Leslie Justin Murray

Ambient-temperature reactivity in biological and industrially relevant reactions is typically attributed from the ground state as a sufficiently large energy gap between ground and excited states of reactants renders the thermal Boltzmann population of excited states as a minor perturbation. However, interactions between locally high-spin centers in polynuclear clusters can afford low-lying excited states with appreciable populations at ambient temperature and provide alternate reaction pathways. Consequently, a mechanistic understanding of the reactivity of such clusters presents challenges. In this perspective, we consider advances towards dissecting the effects of such low-lying states on reactivity.

Nitric oxide (NO), produced by nitrite reductases or nitric oxide synthases, performs vital roles in signaling and the immune response. Iron sulfur (FeS) clusters are known targets for NO induced degradation, serving as sensors to trigger cellular responses. However, this FeS reactivity is proposed as NO specific, with no demonstrated reactivity toward nitrite, a soluble NO storage molecule. We demonstrate that synthetic FeS clusters supported by various ligands undergo facile nitrosylation by nitrite in the presence of a reductant, evidencing the nitrite reductase reactivity for FeS clusters. Moreover, a mononitrosylated Fe₄S₄ cluster, [tempS₃Fe₄S₄(NO)]^2-, can be readily synthesized by this approach, enabling further investigation into the FeS cluster repair and decomposition under NO induced oxidative stress.

Fischer-Tropsch conversion of syngas to hydrocarbons is proposed to begin with CO binding to the iron surface of the catalyst. CO adsorption on various iron facets of relevance to the Fischer-Tropsch process suggest that the Fe(111) surface is the most active for catalysis, and that CO bound to the penultimate layer of Fe atoms or the b-state is the resting state during catalysis. Notably, a μ-1,2 mode was discarded for the b-state due to a lack of exemplar molecular species and expectation that such a mode would have a higher energy infrared (IR) absorption than observed experimentally (viz. 1735-1860 cm^-1). Here, we report the synthesis of a diiron(I/II) complex in which CO binds μ-1,2: (Fe(OTf))(Fe(THF)(μ-1,2-CO))L where L^2- is a bis(β-diketiminate) cyclophane (1). Surprisingly, the observed ν_CO at 1763 cm^-1 for 1 compares well with that reported for b-state. Electron paramagnetic resonance (EPR), Mössbauer, and density functional theory (DFT) results support a weakly coupled s = 3/2 iron(I) and s = 2 iron(II) pair. Reduction of 1 results in C-O cleavage and C-C bond formation to yield a ketenylidene (CCO) complex as a major product observed spectroscopically.

The reactivity of iron clusters with one of more μ-hydrides and in the weak field pertains to catalysis on surfaces and biological metal cluster cofactors. As a model, then, a reactivity survey of the weak-field ligated iron hydride clusters Fe₃H₃L (1) and (FeCO)₂Fe(μ₃-H)L (2) (where L^3- is a tris(β-diketiminate)cyclophanate) with Brønsted acids, organochlorides, acetyl chloride, boron trihalides, and titanium electrophiles is reported. Complex 1 reacts with Brønsted acids H₂O and [Et₃NH][Cl] to afford Fe₃(OH)₃L (3) and Fe₃H₂ClL (4), respectively, consistent with hydridic reactivity. Clusters 1 and 2 react readily with organochlorides, such as CCl₄, CHCl₃, and CH₂Cl₂, with identified intermediates supporting a radical pathway. Complex 1 reacts with trityl chloride (2 equiv) to selectively afford Fe₃HCl₂L (5) with reductive elimination of dihydrogen observed. Mixed-valent complex 2 reacts with AcCl to afford (FeCO)Fe₂HClL (9). The scope of reactivity displayed implicates possible pathways accessible to larger clusters in biology or on metal surfaces.

A pair of cyclophane-encapsulated [Cu₃(μ₃-E)]³⁺ complexes (E = S and Se) were characterized by resonant X-ray diffraction anomalous fine structure (DAFS), revealing unexpected polarization among the three Cu sites attributed to long-range effects of π-stacking interactions with cocrystallized benzene molecules. The resonant K-edge energies of individual Cu sites within the cluster molecules were found to vary as a function of distance from the cocrystallized benzene. This pattern was interpreted in the context of T-shaped, edge-to-face π-stacking with the assistance of theoretical charge density difference calculations.

Iron hydrides are proposed reactive intermediates for N₂ and CO conversion in industrial and biological processes. Here, we report a reactivity study of a low-coordinate di(μ-hydrido)diiron(II) complex, Fe₂(μ-H)₂L, where L^2- is a bis(β-diketiminate) cyclophane, with isocyanides, which have electronic structures related to N₂ and CO. The reaction outcome is influenced by the isocyanide substituent, with 2,6-xylyl isocyanide leading to H₂ loss, to form a bis(μ-1,1-isocyanide)diiron(I) complex, whereas all of the other tested isocyanides insert into the Fe-H bond to give (μ-1,2-iminoformyl) complexes. Steric bulk of the isocyanide substituent determines the extent of insertion (i.e., into one or both Fe-H-Fe units) with tert-butyl isocyanide reacting to yield the mono-(μ-1,2-iminoformyl)diiron(II) complex, exclusively, and isopropyl- and methyl isocyanides affording the bis(μ-1,2-iminoformyl)diiron(II) products. Treatment of Fe₂(μ-1,2-CHNtBu)(μ-H)L with 2,6-xylyl isocyanide (or XylNC) yields Fe₂(μ-XylNC)₂L and tert-butylaldimine as one of the organic products.

Metal-ligand cooperative systems have a long precedent in catalysis, with the classification depending on the site of substrate bond cleavage and formation and on redox state changes. Recently, our group reported the participation of a β-diketiminate ligand in chemical bonding to heterocumulenes such as CO₂ and CS₂ by tricopper complexes, leading to cooperative catalysis. Herein, we report the reactivity of these copper clusters, [Cu₃EL]^- (E = S, Se; L = tris(β-diketiminate) cyclophane ligand), toward other electrophiles, viz. alkyl halides and Brønsted acids. We identified a family of ligand-functionalized complexes, Cu₃EL (R) (R = primary alkyls), and a series of disubstituted products, Cu₃EL (R)₂, through single-crystal X-ray diffraction, mass spectrometry, and infrared and UV-visible spectroscopy. As part of mechanistic studies on these alkylation reactions, we evaluated the acid-base reactivity of these complexes and the influence of the backbone substitution on the reduction potential. Implications of these findings for ligand noninnocence and the relevance of the metal core as a cofactor for the ligand's reactivity are discussed.

Ligand non-innocence is ubiquitous in catalysis with ligands in synthetic complexes contributing as electron reservoirs or co-sites for substrate activation. The latter chemical non-innocence is manifested in H+ storage or relay at sites beyond the metal primary coordination sphere. Reaction of a competent CO2-to-oxalate reduction catalyst, namely, [K(THF)(3)](Cu3SL), where L3- is a tris(beta-diketiminate) cyclophane, with CS2 affords tetrathiooxalate at long reaction times or at high CS2 concentrations, where otherwise an equilibrium is established between the starting species and a complex-CS2 adduct in which the CS2 is bound to the C atom on the ligand backbone. X-ray diffraction analysis of this adduct reveals no apparent metal participation, suggesting an entirely ligand-based reaction controlled by the charge state of the cluster. Thermodynamic parameters for the formation of the aforementioned C-ligand-CS2 bond were experimentally determined, and trends with cation Lewis acidity were studied, where more acidic cations shift the equilibrium toward the adduct. Relevance of such an adduct in the reduction of CO2 to oxalate by this complex is supported by DFT studies, similar effects of countercation Lewis acidity on product formation, and the homocoupled heterocumulene product speciation as determined by isotopic labeling studies. Taken together, this system extends chemical non-innocence beyond H+ to effect catalytic transformations involving C-C bond formation and represents the rarest example of metal-ligand cooperativity, that is, spectator metal ion(s) and the ligand as the reaction center.

Studies of multinuclear metal complexes are greatly enhanced by resonant diffraction measurements, which probe X-ray absorption profiles of crystallographically independent metal sites within a cluster. In particular, X-ray diffraction anomalous fine structure (DAFS) analysis provides data that can be interpreted akin to site-specific XANES, allowing for differences in metal K-edge resonances to be deconvoluted even for different metal sites within a homometallic system. Despite the prevalence of Cu-containing clusters in biology and energy science, DAFS has yet to be used to analyze multicopper complexes of any type until now. Here, we report an evaluation of trends using a series of strategically chosen Cu(I) and Cu(II) complexes to determine how energy dependencies of anomalous scattering factors are impacted by coordination geometry, ligand shell, cluster nuclearity, and oxidation state. This calibration data is used to analyze a formally tricopper(I) complex that was found by DAFS to be site-differentiated due to the unsymmetrical influence on different Cu sites of the electrostatic field from a proximal K⁺ cation.

One route to address climate change is converting carbon dioxide to synthetic carbon‐neutral fuels. Whereas carbon dioxide to CO conversion has precedent in homo‐ and heterogeneous catalysis, deoxygenative coupling of CO to products with C−C bonds—as in liquid fuels—remains challenging. Here, we report coupling of two CO molecules by a diiron complex. Reduction of Fe₂(CO)₂L (2), where L²⁻ is a bis(β‐diketiminate) cyclophane, gives [K(THF)₅][Fe₂(CO)₂L] (3), which undergoes silylation to Fe₂(CO)(COSiMe₃)L (4). Subsequent C‐OSiMe₃ bond cleavage and C=C bond formation occurs upon reduction of 4, yielding Fe₂(μ‐CCO)L. CO derived ligands in this series mediate weak exchange interactions with the ketenylidene affording the smallest J value, with changes to local metal ion spin states and coupling schemes (ferro‐ vs. antiferromagnetism) based on DFT calculations, Mössbauer and EPR spectroscopy. Finally, reaction of 5 with KEt₃BH or methanol releases the C₂O²⁻ ligand with retention of the diiron core