Steric Switching From Photochemical to Thermal N2 Splitting: A Computational Analysis of the Isomerization Reaction {(Cp*)(Am)Mo}2(μ-η1:η1-N2) → {(Cp*)(Am)Mo}2(μ-N)2

Abstract

A μ-η¹:η¹-N₂-bridged Mo dimer, {(η⁵-C₅Me₅)[N(Et)C(Ph)N(Et)]Mo}₂(μ-N₂), cleaves dinitrogen thermally resulting in a crystallographically characterized bis-μ-N-bridged dimer, {(η⁵-C₅Me₅)[N(Et)C(Ph)N(Et)]Mo}₂(μ-N)₂. A structurally related Mo dimer with a bulkier amidinate ligand, ([N(ⁱPr)C(Me)N(ⁱPr)]), is only capable of photochemical dinitrogen activation. These opposing reactivities were rationalized as steric switching between the thermally and photochemically active species. A computational analysis of the geometric and electronic structures of intermediates along the isomerization pathway from Mo₂(μ-η¹:η¹-N₂) to Mo₂(μ-η²:η¹-N₂) and Mo₂(μ-η²:η²-N₂), and finally Mo₂(μ-N)₂, is presented here. The extent to which dispersion affects the thermodynamics of the isomers is evaluated, and it is found that dispersion interactions play a significant role in stabilizing the product and making the reaction exergonic. The concept of steric switching is further explored with theoretical models with sterically even less demanding ligands, indicating that systematic ligand modifications could be used to rationally design the N₂ activation energy landscape. An analysis of electronic excitations in the computed UV-vis spectra of the two complexes shows that a particular type of asymmetric excitations is only present in the photoactive complex.

Keywords: density functional theory; isomerization thermodynamics; molybdenum; nitrogen fixation; theoretical UV-vis spectroscopy.

Figure 1

(A) Overview of Sita's series of N₂-bridged complexes; Zr and Hf complexes are non-planar μ-η²:η²-N₂-bridged whereas all other known complexes have linear μ-η¹:η¹-N₂-bridged cores. (B) Capacity of the molybdenum complexes 1 and 2 for thermal or photochemical dinitrogen activation. (C) Isomerization path predicted computationally (Zhang et al., 2011) for the Ta congener using a simplified ligand framework.

Figure 2

Geometries predicted along the isomerization path of 1 (A) and 2 (B). Color code for atoms is Mo: light blue, N: blue, C: gray, H omitted for clarity.

Figure 3

Key geometric parameters computed for the isomers of 1 (A) and 2 (B), including the core geometries of 1/2_dia in their triplet states. Values found in crystal structures where available are given in blue.

Figure 4

Thermodynamic profile of the isomerization paths for 1 (red) and 2 (blue). Data points for which the electronic structures are calculated with PBE0 and B3LYP are long-dashed and short-dashed, respectively.

Figure 5

Visualization of the number of hydrogen pairs along the isomerization pathway of 1 and 2, (A) in terms of absolute numbers in bins of 1.8–2.2 Å (dark blue), 2.2–2.6 Å (middle blue), 2.6–3.0 Å (light blue), 3.0–3.4 Å (yellow), 3.4–3.8 Å (orange), 3.8–4.2 Å (red), and (B) as a representation of the actual geometric interactions in 1_lin (top left), 1_dia (top right), 2_lin (bottom left) and 2_dia (bottom right); H-H interactions are shown as thin lines in dark blue (1.8–2.2 Å), middle blue (2.2–2.6 Å), and light blue (2.6–3.0 Å). Color code for atoms is Mo: light blue N: blue, C: gray, H light gray.

Figure 6

Effect of omitting dispersion corrections on the thermodynamic profile of the isomerization paths of 2. The thermodynamic profile including dispersion is shown as a solid line for reference; data points for which the electronic structures are calculated with BP are long-dashed-dotted (dark blue), with PBE0 are short-dash-dotted (mid-blue), and with B3LYP short-dashed (light blue).

Figure 7

Thermodynamic profiles for 1, 2, and {Me-Et} (3) and {Ph-H} (4) as two systems with further reduced steric bulk. The top, middle and bottom panel show data for electronic structures calculated with the density functionals BP86, PBE0, and B3LYP, respectively.

Figure 8

Partial MO diagram of the π and δ manifold for the linear cores of complexes 1 and 2.

Figure 9

(A) Experimental UV-vis spectrum of 1_lin (black: digitized, gray: digitized and blue-shifted) and the broadened line spectrum predicted with TDDFT using the LC-BLYP density functional (top), calculated oscillator strengths and intensities of the individual transitions and broadened line spectrum for 1_lin (middle) and 2_lin (bottom). (B) Difference densities for the individual transitions labeled in (A); yellow and red isosurfaces correspond to density loss and gain, respectively.