Metal Complexes of Stable Carbenes*

Wolfgang A. Herrmann    Anorganisch-chemisches Institut der Technischen Universität München, D-85747 Garching bei München, Germany

Thomas Weskamp    Symyx Technologies, Santa Clara, California 95051

Volker P.W. Böhm    Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3290

I INTRODUCTION

A Historic Background

Carbenes—molecules with a neutral dicoordinate carbon atom—have played an important role in organic chemistry ever since their first firm evidence of existence. However, despite the increasing interest in persistent intermediates since the days of Gomberg1,2 and despite the fact that carbenes were introduced into organic chemistry by Doering and Hoffmann in the 1950s3 and into organometallic chemistry by Fischer and Maasböl in the 1960s,4 it was only in the late 1980s and early 1990s that the first carbenes were isolated [Eq. (1)].5–8

  (1)

This discovery resulted in a revival of carbene chemistry, surprisingly more in organometallic chemistry than in organic chemistry.9,10 One explanation for the interest of organometallic chemists in stable free carbenes might be the fact that the metal complexes1 and 2 containing the subsequently isolated N-heterocyclic carbenes were prepared as early as in 1968 by Wanzlick and Schönherr and by Öfele.11,12

That was only 4 years after the preparation of the first Fischer-type carbene complex 3,4 6 years before the first Schrock-type carbene complex 4 was reported,13 and more than 20 years before the isolation of stable imidazolin-2-ylidenes by Arduengo in 1991 [Eq. (1)].7 Once attached to a metal, these Wanzlick- or Arduengo-carbenes have shown a reaction pattern completely different from that of the electrophilic Fischer- and nucleophilic Schrock-type carbene complexes.

This article presents the principles known so far for the synthesis of metal complexes containing stable carbenes, including the preparation of the relevant carbene precursors. The use of some of these compounds in transition-metal-catalyzed reactions is discussed mainly for ruthenium-catalyzed olefin metathesis and palladium-/nickel-catalyzed coupling reactions of aryl halides, but other reactions will be touched upon as well. Chapters about the properties of metal–carbene complexes, their applications in materials science and medicinal chemistry, and their role in bioinorganic chemistry round the survey off. The focus of this review is on N-heterocyclic carbenes, in the following abbreviated as NHC and NHCs, respectively.

B Stable Carbenes

Wanzlick et al. realized in the 1960s that the stability of carbenes should be increased by a special substitution pattern of the disubstituted carbon atom.14–16 Substituents in the vicinal position that provide π-donor/σ-acceptor character to “fill” the p-orbital of the carbene carbon and stabilize the carbene lone pair by a negative inductive effect should reduce the electrophilicity of the singlet carbene and consequently reduce its reactivity (Scheme 1).

Based on this concept and the development of appropriate synthetic methods, many heteroatom-substituted carbenes have been isolated since the first successful attempts by Igau et al.5 and by Arduengo et al.7 The stability of carbenes was originally considered to be limited to cyclic diaminocarbenes (nitrogen provides good π-donor/σ-acceptor character) with steric bulk to prevent dimerization17 and some aromatic character.18 This holds true for imidazolin-2-ylidenes as well as for 1,2,4-triazolin-5-ylidenes. For this family of stable carbenes, many examples have been isolated so far, among them 5–13.9,19–2112 was even reported to be air stable.22 Steric hindrance at the nitrogen substituents does not solely determine whether a carbene can be isolated: the 1,3-dimethylimidazolin-2-ylidene 5 can be distilled for its purification without significant decomposition.23 However, steric parameters certainly influence the long-term stability of NHCs.24,25

Later, imidazolidin-2-ylidenes such as 14, a “saturated,” more electron-rich and nonaromatic version of the imidazolin-2-ylidenes, were isolated.25,26 Isolation of a six-membered tetrahydropyrimid-2-ylidene 1527,28 and of acyclic structures such as 1629,30 was a consequent extension since these compounds still possess two nitrogens vicinal to the carbene carbon, but lack the 6π -electron conjugation.

For all these compounds, the carbene carbon has two nitrogen substituents, which is in complete agreement with the consideration that strong π-donor substituents are an essential requirement for stable carbenes. However, one weaker π -donor substituent, e.g., an alkoxy or alkylsulfido group, can be tolerated as was demonstrated for 17–19.31

C Precursors for Stable Carbenes

1 Precursors for NHCs with Unsaturated Backbone (Imidazolin-2-ylidenes and Benzimidazolin-2-ylidenes)

In many cases the synthesis of NHC complexes starts from N,N'-disubstituted azolium salts. Imidazolium salts as precursors for imidazolin-2-ylidenes are generally accessible by two ways complementing each other: (i) nucleophilic substitution at the imidazole heterocycle or (ii) a multicomponent reaction building up the heterocycle with the appropriate substituents in a one-pot reaction.

Imidazolium salts that can be prepared by the first procedure, the alkylation of imidazole, are easy to obtain and often used for metal complex synthesis. Potassium imidazolide is reacted with the first equivalent of alkyl halide in toluene to give the 1-alkylimidazole.32 Subsequent alkylation in 3-position is achieved by addition of another equivalent of alkyl halide [Eq. (2)].33–35 A variant of this approach employs commercially available N-trimethylsilyl imidazole with 2 equiv of an alkyl chloride, under elimination of volatile Me3SiCl.36 The drawback of these simple routes is the fact that only primary alkyl halides can be reacted in satisfactory yields because secondary and tertiary alkyl halides give substantial amounts of elimination by-products.

  (2)

In order to introduce other substituents at the 1- and 3-positions of the imidazolium salt the reaction of primary amines with glyoxal and formaldehyde in the presence of acid can be used [Eq. (3)].20,37,38 Variation of the amine allows the preparation of imidazolium salt libraries which can be diversified by using different acids in order to change the anion of the imidazolium salt.39 The use of chiral amines in this reaction results in the convenient generation of C2-symmetric imidazolium salts.21 It is possible to generate imidazolium salts with anilines that do not bear a para-substituent in a two-step sequence: synthesis of the bisimine in the first step and subsequent ring closure with formaldehyde and an acid.40,41

  (3)

A method by Gridnev and Mihaltseva allows the combination of both strategies: (i) synthesis of the 1-alkylimidazole by a multicomponent reaction starting from glyoxal, formaldehyde, a primary amine and ammonium chloride, and (ii) subsequent alkylation by a primary alkyl halide to give the imidazolium salt [Eq. (4)] 42

  (4)

Direct coupling of imidazole with aryl iodides in the presence of copper(I) triflate results in 1-aryl-imidazoles, which can be alkylated in a second step [Eq. (5)]. This route represents a variation of the Gridnev method.43

  (5)

The abstraction of a hydride is an additional route for the preparation of benzimidazolium salts: Treatment of 2,3-dihydro-1 H-benzimidazoles with tritylium tetrafluoroborate generates the benzimidazolium salt and triphenylmethane [Eq. (6)].44

  (6)

The reaction of N-alkyl-N-formyl hydrazines with imidoyl chlorides gives 3,4-substituted 1-alkyl-4H-1,2,4-triazolium salts in a one-pot reaction.45

2 Precursors for NHCs with Saturated Backbone (Imidazolidin-2-ylidenes)

The reaction of an ortho-ester, e.g., HC(OEt)3, with a secondary bisamine in the presence of an ammonium salt yields...