Organozirconium Chemistry Arrives

Organozirconium chemistry turns 50 this year. Its golden anniversary was celebrated last month at the American Chemical Society national meeting in Anaheim, Calif., in a symposium that offered a crash course in the history of the field and a sampling of its breadth.

In late October 1952, John M. Birmingham, then a Ph.D. student in the lab of Geoffrey Wilkinson at Harvard University, prepared the first organozirconium compound, zirconocene dibromide. The summer before, Wilkinson and Robert B. Woodward had just described ferrocene--or dicyclopentadienyliron--as a -complex sandwich compound. Ferrocene is extraordinarily stable compared with the organometallic compounds prepared before it. Birmingham got excited by the possibility that compounds like ferrocene would merge the separate worlds of organic and inorganic chemistry. He wanted to be among the first in the emerging field, and he asked to join Wilkinson's group.

Wilkinson and coworkers, including Birmingham, reported the synthesis of zirconocene dibromide (Cp₂ZrBr₂, where Cp = cyclopentadienyl) and nickel, titanium, and vanadium analogs in early 1953. Soon after, Birmingham told C&EN, the recipe was changed to a more convenient one yielding chlorides. Publication was delayed to September 1954, in part because Wilkinson was away on sabbatical leave. The anniversary that's being marked this year is that of the later publication, by Birmingham and Wilkinson [J. Am. Chem. Soc., 76, 4281 (1954)].

Beyond its characterization, not much happened with zirconocene dichloride until 1960, when a patent assigned to David S. Breslow of Hercules Powder Co. disclosed that the compound can be used to catalyze ethylene polymerization. At present, the largest use of organozirconium compounds is as polymerization catalysts. Zirconocenes are a class of metallocenes, which have greatly improved olefin polymerization beyond what could be achieved with traditional Ziegler-Natta catalysts. Metallocenes have enabled unprecedented control of polymerization and are now used to make hundreds of millions of pounds of polymers per year (C&EN, Oct. 22, 2001, page 35). A major metallocene supplier is Boulder Scientific Co., the Colorado company Birmingham founded in 1972.

Contributors to this industrial application of organozirconium compounds include Walter Kaminsky of the University of Hamburg and Hans-Herbert Brintzinger of the University of Konstanz, both in Germany. Kaminsky pioneered the use of zirconocene dichloride as a homogeneous olefin polymerization catalyst with methylaluminoxane as the activator, while Brintzinger first used organozirconium compounds with chiral ligands as olefin polymerization catalysts. Mechanistic aspects of polymerization catalysis by zirconocenes continue to be active areas of research in various labs.

THE USE OF organozirconium compounds to prepare organic fine chemicals began only in the mid-1970s. In 1970, Melbourne, Australia-based chemists Peter C. Wailes and H. Weigold prepared zirconocene hydrochloride. Later they showed that in the presence of alkenes, this compound adds zirconium and hydrogen across the carbon-carbon double bond. That is, the alkene undergoes a hydrozirconation reaction to form an alkylzirconium complex. Jeffrey Schwartz, who in 1972 had just joined Princeton University, then took the hydrozirconation products and converted them to organic compounds.

The groundbreaking work was published on Christmas day in 1974. It showed that alkylzirconium complexes are easy to handle and that their reactivity toward electrophiles makes them excellent intermediates for making alkyl halides. According to Schwartz, his lab prepared "lots of different things," starting with zirconocene hydrochloride, which is now well known as the Schwartz reagent.

Schwartz's chemistry enabled synthetic organic chemists to use the intermediates in single-molecule-type reactions, as opposed to polymerizations, according to Bruce H. Lipshutz, a chemistry professor at the University of California, Santa Barbara, who applies organozirconium chemistry to synthesis. However, he cautions that zirconocene hydrochloride has a short shelf life and that first-time users might conclude that hydrozirconation does not work when it could be that the reagent they are using has deteriorated. For this reason, he and others have come up with ways to generate the compound in situ.

With hydrozirconation as the starting point, organozirconium chemistry has taken varied paths. One major area is use of organozirconium derivatives in transition-metal-catalyzed cross-couplings. These reactions offer the broadest synthetic methodology for forming carbon-carbon bonds. They involve the transformation R¹–M + R²–X →R¹–R², in which M is a metal complex and X is a leaving group, usually a halogen. The usual catalysts are palladium and nickel.

"Any big molecule you want to prepare, you can chop into two pieces," explained Ei-ichi Negishi, an organozirconium chemist at Purdue University. "Put metal on one piece and X on the other, and you can expect them to couple." Negishi is a major developer of cross-coupling technologies and was a co-organizer of the symposium.

One of Negishi's earliest uses of organozirconium chemistry in carbon-carbon bond formation involves hydrozirconation of an alkyne to form an alkenylzirconium. In a nickel-catalyzed reaction, the alkenylzirconium reacts with an aryl halide to form an alkenyl-aryl bond. This chemistry has evolved dramatically, and what is now known as the Negishi coupling comprises palladium- or nickel-catalyzed reactions between organometals containing aluminum, zinc, or zirconium and an organic halide to form a carbon-carbon bond.

THE THEME in cross-couplings is transmetalation--that is, the handing over of an organic group from one metal to another. The ability to do so greatly enhances the versatility of organometallic intermediates. For example, the carbon-zirconium bond has limited carbon-carbon bond-forming reactivity, but it can be used directly to make other carbon-metal bonds. So in palladium-catalyzed organozirconium coupling, palladium takes the place of zirconium. Because the carbon-palladium bond has good carbon-carbon bond-forming reactivity, further reaction is possible.

The chemistry is tremendously enabling, according to Peter Wipf, a chemistry professor at the University of Pittsburgh and a co-organizer of the symposium. The starting materials are alkenes and alkynes, which are readily available and easy to manipulate. Carbon-carbon bonds can be formed between highly functionalized systems under mild conditions, he noted. His lab has used hydrozirconation and transmetalation with dialkylzinc to prepare, among others, allylic amines and alcohols, cyclopropanes and bicyclobutanes, and peptide mimetics.

In hydrozirconation, the organozirconium compound is a stoichiometric reagent. Catalytic applications soon followed.

One of the earliest applications came from Negishi. In 1978, he showed that zirconocene dichloride catalyzes the stereoselective addition of carbon and aluminum across a triple bond--a carboalumination reaction. This reaction has become the method of choice to prepare (E)-methyl-substituted alkenes and has been applied to the syntheses of more than 100 complex natural products, according to Negishi. For example, this chemistry is key to the concise, commercially attractive synthesis of the dietary supplement coenzyme Q₁₀ developed by the Lipshutz group (C&EN, Nov. 18, 2002, page 60).

When applied to alkenes, carboalumination has the potential to create a new stereogenic center. Unfortunately, Negishi's original carboalumination gives low yields with alkenes. However, using a chiral organozirconium compound--dichlorobis(1-neomenthylindenyl)zirconium--the Negishi lab achieved enantioselective carboalumination of simple alkenes in 1995.

"I'm very proud of this chemistry," Negishi told C&EN. The reaction is unique in its ability to form carbon-methyl bonds directly and enantioselectively, he explained. Methyl groups are critical because methyl pendants are found much more frequently in naturally occurring compounds than ethyl or longer alkyl groups, he added.

ONE DRAWBACK is the high cost of the indenyl ligand, but the cost might be lowered by improving the synthesis. Another problem is the low turnover number of the catalyst. "It has generally been more difficult to develop catalytic processes of high turnover numbers with early transition metals such as zirconium compared with late transition metals such as ruthenium and palladium," Negishi noted.

Novel chemistries have sprung from other zirconium complexes. Particularly in the 1980s, several groups developed protocols for generating -complexes of zirconocene and examined their reactivities. Among the reagents that emerged are dibutyl- and diethylzirconocene. According to Negishi, these, among their other uses, serve as precursors to three- and five-membered -complexed zirconacycles, which are the basis of other practical reactions. Recently, for example, Tamotsu Takahashi, at Hokkaido University, in Japan, showed that highly substituted pyridines can be made easily from zirconacyclopentadienes and nitriles.

At the University of Rostock, in Germany, Uwe Rosenthal continues to explore the reactivity of Cp₂Zr(pyridine)[(CH₃)₃SiC₂Si(CH₃)₃]. Named after Rosenthal, this commercially available complex is an excellent source of the highly reactive zirconocene fragment. With diynes, for example, the zirconocene fragment forms unusual structures. T. Don Tilley at the University of California, Berkeley, has used the Rosenthal complex to make functionalized macrocycles with potential for host-guest chemistry and supramolecular chemistry.

At the University of Münster, Germany, Gerhard Erker has been studying zirconium complexes with modified cyclopentadienyl ligands. For example, he first prepared the chiral catalyst for Negishi's enantioselective carboalumination. Recently, his group found that the closely related organozirconium complexes called bis(2-alkenylindenyl)zirconium dichlorides rapidly undergo photochemical cycloaddition to form an unusual cyclobutylene-bridged metallocene framework. The products show catalytic activity for ethene/1-octene copolymerizations.

Many other reactions of organozirconium compounds are now being investigated. Some of the areas of research that were discussed at the symposium include the chemistry of zirconium-heteroatom multiple bonds, alkane activation by -bond metathesis, polar activation of zirconocene derivatives, and migratory insertion chemistry.

In more than 30 years of doing chemistry, Schwartz commented in an interview, his research interests have changed, yet he finds himself going back to zirconium again and again. At present, for example, his group is using organozirconium compounds to attach peptides to surfaces of implant materials. Organozirconium chemistry can be applied in such varied ways, he explained, because of the unique characteristics of zirconium: It resists reduction, it does not mind being coordinatively unsaturated, and it forms strong bonds to oxygen. The last, he added, is key to using organozirconium compounds as linkers to surfaces.

Organozirconium compounds are easy to use, especially in multistep, one-pot reactions, according to Wipf. "You don't need glove boxes and specialized inorganic techniques to handle them." Another advantage is that zirconium is nontoxic, unlike later transition metals.

To date, organozirconium chemistry has branched into applications that Birmingham could not have imagined when he first prepared zirconocene dichloride 50 years ago. Yet researchers still face many challenges. For example, the effect of ligands on the reactivity of zirconium is still poorly understood, according to Wipf. "The mechanistic questions are complex and have not been fully answered," he added.

For now, the pioneers can savor the ordinariness that organozirconium chemistry has achieved. "When I started working in this area, it was considered exotic," Schwartz said. "Now, if you read the organic synthesis literature, it is not unusual to come across organozirconium chemistry as a standard step. You don't take note of it because it is so common. What could be better than that? To me, that's the best thing you could hope for. People use it; it works."