New Strides In Metal-Metal Bonding

The periodic table, after more than a century of valuable service, is still yielding surprises about the chemistry of the elements. Two just-reported examples are the synthesis of a dichromium complex with the shortest metal-metal bond length on record and the synthesis of the first stable magnesium(I) compounds that also happen to contain rare Mg–Mg bonds.

The focus of each research paper reporting these two advances is different, yet the chemistry that made the compounds possible is similar. In the dichromium case, the quintuple-bonding interaction that gives rise to the record short Cr–Cr distance is challenging chemists to understand molecular orbital overlap in transition-metal bonding. In the dimagnesium case, chemists are just getting over the shock that stable Mg(I) compounds actually exist, and now they can decide what to make of them.

Multiple bonding between two transition-metal atoms has long been a source of fascination for inorganic chemists interested in the nitty-gritty of chemical bonding. Until the mid-1960s, chemists largely assumed that the triple bond was the highest possible multiple bond. But in 1964, F. Albert Cotton and coworkers at Texas A&M University surprised the chemistry community with evidence that the [Re₂Cl₈]^2- ion contained a quadruple bond between two metal atoms. Since then, chemists have observed many quadruple-bonded transition-metal compounds.

It took another 40 years to extend the multiple-bond boundary. In 2005, Philip P. Power of the University of California, Davis, and coworkers reported the first quintuple bond in the dichromium complex RCrCrR, where R designates a terphenyl ligand containing diisopropyl substituents (C&EN, Sept. 26, 2005, page 9). This very bulky ligand is used to help stabilize the complex and minimize the number of metal-ligand bonds, which promotes a higher degree of metal-metal bonding, Power tells C&EN.

Power notes that the two Cr(I) atoms, which have a 3d⁵ electron configuration, share five electron pairs in five bonding molecular orbitals. Despite this fivefold bonding interaction, the actual degree of bonding—the calculated bond order—is only 3.52 because of contributions from antibonding molecular orbitals, he explains.

The Cr–Cr bond length in Power's complex is 1.8351 Å. The record at the time, however, still belonged to Cotton, who set the standard of 1.828 Å in 1978 with a quadruple-bonded dichromium complex (Inorg. Chem. 1978, 17, 2084). Cotton's complex relies on four bridging methoxyphenyl ligands to form a cage that restrains the chromium atoms.

In the latest multiple-bonding advance, graduate student Kevin A. Kreisel and chemistry professor Klaus H. Theopold of the University of Delaware and coworkers have synthesized a new type of quintuple-bonded dichromium complex, one that surpasses all known compounds for having the shortest measured metal-metal bond (J. Am. Chem. Soc. 2007, 129, 14162).

Kreisel and Theopold took a page from both Cotton's and Power's notebooks for their dichromium complex: They used restrictive ligands similar to those in Cotton's complex and still were able to achieve high-order bonding similar to that in Power's complex.

The team prepared the complex by using potassium graphite (KC₈) to reduce a dichromium precursor in which each Cr(II) atom is bound by a diazadiene ligand containing diisopropylphenyl groups. The chromium atoms additionally are bridged by two chlorine atoms. In the reduced product, each chromium atom is coordinated to two nitrogen atoms, one each from the diazadiene ligands that reorganized to bridge the chromium atoms. The end result is a tight Cr₂N₄ core structure. The most noticeable feature is the very short 1.8028-Å Cr–Cr bond, which is now the shortest metal-metal bond distance on record for an isolable compound.

In collaboration with Clark R. Landis of the University of Wisconsin, Madison, the researchers supported their experimental observations with calculations to confirm the bonding and electronic structure of the complex, including a bond order of 4.28. Kreisel is now a postdoc in Landis' lab.

"It's really a very nice paper," Power says of the work by Kreisel, Theopold, and Landis. The Cr–Cr distance in the new complex "is quite interesting," he adds. "But the unique aspect, perhaps the most important observation, is clear evidence for five bonding interactions."

Power notes that his group has just completed additional studies of substituent effects in his dichromium complex. With modified ligands, the researchers have reeled in their bond length to 1.8077 Å, just shy of the new record.

Power and Theopold recently discussed the dichromium bonding. The two agree that "there's no reason to suppose that the lower limit in metal-metal bond distances has been reached," Power says. "Eventually, someone will get a shorter one."

The work on dichromium over the years has inspired explorations of metal-metal bonding all around the periodic table. The latest stop is at magnesium in group 2.

Magnesium exists in all its known stable compounds in the 2+ oxidation state. Some Mg(I) compounds, such as HMgMgH, have been studied at low temperature. And the formation of the synthetically important Grignard reagent, RMgX, where R is an organic group and X is a halide, has been proposed to proceed through a Mg(I) intermediate, RMgMgX. But no examples of stable group 2 metal compounds in the 1+ oxidation state have been reported, until now.

Shaun P. Green, Cameron Jones, and Andreas Stasch at Monash University, Victoria, Australia, used known dichromium chemistry and related dizinc chemistry as models in their attempts to make the dimagnesium compounds (Science, DOI: 10.1126/science.1150856).

The researchers started with Mg(II) iodide precursors bearing N,N′-chelating ligands, either a guanidinate ligand or a diketiminate ligand, containing bulky diisopropylphenyl groups. They reduced the precursors using potassium metal to form the Mg–Mg compounds.

But unlike the dichromium complex's fivefold bonding, the Mg–Mg bonds are quite long at about 2.85 Å, owing to a single metal-metal sigma bond. The metal-ligand interactions are predominantly ionic, thus the Mg–Mg center can be viewed as an anion-stabilized Mg₂²⁺ unit, the researchers note.

"Isolation of honest-to-goodness Mg(I) compounds is a spectacular achievement," Theopold says. Although there are many parallels between the syntheses of the dichromium and dimagnesium complexes, the metal-metal bonding couldn't be more different, he adds.

"After getting over the surprise of isolated Mg(I) compounds, most chemists probably will realize that the simple and fairly long sigma metal-metal bond is to be expected, whereas the Cr–Cr complex derives its interest from the nature of the orbital overlap, the very high bond order, and the very short metal-metal bond distance," Theopold says.