Organometallic actinide chemistry

Developing a comprehensive picture of the chemical behavior of the 5f elements

There has been a recent resurgence of interest in understanding the electronic structure of complexes of the actinide elements. This is driven principally by the desire to predict the behavior of actinide species in a variety of applications within the DOE complex, from waste repositories to the design of selective chemical separation methodologies.

Generally, present models describing the electronic structure of the actinides are most relevant to chemistry performed under aerobic conditions and do not provide a comprehensive set of chemical environments in which to probe characteristics of metal-ligand bonding. To assess the validity of theoretical treatments already developed and ensure that chemical behavior is predictable under a wider array of process conditions, it is important to study a broader range of ligands and chemical environments than those commonly examined and their influence on observable chemical behavior.

Nonaqueous organometallic chemistry provides access to the broadest possible set of ligands with which to examine the influence of coordination environment on the electronic structure and chemical reactivity of the actinide elements. We have chosen to focus particular attention on the generation and investigation of actinide-ligand multiple bonds to probe the potential for involvement of metal 5f electrons in chemical bonding, examine the stability of functional groups isoelectronic with the oxo ligand, and demonstrate reaction patterns unique to f elements.

The uranyl ion (1) and the U(VI) bis(imido) complex, (C₅Me₅)₂U(=N-Ph)₂ (2).

Nonaqueous analogues to the uranyl ion

The uranyl ion (1) is the dominant form of uranium in aerobic media and features a linear O≡U≡O fragment with uranium in its highest formal oxidation state, U(VI). More than a decade ago, we prepared the first example of an organometallic U(VI) complex, (C₅Me₅)₂U(=N-Ph)₂ (2). This complex was remarkable in that it also represented the first example of a bis(imido) U(VI) electronic analogue to the uranyl ion, UO₂²⁺

. However, unlike the uranyl ion, this uranium complex displays an unusual nonlinear or cis arrangement of the two imido functional groups (U=N). The fact that complex 2 contained a bent N=U=N core challenged accepted motifs for bonding in actinide complexes. Complex 2 is a 20-electron species that has no transition metal analogue, raising significant questions about the electronic structure of this compound and in particular the participation of valence metal orbitals (5f, 6d, and 7p) in ligand-to-actinide π bonding.

We examined the extent of f-orbital participation in this hexavalent uranium complex through collaborative interactions with both experimental and theoretical scientists. Bruce Bursten of The Ohio State University and P. Jeffrey Hay of Los Alamos conducted theoretical investigations on simplified model complexes and demonstrated that 5f-orbital involvement can contribute to the stabilization of nitrogen donors in a manner that is not possible for the transition metals. Jennifer Green of Oxford University confirmed these results, using photoelectron spectroscopy, which revealed a small but significant f-orbital-based bonding interaction.

Complex 2 can be prepared by a variety of methods as shown in Scheme 1. The most general route is two-electron oxidative atom transfer using organic azides (RN₃) (e.g., 3→2, 5→2). Another method that has proven useful in the synthesis of bis(imido) transition metal complexes is the reductive cleavage of 1,2-disubstituted hydrazines, in which the cleavage of hydrazo compounds (R-NHNH-R) allows for the introduction of a single organoimido functional group.

However, the reaction of complex 4 with 1,2-diphenylhydrazine to produce the bis(imido) complex 2 provides the first example of the reductive cleavage of a hydrazo compound yielding two ogano-imido functional groups at a single metal center. The most unusual chemistry exhibited by these bis(pentamethyl-cyclopentadienyl) uranium complexes is the four-electron reductive cleavage of the double bond in azobenzene (Ph-N=N-Ph) by (C₅Me₅)₂U(Cl)(NaCl) generated by reduction of complex 6 by Na/Hg to give the bis(imido) complex 2, a process not often observed for transition elements. These reactions, shown in Scheme 1, underscore some differences between the transition metals and the actinides and suggest the possible involvement of the f orbitals in chemical reactions of uranium.

Scheme 1. Synthetic scheme for the preparation of the U(VI) bis(imido) complex 2.

In search of reactive actinide-ligand multiple bonds

One of the most exciting discoveries in organometallic chemistry over the past 15 years was the observation that d⁰ transition-metal imido complexes readily react with carbon-hydrogen bonds. Analogous f⁰ actinide imido complexes might be expected to display similar reactivity patterns to their early transition-metal counterparts; however, as in the case of uranyl compounds, the most characteristic chemical property of uranium imido bonds is their decided lack of chemical reactivity.

Recently, we discovered a method to access reactive uranium imido complexes. We found that diazo-alkanes (R₂CN₂) can be used to prepare U(VI) bis(imido) complexes that are reactive toward alkanes, as illustrated in Scheme 2. Reaction of the U(IV) monoimido complex 7 with diphenyldiazomethane generates the U(VI) bis(imido) complex 8. Complex 8 does not lose N₂ to give a U(VI) alkylidene complex (an alkylidene is a complex possessing a metal-carbon, or M=C, functional group). This is in marked contrast to the chemistry observed for isoelectronic N₂O and organoazides (RN₃), which have been exploited for the preparation of U(VI) oxo (U=O) and imido complexes.

The oxidation of the uranium metal center from U(IV) to (VI) is clearly demonstrated by NMR spectroscopy and electronic absorption spectra of complex 8. The electronic absorption spectrum shows no f→f transitions in the near-IR region but does show a broad, intense and featureless charge-transfer band in the visible region, which is consistent with the assignment of an f⁰ U(VI) metal center.

The proton NMR spectra of complex 8 is temperature invariant between –75 and 60 °C but displays anomalous chemical shifts, suggesting that complex 8 is a temperature-independent paramagnet, a characteristic property of organometallic complexes of U(VI). An x-ray crystal structure of complex 8 confirmed the presence of two organoimido groups terminally bound to the uranium center and the overall structure of complex 8.

Scheme 2. Synthetic scheme showing the reaction of the U(IV) monoimido complex 2 with Ph₂CN₂ to generate the bis(imido)U(VI) complex 8, and its thermal transformation into a cyclometallated complex 9.

Proton NMR spectroscopy signaled the quantitative formation of the novel cyclometallated U(IV) bis(amide) complex 9 upon thermolysis of a solution of complex 8; the NMR spectrum of the product is paramagnetically shifted, which indicates that the product is a reduced f² U(IV) species. X-ray crystallography unambiguously ascertained that the C–H bond of one of the t-butyl groups from the U=N-2,4,6-t-Bu₃C₆H₂ fragment in complex 8 has been activated to give complex 9.

This reaction type has no equivalent in transition metal chemistry and represents a new pattern of reactivity available for high-valent actinide imido complexes. The uranium center is bound to three nitrogen atoms. Two exhibit short U–N distances associated with uranium-amide linkages. The remaining U–N length is consistent with a uranium-nitrogen dative interaction. Notably, the U–N(H)-C fragment is bent, which differs significantly from the nearly linear U−N−Cipso unit present in the parent bis(imido) complex 8.

f-electron delocalization in organometallic actinide complexes

We are also interested in understanding the degree to which valence metal orbitals contribute to bonding and reactivity. Covalency (the mixing of orbitals) in metal-ligand bonding is quite prevalent in transition metal complexes but is suggested to play a reduced role in the chemistry of the actinides. We recently discovered several classes of organo-uranium complexes in which the oxidation state of the metal, and thus the location of the valence electrons, is not easily determined. The spectroscopy, reactivity, and structural chemistry of these compounds suggest that the f electrons are far more involved in chemical bonding than previously thought.

We have investigated the reaction chemistry of low-valent organouranium complexes with diazoalkanes, which are known to generate transition metal alkylidene complexes under similar conditions. Reaction of the U(IV) complex 6 with diphenyldiazomethane under reducing conditions generates complex 10 as indicated in Scheme 3. This species is unlike existing transition-metal diazoalkane complexes in that it possesses two diazoalkane molecules bound to a single uranium metal center. One diazoalkane is coordinated to the uranium metal center in an h¹ fashion and forms a uranium imido bond. The second diazoalkane is coordinated to the uranium in an h² manner, possibly through the N=N π system.

The exceedingly long N--N bond distance indicates that substantialU fragment and an empty orbital on the diazoalkane unit, which has antibonding character, resulting in a lengthening of the N-N bond. Interestingly, theoretical calculations reveal an f¹π* electronic configuration and suggest delocalization of a 5f electron throughout the N=N=C framework of the h²-coordinated diazoalkane.

Upon standing in solution, the bis(diazo-alkane) complex 10 transforms into the novel uranium bis(ketimido) complex 11, via loss of N₂from two diazoalkane fragments. The U−N bond distances are significantly shorter than other structurally characterized U(IV) bis(amido) complexes such as compound 9.

Scheme 3. Synthetic scheme for the preparation of uranium diazoalkane complex 10 and its transformation into the U(IV) bis(ketimido) complex 1.

The nearly linear U−N−C fragments are consistent with the formulation of the ligands as ketimido groups with the metal center accepting additional electron density from the nitrogen lone pairs. Spectroscopic data suggest donation of electronic density from the metal center onto the ketimido ligands in this system.

Complex 11 is reminiscent of high-valent U(VI) bis(imido) complexes that have significant U-N multiple bond character as typified by complex 2. The metal center in complex 11 is not U(VI), which is evident upon comparison of the U-N distances in the two systems as well as the NMR and UV-visible-near-IR spectra. The complex is formally an f² system, yet it does not behave like known U(IV) amido complexes. The uranium ketimido complex is surprisingly unreactive, displays unusual electronic properties, and is able to support the uranium metal center in a variety of oxidation states ranging from (III) to (V).

The physical properties and chemical stability of this U(IV) complex indicate higher U–N bond order due to significant ligand-to-metal π bonding in the uranium ketimido interactions. The combined data suggest electronic delocalization throughout the N-U-N core and indicate that the f electrons in midvalent organouranium complexes might be far more involved in chemical bonding and reactivity than previously thought.

The role of f orbitals in reactions of actinides

The role of the 5f orbitals in bonding and reactivity is still one of the most intriguing questions in actinide chemistry and continues to be a topic of considerable debate among theoretical and experimental chemists. The high nodality of the 5f orbitals and the expanded electron count of many organoactinide complexes provide the opportunity for the actinide metals to engage in bonding schemes that are not available to the transition metals, leading to new patterns of reactivity. While there is spectroscopic and physical evidence to suggest that f orbitals are key to the stabilization of many high-valent (V, VI) organo-actinide complexes, fewer examples of 5f orbitals directing reactivity are known in tetravalent actinide chemistry.

In our studies, we have found that diazoalkanes readily insert into both M-C bonds of bis(alkyl) complexes (C₅Me₅)₂MR₂ to yield bis(hydrazonato) complexes (see compound 13 as shown in Scheme 4). These 22-electron complexes have no transition metal analogues; the reaction of diphenyldiazomethane with zirconium derivatives affords only mono insertion products. Because transition metals do not have valence f orbitals, these uranium and thorium bis(hydrazonato) complexes provide a unique opportunity to investigate f-orbital participation in the reaction chemistry of organometallic actinide compounds. Theoretical calculations on related model complexes reveal that the f orbitals interact to a small extent with the N–N π orbitals and the N–N s orbitals. These results suggest that 5f orbitals can contribute to the stabilization of side-bound N–N units for which no appropriate symmetry match exists for transition metals.

Scheme 4. Synthetic route to the Th(IV) bis(hydrazonato) complex 13.

Future directions

The use of novel ancillary ligand frameworks has provided the opportunity to access complexes in a variety of oxidation states that feature examples of actinide-ligand multiple bonding, yielding a much deeper understanding of 5f-orbital energetics and their involvement in stabilizing metal-ligand multiple bonds. In particular, the observation of "valence-ambiguous" behavior in organo-uranium complexes challenges the once common belief that 5f orbitals are energetically inaccessible and serve merely to house unpaired electrons. With our knowledge of the role of both oxidation state and ancillary ligands in stabilizing metal-ligand multiple bonds, it will be possible to extend this work to other as yet unrealized molecular targets, such as species with metal-carbon multiple bonds.

As in the case of actinyl compounds, species containing multiply bound functional groups have most often been found to be unreactive. Recent advances, however, shed light on the means to enhance the reactivity of actinide-ligand multiple bonds, revealing metal-mediated transformations uniquely enabled by 5f orbitals. Further studies will explore methods to rationalize and exploit this behavior.

The development of the chemistry of transition metals with multiply bound ligands altered chemists' views of the scope of reactivity possible for these metals. Development of the metal-ligand multiple bond chemistry of the actinides will similarly expand the reach of modern f-element chemistry, while also setting the organometallic chemistry of the actinides apart from that of the transition metals.

This article was contributed by Los Alamos researchers Jaqueline L. Kiplinger of the Chemistry Division and Carol J. Burns of the Office of the Deputy Director for National Security.

NMT | LANL | DOE
Phone Book | Search | Help/Info

L O S  A L A M O S  N A T I O N A L   L A B O R A T O R Y
Operated by the University of California for the US Department of Energy