Back to Table of Contents

47013-G3
Molecular Mechanism of Hydrogen-formation in Fe-only Hydrogenases

Nicolai Lehnert, University of Michigan

After starting as an assistant professor at the University of Michigan, I became aware of the strong program in energy-related research at this institution, fostered by the Michigan Memorial Phoenix Energy Institute (MMPEI). My interactions with other scientists through this institute got me interested in hydrogen-based approaches for alternative energies. The PRF-G funding that I received in 2007 ultimately allowed me to start research in this area by providing funds for a postdoctoral fellow, Dr. Grace Galinato. Hence, the PRF-G grant has laid the foundation for my engagement in energy-related research, and my research program in this area has since been growing. For example, I have obtained funding from the University of Michigan to start additional research in this area, and I am participating in multi-PI applications in energy research spearheaded by the MMPEI. Aside from the postdoc, I also had two students working on this project in internships. In conclusion, the PRF-G grant has had an important impact on my career as it allowed me to pursue a second research thrust in my laboratory besides my studies on Nitric Oxide Reductases.

The research performed in the first year was focused on the investigation of the spectroscopic properties and electronic structures of the terminal and bridging hydride isomers of [Fe(pdt)(dppv)₂(CO)₂(H)]⁺ in collaboration with Prof. Thomas Rauchfuss (University of Illinois at Urbana-Champagne). These complexes are models for the key protonated intermediate of the active site in Iron-only hydrogenases. Our central hypothesis is that only the terminal hydride isomer is catalytically active leading to production of H₂. We have studied both isomers of [Fe₂(pdt)(dppv)₂(CO)₂(H)]⁺ using (resonance) Raman and IR spectroscopy, and DFT calculations. These studies constitute the first vibrational investigations on protonated hydrogenase model complexes, and hence, our results provide the necessary groundwork to understand the spectroscopic and electronic-structural differences between the two isomers in detail.

AIM #1: Spectroscopic Investigation of [Fe₂(pdt)(dppv)₂(CO)₂(H)]⁺

Resonance Raman spectra of the solid and solution forms of the terminal (abbreviation: H-term) and bridging (μ-H) hydride complexes [Fe₂(pdt)(dppv)₂(CO)₂(H)]⁺ and corresponding deuterated compounds were obtained by laser excitation at multiple wavelengths between 457 and 647 nm. The optimum conditions for Raman investigations required saturated CH₂Cl₂ solutions and 568 nm excitation. Distinct differences in the low-energy region of the H-term and μ-H spectra, particularly in the ν(Fe-CO) stretching and δ(Fe-C-O) bending region (440 - 520 cm^-1), are observed. Additional ¹³CO labeling experiments will follow in future studies to assist in spectral assignments. Since the CO ligands serve as probes for the ‘electron-richness’ of the metal centers, these changes indicate important differences in the electronic structures of H-term and μ-H. These data are currently analyzed using our QCC-NCA procedure. DFT calculations predict the ν(Fe-CO) stretches of H-term between 400 and 490 cm^‑1, whereas μ-H shows these modes at 527 cm^-1 and 541 cm^-1.

Despite these interesting results, it is noteworthy that the overall Raman scattering of [Fe₂(pdt)(dppv)₂(CO)₂(H)]⁺ is weak. In comparison, very strong Raman signals are obtained for the precursor [Fe(pdt)(CO)₆] on our instrumental setup. This weakness in Raman scattering is unfortunate, because it prevents identification of the important ν(Fe-H) stretch in both isomers. DFT calculations predict ν(Fe-H) at 2000 cm^-1 for H-term and at 1294/1351 cm^-1 for μ-H.

Since ν(Fe-H) is not available from Raman experiments, diffuse reflectance FT-IR spectroscopy, which uses solids for the measurements, was utilized next. This prevents isomerization of H-term to μ-H in the process of making the usual KBr disks for IR spectroscopy. The obtained data show interesting differences in the ν(C=O) stretching region, which are currently analyzed. Unfortunately, the IR spectra of [Fe₂(pdt)(dppv)₂(CO)₂(H)]⁺ are dominated by vibrations of the eight phenyl substituents of the dppv ligands, which overshadow smaller signals. Correspondingly, we were not able to identify ν(Fe-H) from the spectra. We will therefore turn to the analogous compound [Fe₂(edt)(PMe₃)₄(CO)₂(H)]⁺ for future studies where the large dppv ligands are replaced by simple PMe₃.

AIM #2: DFT Calculations

For the DFT calculations, we have focused on the hydride-binding isomers of [Fe₂(edt)(PMe₃)₄(CO)₂(H)]⁺, because the small PMe₃ ligands allow for the application of better computational methods. The results show that μ-H is 5 – 12 kcal/mol (depending on the functional) more stable than H-term, which suggests that the lower reactivity of μ-H towards acids could simply be a consequence of the lower total energy of this species. A comparison of the electronic structures of both isomers indicates that the atomic charge of hydride is similar in both cases. This implies that the total charge of the bound hydride does not contribute to the difference in reactivity (no charge control). On the other hand, the H-term isomer has a key molecular orbital at relatively high energy that shows a strong hydride(1s) contribution (23%), which is lacking in the μ-H complex. This indicates a possible orbital control of the reaction of the complexes with acid. This aspect requires further study.

Back to top