59th Annual Report on Research 2014

The goal of this project is to harness redox cooperativity within metal clusters to carry out challenging multielectron redox reactions, such as dinitrogen reduction and O₂ activation. These reactions are pivotal for renewable fuel synthesis (e.g., CO₂ reduction) and hydrocarbon functionalization (e.g., methane oxidation to methanol). Based on precedence in biology, redox cooperativity can allow access to significantly lower energy pathways for these reactions as compared to current methodologies. Our approach is based on the design of modular macrobicyclic ligands to template all parameters of cluster assembly, which in turn will allow for a systematic investigation how electronic and steric effects dictate reactivity. Our long term goal is to develop a mechanistic understanding of the parameters that govern redox cooperativity within metal clusters.

Prior to the reporting period, we communicated the synthesis of a tris(diketimine) cyclophane ligand (H₃L) and a hexacarboxamide cryptand (H₆L’), which both template trinuclear clusters. This designed approach to clusters using macrobicycles was demonstrated in the synthesis of an (FeCl₂)₃ cryptate complex, as well as that of the triiron(II) tri(bromide) and a trimanganese(II) tri(bromide) cyclophane complexes. We have primarily focused on complexes of the cyclophane ligand. One of our goals for the proposal is to understand the role of metal ion type on reactivity. Therefore, we synthesized the corresponding trichromium(II), tricobalt(II), trinickel(II), tricopper(I), trizinc(II), trigermanium(II), tritin(II), and trialuminum(III) complexes using the cyclophane ligand (Figure 1).

Figure 1.

Our results on copper cluster chemistry are summarized here. We reported the anionic chloride-bridged tricopper(I) compound, which exhibits strong luminescence upon excitation in the near UV. By using Cu(OTf) as our metal source, we were able to crystallize the first molecular complex containing Cu–N₂ interactions, i.e., Cu₃(N₂)L. This adduct is possible because of the unparalleled size and shape selectivity afforded by our cyclophane; the steric constraints exclude solvent ligands to the metal centers, leaving N₂ as the only available ancillary donor. From vibrational spectroscopy, the N–N stretching mode is lower in energy than expected for an unactivated N₂ (i.e., 1952 cm^-1 for ¹⁴N₂). Computationally, minor changes to the orientation of the ligand arms results in a surprisingly large decrease (~200 cm^-1) for the N–N vibration. This result is important because it shows that N₂ reduction using Cu may be possible despite the poor π-backbonding for Cu. The fluorescence of this N₂ adduct differs significantly from the chloride-bridged complex, and the guest dependent luminescence is currently under investigation. The N₂ complex, Cu₃(N₂)L, readily reacts with S₈ to afford the first example of a coordinatively unsaturated sulfide-bridged copper cluster in which all metal centers are in a nitrogen-rich environment, Cu₃SL (Figure 2). This cluster bears strong resemblance to the copper-sulfide cluster, CuZ, in nitrous oxide reductase, which activates N₂O (a potent greenhouse gas) to liberate N₂ and water. Our synthetic analogue is readily reduced by one electron, and further reactivity studies of this cluster are in progress. In addition, Cu₃(N₂)L reacts directly with N₂O and other O-atom sources to generate an oxo-bridged cluster, Cu₃OL, with Se to afford the selenide-bridged tricopper compound, or with N-atom sources (e.g.,azide) to yield the anionic nitride-bridged tricopper compound. Reaction of Cu₃(N₂)L with O₂ in the presence of hydrocarbon substrates (e.g., toluene, ethylbenzene) results in O-atom insertion from O₂ into substrate C–H bonds. Similar reactions of our tricopper(I) cluster with N-atom sources (e.g., azide) leads to the formation of a complex consistent with a nitride-bridged cluster. Reactivity studies and further characterization of these compounds is currently ongoing.

Figure 2.

Reaction of Fe₃Br₃L with KEt₃BH or Ph₃CS^- affords the trihydride complex, Fe₃H₃L, or the triferric tri(sulfide), Fe₃S₃L, clusters. The hydride structure contains unusually long Fe–H bond distances and acute N–Fe–N bite angles, which highlights how the structural constraints imposed by the ligand can perturb the cluster geometry. Metal hydrides are important species in industrial and biological catalysis, and these compounds provide an opportunity to understand the reactivity of hydride clusters in a rational manner. The sulfide-bridged compound has an unprecedented planar arrangement of the metal and chalcogenide ions. This geometry means that we can examine the effect of cluster geometry (as compared to cubane-type clusters) on redox properties and reactivity. Metal-sulfide clusters are prevalent cluster in biology as well as being important materials for light harvesting (e.g., MoS₂) and industrial catalysis. Fe₃H₃L reacts selectively with CO₂ to give Fe₃H₂(OOCH)L (Figure 2), but the trihydride is unreactive towards all other substrates tested (inc., CH₃CN, CS₂, C₂H₂) at room temperature. This result is unusual and sharply contrasts the reactivity of the mononuclear hydride complexes. The corresponding tricobalt and trizinc tri(hydride) complexes also demonstrate similar selectivity for CO₂, suggesting that steric effects may dominate reactivity. Reduction of Fe₃Br₃L with 6 equiv. of KC₈ under an N₂ atmosphere results in cleavage of the N=N as evidenced by reactions using isotopically-enriched ¹⁵N₂, the release of ammonia upon acid decomposition of the product, and the solid-state structure of the isolated product (Figure 2). From both THF and toluene reactions, the isolated product Fe₃(NH)₃L contains three protonated N-atom bridges, which suggests that intercomplex cooperativity and X–H bond activation (X = C or O) or protonation likely occur. This complex is the first designed metal cluster competent for N₂ activation.

Finally, we have demonstrated the selective synthesis of dinuclear clusters with proximal hydrogen bonding interactions by partial metalation of our hexacarboxamide cryptand with either nickel(II) or copper(II) (Figure 3). The elaborate hydrogen bonding network within the internal cavity of ligand is reminiscent of secondary coordination spheres in enzyme active sites. We have observed that the di(hydroxonickel) complex reacts rapidly and irreversibly with CO₂ from air to generate a bicarbonate complex. Our inability to displace CO₂ either under vacuum or upon sparging differs from comparable monometallic compounds and suggests a similar strategy can be applied to design high affinity CO₂ sequestration materials.

Figure 3.