Incorporation of cyanide ligands into models of [FeFe]- and [NiFe]-hydrogenase

Manor, Brian

Permalink

https://hdl.handle.net/2142/72937 Copy

Description

Title

Incorporation of cyanide ligands into models of [FeFe]- and [NiFe]-hydrogenase

Author(s)

Manor, Brian

Issue Date

2015-01-21

Director of Research (if dissertation) or Advisor (if thesis)

Rauchfuss, Thomas B.

Doctoral Committee Chair(s)

Rauchfuss, Thomas B.

Committee Member(s)

• Lu, Yi
• Suslick, Kenneth S.
• Girolami, Gregory S.

Department of Study

Chemistry

Discipline

Chemistry

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Hydrogenase
• Hydrogen
• Nickel
• Iron
• Cyanide
• Borane
• Hydrogen Oxidation

Abstract

Hydrogen is processed in Nature by a class of metalloenzymes called hydrogenases. Hydrogenase enzymes catalyze the reduction of protons into H2 or oxidized H2 into protons. The catalytic rates and overpotentials of hydrogenase enzymes are comparable to Pt catalysts. The most prevalent hydrogenase enzymes are bimetallic with FeFe or NiFe active sites. The metals of the active sites are bound to carbonyl and cyanide ligands with the metal centers bridged by thiolates. The use of inexpensive and readily available elements by hydrogenase enzymes has been a driving force for the development of compounds that model both the structure and activity of the enzymes. Though functional models of hydrogenase active sites have been developed, many occur at lower rates and higher overpotentials than the enzymes. The catalytic cycles of these models often differ from the catalytic cycles of hydrogenase. The work presented focuses on the development towards functional hydrogenase models with a biologically accurate coordination spheres particularly the presence of cyanide on the Fe centers. The presence of a biomimetic coordination sphere may develop models that function by more biomimetic catalytic cycles. Chapter 1 highlights the current understanding of the structural features and catalytic cycles of hydrogenase enzymes. An overview of hydrogenase models is presented with a particular focus the structural features of the models that impact production and oxidation of H2. Chapter 2 focuses on modeling the hydrogen-bonding environments found in [FeFe]-H2ase by coordination of triarylboranes to the [FeFe]-H2ase models [Fe2(pdt)(CO)4(CN)2]2‒ and [Fe2(adt)(CO)4(CN)2]2‒. Treatment of two equivalents of borane to [Fe2(pdt)(CO)4(CN)2]2‒ formed the 2:1 adduct [Fe2(pdt)(CO)4(CNBR3)2]2‒. Attempts to model the inequivalent hydrogen bonding environment of [FeFe]-H2ase with BPh3 and B(C6F5)3 were unsuccessful. The use of the bulky borane B(C6F4-o-C6F5)3 (BArF*3) with K2[Fe2(pdt)(CO)4(CN)2]2‒ were found to form the 1:1 adduct K2[Fe2(pdt)(CO)4(CN)(CNBArF*3]2‒. The 2:1 borane adducts [Fe2(pdt)(CO)4(CNBR3)2]2‒ could be readily protonated to form the stable bridging hydrides [Fe2(μ-H)(pdt)(CO)4(CNBR3)2]2‒. The adducts [Fe2(pdt)(CO)4(CNBR3)2]2‒ were found to undergo quasi-reversible oxidation around ‒0.3 V vs Fc0/+ while the hydrides undergo quasi-reversible reductions around ‒1.7 V vs Fc0/+. The coordination of the boranes were found to have a large effect on the electron density of the Fe centers as evidence by higher νCO, more acidic pKa’s, and more negative oxidation and reduction potentials compared to [Fe2(pdt)(CO)4(CN)2]2‒. Comparison of certain parameters indicate that PMe3 is a more basic ligand than CNBR3‒ (νCO, pKa) though the presence of CNBR3‒ ligands resulted in more negative oxidation and reduction potentials than PMe3. A simplified synthesis of [Fe2(pdt)(CO)4(CN)2]2‒, [Fe2(adt)(CO)4(CN)2]2‒, and [Fe2(pdt)(CO)5(CN)]‒ was developed entailing the reaction of the diiron hexacarbonyl complexes with KCN in MeCN. With the success of coordination of boranes to [Fe2(pdt)(CO)4(CN)2]2‒, the coordination of boranes to the Fe-CN of the [NiFe]-H2ase model [HFe(CO)3(CN)2]‒ was explored in Chapter 3. The compound [HFe(CO)3(CN)2]‒ is a possible synthon for a one-pot synthesis of [NiFe]-H2ase models as the Fe center is coordinated by carbonyl, cyanide, and a hydride ligands. Reactions of [HFe(CO)3(CN)2]‒ with typical Ni reagents for [NiFe] H2ase models did not yield any hydride-containing products. In order to allow for facile ligand substitution of [HFe(CO)3(CN)2]‒ by displacement of CO, two equivalents of boranes were coordinated to [HFe(CO)3(CN)2]‒ to form the 2:1 adducts [HFe(CO)3(CNBR3)2]‒. The coordination of boranes was found to significantly decrease the pKa of the hydride by at least 13 pKa units. Treatment of [HFe(CO)3(CNBArF3)2]‒ with Ni(pdt)(dppe) resulted initially in deprotonation to form [Fe(CO)3(CNBArF3)2]2‒ before ligand redistribution forms the Fe-H [HFe(CO)(CNBArF3)2(dppe)]‒. The treatment of the less acidic [HFe(CO)3(CNBPh3)2]‒ with Ni(pdt)(dxpe) results initially in partial deprotonation of the hydride. The use of the diphosphine dppe forms a mix of mono Fe-H [HFe(CO)(CNBPh3)2(dppe)]‒ and the NiFe hydride (CO)2(CNBPh3)Fe(H)(pdt)Ni(dppe) while the diphosphine dcpe forms only (CO)2(CNBPh3)Fe(H)(pdt)Ni(dcpe) as the only hydride-containing product. An alternative route for the synthesis of functional cyanide-containing [NiFe]-H2ase models are examined in Chapter 4 by decarbonylation the [NiFe]-H2ase models (CO)2(CN)2Fe(pdt)Ni(dxpe). The compound (CO)2(CN)2Fe(pdt)Ni(dppe) was present as only the cis-CO, trans-CN isomer. Using the more basic diphosphine dcpe formed both the cis-CO, trans-CN isomer and cis-CO, cis-CN isomers of (CO)2(CN)2Fe(pdt)Ni(dcpe). Attempts to decarbonylate the dicarbonyl complexes by photolysis resulted in decomposition to insoluble precipitates. To facilitate decarbonylation, two equivalents of triarylboranes were added to (CO)2(CN)2Fe(pdt)Ni(dxpe) forming the 2:1 adducts (CO)2(CNBR3)2Fe(pdt)Ni(dxpe). Irradiation of the adducts resulted in isomerization instead of decarbonylation. Studies of the photoisomerization of the adducts found the isomerization to be slightly inhibited by CO while photolysis under 13CO was found to undergo CO exchange. Both of these results are consistent with the formation of an open coordination site which is thought to occur from CO dissociation. The formation of NiFe hydrides from hydrogen is highlighted in Chapter 5. Using the models described in Chapter 4, the treatment of the dicarbonyl adducts (CO)2(CNBArF3)2Fe(pdt)Ni(dxpe) with a decarbonylating agent followed by either BH4‒ or H2 formed the biomimetic hydrides [(CO)(CNBArF3)2Fe(H)Ni(dxpe)]‒. The crystal structure of [(CO)(CNBArF3)2Fe(H)(pdt)Ni(dppe)]‒ was found to have asymmetric metal-hydride distances (Fe-H: 1.51Å; Ni-H: 1.71Å). Although bridging hydrides typically have protic character, the hydrides [(CO)(CNBArF3)2Fe(H)Ni(dxpe)]‒ were hydric in nature. Treatment of [(CO)(CNBArF3)2Fe(H)Ni(dxpe)]‒ with stronger acids (eg. anilinium, HCl) was found to produce H2. Furthermore, the addition of weak acids (eg. HNMe3+, pyrollidinium) were found to form a dihydrogen bond with the hydride as indicated by 1H NMR experiments. Given their hydric nature, the hydrides [(CO)(CNBArF3)2Fe(H)Ni(dxpe)]‒ are best described as terminal Fe-H hydrides with a weak Ni-H interaction. The hydrides were oxidized at mild potentials (~0.1 V vs Fc0/+) though only a 20 mV difference in oxidation potential was observed between dppe and dcpe ligands, indicative of oxidation at Fe, reverse of [NiFe]-H2ase. The hydride [(CO)(CNBArF3)2Fe(H)Ni(dppe)]‒ was found to serve an electrocatalyst for the oxidation of H2, a first for [NiFe]-H2ase models.

Graduation Semester

2014-12

Permalink

http://hdl.handle.net/2142/72937

Copyright and License Information

Copyright 2014 Brian Manor

Owning Collections