2HAsO4, Na2TeO3. Each of these oxyanions is soluble in water and minimally solubility in organic solvents, complicating their reduction by FeII-OTf in acetonitrile. A variety of techniques were used to increase the solubility of these anions, such as addition of the solubilizing agents benzo-15-crown-5, 18-crown-6, or crypt-222 or heating the reaction mixture, but to no avail. In each case, we were unable to identify if lack of reactivity was due to the poor solubility or an unfavorable reduction potential of the oxyanion.

3.7 Conclusions The scope of oxyanions reduced by FeII-OTf was expanded to include the dianionic oxyanions selenite, selenate, and chromate. The deoxygenation of selenite proceeded almost quantitatively, generating elemental selenium and FeIII-O. The elemental selenium was identified 31 and quantified through the formation of Se=PPh3. P NMR spectroscopy aided in the characterization of Se=PPh3, which was readily identified by the satellites from the 7.6% natural abundance of the NMR-active selenium-77 isotope coupling to the phosphorus-31 nucleus. The reductions of selenate and chromate did not proceed as smoothly because neither reaction went to completion. We were unable to isolate the resulting selenium- or chromium- containing products and thus could not propose a mechanism for the reduction of either oxyanion. II The addition of SeO2 to Fe -OTf did not furnish any insight into the reduction of selenate. The reduction of selenate did, however, furnish Se=PPh3 in 40% yield, indicating that some selenate

59 was fully deoxygenated. In the reduction of chromate, we used the amount of FeIII-O formed to estimate the number of deoxygenation steps that occurred to be two. Finally, the oxyanions that were unable to be deoxygenated by FeII-OTf, due to either poor solubility in acetonitrile or an unfavorable reduction potential, were outlined.

3.8 Experimental General considerations. To avoid contact with oxygen and water, all manipulations were carried out under an atmosphere of nitrogen in an MBraun inert atmosphere drybox or using standard Schlenk techniques. Solvents for air- and moisture-sensitive manipulations were dried and deoxygenated using a Glass Contour System and stored over 4 Å molecular sieves prior to use. Celite 545 was heated to 150 °C under dynamic vacuum for 24 h prior to use in the drybox. All reagents were purchased from commercial sources and used as received unless otherwise noted. 1,2-Diphenylhydrazine was recrystallized from a saturated solution in diethyl ether layered with hexane and stored at -35 °C. Sodium selenite, sodium selenate, potassium chromate, and potassium selenate were recrystallized from a saturated aqueous solution layered with methanol and cooled to 0 °C. Triethylamine was distilled and stored over 4 Å molecular sieves at -35 °C prior to use. Cy 17 14 N(afa )3Fe(OTf)2 and amorphous red selenium were synthesized according to literature procedure. NMR solvent (acetonitrile-d3) was degassed and stored over 4 Å molecular sieves prior to use. NMR spectra were recorded at ambient temperature on a Varian spectrometer operating at 500 MHz (1H NMR), 202 MHz (31P NMR), and 95 MHz (77Se NMR) and referenced to the peak of the residual solvent (δ parts per million and J in Hz). Solid-state infrared spectra were measured using a PerkinElmer Frontier FT-IR spectrophotometer equipped with a KRS5 thallium bromide/iodide universal attenuated total reflectance accessory. Analysis by inductively coupled plasma optical emission spectroscopy was performed by the University of Illinois at Urbana- Champaign School of Chemical Sciences Microanalysis Laboratory in Urbana, IL.

Preparation of Se=PPh3. A 20 mL scintillation vial was charged with black elemental selenium

(0.0151 g, 0.191 mmol), PPh3 (0.0500 g, 0.191 mmol), and 4 mL acetonitrile. The solution was stirred 4 h, during which time the black selenium powder gradually went into solution. Volatiles were removed under reduced pressure and the resulting white powder was analyzed by 31P NMR

60 spectroscopy (Figure 3.5), revealing quantitative formation of Se=PPh3 (0.0652 g, 0.191 mmol, 100%).

Selenite reduction and the quantification of selenite reduction products. A 20 mL scintillation Cy vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO3 (0.0035 g, 0.0200 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h, with the

II 3 Fe -OTf + Na SeO in MeCN/MeOH (3:1) 2 3 ◼ ◼ * *

◼ ◼ II 3 Fe -OTf + Na SeO + 2 benzo-15-crown-5 2 3 * *

II Fe -OTf + excess Na SeO 2 3 * *

Cy OTf OTf Cy Cy Cy HN Cy O CF3 Cy NH HN S NH HN NH O O O Fe N N N N Fe N N

N N * ◼ Figure 3.7 1H NMR spectra of selenite reduction reactions demonstrating the formation of III Fe -O (CD3CN, 21 °C).

61 color gradually changing from yellow-orange to red-brown, concomitant with formation of a red precipitate. Analysis of the crude reaction mixture by 1H NMR spectroscopy revealed a small Cy Cy amount of unreacted N(afa )Fe(OTf)2 and the formation of [N(afa )3FeO]OTf. Earlier iterations Cy of the reaction included the combination of N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), excess

Na2SeO3 (approximately 15 mg), 3 mL acetonitrile, and 1 drop methanol and the combination of Cy N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO3 (0.0035 g, 0.0200 mmol), benzo-15-crown- 5 (0.0107 g, 0.0400 mmol), and 3 mL acetonitrile. The quantification of selenite reduction products was performed as follows; a 20 mL scintillation Cy vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO3 (0.0035 g, 0.0200 mmol), PPh3 (0.0210 g, 0.0800 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h before the volatiles were removed under reduced pressure. The resulting brown powder was triturated with 4 mL diethyl ether and filtered over a pad of Celite into a clean 20 mL scintillation vial to isolate the phosphorus-containing products. The remaining brown solid was collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. Both solutions were evaporated to dryness in vacuo. The diethyl ether wash was analyzed by 31P NMR spectroscopy

(S=PPh3 was added as an internal standard; 0.0059 g, 0.0200 mmol) and revealed the formation of

PPh3

S=PPh3 Se=PPh3

O=PPh3

31 Figure 3.8 Quantification by P NMR spectroscopy of the Se=PPh3 produced from the II reaction of 4 equiv Fe -OTf, 1 equiv Na2SeO3, and 4 equiv PPh3 (CH2Cl2, 21 °C).

62 Se=PPh3 (0.93 equiv; 1.00 equiv expected) and O=PPh3 (0.88 equiv) and unreacted PPh3 (2.14 equiv). The dried dichloromethane wash was redissolved in 4 mL acetonitrile and stirred with 1,2- diphenylhydrazine (0.0111 g, 0.0600 mmol) for 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2-diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 30% conversion of 1,2-diphenylhydrazine to Cy azobenzene, which is equivalent to the reduction of 1.81 equiv [N(afa )3FeO]OTf (3.00 equiv Cy expected); this low yield may be due to the formation of O=PPh3 from [N(afa )3FeO]OTf during the reduction of selenite in the presence of PPh3.

H N N H N N

Figure 3.9 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR spectroscopy revealed 1.81 equiv FeIII-O were produced from selenite reduction in the presence of PPh3 (CD3CN, 21 °C).

When the reaction was repeated without the addition of PPh3 the quantification of Cy [N(afa )3FeO]OTf was much closer to the expected value, but the quantification of elemental Cy selenium was much lower. A 20 mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO3 (0.0035 g, 0.0200 mmol), 3 mL acetonitrile, and 1 mL methanol. The

63 solution was stirred 16 h before the solution was filtered over a pad of Celite into a clean 20 mL scintillation vial to remove as much colloidal elemental selenium as possible. The elemental selenium and the Celite were emptied back into the reaction vial and stirred with PPh3 (0.0052 g, 0.0200 mmol) in 4 mL acetonitrile. The red solid gradually went into solution over the course of 3 h. The mixture was then filtered, dried in vacuo, and analyzed by 31P NMR spectroscopy with

S=PPh3 was added as an internal standard (0.0059 g, 0.0200 mmol). The filtrate of the selenite reduction reaction was evaporated to dryness before being redissolved in 4 mL acetonitrile and stirred with 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol) for 16 h. Additional elemental selenium crashed out during this time and was collected via filtration and reacted with PPh3 as previously described (we do not believe this selenium was produced from the reaction of residual

Na2SeO3 with 1,2-diphenylhydrazine, as a control reaction of these reagents did not yield any elemental selenium). The amount of Se=PPh3 from both filtrations added to 0.63 equiv. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2-diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene

H N N N N H

Figure 3.10 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR III spectroscopy revealed 2.98 equiv Fe -O were produced from selenite reduction without PPh3 present (CD3CN, 21 °C).

64 was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 49% conversion of 1,2- diphenylhydrazine to azobenzene, which is equivalent to the reduction of 2.98 equiv Cy [N(afa )3FeO]OTf (3.00 equiv expected).

S=PPh3

Se=PPh3

PPh3

st 1 Filtration

S=PPh3

PPh3

Se=PPh3 nd 2 Filtration

Figure 3.11 Quantification by 31P NMR spectroscopy of the elemental selenium isolated from II the reaction of 3 equiv Fe -OTf and Na2SeO3 (CH2Cl2, 21 °C).

65 Cy Control reaction of [N(afa )3FeO]OTf and PPh3. A 20 mL scintillation vial was charged with Cy [N(afa )3FeO]OTf (0.0482 g, 0.0600 mmol), PPh3 (0.0210 g, 0.0800 mmol), 3 mL acetonitrile, and 1 mL methanol. The mixture was stirred for 16 h and followed by removal of volatiles under reduced pressure. The resulting brown residue was treated with 3 mL diethyl ether, filtered over a Cy 31 pad of Celite to isolate the phosphorus-containing products and N(pi )3Fe, and analyzed by P NMR spectroscopy and IR spectroscopy, which showed a C=N stretch corresponding to Cy -1 N(pi )3Fe at 1581 cm . The remaining brown solid was eluted with 3 mL dichloromethane and 1 Cy evaporated to dryness. The H NMR spectrum showed only [N(afa )3FeO]OTf.

PPh3

S=PPh3

O=PPh3

31 III Figure 3.12 P NMR spectrum of the reaction of Fe -O with PPh3 (CH2Cl2, 21 °C).

Selenate reduction and quantification of selenate reduction products. A 20 mL scintillation Cy vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), K2SeO4 (0.0033 g, 0.0150 mmol), crypt-222 (0.0113 mg, 0.0300 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h, with the color gradually changing from yellow-orange to red-orange. Analysis of the crude reaction mixture by 1H NMR spectroscopy revealed a mixture of unreacted Cy Cy N(afa )Fe(OTf)2 and [N(afa )3FeO]OTf. Earlier iterations of the reaction included the Cy combination of N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO4 (0.0028 g, 0.0150 mmol),

66 Cy 3 mL acetonitrile, and 1 mL methanol and the combination of N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Ag2SeO4 (0.0054 g, 0.0200 mmol), and 4 mL acetonitrile in the dark.

◼ ◼

II 4 Fe -OTf + K SeO + 2 crypt-222 2 4 * * ◼

◼

II 4 Fe -OTf + Ag SeO 2 4 * * ◼

◼

II 4 Fe -OTf + Na SeO 2 4 * *

Cy OTf OTf Cy Cy Cy HN Cy O CF3 Cy NH HN S NH HN NH O O O Fe N N N N Fe N N

N N * ◼ Figure 3.13 1H NMR spectra of selenate reduction reactions demonstrating the formation of III Fe -O (CD3CN, 21 °C).

The quantification of products from the partial reduction of Na2SeO4 was performed as follows; a Cy 20 mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), Na2SeO4

67 (0.0028 g, 0.0150 mmol), PPh3 (0.0039 mg, 0.0150 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h before the volatiles were removed under reduced pressure. The resulting brown powder was triturated with 4 mL diethyl ether and filtered over a pad of Celite into a clean 20 mL scintillation vial to isolate the phosphorus-containing products. The remaining brown solid was collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. Both solutions were evaporated to dryness in vacuo. The diethyl ether wash was analyzed by 31P NMR spectroscopy (S=PPh3 was added as an internal standard; 0.0044 g, 0.0150 mmol) and revealed the formation of Se=PPh3 (0.14 equiv; 1.00 equiv expected).

S=PPh3

PPh3

Se=PPh3

Figure 3.14 Quantification by 31P NMR spectroscopy of the elemental selenium isolated from II the reaction of 4 equiv Fe -OTf and Na2SeO4 (CH2Cl2, 21 °C).

The dried dichloromethane wash was redissolved in 4 mL acetonitrile, and stirred with 1,2- diphenylhydrazine (0.0111 g, 0.0600 mmol) for 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2-diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 37% conversion of 1,2-diphenylhydrazine to Cy azobenzene, which is equivalent to the reduction of 1.49 equiv [N(afa )3FeO]OTf (4.00 equiv expected).

68 Fc

H N N H

N N

Figure 3.15 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR spectroscopy revealed 1.49 equiv FeIII-O were produced from the reaction of 4 equiv FeII-OTf and Na2SeO4 (CD3CN, 21 °C).

The quantification of products from the partial reduction of K2SeO4 was performed as follows; a Cy 20 mL scintillation vial was charged with N(afa )2Fe(OTf)2 (0.0561 g, 0.0600 mmol), K2SeO4

(0.0033 g, 0.0150 mmol), crypt-222 (0.0113 g, 0.0150 mmol), PPh3 (0.0197 mg, 0.0750 mmol), and 3 mL acetonitrile. The solution was stirred 16 h before the volatiles were removed under reduced pressure. The resulting brown powder was triturated with 4 mL diethyl ether and filtered over a pad of Celite into a clean 20 mL scintillation vial to isolate the phosphorus-containing products. The remaining brown solid was collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. Both solutions were evaporated to dryness in vacuo. The diethyl ether wash was 31 analyzed by P NMR spectroscopy (S=PPh3 was added as an internal standard; 0.0044 g, 0.0150 mmol) and revealed the formation of Se=PPh3 (0.26 equiv; 1.00 equiv expected). When the reaction was repeated with the addition of 1 mL methanol, the yield of Se=PPh3 increased to 0.40 equiv.

69 PPh3

With MeOH

S=PPh3

Se=PPh3

O=PPh3

PPh3

Without MeOH

S=PPh3

Se=PPh3 O=PPh3

Figure 3.16 Quantification by 31P NMR spectroscopy of the elemental selenium isolated from II the reaction of 4 eq Fe -OTf and K2SeO4 with (top) and without MeOH (bottom; CH2Cl2, 21 °C).

70 Cy The quantification of [N(afa )3FeO]OTf was performed using the dried dichloromethane wash of the K2SeO4 reduction in acetonitrile and methanol without the addition of PPh3. The brown residue obtained was redissolved in 4 mL acetonitrile and stirred with 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol) for 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate azobenzene and 1,2-diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 62% conversion of 1,2-diphenylhydrazine to azobenzene, which is equivalent to the Cy reduction of 2.46 equiv [N(afa )3FeO]OTf (4.00 equiv expected).

H N N H N N

Figure 3.17 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR spectroscopy revealed 2.46 equiv FeIII-O were produced from the reaction of 4 equiv FeII- OTf and K2SeO4 (CD3CN, 21 °C).

Selenium dioxide reduction and quantification of selenium dioxide reduction products. A 20 Cy mL scintillation vial was charged with N(afa )3Fe(OTf)2 (0.0561 g, 0.0600 mmol), SeO2 (0.0034 g, 0.0300 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h before the volatiles were removed under reduced pressure. The resulting residue was analyzed by 1H NMR Cy Cy spectroscopy, revealing a mixture of unreacted N(afa )3Fe(OTf)2 and N(pi )3 • 3 HOTf.

71 II 2 Fe -OTf + SeO + NEt 2 3 ◼ ◼ * * ●

II 2 Fe -OTf + SeO 2 ◼ ◼

Cy Cy 3 OTf OTf Cy Cy OTf Cy Cy Cy O CF HN HN Cy 3 Cy NH HN NH HN S NH HN NH O O O Fe N N N NH HNHN N Fe N N

N N N * ● ◼ 1 II Figure 3.18 H NMR spectra of the reaction of 2 equiv Fe -OTf and SeO2 with (top) and without (bottom) the addition of NEt3 (CD3CN, 21 °C).

When the reaction was repeated with the addition of NEt3 (0.0061 g, 0.0600 mmol), the formation Cy of N(pi )3 • 3 HOTf was precluded and a red precipitate was observed. The reaction mixture was filtered over a pad of Celite to isolate as much of the red material as possible. The elemental selenium and the Celite were emptied back into the reaction vial and stirred with PPh3 (0.0236 g, 0.0900 mmol) in 4 mL acetonitrile. The red solid gradually went into solution over the course of 3 h. The mixture was then filtered, dried in vacuo, and analyzed by 31P NMR spectroscopy with

S=PPh3 was added as an internal standard (0.0088 g, 0.0300 mmol). The yield of Se=PPh3 formed from elemental selenium was found to be 0.26 equiv after subtracting the amount of Se=PPh3 formed from residual SeO2 (calculated from the amount of O=PPh3).

72 S=PPh3

PPh3

Se=PPh3

O=PPh3

Figure 3.19 Quantification by 31P NMR spectroscopy of the elemental selenium isolated from II the reaction of 2 equiv Fe -OTf and 1 equiv SeO2 (CH2Cl2, 21 °C).

The filtrate of the selenium dioxide reaction was evaporated to dryness before being redissolved in 4 mL acetonitrile and stirred with 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol) for 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate the azobenzene and 1,2- diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 39% conversion of 1,2-diphenylhydrazine to azobenzene, which is equivalent to the reduction of Cy) 0.79 equiv [N(afa 3FeO]OTf (2.00 equiv expected).

73 Fc

H N N H N N

Figure 3.20 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR spectroscopy revealed 0.78 equiv FeIII-O were produced from the reaction of 2 equiv FeII- OTf and 1 equiv SeO2 (CD3CN, 21 °C).

Control reaction of selenium dioxide and PPh3. A 20 mL scintillation vial was charged with

SeO2 (0.0022 g, 0.0200 mmol), PPh3 (0.0157 g, 0.0600 mmol), and 3 mL acetonitrile. The solution was stirred 16 h before being evacuated to dryness in vacuo. The white powder obtained was 31 analyzed by P NMR spectroscopy (S=PPh3 was added as an internal standard; 0.0177 g, 0.0600 mmol), revealing the formation of 0.73 equiv Se=PPh3 and 2.14 equiv O=PPh3.

74 S=PPh3

O=PPh3

Se=PPh3

31 Figure 3.21 P NMR spectrum of the reaction of 1 equiv SeO2 and 3 equiv PPh3 (CH2Cl2, 21 °C).

Chromate reduction and quantification of chromate reduction products. A 20 mL scintillation Cy vial was charged with N(afa )3Fe(OTf)2 (0.0561 g, 0.0600 mmol), K2CrO4 (0.0029 g, 0.0150 mmol), crypt-222 (0.0113 g, 0.0300 mmol), 3 mL acetonitrile, and 1 mL methanol. The solution was stirred 16 h and evacuated to dryness in vacuo. The residue was analyzed by 1H NMR Cy Cy spectroscopy, which showed the residue contained [N(afa )3FeO]OTf, N(pi )3 • 3 HOTf, and Cy unreacted N(afa )3Fe(OTf)2.

75 ●

◼ * * ◼

Cy Cy 3 OTf OTf Cy Cy OTf Cy Cy Cy O CF HN HN Cy 3 Cy NH HN NH HN S NH HN NH O O O Fe N N N NH HNHN N Fe N N

N N N * ● ◼ 1 II Figure 3.22 H NMR spectrum of the reaction of 4 equiv Fe -OTf and 1 equiv K2CrO4, III Cy which furnished Fe -O and N(pi )3 • 3 HOTf (CD3CN, 21 °C).

Repeating the reaction in the presence of NEt3 (0.0061 g, 0.0600 mmol) precluded the formation Cy Cy of N(pi )3 • 3 HOTf. The amount of [N(afa )3FeO]OTf furnished from the reaction was determined by triturating the dried crude reaction mixture with diethyl ether, redissolving the residue in 4 mL acetonitrile, adding 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol), and stirring the solution for 16 h. Volatiles were removed under reduced pressure and the dark red powder obtained was washed with 3 mL diethyl ether and filtered over a pad of Celite to isolate the azobenzene and 1,2-diphenylhydrazine. The diethyl ether wash was dried under vacuum and analyzed by 1H NMR spectroscopy (ferrocene was added as an internal standard; 0.0056 g, 0.0300 mmol), revealing a 31% conversion of 1,2-diphenylhydrazine to azobenzene, which is equivalent Cy) to the reduction of 2.47 equiv [N(afa 3FeO]OTf (4.00 equiv expected).

76 Fc

H N N H N N

Figure 3.23 Quantification of the ratio of azobenzene:1,2-diphenylhydrazine by 1H NMR spectroscopy revealed 2.47 equiv FeIII-O were produced from the reaction of 4 equiv FeII- OTf and 1 equiv K2CrO4 (CD3CN, 21 °C).

3.9 References (1) Stolz, J. F.; Basu, P.; Santini, J. M.; Oremland, R. S. Arsenic and Selenium in Microbial Metabolism. Annu. Rev. Microbiol. 2006, 60, 107-130.

(2) Burriss, D.; Zou, W.; Cremer, D.; Walrod, J.; Atwood, D. Removal of Selenite from Water Using a Synthetic Dithiolate: An Experimental and Quantum Chemical Investigation. Inorg. Chem. 2014, 53, 4010-4021.

(3) Nancharaiah, Y. V.; Lens, P. N. L. Ecology and Biotechnology of Selenium-Respiring Bacteria. Microbiol.Mol. Biol. Rev. 2015, 79, 61-80.

(4) Maher, M. J.; Santini, J.; Pickering, I. J.; Prince, R. C.; Macy, J. M.; George, G. N. X-ray Absorption Spectroscopy of Selenate Reductase. Inorg. Chem. 2004, 43, 402-404.

(5) Wang, J.-J.; Tessier, C.; Holm, R. H. Analogue Reaction Systems of Selenate Reductase. Inorg. Chem. 2006, 45, 2979-2988.

(6) Tan, L. C.; Nancharaiah, Y. V.; van Hullebusch, E. D.; Lens, P. N. L. Selenium: Environmental Significance, Pollution, and Biological Treatment Technologies. Biotechnology Advances 2016, 34, 886-907.

77 (7) Anger, G.; Halstenberg, J.; Hochgeschwender, K.; Scherhag, C.; Korallus, U.; Knopf, H.; Schmidt, P; Ohlinger, M. Chromium Compounds. In Ullmann's Encyclopedia of Industrial Chemistry, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2000; pp 157–191.

(8) Thatoi, H.; Das, S.; Mishra, J.; Rath, B. P.; Das, N. Bacterial Chromate Reductase, a Potential Enzyme for Bioremediation of Hexavalent Chromium: A Review. J. Environ. Manage. 2014, 146, 383-399.

(9) Ackerley, D. F.; Gonzalez, C. F.; Keyhan, M. Blake, R.; Matin, A. Mechanism of Chromate Reduction by the Escherichia coli Protein, NfsA, and the Role of Different Chromate Reductases in Minimizing Oxidative Stress During Chromate Reduction. Environ. Microbiol. 2004, 6, 851-860.

(10) Cheung, K. H.; Gu, J.-D. Mechanism of Hexavalent Chromium Detoxification by Microorganisms and Bioremediation Application Potential: A Review. Int. Biodeterior. Biodegradation 2007, 59, 8-15.

(11) Laitinen, R. S.; Pakkanen, T. A. 77Se NMR Study of the Sulphur-Selenium Binary System. J. Chem. Soc., Chem. Commun. 1986, 0, 1381-1382.

(12) Duddeck, H. Selenium-77 Nuclear Magnetic Resonance Spectroscopy. In Progress in NMR Spectroscopy; Emsley, J. W., Feeney, J., Sutcliffe, L. H., Eds.; Elsevier: London, 1995; Vol. 27, pp 1-323.

(13) Chen, Y.-W.; Li, L.; D’Ulivo, A.; Belzile, N. Extraction and Determination of Elemental Selenium in Sediments – A Comparative Study. Analytica Chimica Acta 2006, 577, 126- 133.

(14) Hassan, C. E.; Webster, T. J. The Effect of Red-Allotrope Selenium Nanoparticles on Head and Neck Squamous Cell Viability and Growth. Int. J. Nanomedicine 2016, 11, 3641-3654.

(15) Allen, D. W.; Taylor, B. F. The Chemistry of Heteroarylphosphorus Compounts. Part 15. Phosphorus-31 Nuclear Magnetic Resonance Studies of the Donor Properties of Heteroarylphosphines Towards Selenium and Platinum(II). J. Chem. Soc., Dalton Trans. 1982, 0, 51-54.

(16) Larsen, P. L.; Gupta, R.; Powell, D. R.; Borovik, A. S. Chalcogens as Terminal Ligands to Iron: Synthesis and Structure of Complexes with FeIII-S and FeIII-Se Motifs. J. Am. Chem. Soc. 2004, 126, 6522-6523.

(17) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of Iron(II) Aqua and Hydroxyl Complexes Featuring a Tripodal H-Bond Donor and Acceptor Ligand. Inorg Chem 2014, 53, 4450–4458.

78 CHAPTER 4: INVESTIGATIONS INTO THE SYNTHESIS AND REDUCTION OF A NON-HEME IRON(III)-OXO†

4.1 Introduction Heme and non-heme iron centers play a critical role in facilitating oxidation reactions in biological systems.1 Many of these iron metalloenzymes are purported to form a high-valent terminal iron-oxo as a catalytic intermediate, particularly when dioxygen serves as the oxidant.2 The synthesis of heme and non-heme model systems has greatly expanded the understanding of the reactivity and spectroscopic properties of high-valent iron-oxo species.3,4 While many biomimetic iron(IV)-oxo complexes have been synthesized and characterized, a paucity of terminal iron(III)-oxo complexes has been reported. Borovik and co-workers reported the first structurally characterized terminal iron(III)-oxo complex, which was generated from the activation of dioxygen.5 The incorporation of a secondary coordination sphere into the bioinspired iron complex stabilized the oxo moiety and allowed for its isolation.6 In addition, high-valent iron-oxo species mediate hydrogen atom transfer reactions in metalloenzymes.7 Notable examples of enzymes whose mechanism invoke hydrogen atom transfer include lipoxygenase, the oxygen evolving complex of photosystem II, and methane monooxygenase.7 In synthetic systems, hydrogen atom abstraction with substrates having known X–H bond energies provides a measure of the O–H bond dissociation energy in metal-hydroxo species.8 Inspired by Nature and synthetic biomimetic systems that model the formation of high- valent iron-oxo species, our lab synthesized a non-heme iron platform capable of tautomerization to include hydrogen-bond donors or acceptors in the secondary sphere.9 We previously reported Cy II the reduction of nitrite by the iron(II)-triflate species N(afa )3Fe(OTf)2 (Fe -OTf) to generate the Cy III triflate salt of the iron(III)-oxo complex [N(afa )3FeO]OTf (Fe -O) with concomitant release of nitric oxide (Figure 4.1).10 Characterization of FeIII-O confirmed its iron(III) formulation and solid-state structural analysis revealed an Fe–O bond length (1.8079(9) Å) comparable to that reported by Borovik and co-workers (1.813(3) Å).5 The iron(III)-oxo complex FeIII-O was also furnished from the one electron oxidation of Cy Cy II the iron(II)-hydroxo complex N(afa )2(pi )FeOH (Fe -OH) using silver or ferrocenium salts or

† The collection and integration of the solid-state structural data of [FeII-OH]OTf was completed by Dr. Chun-Hsing Chen.

79 electrochemically (Figure 4.1).11 In these oxidation reactions, a proton must be transferred from the bound hydroxo moiety to the ligand platform, resulting in tautomerization of the pyrrole-imine arm in FeII-OH to the azafulvene-amine tautomer. This process was also chemically reversible, as the addition of potassium graphite to FeIII-O regenerated FeII-OH.11 The formation of FeIII-O from the addition of oxyanion salts to FeII-OTf was described in Chapters 2 and 3.

Cy Cy Cy HN N HN FcOTf or AgOTf H O - Fc or Ag Fe N N N

N FeII-OH OTf Cy OTf Cy Cy Cy Cy O CF3 Cy HN S NH NH HN HN NH O O O [NBu4][NO2] - NO Fe N N Fe N N N N

N N III FeII-OTf Fe -O

OTf Cy Cy O CF3 Cy S NH HN NH O O halogen or selenium oxyanions N Fe N N

N FeII-OTf Figure 4.1 Summary of synthetic routes to access FeIII-O.

Because of the dearth of terminal iron(III)-oxo complexes, we were interested in further exploring the reactivity of FeIII-O. In order to catalytically reduce perchlorate, we utilized the traditional hydrogen atom abstraction reagent 1,2-diphenylhydrazine (DPH) to reduce FeIII-O and regenerate the active catalyst. The reactivity of FeIII-O with DPH is detailed below. Upon the reduction of FeIII-O by 0.5 equiv DPH, a new iron(II)-hydroxo species was generated, Cy II [N(afa )3FeOH]OTf ([Fe -OH]OTf). The solid-state structural characterization and alternative synthetic routes to access this complex are presented herein, along with a route to access FeIII-O from an octahedral iron(III) complex.

Cy 4.2 Reduction of [N(afa )3FeO]OTf with 1,2-Diphenylhydrazine During our studies of oxyanion reduction by the iron(II)-triflate species FeII-OTf, we sought a method to quantify the iron complexes that were produced from deoxygenation of the

80 oxyanion substrate. Generally, a mixture of the iron(III)-oxo complex FeIII-O and an iron(II) complex, either unreacted FeII-OTf or an iron(II)-halide in the case of halogen oxyanion reduction, were obtained upon the addition of an oxyanion to FeII-OTf. The solubility of these iron species in polar organic solvents were similar enough to make their isolation sufficiently difficult and to warrant a different method of quantification. Moreover, attempts to quantify the mixture of products by spectroscopic means were unsuccessful; the direct integration of the 1H NMR spectra of the crude oxyanion reduction mixtures was precluded due to the poor solubility of the iron(II)- halide complexes and quantification by electronic absorption spectroscopy was also ineffectual. Instead, we relied on the reaction of FeIII-O with 1,2-diphenylhydrazine (DPH) to yield azobenzene to quantify the amount of FeIII-O and iron(II) products that resulted from the oxyanion reduction. Because the iron(II) complexes, both FeII-OTf and the series of iron(II)-halide complexes, did not react with DPH, the amount of DPH that was converted to azobenzene measured the amount of FeIII-O that was present. One drawback of employing a chemical method to quantify the formation of FeIII-O was the possibility of side reactions between the oxyanion substrate and DPH. Thus, control reactions were necessary to determine whether DPH was compatible under the reaction conditions. In the case where reactivity between the oxyanion substrate and DPH was detected, the oxyanion reduction was completed first and DPH was added after the work-up of the crude reaction mixture to remove any residual oxyanion salt. Because the quantification of FeIII-O proceeded differently depending upon whether the addition of DPH was with the oxyanion substrate or in a second step following oxyanion reduction both methods will be discussed in detail. When DPH was added along with the oxyanion substrate for in situ quantification, DPH behaved as a 2 H+/2 e- source (Figure 4.2). We hypothesized that one electron went toward the oxygen atom transfer from the oxyanion substrate to FeII-OTf to generate FeIII-O. The remaining one electron reduced the iron(III) center and the oxo moiety was protonated by the two protons, resulting in the formation of water. This aquo ligand was then released from the iron(II) center, regenerating FeII-OTf. The production of water was confirmed via Karl Fischer titration for the stoichiometric reduction of perchlorate (Chapter 2). Furthermore, the 1H NMR spectra of the perchlorate reduction reaction confirmed the generation of FeII-OTf, along with the iron(II)- Cy II 1 chloride complex [N(afa )3FeCl]OTf (Fe -Cl). H NMR spectroscopy also allowed for the ratio of azobenzene to DPH to be determined. In this procedure, the amount of azobenzene formed

81 corresponded directly to the amount of FeIII-O that was generated during the course of the oxyanion reduction.

OTf Cy Cy Cy Cy OTf Cy O CF3 Cy S NH HN HN NH HN NH - - O O + EOx + e O Fe - EO - Fe N N N x-1 N N N E = Cl or Br N N II Fe -OTf FeIII-O

+ 2H+ + e-

- H2O

2H+ / 2e- = 1,2-diphenylhydrazine Figure 4.2 Proposed mechanism of in situ reduction of FeIII-O by 1,2-diphenylhydrazine.

Alternatively, when DPH was added at the conclusion of oxyanion reduction, it behaved as a 1 H+/1 e- source per iron center, as evidenced by the formation of a new iron(II) species. Further characterization of this complex demonstrated it to be an iron(II)-hydroxo species with dative coordination to the azafulvene-amine tautomeric form of the ligand and an outer-sphere Cy II II triflate anion, [N(afa )3FeOH]OTf ([Fe -OH]OTf). Moreover, [Fe -OH]OTf was the expected product from the addition of one electron and one proton to FeIII-O (Figure 4.3).

Cy Cy Cy Cy OTf Cy Cy OTf HN HN NH HN NH HN O HO + H+ + e- Fe Fe N N N N N N

N N FeIII-O [FeII-OH]OTf H+ / e- = 0.5 equiv 1,2-diphenylhydrazine Figure 4.3 Reduction of isolated FeIII-O by 1,2-diphenylhydrazine.

The formation of [FeII-OH]OTf from the addition of DPH to FeIII-O was optimized using independently prepared FeIII-O.10 While the stoichiometry of this transformation suggests 0.5 equiv DPH were requisite, we determined that an excess of DPH was required to elicit full conversion of FeIII-O to [FeII-OH]OTf. Complete consumption of FeIII-O was achieved upon the addition of 1 equiv DPH. The 1H NMR spectrum of this reaction demonstrated a 1:1 ratio of azobenzene to unreacted DPH, confirming the stoichiometry of the reaction was 0.5 equiv DPH for each equiv FeIII-O. Furthermore, the only paramagnetic iron-containing product observed in the 1H NMR spectrum was [FeII-OH]OTf, which was isolated in 94% yield. To further corroborate

82 the proposed stoichiometry, the reaction of 2 equiv DPH with 1 eq FeIII-O was completed. As predicted, the 1H NMR spectrum showed the conversion of 0.5 equiv DPH to azobenzene, with 1.5 equiv DPH remaining unreacted. Thus, when quantifying the amount of FeIII-O generated from oxyanion reduction, 1 equiv DPH was added for each equiv FeII-OTf used in the first oxyanion reduction step. Complete deoxygenation of the oxyanion substrate would result in quantitative formation of FeIII-O. Half of the added DPH would then be converted to azobenzene by FeIII-O. Therefore, complete deoxygenation of any given oxyanion substrate would result in a ratio of 0.5 equiv azobenzene to 0.5 equiv DPH using this method of quantification. We also investigated the reactivity of FeIII-O with more biologically relevant 1 H+/1 e- III sources: thiols. The reaction of triphenylmethanethiol (Ph3CSH) with Fe -O proceeded to form II [Fe -OH]OTf. Conversion was only complete, however, when 2 equiv Ph3CSH were added and the reaction was heated to 50 °C. Due to the overlapping resonances in the 1H NMR spectrum of the thiol (Ph3CSH) and the oxidized disulfide product (Ph3CSSCPh3), we were unable to use this method to quantify FeIII-O. Thus, we were interested in the use of 4-fluorothiophenol because of the possibility of quantifying FeIII-O by 19F NMR spectroscopy. Unfortunately, the addition of 4- fluorothiophenol to FeIII-O did not result in the formation [FeII-OH]OTf. The protodemetallated Cy 1 ligand, N(pi )3 • 3 HOTf, was observed in the H NMR spectrum, an indication that the acidic thiol had protonated the iron(III) complex instead of participating in hydrogen atom abstraction.

Cy 4.3 Synthesis and characterization of [N(afa )3FeOH]OTf The synthesis of [FeII-OH]OTf was achieved through two alternative routes (Figure 4.4). The iron(II) species was furnished from the addition of 1 equiv acid to the iron(II)-hydroxide Cy Cy II complex N(afa )2(pi )FeOH (Fe -OH). Also, the addition of 2 equiv acid to the potassium salt Cy II II of the iron(II)-aquo complex, K[N(pi )3FeOH2] (K[Fe -OH2]), generated [Fe -OH]OTf. II II Similar to the previous work in our lab demonstrating the synthesis of Fe -OH from K[Fe -OH2] upon addition of 1 equiv acid9, the formation of [FeII-OH]OTf from both FeII-OH and K[FeII-

OH2] requires tautomerization of the ligand platform, demonstrating its flexibility. In addition, [FeII-OH]OTf was observed when water was added to FeII-OTf, although the reaction was not quantitative.

83 + 2H+

Cy K Cy OTf Cy Cy Cy Cy Cy Cy Cy N HN HN N N N HN NH HN H H O H O HO + H+ + H+ Fe Fe Fe N N N N N N N N N

N N N

II II II K[Fe -OH2] Fe -OH [Fe -OH]OTf II II II Figure 4.4 Synthesis of [Fe -OH]OTf via addition of acid to K[Fe -OH2] or Fe -OH.

Crystals of [FeII-OH]OTf that were suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a saturated solution of dichloromethane and tetrahydrofuran (1:1). Refinement of the data revealed an iron(II) center in trigonal bipyramidal geometry with an axial hydroxide ligand and an outer-sphere triflate anion (Figure 4.5). The hydrogen atom of the hydroxide ligand and the hydrogen atoms of two amine moieties of the secondary sphere were located in the difference map and independently refined. The remaining hydrogen atom of the secondary sphere was refined in its idealized position. While the Fe–O bond was shorter than our 11 previously reported iron(II)-hydroxo complexes , the solution magnetic moment of 5.42(39) µB was consistent with a high-spin iron(II) center. The IR spectrum of the complex also aided in confirming its formulation: a stretching frequency at 3539 cm-1 corresponded to the O–H moiety and a stretching frequency at 1649 cm-1 was consistent with the C=N bond of the azafulvene-amine tautomer of the ligand platform.9

H5 H1 H6 O1 H7 Fe1

[FeII-OH]OTf

Figure 4.5 Molecular structure of [FeII-OH]OTf shown with 50% probability ellipsoids. Select hydrogen atoms and the outer-sphere triflate anion have been emitted for clarity.

84 Table 4.1 Selected structural parameters of [FeII-OH]OTf. [FeII-OH]OTf

Fe–O (Å) 1.857(4) Fe–N (Å) 2.304(4)

Fe–Nafa (Å) 2.056(5) 2.040(5) 2.069(4) O–H stretch (cm-1) 3539 N–H stretch (cm-1) 3233, 3188 C=N stretch (cm-1) 1649

µeff (µB) 5.42(39)

The reactivity of [FeII-OH]OTf was briefly explored. The reaction of [FeII-OH]OTf with AgOTf furnished the one electron oxidized product, FeIII-OH, and a new complex with a characteristic resonance in the 1H NMR spectrum at 48 ppm. Isolation of this species to confirm its formulation was unsuccessful. However, this reactivity accounted for the formation of the II unidentified species from the reaction of Ag2SeO4 with Fe -OTf (Chapter 3). Residual water from the selenate salt could have formed [FeII-OH]OTf from FeII-OTf, which could then react with the silver cations in solution.

Cy Cy 4.4 Synthesis of [N(afa )3FeO]OTf from N(pi )3Fe III Cy The alternative preparation of Fe -O from the octahedral iron(III) complex N(pi )3Fe Cy was explored. N(pi )3Fe was furnished from the addition of FeCl3 to the triply deprotonated Cy 9 potassium salt of the ligand (K3[N(pi )3]). This molecule was previously thought to be Cy unreactive; however, when 1 equiv N(pi )3Fe and 1 equiv acid were combined and stirred for 16 h, FeIII-O was formed (Figure 4.6). We hypothesized that the oxo ligand originated from adventitious water and that the iron(III) center dissociated from the imine moieties of the ligand to allow for installation of an aquo ligand. This aquo ligand and the proton from the added acid provided the requisite three hydrogen atoms and one oxygen atom necessary to furnish FeIII-O, upon tautomerization of all three arms of the ligand platform from the pyrrole-imine tautomer to

85 the azafulvene-amine tautomer. Overall, this transformation demonstrated the flexibility of our ligand scaffold and the propensity of the system to support the iron(III)-oxo complex FeIII-O.

OTf Cy Cy Cy Cy Cy Cy N HN N N NH HN Fe O + H+ Fe N N N + H2O N N N

N N

Cy III N(pi )3Fe Fe -O III Cy Figure 4.6 Synthesis of Fe -O from N(pi )3Fe.

4.5 Conclusions The reduction of high-valent iron-oxos is fundamental to the functioning of biological systems. In this chapter, we explored the reduction of FeIII-O and discovered a new method to furnish FeIII-O from an octahedral iron(III) complex. While we sought to explore the reduction of FeIII-O as a means to quantify the products we observed in oxyanion reduction, studying this transformation contributes to the understanding of complex biological systems and provides a means to improving future catalytic oxyanion efforts involving our non-heme system. Furthermore, the iron(II)-hydroxo species furnished from the reduction of FeIII-O by DPH, [FeII- OH]OTf, was characterized by a solid-state structural analysis and IR spectroscopy. Insights into the reactivity of the iron(II)-hydroxo species were garnered from the addition of oxidant to [FeII- OH]OTf.

4.6 Experimental General considerations. To avoid contact with oxygen and water, all manipulations were carried out under an atmosphere of nitrogen in an MBraun inert atmosphere drybox or using standard Schlenk techniques. Solvents for air- and moisture-sensitive manipulations were dried and deoxygenated using a Glass Contour System and stored over 4 Å molecular sieves prior to use. Celite 545 was heated to 150 °C under dynamic vacuum for 24 h prior to use in the drybox. All reagents were purchased from commercial sources and used as received unless otherwise noted. 1,2-Diphenylhydrazine was recrystallized from a saturated solution in diethyl ether layered with hexane and stored at -35 °C. Formic acid (88%) was dried by refluxing with phthalic anhydride 12 Cy 9 for 6 h followed by distillation. Ferrous trifluoromethanesulfonate, N(afa )3Fe(OTf)2,

86 Cy 9 Cy 10 N(pi )3Fe, and [N(afa )3FeO]OTf were synthesized according to literature procedure. NMR solvent (acetonitrile-d3) was degassed and stored over 4 Å molecular sieves prior to use. NMR spectra were recorded at ambient temperature on a Varian spectrometer operating at 500 MHz (1H NMR) and referenced to the peak of the residual solvent (δ parts per million and J in Hz). Solid- state infrared spectra were measured using a PerkinElmer Frontier FT-IR spectrophotometer equipped with a KRS5 thallium bromide/iodide universal attenuated total reflectance accessory. UV-vis spectra were recorded using a HP 8453 UV-vis spectrophotometer. Elemental analyses were performed by the University of Illinois at Urbana-Champaign School of Chemical Sciences Microanalysis Laboratory in Urbana, IL. Solid-state structural characterization was done using the Advanced Photon Source at Argonne National Laboratory. WARNING: Perchlorate salts are explosive and should be handled with care in small quantities using a plastic spatula.

Figure 4.7 Electronic absorption spectrum of FeIII-O (0.01 mM in MeCN, 21 °C).

Cy Quantification of [N(afa )3FeO]OTf in situ using 1,2-diphenylhydrazine. This reaction was carried out as outlined in Chapter 2 for the quantification of the products of chlorine and bromine oxyanion reduction. Upon separating the iron-containing products from the azobenzene and unreacted 1,2-diphenylhydrazine by triturating the dried crude reaction mixture with diethyl ether, the remaining residue was examined by 1H NMR spectroscopy. The 1H NMR spectrum showed a Cy Cy mixture of N(afa )3Fe(OTf)2 and the corresponding iron(II)-halide complex, [N(afa )3FeX]OTf X = Cl or Br.

87 OTf Cy OTf Cy Cy Cy Cy O CF3 Cy NH S NH HN NH HN NH Cl O O Fe N N Fe N N N N ◼ ◼ N N ◼ ▲ ▲ ▲ ▲

1 II Figure 4.8 H NMR spectrum of the reaction of 5 equiv Fe -OTf, 1 equiv [NBu4][ClO4], and II II 5 equiv 1,2-diphenylhydrazine, demonstrating a mixture of Fe -OTf and Fe -Cl (CD3CN, 21 °C).

Cy Quantification of [N(afa )3FeO]OTf by the addition of 1,2-diphenylhydrazine following oxyanion reduction. The detailed description of the reaction conditions is outlined in Chapter 2 and Chapter 3 for the quantification of the products of iodine and selenium oxyanion reduction. Upon separating the iron-containing products from the azobenzene and unreacted 1,2- diphenylhydrazine by triturating the dried crude reaction mixture with diethyl ether, the remaining residue was examined by 1H NMR spectroscopy. The 1H NMR spectrum showed the formation of Cy [N(afa )3FeOH]OTf.

1 II Figure 4.9 H NMR spectrum of [Fe -OH]OTf (CD3CN, 21 °C).

88 Cy Optimization of the reaction of [N(afa )3FeO]OTf and 1,2-diphenylhydrazine. A 20 mL Cy scintillation vial was charged with [N(afa )3FeO]OTf (0.0482 g, 0.0600 mmol), 1,2- diphenylhydrazine (0.5 equiv = 0.0055 g, 0.0300 mmol; 1.0 equiv = 0.0111 g, 0.0600; 2.0 equiv =

III Fe -O + DPH H N N N N H

H N III N Fe -O + 2 DPH H

N N

Figure 4.10 1H NMR spectra demonstrating the expected ratio of azobenzene:1,2- diphenylhydrazine upon the addition of 1 equiv (top) or 2 equiv (bottom) 1,2- III diphenylhydrazine to Fe -O (CD3CN, 21 °C).

89 0.0221 g, 0.120 mmol), and 4 mL acetonitrile. The solution was stirred for 16 h and evaporated to dryness in vacuo. The resulting red powder was washed with 4 mL diethyl ether and filtered over a pad of Celite to isolate the azobenzene and 1,2-diphenylhydrazine. The remaining red solid was collected into a clean 20 mL scintillation vial with 4 mL dichloromethane. Both the diethyl ether solution and the dichloromethane solution were evaporated to dryness and examined by 1H NMR spectroscopy.

Cy Reactivity of [N(afa )3FeO]OTf with triphenylmethanethiol. A 15 mL Schlenk bomb was Cy charged with [N(afa )3FeO]OTf (0.0241 g, 0.0300 mmol), Ph3CSH (0.0166 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was heated to 50 °C and stirred for 16 h. The reaction was then dried in vacuo and analyzed by 1H NMR spectroscopy.

1 III Figure 4.11 H NMR spectrum of the reaction of Fe -O and 2 equiv Ph3CSH showing the II formation of [Fe -OH]OTf (CD3CN, 21 °C).

Cy Reactivity of [N(afa )3FeO]OTf with 4-fluorothiophenol. A 20 mL scintillation vial was Cy charged with [N(afa )3FeO]OTf (0.0241 g, 0.0300 mmol), 4-fluorothiophenol (0.0.038 g, 0.0300 mmol), and 4 mL acetonitrile. The solution was stirred for 3 h before volatiles were removed under reduced pressure and the resulting residue was analyzed by 1H NMR spectroscopy, which showed Cy formation of N(pi )3 • 3 HOTf.

90 Cy OTf Cy 3 OTf Cy 2 OTf Cy OTf Cy Cy Cy Cy Cy Cy Cy ● Cy HN HN HN NH NH HN NH HN HN NH HN NH O HO Cl

Fe Fe Fe N N N N NH HNHN N N N N N

N N N N ●

●

Figure 4.12 1H NMR spectrum of the reaction of FeIII-O with 4-fluorothiophenol, which Cy formed N(pi )3 • 3 HOTf (CD3CN, 21 °C).

Cy Preparation of [N(afa )3FeOH]OTf. A 20 mL scintillation vial was charged with Cy [N(afa )3FeO]OTf (0.0482 g, 0.0600 mmol), 1,2-diphenylhydrazine (0.0111 g, 0.0600 mmol), and 4 mL acetonitrile. The solution was stirred for 16 h. Volatiles were removed under reduced pressure. The resulting red powder was washed with 4 mL diethyl ether and filtered over a pad of Celite to remove azobenzene and 1,2-diphenylhydrazine. The remaining red solid was collected in a clean 20 mL scintillation vial with 4 mL dichloromethane. This solution was evaporated to Cy dryness, yielding the product [N(afa )3FeOH]OTf (0.0453 g, 0.0564 mmol, 94% yield). Crystals suitable for X-ray diffraction were grown from the vapor diffusion of diethyl ether into a saturated solution of the target molecule in dichloromethane and tetrahydrofuran (1:1). Analysis for

C37H52N7FeSO4F3: Calcd. C, 55.29; H, 6.52; N, 12.20. Found C, 55.01; H, 6.35; N, 12.18%. IR = -1 -1 -1 3539 cm (O-H), 3233 and 3188 cm (N-H), 1649 cm (C=N). µeff = 5.42(39) µB.

Cy Cy Cy Alternative synthesis of [N(afa )3FeOH]OTf from N(pi )(afa )2FeOH. A 20 mL Cy Cy scintillation vial was charged with N(pi )(afa )2FeOH (0.0261 g, 0.0400 mmol) and 4 mL acetonitrile. The solution was cooled to -35 °C before HOTf (0.0060 g, 0.0400 mmol) was weighed by difference and added to the pale brown solution, resulting in a color change to deep red. After stirring for 30 min, volatiles were removed under reduced pressure and the formation of Cy 1 [N(afa )3FeOH]OTf was confirmed by H NMR spectroscopy.

91 Cy Cy Alternative synthesis of [N(afa )3FeOH]OTf from K[N(pi )3Fe(OH2)]. A 20 mL scintillation Cy vial was charged with K[N(pi )3Fe(OH2)] (0.0277 g, 0.0400 mmol) and 4 mL tetrahydrofuran. After the solution was cooled to -35 °C, HOTf (0.0120 g, 0.0800 mmol) was weighed by difference and added to the stirring orange solution, which gradually became deep red. After stirring for 1 h, the solution was evaporated to dryness and analyzed by 1H NMR spectroscopy, confirming the Cy formation of [N(afa )3FeOH]OTf.

Cy Addition of water to N(afa )3Fe(OTf)2. A 50 mL Schlenk flask was charged with Cy N(afa )3Fe(OTf)2 (0.0187 g, 0.0200 mmol) and 3 mL acetonitrile under a nitrogen atmosphere. In a separate Schlenk flask, a 0.2 M solution of water in tetrahydrofuran was degassed. Using a gas-tight syringe, 0.1 mL of the water solution was transferred to the acetonitrile solution. The reaction was stirred under a flow of nitrogen for 3 h. An aliquot of the reaction mixture was removed using a clean gas-tight syringe, evaporated to dryness under an inert atmosphere, and 1 Cy analyzed by H NMR spectroscopy. The formation of a small amount of [N(afa )3FeOH]OTf was noted.

OTf Cy Cy Cy Cy OTf ◼ ◼ Cy O CF3 Cy HN S NH NH HN HN NH O O HO Fe N Fe N N N N N

N N ◼ ▲

▲ ▲ ▲ ▲

1 II Figure 4.13 H NMR spectrum of the addition of 1 equiv H2O to Fe -OTf (CD3CN, 21 °C).

92 Cy Oxidation of [N(afa )3FeOH]OTf with AgOTf. A 20 mL scintillation vial was charged with Cy [N(afa )3FeOH]OTf (0.0300 g, 0.0373 mmol) and 3 mL dichloromethane. The solution was cooled to -35 °C before AgOTf (0.0096 g, 0.0373 mmol) was weighed by difference and added. The solution was stirred 3 h, filtered to remove silver metal, and evaporated to dryness in vacuo. The resulting residue was analyzed by 1H NMR spectroscopy and revealed the formation of two Cy paramagnetic products: [N(afa )3FeOH](OTf)2 and an unidentified product with a characteristic resonance at 48 ppm.

Cy 2 OTf Cy Cy HN HN NH HO Fe N N N

N ● ● ● ● ● ●

1 II Figure 4.14 H NMR spectrum of the reaction of [Fe -OH]OTf and AgOTf (CD3CN, 21 °C).

Cy Cy Synthesis of [N(afa )3FeO]OTf from N(pi )3Fe. A 20 mL scintillation vial was charged with Cy N(pi )3Fe (0.0254 g, 0.0400 mmol) and 4 mL tetrahydrofuran. The solution was cooled to -35 °C and HOTf (0.0060 g, 0.0400 mmol), also cooled to -35 °C, was weighted by difference and added to the forest green solution. The reaction was brought to room temperature and stirred 16 h. Volatiles were removed under reduced pressure and the brown residue was analyzed by 1H NMR Cy spectroscopy, revealing formation of [N(afa )3FeO]OTf.

1 III Cy Figure 4.15 H NMR spectrum of Fe -O synthesized from the addition of acid to N(pi )3Fe (CD3CN, 21 °C).

93 Table 4.2 Crystallographic parameters of [FeII-OH]OTf. Cy [N(afa )3FeOH]OTf

Empirical Formula C37H52F3FeN7O4S Formula Weight 803.78 Temperature 100 K Wavelength 0.71073 Crystal System Triclinic Space Group P-1 a = 12.9155(13) Å Unit Cell Dimensions b = 13.4401(13) Å

c = 13.8397(13) Å

α = 63.7230(10)°

β = 75.002(2)°

γ = 65.480(2)° Volume 1951.5(3) Z 2 Reflections Collected 19243 Independent Reflections 6596 Goodness-of-Fit on F2 1.031 R1 = 00686 Final R Indices [I>2sigma(I)] wR2 = 0.1784 R1 = 0.1277 R Indices wR2 = 0.2144

4.7 References (1) Groves, T. J. High-Valent Iron in Chemical and Biological Oxidations. J. Inorg. Biochem. 2006, 100, 434-447.

94 (2) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L. Dioxygen Activation at Mononuclear Nonheme Iron Active Sites: Enzymes, Models, and Intermediates. Chem. Rev. 2004, 104, 939-986.

(3) Que, L. The Road to Non-Heme Oxoferryls and Beyone. Acc. Chem. Res. 2007, 40, 493- 500.

(4) Cook, S. A.; Borovik, A. S. Molecular Designs for Controlling Local Environments Around Metal Ions. Acc. Chem. Res. 2015, 48, 2407-2414.

(5) MacBeth, C. E.; Golombek, A. P.; Young, V. G.; Yang, C.; Kuczera, K.; Hendrick, M. P.; Borovik, A. S. O2 Activation by Nonheme Iron Complexes: A Monomeric Fe(III)-Oxo Complex Derived from O2. Science 2000, 289, 938-941.

(6) Borovik, A. S. Bioinspired Hydrogen Bond Motifs in Ligand Design: The Role of Noncovalent Interactions in Metal Ion Mediated Activation of Dioxygen. Acc. Chem. Res. 2005, 38, 54-61.

(7) Gupta, R.; Borovik, A. S. Monomeric MnIII/II and FeIII/II Complexes with Terminal Hydroxo and Oxo Ligands: Probing Reactivity via O–H Bond Dissociation Energies. J. Am. Chem. Soc. 2003, 125, 13234-13242.

(8) Mayer, J. M. Hydrogen Atom Abstraction by Metal-Oxo Complexes: Understanding the Analogy with Organic Radical Reactions. Acc. Chem. Res. 1998, 31, 441-450.

(9) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of Iron(II) Aqua and Hydroxyl Complexes Featuring a Tripodal H-Bond Donor and Acceptor Ligand. Inorg. Chem. 2014, 53, 4450-4458.

(10) Matson, E. M.; Park, Y. J.; Fout, A. R. Facile Nitrite Reduction in a Non-Heme Iron System: Formation of an Iron(III)-Oxo. J. Am. Chem. Soc. 2014, 136, 17398-17401.

(11) Gordon, Z.; Drummond, M. J.; Matson, E. M.; Bogard, J. A.; Schelter, E. J.; Lord, R. L.; Fout. A. R. Tuning the Fe(II/III) Redox Potential in Nonheme Fe(II)-Hydroxo Complexes Through Primary and Secondary Sphere Modifications. Inorg. Chem. 2017, 56, 4852-4863.

(12) Hagen, K. S. Iron(II) Triflate Salts as Convenient Substitutes for Perchlorate Salts: Crystal Structures of [Fe(H2O)6](CF3SO3)2 and Fe(MeCN)4(CF3SO3)2. Inorg. Chem. 2000, 39, 5867-5869.

95 CHAPTER 5: EFFECTS OF VARYING SECONDARY SPHERE INTERACTIONS: SYNTHESIS OF COBALT AND IRON COMPLEXES OF A TRIPODAL LIGAND FEATURING TWO HYDROGEN-BOND DONORS OR ACCEPTORS†

5.1 Introduction In biological systems, the function of a metalloenzyme is determined by both the primary and secondary coordination spheres surrounding the active site. The primary sphere influences the electronic properties of the metal center, whereas the secondary sphere imparts substrate selectivity.1 Synthetic systems can readily mimic the primary coordination sphere of many metalloenzymes, and a myriad of investigations have delineated the influence of the primary sphere on metal reactivity. Studies in which the secondary sphere is incorporated into a synthetic system are relatively scarce due to the difficulty in controlling intramolecular hydrogen-bonding interactions.2 There are even fewer reports of the effects of varying the secondary sphere interactions on the reactivity of a metal center. Pioneering work by Borovik and co-workers incorporated hydrogen-bond donors into a rigid tripodal ligand framework and allowed for the isolation of the first terminal iron(III)-oxo complex.3 The importance of the secondary sphere in the reactivity was further underlined in a later study by Borovik and co-workers that systematically reduced the number of hydrogen-bond donors present in a series of metal complexes (Figure 5.1).4 The electronic structure of the cobalt(II) centers was maintained through substitution of the urea functionalities with carboxamide moieties and confirmed in the series of cobalt(II) complexes through electronic absorption spectroscopy and EPR spectroscopy. Upon the addition with dioxygen to form the cobalt(III)- hydroxo complexes, stark differences were observed in the reactivity of the cobalt(II) complexes. When two or three hydrogen-bond donors were present, the complex readily reacted with 0.5 equiv dioxygen to furnish the cobalt(III)-hydroxo complex. If only one hydrogen-bond donor was present, an excess of dioxygen was required to synthesize the cobalt(III) product, and when no hydrogen-bond donors were present, no reaction was observed. Thus, the number of hydrogen- bond donors present in the secondary sphere had implicit involvement in the dioxygen activation at the cobalt(II) centers.

† The synthesis and solid-state structural analysis of PyMn-OTf was completed by Yun Ji Park.

96 But But But NH NH But NH O O O iPr But HN But N O N iPr N iPr N iPr NH NH O Co N Co N iPr Co N iPr Co N O N O N N N N N O N O N O O O

0.5 eq O 0.5 eq O xs O 2 2 2 Xxs O2

But But But NH t NH NH O Bu O O iPr t t Bu OH HN Bu OH OH OH N O N iPr N iPr N iPr NH NH O Co N Co N iPr Co N iPr Co N O N O N N N N N O N O N O O O Figure 5.1 Dioxygen activation in a series of cobalt(II) urea/carboxamido hybrid complexes reported by Borovik and co-workers.4

A similar modular approach was taken in developing a series urea/sulfonamido hybrid ligands, which contained a mixture of hydrogen-bond donors and acceptors in the secondary sphere.5 Comparing the series of cobalt(II)-hydroxo complexes with varied secondary coordination spheres revealed differences in the Co–O bond length and the strength of the hydrogen-bonding interaction between the hydroxo ligand and the sulfonamido of the ligand, measured by IR spectroscopy. However, the variation in the ligand field strength of the urea versus the sulfonamide groups changed the electronic structure of the cobalt(II) center across the series and made discerning the effects of modulating the secondary sphere onerous. To further contribute to this area of work, our group also set out to investigate the effects of tuning the primary and secondary coordination sphere on the reactivity of a metal center. The primary coordination sphere of our tripodal non-heme system was brominated on the pyrrole backbone, resulting in an increased iron(II/III) redox couple (Figure 5.2).6 In addition, exchanging the capping group from a cyclohexyl moiety to an electron-withdrawing phenyl group also increased the iron(II/III) redox couple and the effects were additive, as observed in the case of the brominated ligand with a phenyl capping group. These studies, however, maintained the number of hydrogen-bond donors or acceptors in the secondary coordination sphere.

97 Cy Cy Ph Cy Cy Cy Cy Ph Ph Ph Ph Ph HN HN HN HN N HN N HN N HN N HN H H H H O O O O Fe Fe Fe Fe N N N N N N N N N N N N Br Br N N N N Br Br Br Br Iron(II/III) couple - 0.69 V - 0.49 V - 0.51 V - 0.29 V vs Fc/Fc+

Figure 5.2 Effect of primary and secondary sphere modifications on the iron(II/III) couple.

In order to expand our knowledge of the secondary sphere in our system, we synthesized a new non-heme ligand scaffold containing only two hydrogen-bond donors or acceptors to probe the effect on the resulting metal complexes. This chapter details the synthesis of the new ligand, where one pyrrole-imine arm has been replaced by a pyridyl donor, thus allowing for neutral coordination to the metal center while removing one hydrogen-bond donor or acceptor from the secondary coordination sphere. We next synthesized a family of cobalt(II) and iron(II) complexes, featuring both anionic and dative coordination to the ligand. The secondary sphere of these complexes was characterized in the solid-state using X-ray crystallography and IR spectroscopy. Finally, the iron(II) complexes were oxidized to compare their reactivity to that of the tris(5- Cy 7 cyclohexylimino-pyrrol-2-ylmethyl)amine (N(pi )3) platform.

Py Cy 5.2 Synthesis of a ligand featuring two secondary sphere interactions: N(pi )2 Py Cy The bis(5-cyclohexyliminopyrrol-2-ylmethyl)-2-pyridylmethylamine ( N(pi )2) ligand platform was designed to fulfill two goals: (1) to increase substrate accessibility to a metal center while preserving a rigid secondary sphere and (2) to investigate the effects of varying the number of hydrogen-bond donors or acceptors present in the secondary coordination sphere. By installing Cy an equatorial pyridine donor in place of one pyrrole-imine arm of the N(pi )3 system, trigonal bipyramidal geometry would be maintained about a bound metal center and neutral coordination of the ligand platform could still be achieved. This ligand design would thus allow for direct Py Cy Cy comparisons between metal complexes of the N(pi )2 and N(pi )3 ligand platforms, providing insight into the structure-function relationship of the secondary coordination sphere in our system. Py Cy Synthesis of N(pi )2 began with the reductive amination of 2 equiv pyrrole-2- carboxaldehyde and 1 equiv 2-picolylamine with sodium triacetoxyborohydride to form

98 bis(pyrrol-2-ylmethyl)-2-pyridylmethylamine. We followed the modular approach previously established in our lab to generate the pyrrole-imine functionality; a Vilsmeier-Haack reaction resulted in 5′ formylation of the pyrrolyl groups and subsequent condensation of cyclohexylamine 6,7 Py Cy 1 furnished the ligand. Analysis of the N(pi )2 ligand by H NMR spectroscopy demonstrated the expected 1:2 ratio of the pyridyl to pyrrolyl C–H resonances. A singlet at 8.04 ppm confirmed the successful installation of the secondary imine and the C=N stretching frequency of 1630 cm-1 Cy was measured by IR spectroscopy. As with the N(pi )3 ligand and its derivatives, the isolated Py Cy N(pi )2 ligand was a hydrate and could be made anhydrous by the dissolution of the ligand in diethyl ether over 4 Å molecular sieves.6,7 Cy Cy O N O N

1. POCl3 // DMF HN HN HN N HN 2. NaOAc N HN H2NCy N HN - H O N N 2 N

Py Cy Figure 5.3 Synthesis of N(pi )3.

5.3 Synthesis of cobalt(II) complexes Py Cy Following the successful synthesis of N(pi )2, we explored metalation with cobalt(II) Cy 6 salts (Figure 5.4). Emulating the anionic metalation of N(pi )3, the two pyrrolyl moieties of the ligand were deprotonated by the addition of 2.1 equiv KH to a tetrahydrofuran solution of Py Cy N(pi )2. After stirring for 2 h, the mixture was filtered over Celite into a slurry of CoCl2 in acetonitrile, generating a deep red solution. Solid-state structural characterization of the product Py Cy Py revealed a cobalt(II)-aquo complex in trigonal bipyramidal geometry, N(afa )2Co(OH2) ( Co-

OH2), with the aquo ligand bound trans to the apical nitrogen of the ligand platform (Figure 5.5). The electron density of the aquo hydrogen atoms were located in the difference map of the refined data and were within hydrogen-bonding distance of the imine nitrogen atoms of the secondary sphere of the ligand.8 The O1–N5 and O1– N6 distances were 2.6894(14) Å and 2.6792(14) Å, respectively, and the O1–H1A–N5 and O1–H1B–N6 bond angles were 168.2(18)° and 172.4°, respectively. The Co1–O1 bond distance, 2.0352(10) Å, was within the range of other cobalt(II)- aquo complexes (2.018(3) – 2.1538(11) Å).9 The C=N stretching frequency of the ligand was measured by solid-state IR spectroscopy to be 1634 cm-1.

99 Cy Cy N N H H O 2.1 eq KH M MCl2 N N N Py M-OH2 N

Cy Cy OTf Cy Cy N HN N H HN O M(OTf)2 xs Li O M HN 2 N Py N HN N N M-OH N N

Cy HN OTf O F3C Cy S O N O NH M(OTf)2 O Py M M-OTf N N M = Fe or Co N

Py Cy Figure 5.4 Synthesis of metal(II) complexes of N(pi )2.

We also sought to characterize cobalt(II) complexes that exhibited dative coordination to the ligand platform. Upon the addition of Co(OTf)2(MeCN)2 to wet ligand, two species were Py Cy Py formed: a cobalt(II)-triflate species, N(afa )2Co(OTf)2 ( Co-OTf), and a cobalt(II)-hydroxo Py Cy Py Py species, [ N(afa )2CoOH]OTf ( Co-OH). Crystals of Co-OH that were suitable for X-ray diffraction were grown from the mixture of products. Refinement of the data revealed a cobalt(II) center in trigonal bipyramidal geometry (Figure 5.5). The hydroxo moiety was engaged in hydrogen-bonding with both amines in the secondary coordination sphere and the outer-sphere triflate anion, which was confirmed through the distances between the hydrogen-bond donors and acceptors and the nearly linear angles of the three atoms involved in each hydrogen-bond.8 The O1–O2 distance was 2.917(2) Å and the O1–H1–O2 angle was 171(3)°; the N5–O1 and N6–O1 distances were 2.614(2) Å and 2.706(2) Å, respectively, and the N5–H5–O1 and N6–H6–O1 angles were 175(3)° and 176(3)°, respectively. The Co1–O1 bond distance, 1.9932(15) Å, was within the range of other structurally characterized five-coordinate cobalt(II)-hydroxo complexes (1.938(2) – 2.053(3) Å).10 Analysis of PyCo-OH by solid-state IR spectroscopy revealed an O–H

100 stretch at 3503 cm-1 and a C=N stretch at 1659 cm-1. The shortened Co1–O1 bond distance Py compared to Co-OH2 and the blue-shift in the C=N stretching frequency observed by IR spectroscopy were consistent with coordination of an anionic hydroxide ligand and neutral coordination of the ligand platform to the cobalt(II) center.7

N6 N6 H1B H1 H6 N5 N5 O1 H1A O1 H5

Co1 Co1

Py Py Co-OH2 Co-OH

Py Py Figure 5.5 Molecular structures of Co-OH2 and Co-OH drawn with 50% probability ellipsoids. Select hydrogen atoms and outer-sphere anions have been omitted for clarity.

Py Py Table 5.1 Select structural parameters of Co-OH2 and Co-OH. Py Py Co-OH2 Co-OH Co–O (Å) 2.0351(10 1.9926(15)

Co–Naxial (Å) 2.2564(11) 2.2418(18) 2.0268(10) 2.0493(18) Co–N (Å) 2.0297(10) 2.0457(18)

Co–NPy (Å) 2.0771(11) 2.1019(18) C=N stretch (cm-1) 1626 1659

Both PyCo-OTf and PyCo-OH were independently synthesized. When dried ligand was Py added to Co(OTf)2(MeCN)2, Co-OTf was the sole product formed. Crystals suitable for X-ray diffraction were unable to be grown; however, the analogous Mn(II)-triflate species, Py Cy Py N(afa )2Mn(OTf)2 ( Mn-OTf) was characterized crystallographically (Figure 5.6). The refined data featured an octahedral Mn(II) center, with axial coordination of a triflate anion. A

101 tetrahydrofuran solvate was bound trans to the pyridyl nitrogen, leading to trans coordination of the azafulvene nitrogen atoms. Based on the elemental analysis of PyCo-OTf, which included a tetrahydrofuran solvate, we proposed an analogous coordination environment about the cobalt(II) center. Attempts to oxidize PyCo-OTf were unsuccessful. PyCo-OH was synthesized from the Py Cy reaction of N(pi )2, Co(OTf)2(MeCN)2, and excess Li2O in tetrahydrofuran overnight. The reaction initially turned deep red, indicative of the formation of PyCo-OTf, but gradually turned green as the red species was converted to PyCo-OH.

PyMn-OTf

Figure 5.6 Molecular structure of PyMn-OTf drawn with 50% probability ellipsoids. Select hydrogen atoms and the outer-sphere anion have been omitted for clarity.

5.4 Synthesis of iron(II) complexes Py Cy Py Py Cy Py The iron(II) species N(afa )2Fe(OH2) ( Fe-OH2), [ N(afa )2FeOH]OTf ( Fe-OH), Py Cy Py and N(afa )2Fe(OTf)2 ( Fe-OTf) were synthesized analogously to the cobalt(II) complexes in Py Cy order to directly compare the reactivity of the N(pi )2 ligand system with the reported iron Cy 7 Py chemistry of the N(pi )3 ligand system (Figure 5.4). Fe-OH2 was generated from the addition Py Cy of the deprotonated ligand salt K2[ N(pi )2] to FeCl2. Crystals suitable for X-ray diffraction were grown from a concentrated solution of diethyl ether at -35 °C. Refinement of the data showed an iron(II) center in trigonal bipyramidal geometry with an axial aquo ligand (Figure 5.7). The hydrogen atoms of the aquo ligand were located in the difference map and were within hydrogen- bonding distance of the imine moieties of the ligand platform.8 The O1–N5 and O1–N6 distances were 2.6990(17) Å and 2.7015(18) Å, respectively, and the O1–H1A–N5 and O1–H1B–N6 angles were 173(2)° and 169(2)°, respectively. The Fe1–O1 bond distance, 2.0587(11) Å, was within

102 range of reported iron(II)-aquo complexes (1.976(6) – 2.123(3) Å) and shorter than the Fe1–O1 Cy II 7,11 distance reported for K[N(pi )3FeOH2] (Fe -OH2; 2.080(2) Å). Py Fe-OH was synthesized from both the addition of wet ligand to Fe(OTf)2(MeCN)2 and the reaction of dry ligand, Fe(OTf)2(MeCN)2, and excess Li2O. Crystals suitable for X-ray diffraction were grown from a saturated solution of tetrahydrofuran layered with pentane. The structure contained a trigonal bipyramidal iron(II)-hydroxo with an outer-sphere triflate anion (Figure 5.7). The hydrogen atoms of the hydroxo and amine moieties were located in the difference map and independently refined. Hydrogen-bonding interactions between the amines of the ligand, the hydroxo moiety, and the outer-sphere triflate were observed. The O1–O2 distance was 2.950(4) Å and the O1–H1–O2 angle was 170(4)°. The N5–O1 and N6–O1 distances were 2.645(4) Å and 2.734(4) Å, respectively, and the N5–H5–O1 and N6–H6–O1 angles were 176(4)° and 177(4)°, respectively. The Fe1–O1 bond was 1.978(2) Å, which was in the range of reported iron(II)- hydroxo complexes(1.830(8) – 2.051(3) Å).12 Moreover, the bond was elongated compared to that Py of Fe-OH2.

N6 H1B N6 N5 H1 H6 O1 N5 H1A O1 H5

Fe1 Fe1

Py Py Fe-OH2 Fe-OH

Py Py Figure 5.7 Molecular structures of Fe-OH2 and Fe-OH drawn with 50% probability ellipsoids. Select hydrogen atoms and outer-sphere anions have been omitted for clarity.

103 Py Py Table 5.2 Selected structural parameters of Fe-OH2 and Fe-OH. Py Py Fe-OH2 Fe-OH Fe–O (Å) 2.0587(11) 1.978(2)

Fe–Naxial (Å) 2.2911(13) 2.278(3) 2.0563(13) 2.078(3) Fe–N (Å) 2.0564(13) 2.076(3)

Fe–NPy (Å) 2.1274(13) 2.143(3) C=N stretch (cm-1) 1654

Cy Py When compared to the N(pi )3 ligand system, the Fe–O bond length of Fe-OH fell Cy Cy II between that of the two characterized iron(II)-hydroxo complexes: N(pi )(afa )2FeOH (Fe - 7 Cy II OH; 2.0339(12) Å) and [N(afa )3FeOH]OTf ([Fe -OH]OTf; 1.857(4) Å). The Fe–Nafa bond distances of PyFe-OH were also within the range of those reported for FeII-OH and [FeII-OH]OTf: 2.076(3) – 2.078(3) Å,7 2.1264(14) – 2.1020(14) Å, and 2.040(5) – 2.069(4) Å, respectively. The C=N stretch of PyFe-OH at 1651 cm-1 was indicative of neutral coordination of the azafulvene- amine tautomer of the ligand to the iron(II) center.7 This value is within error of the azafulvene- amine C=N stretches of FeII-OH (1655 cm-1)7 and [FeII-OH]OTf (1649 cm-1). Furthermore, the 1H NMR spectrum of PyFe-OH closely resembled that of [FeII-OH]OTf, particularly the three resonances located between 80 and 100 ppm (Figure 5.8). Based on the structural similarities Py II II Py Cy between Fe-OH, Fe -OH, and [Fe -OH]OTf we believed the N(pi )2 ligand appropriately Cy modeled the coordination environment of the N(pi )3 ligand with the removal of one hydrogen- bond donor or acceptor in the secondary sphere. Py Fe-OTf was furnished from the addition of Fe(OTf)2(MeCN)2 to a tetrahydrofuran solution of dry ligand. Crystals suitable for X-ray diffraction were unable to be grown of the pale yellow product, precluding detailed structural comparisons between PyFe-OTf and the analogous Cy Cy II species of the N(pi )3 system: N(afa )3Fe(OTf)2 (Fe -OTf). However, as was the case with PyCo-OTf, elemental analysis suggested the formulation of the iron(II)-triflate product contained a tetrahydrofuran solvate. The binding of both arms of the ligand in the azafulvene-amine

104 tautomeric form was confirmed in the solid-state IR spectrum, where the C=N stretch was observed at 1654 cm-1.7

Py Fe-OH

II [Fe -OH]OTf

1 Py II Figure 5.8 H NMR spectra of Fe-OH and [Fe -OH]OTf (CD3CN, 21 °C).

5.5 Oxidation of the iron(II) complexes We were eager to examine the oxidative reactivity of the iron(II) complexes, particularly PyFe-OH and PyFe-OTf, in order to compare their reactivity to that of FeII-OTf. The addition of nitrite to a solution of PyFe-OTf resulted in an immediate color change from yellow to brown. The 1H NMR spectrum of the reaction revealed broad paramagnetic resonances between 30 and 0 ppm. Solid-state structural characterization of the product, crystallized by the vapor diffusion of diethyl ether into a saturated solution of tetrahydrofuran, revealed a homodimeric species consisting of oxo-bridged iron(III)-hydroxo centers rotated 90° from one another, [µ- Py Cy Py O( N(afa )2FeOH)2](OTf)2 (µ-O[ Fe-OH]2; Figure 5.9). Even though the hydrogen atoms of the secondary sphere were not located within the difference map, their calculated positions placed them within the range of hydrogen-bonding interactions.8 The distances between the amine nitrogen atoms of the ligand platform and the hydroxo oxygen atoms ranged from 2.677(8) Å to 2.937(12) Å, and the N–H–O angles ranged from 166.6° to 176.4°. A weak hydrogen-bonding interaction was observed between one hydroxo moiety of one iron center and the azafulvene nitrogen of the ligand bound to the second iron center. The O–N distance was 3.459(5) Å and the O–H–N angle was 154.6°. The Fe–O distances of the iron(III)-hydroxos were 1.953(3) Å and 1.933(3) Å, slightly contracted from that of PyFe-OH. The Fe–O distances of the bridging oxo

105 were 1.784(3) Å and 1.782(3) Å, similar to the Fe–O distance of the terminal iron(III)-oxo species Cy III 13 [N(afa )3FeO]OTf (Fe -O), 1.8079(9) Å.

N6 H6 N11 O1 H11 N5 H5 H1 Fe1 H2 O2 H12 O3 Fe2 N12

Py μ-O[ Fe-OH]2 Py Figure 5.9 Molecular structure of µ-O[ Fe-OH]2 drawn with 50% probability ellipsoids. Select hydrogen atoms, cyclohexyl groups, and outer-sphere anions have been omitted for clarity.

Py Table 5.3 Selected structural parameters of µ-O[ Fe-OH]2. Py µ-O[ Fe-OH]2

Fe–Ohydroxo (Å) 1.953(3) 1.933(3)

Fe–Ooxo (Å) 1.784(3) 1.782(3)

Fe–Naxial (Å) 2.220(3) 2.2221(4) 2.143(4) 2.138(3) Fe–N (Å) 2.136(3) 2.158(6)

Fe–NPy (Å) 2.225(3) 2.218(4) C=N stretch (cm-1) 1651 O–H stretch (cm-1) 3554

The solid-state IR spectrum of the product contained an O–H stretch at 3554 cm-1 and a C=N stretch at 1651 cm-1, confirming the formulation of the product as containing a hydroxo moiety and possessing solely the azafulvene-amine tautomer of the ligand.7 We proposed that adventitious water provided the additional oxygen atom needed for the formation of the oxo-

106 bridged product. Moreover, the propensity of PyFe-OTf to form the dimeric iron(III) product represents a significant limitation to this system. The decreased steric profile of the pyridyl moiety compared to that of the azafulvene-amine arms of the ligand could have contributed to the synthesis of the dimer instead of a monomeric iron complex. The reduction of the number of Cy Py Cy hydrogen-bond donors from three in the N(pi )3 ligand to two in the N(pi )2 ligand may have also contributed to the inability to observe a monomeric iron(III) product. The reactivity of two-electron oxidants was also investigated. The addition of iodosobenzene (PhIO) to a solution of PyFe-OTf resulted in a color change from yellow to brown, similar to the addition of nitrite; however, the 1H NMR spectrum of the brown product differed Py from that of µ-O[ Fe-OH]2. No O–H stretch was observed in the solid-state IR spectrum of the product and the C=N stretch at 1649 cm-1 was consistent with azafulvene-amine coordination of the ligand.7 Crystallization of the product for solid-state structural analysis was unsuccessful. Py Instead, solutions of the product converted to µ-O[ Fe-OH]2, which readily crystallized from solution.

5.6 Conclusion We synthesized a new ligand system and its dative and anionic cobalt(II) and iron(II) Cy complexes. Comparison of the iron(II) complexes to those of the N(pi )3 system revealed high structural fidelity in the new system while reducing the number of hydrogen-bonding interactions possible in the secondary sphere from three to two. This reduction in hydrogen-bonding, however, meant a reduction in the steric profile of the ligand. The effects of this reduced steric profile were manifested in the formation of a homodimeric product with a bridging oxo upon oxidation of the iron(II)-triflate species. Thus, it was unclear whether altering the secondary sphere played a major role in the reactivity observed, or if the reactivity was due to the change in sterics.

5.7 Experimental General considerations. To avoid contact with oxygen and water, all air- and moisture-sensitive manipulations were carried out under an atmosphere of nitrogen in an MBraun inert atmosphere drybox or using standard Schlenk techniques. Solvents for air- and moisture-sensitive manipulations were dried and deoxygenated using a Glass Contour System and stored over 4 Å molecular sieves prior to use. Celite 545 was heated to 150 °C under dynamic vacuum for 24 h

107 prior to use in the drybox. All reagents were purchased from commercial sources and used as received unless otherwise noted. Ferrous trifluoromethanesulfonate14 and cobaltous trifluoromethanesulfonate15 were synthesized according to literature procedure. NMR solvents

(acetonitrile-d3, tetrahydrofuran-d8, and chloroform-d1) were degassed and stored over 4 Å molecular sieves prior to use. NMR spectra were recorded at ambient temperature on a Varian spectrometer operating at 500 MHz (1H NMR), 126 MHz (13C NMR), and 470 MHz (19F NMR) and referenced to the peak of the residual solvent (δ parts per million and J in Hz). Solid-state infrared spectra were measured using a PerkinElmer Frontier FT-IR spectrophotometer equipped with a KRS5 thallium bromide/iodide universal attenuated total reflectance accessory. Elemental analyses were performed by the University of Illinois at Urbana-Champaign School of Chemical Sciences Microanalysis Laboratory in Urbana, IL. Solid-state structural characterization was done using a Bruker D8 Venture Duo or Bruker X8ApexII diffractometer at the George L. Clark X-Ray Facility and 3M Material Laboratory at the University of Illinois at Urbana-Champaign.

Preparation of dipyrrolylpicolylamine (dppa). In a 500 mL round-bottom flask under an N2 atmosphere, pyrrole-2-carboxaldehyde (4.00 g, 0.0526 mol) and picolylamine (2.84 g, 0.0263 mol) were combined in 50 mL 1,2-dichloroethane and cooled to 0 °C. Following dissolution of the pyrrole-2-carboxaldehyde, sodium triacetoxyborohydride (17.83 g, 0.0841 mol) was added slowly to the reaction mixture and stirring was continued for 1.5 h. The resulting bright orange solution was quenched with aqueous sodium bicarbonate and vigorously stirred until effervescence subsided. The product was extracted into diethyl ether, dried over Na2SO4, and evaporated to dryness. The pale orange solid was dissolved in dichloromethane and hexanes was added until a red oil formed. The colorless supernatant was decanted and dried in vacuo, yielding the product as 1 a beige solid (4.30 g, 0.0162 mol, 62% yield). H NMR (500 MHz, CDCl3, 25°C) δ = 9.50 (br s, 2H, NH), 8.62 (ddd, J = 4.9, 1.7, 0.9 Hz, 1H, Py-CH), 7.71 (td, J = 7.7, 1.8 Hz, 1H, Py-CH), 7.34 (dt, J = 7.8, 1.1 Hz, 1H, Py-CH), 7.24 (ddd, J = 7.5, 5.0, 1.1 Hz, 1H, Py-CH), 6.81 (ddd, J = 2.6, 2.6, 1.5 Hz, 2H, Pyrr-CH), 6.16 (app q, J = 2.8 Hz, 2H, Pyrr-CH), 6.06 (app br s, 2H, Pyrr-CH), 13 3.65 (s, 2H, Py-CH2-N), 3.60 (s, 4H, Pyrr-CH2-N). C NMR (126 MHz, CDCl3, 25°C) δ = 158.96 (Py-C), 148.76 (Py-C), 136.98 (Py-C), 128.32 (Pyrr-C), 124.85 (Py-C), 122.46 (Py-C), 117.50

108 (Pyrr-C), 108.00 (Pyrr-C), 107.94 (Pyrr-C), 57.47 (Pyrr-CH2-N), 49.74 (Pyrr-CH2-N). IR = 3292 -1 + cm (N-H). HRMS (ESI-TOF) m/z: [M + H] Calcd. for C16H19N4 267.1610; Found 267.1606.

1 13 Figure 5.10 H (top) and C (bottom) NMR spectra of dppa (CDCl3, 21 °C).

Preparation of dppaCO. A 500 mL three-neck round-bottom flask was charged with dimethylformamide (3.5 mL, 0.0451 mol) and cooled to 0 °C in an ice bath under N2. POCl3 (3.5 mL, 0.0376 mol) was diluted in 15 mL 1,2-dichloroethane and added dropwise to the dimethylformamide via an addition funnel over 30 min with vigorous stirring. The solution was removed from the ice bath and stirred at room temperature for 15 min. Once the reaction mixture was cooled back to 0 °C, the addition funnel was washed with 3 mL 1,2-dichloroethane. Dipyrrolylpicolylamine (4.00 g, 0.0150 mol) was dissolved in 20 mL 1,2-dichloroethane and 3 mL dimethylformamide and then added dropwise over 30 min to the reaction mixture. The resulting dark red solution was stirred at room temperature for 1 h. A saturated aqueous solution of NaOAc

109 (15.40 g, 0.188 mol) was added and the cloudy red-brown solution was heated to 45 °C for 30 min before neutralization by slow addition of Na2CO3. The product was extracted into dichloromethane

(3 x 25 mL) and dried over Na2SO4. Removal of volatiles under reduced pressure yielded a red oil, which was treated with acetonitrile and diethyl ether (1:2) to precipitate the product dppaCO (1.58 1 g, 0.00490 mol, 32.7%). H NMR (500 MHz, CDCl3, 25°C) δ = 11.54 (br s, 2H, NH), 9.50 (s, 2H, O=CH), 8.54 (d, J = 4.7 Hz, 1H, Py-CH), 7.43 (td, J = 7.5, 1.5 Hz, 1H, Py-CH), 7.33 (d, J = 7.8 Hz, 1H, Py-CH), 7.10 (ddd, J = 7.5, 5.0, 1.2 Hz, 1H, Py-CH), 6.94 (dd, J = 3.6, 2.0 Hz, 2H, Pyrr-

CH), 6.19 (dd, J = 3.6, 2.2 Hz, 2H, Pyrr-CH), 3.73 (s, 4H, Pyrr-CH2-N), 3.69 (s, 2H, Py-CH2-N). 13 C NMR (126 MHz, CDCl3, 25°C) δ = 179.35 (O=CH), 158.17 (Py-C), 149.10 (Py-C), 139.08 (Pyrr-C), 136.75 (Py-C), 133.16 (Pyrr-C), 123.59 (Pyrr-C), 122.73 (Py-C), 122.50 (Py-C), 111.13 -1 -1 (Pyrr-C), 59.45 (Py-CH2-N), 51.12 (Pyrr-CH2-N). IR = 3314 cm (N-H); 1655 cm (C=O). HRMS + (ESI-TOF) m/z: [M + H] Calcd. for C18H19N4O2 323.1508; Found 323.1507.

1 13 CO Figure 5.11 H (top) and C (bottom) NMR spectra of dppa (CDCl3, 21 °C).

110 Py Cy CO Preparation of N(pi )2. dppa (2.00 g, 0.00621 mol) was added to 20 mL acetonitrile in a 250 mL round-bottom flask. Cyclohexylamine (1.54 g, 0.0155 mol) was added and stirring was continued for 8 h. The product precipitated from solution as a beige solid and was isolated by filtration. The filtrate was cooled to -10 °C for 3 h, after which more product had crystallized from the solution. The isolated material was dissolved in 10 mL diethyl ether and dried over 4 Å molecular sieves. Following filtration and removal of volatiles under reduced pressure, the product 1 was obtained as a white powder. H NMR (500 MHz, CDCl3, 25°C) δ = 9.96 (br s, 2H, NH), 8.63 (d, J = 4.8 Hz, 1H, Py-CH), 8.04 (s, 2H, N=CH), 7.66 (t, J = 7.7 Hz, 1H, Py-CH), 7.34 (d, J = 7.8 Hz, 1H, Py-CH), 7.19 (dd, J = 7.5, 4.9 Hz, 1H, Py-CH), 6.36 (d, J = 3.4 Hz, 2H, Pyrr-CH), 6.07

(d, J = 3.5 Hz, 2H, Pyrr-CH), 6.00 (s, 1H), 3.69 (s, 2H, Py-CH2-N), 3.61 (s, 4H, Pyrr-CH2-N), 3.09 13 (m, 2H, Cy-CH), 1.24-1.84 (m, 20H, Cy-CH2). C NMR (126 MHz, CDCl3, 25°C) δ = 158.74 (Py-C), 149.10 (Py-C), 149.00 (N=CH), 136.76 (Py-C), 132.21(Pyrr-C), 130.65 (Pyrr-C), 124.13

(Py-C), 122.29 (Py-C), 113.41 (Pyrr-C), 109.67 (Pyrr-C), 69.23 (Cy-C), 58.73 (Py-CH2-N), 50.40 -1 -1 (Pyrr-CH2-N), 34.88 (Cy-C), 25.89 (Cy-C), 25.02 (Cy-C). IR = 3284 cm (N-H); 1630 cm + (C=N). HRMS (ESI-TOF) m/z: [M + H] Calcd. for C30H41N6 485.3393; Found 485.3398.

111 1 13 Py Cy Figure 5.12 H (top) and C (bottom) NMR spectra of N(pi )2 (CDCl3, 21 °C).

Py Cy Py Cy Preparation of N(pi )2Co(OH2). A 20 mL scintillation vial was charged with N(pi )2 (0.0500 g, 0.103 mmol), KH (0.0087 g, 0.217 mmol), and 5 mL tetrahydrofuran. The solution was vigorously stirred until effervescence subsided, approximately 2 h. It was then filtered over a pad of Celite into a vial containing CoCl2 (0.0134 g, 0.103 mmol) and 2 mL acetonitrile. Stirring was continued for 16 h. The red-purple solution was dried in vacuo. Diethyl ether was added to the resulting residue and the red solution of the product was filtered over a pad of Celite and dried (0.0216 g, 0.0386 mmol, 37%). Crystals suitable for X-ray anaylsis were grown from a saturated solution of acetonitrile.

112 1 Py Figure 5.13 H NMR spectrum of Co-OH2 (THF-d8, 21 °C).

Py Cy Py Cy Preparation of [ N(afa )2CoOH]OTf. A 20 mL scintillation vial was charged with N(pi )2

(0.0500 g, 0.103 mmol), Co(OTf)2(MeCN)2 (0.0453 g, 0.103 mmol), Li2O (0.0062 g, 0.206 mmol), and 5 mL tetrahydrofuran. The solution was stirred 16 h, during which time the color changed from deep red to pale green. Volatiles were removed under reduced pressure and the resulting residue was triturated with diethyl ether and tetrahydrofuran. The product was recrystallized from a saturated solution of dichloromethane layered with hexanes (0.0373 g, 0.0526 mmol, 51% crystalline yield). Crystals suitable for X-ray diffraction were grown from the vapor diffusion of pentane into a concentrated tetrahydrofuran solution. IR = 3503 cm-1 (O–H), 1659 cm-1 (C=N).

1 Py Figure 5.14 H NMR spectrum of Co-OH (THF-d8, 21 °C).

Py Cy Preparation of N(afa )2Co(OTf)2. A 20 mL scintillation vial was charged with Py Cy Co(OTf)2(MeCN)2 (0.0226 g, 0.0516 mmol) and 6 mL tetrahydrofuran. N(pi )2 (0.0250 g, 0.0516 mmol) was weighed by difference and added to the tetrahydrofuran solution. The dark pink solution was stirred 1 h before volatiles were removed under reduced pressure. The product was recrystallized from a saturated solution of tetrahydrofuran layered with pentane. Analysis for

113 C32H40N6CoF6O6S2•C4H8O: Calcd. C, 47.31; H, 5.29; N, 9.20. Found C, 47.02; H, 5.21; N, 9.30%. IR = 3210 cm-1 (N–H); 1643 cm-1 (C=N).

1 Py Figure 5.15 H NMR spectrum of Co-OTf (THF-d8, 21 °C).

Py Cy Py Cy Preparation of N(pi )2Fe(OH2). A 20 mL scintillation vial was charged with N(pi )2 (0.0500 g, 0.103 mmol), KH (0.0087 g, 0.217 mmol), and 5 mL tetrahydrofuran. The solution was vigorously stirred until effervescence subsided, approximately 2 h. It was then filtered over a pad of Celite into a vial containing FeCl2 (0.0131 g, 0.103 mmol) and 2 mL acetonitrile. Stirring was continued for 16 h. The orange solution was dried in vacuo. Diethyl ether was added to the resulting residue and the red solution of the product was filtered over a pad of Celite and dried. Crystals suitable for X-ray anaylsis were grown from a saturated solution of diethyl ether layered with hexane and stored at -35 °C.

1 Py Figure 5.16 H NMR spectrum of Fe-OH2 (THF-d8, 21 °C).

114 Py Cy Py Cy Preparation of [ N(afa )2FeOH]OTf. A 20 mL scintillation vial was charged with N(pi )2

(0.0500 g, 0.103 mmol), Fe(OTf)2(MeCN)2 (0.0450 g, 0.103 mmol), Li2O (0.0062 g, 0.206 mmol), and 5 mL tetrahydrofuran. The solution was stirred 16 h, during which time the solution became bright orange. Volatiles were removed under reduced pressure and the resulting residue was triturated with diethyl ether and tetrahydrofuran. The product was recrystallized from a saturated solution of dichloromethane layered with hexanes. Crystals suitable for X-ray diffraction were grown from a concentrated solution of tetrahydrofuran layered with pentane. Analysis for

C31H41N6FeF3O4S: Calcd. C, 52.62; H, 5.98; N, 11.88. Found C,52.56; H, 5.72; N, 11.77%. IR = 3506 cm-1 (O–H), 1655 cm-1 (C=N).

1 Py Figure 5.17 H NMR spectrum of Fe-OH (CD3CN, 21 °C).

Py Cy Preparation of N(afa )2Fe(OTf)2. A 20 mL scintillation vial was charged with Py Cy Fe(OTf)2(MeCN)2 (0.0225 g, 0.0516 mmol) and 6 mL tetrahydrofuran. N(pi )2 (0.0250 g, 0.0516 mmol) was weighed by difference and added to the tetrahydrofuran solution. The yellow solution was stirred 1 h before volatiles were removed under reduced pressure. The product was recrystallized from a saturated solution of tetrahydrofuran layered with hexamethyldisiloxane

(0.0402 g, 0.0480 mmol, 93%). Analysis for C32H40N6FeF6O6S2•C4H8O: Calcd. C, 47.47; H, 5.31; N, 9.23. Found C, 46.95; H, 5.14; N, 9.44%. IR = 1638 (C=N).

115 1 Py Figure 5.18 H NMR spectrum of Fe-OTf (THF-d8, 21 °C).

Py Cy Preparation of [µ-O( N(afa )2FeOH)2](OTf)2. A 20 mL scintillation vial was charged with Py Cy N(afa )2Fe(OTf)2 (0.0200 g, 0.0239 mmol) dissolved in 4 mL tetrahydrofuran. [NBu4][NO2] (0.0069 g, 0.0239 mmol) was weighed by difference and added to the stirring solution, which turned red-brown in color. The solution was stirred 2 h and volatiles were removed under reduced pressure. Crystals suitable for X-ray diffraction were grown from vapor diffusion of diethyl ether into a concentrated tetrahydrofuran solution. IR = 3554 cm-1 (O–H); 1651 cm-1 (C=N).

1 Py Figure 5.19 H NMR spectrum of µ-O[ Fe-OH]2 (CD3CN, 21 °C).

Py Cy Oxidation of N(afa )2Fe(OTf)2 by PhIO. A 20 mL scintillation vial was charged with Py Cy N(afa )2Fe(OTf)2 (0.0200 g, 0.0239 mmol) dissolved in 5 mL acetonitrile. Iodosobenzene (0.0052 g, 0.0239 mmol) was weighed by difference and added as a solid to the solution, resulting in an immediate color change to brown. The solution was stirred 2 h, after which time volatiles

116 were removed under reduced pressure and the brown solid obtained was triturated with benzene. IR = 1649 cm-1 (C=N).

1 Py Figure 5.20 H NMR spectrum of the reaction Fe-OTf and PhIO (CD3CN, 21 °C).

117 Py Py Table 5.4 Crystallographic parameters of Co-OH2 and Co-OH. Py Cy Py Cy N(pi )2Co(OH2) [ N(afa )2CoOH]OTf

Empirical Formula C30H40CoN6O C31H41CoF3N6O4S Formula Weight 559.61 709.69 Temperature 100 K 100 K Wavelength 0.71073 0.71073 Crystal System Monoclinic Monoclinic Space Group P 2(1)/n P 2(1)/n a = 13.1851(8) Å a = 11.7088(3) Å b = 16.5369(9) Å b = 17.9610(4) Å c = 13.5589(9) Å c = 15.8615(3) Å Unit Cell Dimensions α = 90° α = 90° β = 103.516(2)° β = 100.0523(11)° γ = 90° γ = 90° Volume 2874.5(3) 3284.49(13) Z 4 4 Reflections Collected 89972 106252 Independent Reflections 7155 8196 Goodness-of-Fit on F2 1.076 1.080 R1 = 0.0282 R1 = 0.0465 Final R Indices [I>2sigma(I)] wR2 = 0.0750 wR2 = 0.1218 R1 = 0.0331 R1 = 0.0536 R Indices wR2 = 0.0772 wR2 = 0.1267

118 Py Py Table 5.5 Crystallographic parameters of Fe-OH2 and Fe-OH. Py Cy Py Cy N(pi )2Fe(OH2) [ N(afa )2FeOH]OTf

Empirical Formula C30H40FeN6O C31H41F3FeN6O4S Formula Weight 556.53 706.61 Temperature 100 K 100 K Wavelength 0.71073 0.71073 Crystal System Monoclinic Monoclinic Space Group P 2(1)/n P 2(1)/n a = 13.2301(4) Å a = 11.768(3) Å b = 16.5659(6) Å b = 17.945(5) Å c = 13.5329(5) Å c = 16.019(4) Å Unit Cell Dimensions α = 90° α = 90° β = 103.8830(10)° β = 100.322(9)° γ = 90° γ = 90° Volume 2879.34(17) 3328.3(15) Z 4 4 Reflections Collected 104979 35682 Independent Reflections 7165 6116 Goodness-of-Fit on F2 1.144 1.014 R1 = 0.0373 R1 = 0.0508 Final R Indices [I>2sigma(I)] wR2 = 0.0905 wR2 = 0.1181 R1 = 0.0480 R1 = 0.0881 R Indices wR2 = 0.0955 wR2 = 0.1401

119 Py Table 5.6 Crystallographic parameters of µ-O[ Fe-OH]2. Py Cy [µ-O( N(afa )2Fe(OH))2](OTf)2

Empirical Formula C62H82F6Fe2N12O9S2 Formula Weight 1429.21 Temperature 100 K Wavelength 0.71073 Crystal System Triclinic Space Group P-1 a = 13.2447(7) Å Unit Cell Dimensions b = 16.0732(8) Å

c = 18.8893(9) Å

α = 76.550(2)°

β = 71.234(2)°

γ = 67.038(2)° Volume 3479.2(3) Z 2 Reflections Collected 64032 Independent Reflections 12809 Goodness-of-Fit on F2 1.068 R1 = 0.0760 Final R Indices [I>2sigma(I)] wR2 = 0.2115 R1 = 0.0853 R Indices wR2 = 0.2205

5.8 References (1) Shook, R. L.; Borovik, A. S. Role of the Secondary Coordination Sphere in Metal- Mediated Dioxygen Activation. Inorg. Chem. 2010, 49, 3646-3660.

120 (2) Borovik, A. S. Bioinspired Hydrogen Bond Motifs in Ligand Design: The Role of Noncovalent Interactions in Metal Ion Mediated Activation of Dioxygen. Acc. Chem. Res. 2005, 38, 54-61.

(3) MacBeth, C. A.; Golombek, A. P.; Young, V. G.; Yang, C.; Kuczera, K.; Hendrich, M. P.; Borovik, A. S. O2 Activation by Nonheme Iron Complexes: A Monomeric Fe(III)-Oxo Complex Derived from O2. Science 2000, 289, 938-941.

(4) Lucas, R. L.; Zart, M. K.; Murkerjee, J.; Sorrell, T. N.; Powell, D. R.; Borovik, A. S. A Modular Approach toward Regulating the Secondary Coordination Sphere of Metal Ions: Differential Dioxygen Activation Assisted by Intramolecular Hydrogen Bonds. J. Am. Chem. Soc. 2006, 128, 15476-15489.

(5) Jones, J. R.; Ziller, J. W.; Borovik, A. S. Modulating the Primary and Secondary Coordination Spheres within a Series of CoII–OH Complexes. Inorg. Chem. 2017, 56, 1112-1120.

(6) Gordon, Z.; Drummond, M. J.; Matson, E. M.; Bogard, J. A.; Schelter, E. J.; Lord, R. L.; Fout, A. R. Tuning the Fe(II/III) Redox Potential in Nonheme Fe(II)-Hydroxo Complexes through Primary and Secondary Coordination Sphere Modifications. Inorg. Chem. 2017, 56, 4852-4863

(7) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of Iron(II) Aqua and Hydroxyl Complexes Featuring a Tripodal H-Bond Donor and Acceptor Ligand. Inorg Chem 2014, 53, 4450–4458.

(8) Steiner, T. The Hydrogen-Bond in the Solid State. Angew. Chem. Int. Ed. 2002, 41, 48-76.

(9) (a) Cruz, D. M. P.; Woodruff, D. N.; Jeon, I.-R.; Bhowmick, I.; Secu, M.; Hillard, E. A.; Dechambenoit, P.; Clérac, R. Switching Off the Single-Molecule Magnet Properties of the II 2+ III − [Co (Me6tren)(OH2)] Module by Complexation with Trans- [Ru (Salen)(CN)2] . New J. Chem. 2014, 38, 3443–3448. (b) Meyer, F.; Heinze, K.; Nuber, B.; Zsolnai, L. Tuning the Metal–Metal Separation in Pyrazolate-Based Dinuclear Complexes by the Length of Chelating Side Arms. J. Chem. Soc., Dalton Trans. 1998, 2, 207–214. (c) Reiter, M.; Altenbuchner, P. T.; Kissling, S.; Herdtweck, E.; Rieger, B. Amine- Bis(Phenolato)Cobalt(II) Catalysts for the Formation of Organic Carbonates From Carbon Dioxide and Epoxides. European Journal of Inorganic Chemistry 2015, 2015, 1766–1774. (d) Baker, A. T.; Ferguson, N. J.; Goodwin, H. A. Five-Coordination in Complexes of 2- (3,5-Dimethylpyrazol-1-yl)Quinoline, a Sterically Hindered Diimine System. Aust. J. Chem. 1984, 37, 2421–2429.

(10) (a) MacBeth, C. E.; Hammes, B. S.; Young, V. G.; Borovik, A. S. Hydrogen-Bonding Cavities About Metal Ions: Synthesis, Structure, and Physical Properties for a Series of Monomeric M−OH Complexes Derived From Water. Inorg. Chem. 2001, 40, 4733–4741. (b) Hammes, B. S.; Young, V. G; Borovik, A. S. Hydrogen-Bonding Cavities About Metal Ions: a Redox Pair of Coordinatively Unsaturated Paramagnetic Co-OH Complexes.

121 Angew. Chem. Int. Ed. 1999, 38, 666-669. (c) Zinn, P. J.; Sorrell, T. N.; Powell, D. R.; Day, V. W.; Borovik, A. S. Acetonitrile Hydration and Ethyl Acetate Hydrolysis by Pyrazolate-Bridged Cobalt(II) Dimers Containing Hydrogen-Bond Donors; American Chemical Society, 2007; Vol. 46, pp. 10120–10132. (d) Ng, G. K. Y.; Ziller, J. W.; Borovik, A. S. Preparation and Structures of Dinuclear Complexes Containing MII–OH Centers. Chem. Commun. 2012, 48, 2546–2548.

(11) (a) Chen, K.; Que, L. J. Am. Chem. Soc. 2001, 123, 6327−6337. (b) Barkigia, K. M.; Palacio, M.; Sun, Y.; Nogues, M.; Renner, M. W.; Varret, F.; Battioni, P.; Mansuy, D.; Fajer, J. Inorg. Chem. 2002, 41, 5647−5649. (c) Song, D.; Lim, J. M.; Cho, S.; Park, S.-J.; Cho, J.; Kang, D.; Rhee, S. G.; You, Y.; Nam, W. Chem. Commun. 2012, 48, 5449−5451. (d) Kodera, M.; Itoh, M.; Kano, K.; Funabiki, T.; Reglier, M. Angew. Chem., Int. Ed. 2005, 44, 7104−7106.

(12) (a)Kendall, A. J.; Zakharov, L. N.; Gilbertson, J. D. Synthesis and Stabilization of a Monomeric Iron(II) Hydroxo Complex via Intramolecular Hydrogen Bonding in the Secondary Coordination Sphere. Inorg. Chem. 2010, 49, 8656−8658. (b) Soo, H. S.; Komor, A. C.; Iavarone, A. T.; Chang, C. J. A Hydrogen-Bond Facilitated Cycle for Oxygen Reduction by an Acid- and Base- Compatible Iron Platform. Inorg. Chem. 2009, 48, 10024-10035. (c) Hikichi, S.; Ogihara, T.; Fujisawa, K.; Kitajima, N.; Akita, M.; Moro- oka, Y. Synthesis and Characterization of the Benzoylformato Ferrous Complexes with the Hindered Tris(pyrazolyl)borate Ligand as a Structural Model for Mononuclear Non-Heme Iron Enzymes. Inorg. Chem. 1997, 36, 4539-4547. (d) Beńsivy,L.; Halut,S.; Donnadieu,B.; Tuchagues,J.-P.; Chottard, J.-C.; Li, Y. Monomeric Iron(II) Hydroxo and Iron(III) Dihydroxo Complexes Stabilized by Intermolecular Hydrogen Bonding. Inorg. Chem. 2006, 45, 2403−2405. (e) MacBeth, C. E.; Hammes, B. S.; Young, V. G.; Borovik, A. S. Hydrogen-Bonding Cavities about Metal Ions: Synthesis, Structure, and Physical Properties for a Series of Monomeric M-OH Complexes Derived from Water. Inorg. Chem. 2001, 40, 4733−4741.

(13) Matson, E. M.; Park, Y. J.; Fout, A. R. Facile Nitrite Reduction in a Non-Heme Iron System: Formation of an Iron(III)-Oxo. J. Am. Chem. Soc. 2014, 136, 17398-17401.

(14) Hagen, K. S. Iron(II) Triflate Salts as Convenient Substitutes for Perchlorate Salts: Crystal Structures of [Fe(H2O)6](CF3SO3)2 and Fe(MeCN)4(CF3SO3)2. Inorg. Chem. 2000, 39, 5867-5869.

(15) Seitz, M.; Kaiser, A.; Powell, D. R.; Borovik, A. S.; Reiser, O. Predetermined Helical

Chirality in Octahedral Complexes with a Novel Pentadentate C2-Symmetrical Chiral Bis(oxazoline) Ligand. Adv. Synth. Catal. 2004, 346, 737-741.

122