вход по аккаунту


N-Heterocyclic Carbenes.

код для вставкиСкачать
A downhill energy pathway
leads from methylene to isolable heterocyclic carbenes
such as 1,3,4-triphenyI-4,5dihydro-1H-l,2,4-triazol-5-ylidene and 1,3-di-R-dihydroimidazolylidene.
These heterocyclic carbenes
form complexes or covalent
compounds with the elements
marked in the periodic table.
Several of the metal complexes can be successfully used
as homogeneous catalysts,
as indicated by the schematic
representation of the energy
Mn Fe
Ga Ge As
Mo Tc Ru Rh Pd Ag Cd
Pt Au Hg
Cu Zn
Hf Ta
Re 0 s
Mt Unn Unu
Nd Pm Sm Eu
Tm Yb
Pu Am Cm Bk
Es Fm
Md No
N-Heterocyclic Carbenes**
Wolfgang A. Herrmann" and Christian Kocher
Dedicated to Professor Heribert Offermanns on the occasion of his 60th birthday
The chemistry of N-heterocyclic carbenes has long been limited to metal coordination compounds derived from
azolium precursors, a development that
was started by Ofele and Wanzlick in
1968. Since free carbenes are now available through the work of Arduengo
(1991), a renaissance in this little-recognized area of chemistry has occurred. A
leading motive is the advantage of Nheterocyclic carbenes as ligands in
organometallic catalysts, where they
extend the scope of application reached
by phosphanes (functionalized, chiral,
water-soluble, and immobilized derivatives). The present review summarizes
the state of the art with regard to synthesis, structure, bonding theory, metal coordination chemistry, and catalysis.
Chelating, functionalized, chiral, and
immobilized ligands can be generated
and attached to metal centers in
straightforward procedures under mild
conditions. A wealth of new chemistry is
1. Introduction
Carbenes have played an important role in organic chemistry
ever since the first firm evidence of their existence. While much
of the early chemistry of these laboratory curiosities was established in the 1950s by Skell,['] it was Fischer and his students
who introduced carbenes into inorganic and organometallic
chemistry in 1964.12]Suffice it to say that metal carbenes have
become quite significant, particularly in organic synthesis, catalysis, and macromolecular c h e m i ~ t r y-.6~1 ~Shortly after the discovery of the first authentic metal-carbene complexes of type 1,
two little-recognized reports by Ofele (Technische Universitat
Miinchen)['] and Wanzlick et al. (Technische Universitat
Prof. Dr. W. A . Heirmann, Dr. C. Kocher
Anorganisch-chemisches Institut der Technischen Universitat Miinchen
Lichtenbergstrasse 4, D-85748 Garching (Germany)
Fax: Int. code +(89)2891-3473
e-mail : herrmannia zaphod.anorg
Essays on Organoinetallic Chemistry, Part 9. Part 8 : W. A. Herrmann, B
Cornils. A n p i c . Ckcm. 1997, 109, 1074; Angen Chem. hi.Ed. Engl. 1997, 36,
Angen. Chem. Int. Ed. Engi. 1997,36,2162-2187
thus opened. It is also shown how
carbenes derived from imidazoles and
triazoles behave as ligands in catalysis.
It is reasonable to assume that N-heterocyclic carbenes surpass the ubiquitous
phosphanes as ligands in a number of
organometallic catalytic reactions.
Keywords: carbenes heterocycles * homogeneous catalysis - metal complexes
Berlin)['] appeared that described complexes 2 and 3 with Nheterocyclic carbenes as ligands. These complexes, which at that
time were highly unusual compounds, were both obtained from
imidazolium salts and metal-containing precursors of sufficient
basicity to deprotonate the organic substrate. In the first case a
carbonylmetalate was used, while the mercury complex resulted
simply from the metal acetate [Eqs. (I), (2)].
Ofele and his students established several basic features of this
chemistry before Arduengo opened the access to free, isolable
According to Equation (3),
N-heterocyclic carbenes in 1991
NaH. cat. DMSO anion
'- -
& i ]
Q WILEY-VCH Verlag GmbH. D-69451 Weinheim. 1997
* [>c:
0570-083319713620-2163 $17.50+ .50/0
W. A. Herrmann and C. Kocher
deprotonation of the imidazolium ion is effected by NaH and
catalytic amounts of KOtBu or the DMSO anion. Although this
methodology has its limitations with regard to substituent R,
Arduengo's work proved that a carbene of type 4 is in principle
thermodynamically stable. At roughly the same time we synthesized the crystalline divalent germanium(I1) congener 5 from
diimine tBuN=CH-CH=NtBu, lithium, and germanium(I1)
chloride."" As a result of this discovery Denk et al. and Lappert et al. successfully synthesized the stable silylenes 6 by reduction of the corresponding silicon(1v)cyclodiamides with potassiurn.["] Compound 5 proved to be a successful precursor for
chemical vapor deposition (CVD) of amorphic germanium,[134-
synthetic chemistry.['. 16* 1 7 ] Carbenes are uncharged compounds with a divalent carbon atom and two unshared electrons, which can be assigned to two nonbonding orbitals in
different ways. For this reason, the chemistry of carbenes, nitrenes, and arynes is similar in many respects.
1.1.1. Hybridization and Structure
Depending on spin multiplicity and ground state, the (electronic) structure of a given carbene varies as follows: If the
carbene has a nonlinear structure with an sp2-hybridized carbon
atom, the structures of the singlet ground state and the triplet
state compare to that of carbenium ions (CR,)' and free radicals 'CR,, respectively. For the triplet state and excited singlet
states linear structures with sp-hybridized carbon atoms also
need to be considered."'] A nonlinear triplet ground state is
experimentally established for the majority of carbenes; singletstate dihalocarbenes and carbenes with oxygen, sulfur, or nitrogen substituents are exceptions.
1.1.2. Triplet Carbenes
Based on earlier work in the Munich laboratory,['2a-el we
suggested that N-heterocyclic carbenes 4 would exhibit ligand
properties similar to those known from electron-rich phosphanes. This led to the conclusion that an equally rich coordination and catalysis chemistry should arise from these pronouncedly C-nucleophilic compounds.['31 Experimental evidence has increasingly been keeping track with our predictions.
1.1. The Way to Isolable Carbenes
There has been continued and growing interest in persistent
intermediates since the days of G ~ m b e r g . ~ ' ~ .However, almost no efforts were made to isolate carbenes, and stable
"ylidene carbenes" were only recently isolated. Carbene intermediates such as dihalocarbenes are now well established in
Electrophilic carbenes :CR'R2 with substituents R' and RZ
that are not electron donors were originally regarded as being
too reactive to be isolated. As these triplet carbenes cannot be
stabilized through thermodynamic effects, they have to be rendered unreactive by steric protection. Considerable progress
was made by Tomioka et al., who stabilized diphenylcarbenes
with methyl or chloro substituents in ortho positions. Whereas
laser flash photolysis (LFP) of pheny1(2,4,6tris-tert-butylpheny1)diazomethane yielded a
nascent singlet carbene which intramolecuQBr
larly inserted into a tert-butyl C-H bond beBr
fore undergoing intersystem crossing, LFP
in the presence of the triplet sensitizer benzophenone produced a transient triplet
~arbene.['~IIn 1995 they synthesized
2,2',4,4',6,6'-hexabromodiphenylcarbene (7),
Wolfgang A . Herrmann, born in 1948, recieved his organometallic
academic education with E. 0. Fischer (Munich, Germany), H.
Brunner (Regensburg, Germany), and P. S. Skell (Pennsylvania,
U S A ) before becoming a professor at Regensburg (1979) and
Frankfurt (1982). In 1985 he was appointed the chair professor of
inorganic chemistry at the Technische Universitat Miinchen as the
successor to E. 0. Fischer. His research focuses on homogeneous
catalysis with a strong emphasis on synthesis, spectroscopy, and
structures. He has authored about 450 scientific publications, 38
patents, and a number of review articles. He holds honorary doctoral
degrees and has received numerous scientific awards. In 1996 he
became president of the Technische Universitat Miinchen.
W. A. Herrmann
C. Kocher
Christian Kucher, born in 1968 in Munich, studied chemistry at the Technische Universitat Miinchen. For his Diplomarbeil in
the group of W A . Herrmann he investigatedfunctionalizedderivatives of C,, . This year hefinished his dissertation "New Ways
to N-Heterocyclic Carbenes and their Transition Metal Complexes-Applications in Homogeneous Catalysis". He is coauthor
of eight publications and five international patents.
Angew. Chem.
In!. Ed. Engl. 1997, 36, 2162-2187
N-Heterocyclic Carbenes
triplet carbene (ESR spectroscopy), again by photolysis of the
parent diazo compound. In glassy (77 K) and completely fluid
(120 K) 2-methyltetrahydrofuran, 7 is stable without significant
decomposition. The crystalline compound can be kept at room
tempera ture.[201
1.1.3. Singlet Carbenes: Heteroatoms at the Divalent Carbon
Heteroatom donor groups on a carbene center render the
originally degenerate orbitals on carbon unequal in energy, thus
enhancing the nucleophilicity of the carbon atom and the thermodynamic stability. The singlet- triplet splitting correlates
with increasing electronegativity of the n-donor substituents X
and Y in carbenes of the type :CXY. Although several combinations of heteroatolns (0,S, N) are conceivable, only singlet
carbenes with two nitrogen atoms (amino groups) were isolated
as crystalline compounds so far (“carbon diamides”). These
singlet carbenes have a pronounced low-energy HOMO and a
high-energy LUM0.[21]Because of the lower electronegativity
of carbon, they are stronger electron-pair donors than amines.
The electron-accepting capability is more significant than that
of boranes (electronegativity!). Since the amino groups are ndonating (mesomeric) and o-withdrawing (inductive), 2,3-dihydro-1H-imidazol-2-ylidenes benefit from a “push-pull” effect.[22,2 3 1 Although the methoxy and trifluoromethyl groups
also display push- pull stabilization, methoxy(trifluoromethy1)carbene, CH,O(CF,)C:, is unstable at room temperature and
electronically indiscriminant upon reaction with different alk e n e ~ . ‘ ~Dialkoxycarbenes
could be generated in situ by
thermolysis of, for example, 2,2-dialkoxy-d ,-l ,3,4-oxadiazolines,[251and amino(ch1oro)carbenes by deprotonation of
Vilsmeier reagents such as chloromethylene iminium salts;[261
however, attempts at their isolation resulted in dimeric or
oligomeric producrs.
Therefore, the extent of “stability” and “aromaticity” (for
6n-electron systems) of free carbenes and their homologues depends on the point of view. For preparative chemistry kinetic
stability is crucial. Stable N-heterocyclic carbenes are being
studied for several reasons. Structure, reactivity, and theoretical
understanding of these highly Lewis basic (one of the strongest
known neutral bases) and nucleophilic molecules are of interest
per se. Moreover, stable ylidene carbenes are used to prepare
main group and transition metal complexes, although there are
several in situ methods for synthesizing metal ylidene complexes
without having to isolate free carbenes or their equivalents.
1.1.4. N-Heterocyclic Singlet Carbenes
In the early 1960s Wanzlick et al. investigated saturated and
unsaturated heterocyclic carbene~.[~’]
The focus was first put on
C-C-saturated molecules of type 9. Imidazoline 8, one possible
precursor of 9, was obtained from the reaction of 1,2-dianilinoethane with chloral; it underwent thermal cr-elimination of
chloroform. The resulting product dimerized, however, to yield
the electron-rich olefin 8a as a colorless crystalline compound
[Eq. (4)]. Reactions with oxygen, water, nitromethane, and
cyclopentanone confirmed the nucleophilic properties of what
was incorrectly described as a “monomeric carbene”. CrossAngew. Chem Inr. Ed. Engl 1997, 36, 2162-2187
coupling experiments with differently substituted “dimers”
showed that they are not in equilibrium with their monomers.
Most notably, Lappert et al. developed a general synthesis of
metal-carbene complexes by treatment of electron-rich (Wanzlick) olefins with transition metal complexes such as chlorotris(triphenylphosphane)rhodium(r) . It is interesting that this
particular rhodium complex catalyzes the metathetical cleavage
of electron-rich olefins.
Numerous five- and six-membered heterocyclic cations display CH-acidic properties and base-catalyzed H - D exchange.
The N-methylbenzthiazolium cation was most easily deprotonated by triethylamine in N,N-dimethylformamide or acetone,
as shown by Wanzlick, Vorsanger, and Hiinig. Once again, only
the electron-rich olefin could be isolated in place of the free
carbene,cz8]but numerous transition metal complexes have become known since.
Wanzlick et al. recognized that aromatic resonance structures in unsaturated N-heterocyclic five-membered rings contribute to carbene stability. Free carbenes, such as 1,3,4,5tetraphenyl-2,3-dihydro-lH-imidazol-2-ylidene
(lo), were generated by deprotonation of imidazolium salts with KOtBu
[Eq. (5)] and then allowed to react with isothiocyanates or
metal-containing p r e c ~ r s o r s . 301
[ ~ ~Although
Wanzlick et al.
were so close to carbenes, they lacked the luck needed to isolate
them.[311Gleiter and Hoffmann pointed out that the ground
state of a carbene will be a singlet state if the energy difference
between the 5 and x orbitals is greater than 2 eV, and that this
splitting may be achieved by interaction of the unoccupied carbene p orbital with an aromatic system of (4n + 2) n electrons. Synthesis and NMR Spectroscopy
In 1895 Nef announced that the isolation of methylene would
be his next challenge.[321After he and many others failed, however, no one believed any longer that carbenes would ever be
isolated. Only the careful investigations by Arduengo et al.
yielded the surprise of stable, crystalline (ylidic) carbenes almost
one hundred years later.1331For example, 1,3-diadamantyl-2,3dihydro-1H-imidazol-2-ylidene (11 ; Figure 1) was obtained as
W. A. Herrmann and C. Kocher
= 7.2
6(C)= 136
extended n delocalization
weaker n delocalization
Figure 1. Geometric and spectroscopic parameters of the first isolated free carbene
11 and the corresponding imidazolium ion.
a colorless, crystalline solid by deprotonation of 1,3-diadamantylimidazolium iodide with NaH in the presence of catalytic amounts of DMSO anion [see Eq. (3)].
The thermal stability of solid 11 is surprising (melting point
240-241 "C without decomposition). In contrast, decomposition occurs rather quickly at room temperature in solution. To
obtain reasonable rates for the deprotonation, catalytic
amounts of either KOtBu or the DMSO anion have to be applied, since sodium hydride and potassium hydride (and in most
cases also the imidazolium salt) are insoluble in T H E Alternatively, stoichiometric amounts of KOtBu may be used, but the
less volatile tevt-butyl alcohol is generated in place of hydrogen.
Major progress in the synthesis was achieved with liquid ammonia as solvent: Deprotonation with sodium hydride or potassium amide occurred quickly and quantitatively in a homogeneous phase, and provided carbenes that were not previously
accessible, such as oxygen-, nitrogen-, and diarylalkylphosphino-functionalized or chiral c a r b e n e ~ . ~ ~ ~ ]
Kuhn et al. reported a versatile two-step approach to therTo
mally stable, alkyl-substituted N-heterocyclic ~arbenes.[~']
give an example, reducing 1,3,4,5-tetramethylimidazol-2(3H ) thion with potassium in boiling THF yielded analytically
pure 1,3,4,5-tetramethyl-2,3-dihydro-lH-imidazol-2-ylidene
Eq. (6)] after filtration and removal of the solvent under vacuum.
The thiones were accessible through condensation of N,Ndialkylthioureas with 3-hydroxy-2-butanone. 1,3,4,5-Tetramethylimidazol-2(3H)-thione was obtained in a stoichiometric
mixture with the pentamethylimidazolium ion when 1,3,4,5tetramethyl-2-(methylthio)imidazolium iodide was reduced
with potassium. The only drawback to this technique is the
thermal lability of 2,3-dihydro-I H-imidazol-2-ylidenes in solution, which is dependant on the nature of the substituents R on
nitrogen. The same problems are encountered upon generation
of free carbenes under "Arduengo conditions".
1,3-Disubstituted ylidenes with aryl residues fp-tolyl, p chlorophenyl, and 2,4,6-trimethylphenyl; m.p. 157 (decomp),
154 (decomp), and 153 "C (decomp), respectively) or higher
alkyl groups (tert-butyl, cyclohexyl, adamantyl) and 1,3,4,5-tetramethyl-2,3-dihydro-lH-imidazol-2-ylidene
(m.p. 109 "C) are
colorless crystalline solids; 1,3-dimethy1-2,3-dihydro-l
H-imidaUnder an inert-gas atmosphere
zol-2-ylidene is an oily
they can be kept in the solid state for months at - 30 "C. 1,3-Dicyclohexyl-2,3-dihydro-lH-imidazol-2-ylidene
does not fully
decompose upon heating at 160 "C for 24 h. In solution they are
stable to approximately 50 "C for several hours; this is even
valid for 1,l'-( 1,2-ethylene)-3,3'-dimethylbis(2,3-dihydro-l
Himidaz01)-2-ylidene.[~~IAS long as the solid carbenes or
their solutions are colorless, they are very pure and contain no decomposition products. (Slightly colored solutions, from yellow to orange, are in most cases NMR spectroscopically pure.)
It was clear from the constitution of the stable, sterically
unencumbered ylidenes named above that not kinetic (steric
hindrance) but rather thermodynamic stabilization (electronic
structure) is the essential feature. Two electronic effects deserve
consideration. Firstly, the difference in electronegativity of nitrogen and carbon accounts for the (- I) inductive (3 effect,
which stabilizes the pair of unshared electrons in the in-plane
carbene orbital. Secondly, the unoccupied p orbital yields a
x-resonance interaction; in the ylidic resonance structures the
nitrogen atoms donate their lone pairs to carbon (+ M effect).
Steric hindrance quite certainly contributes to stability, but
the electronic situation in the N-C-N moiety seems to be
Imidazolium salts, which are promising solvents for twophase catalysis,[38-401 are electron-rich heteroaromatic compounds. Hence, the I3C NMR shift of the N-C-N sp2 carbon
(which later becomes the carbene center) appears at 6 z 137.,The
H4(H5) ring protons and the CH-acidic H2 proton resonate at
6 =7.9-7.3 and 8-10, respectively (Figure 1).
Free 2,3-dihydro-l H-imidazol-2-ylidenes are also diamagnetic. The divalent carbon atoms resonates at rather low field
(6 = 210-220 in [DJTHF). The substituents on nitrogen
exert only little influence(!) on the chemical shift. The H4(H5)
ring protons typically resonate at 6 = 6.9-7.0. However,
in 2,3-dihydro-lH-irnidazol-2-ylideneswith aryl residues
adopting a roughly coplanar arrangement (no ortho substituents!), anisotropic effects of the aryl substituents account
for chemical shifts at 6 =7.6-7.8.[361 Deprotonation of the
imidazolium salts to the corresponding 2,3-dihydro-I H-imidazol-2-ylidenes entails a significant high-field shift of the
H4(H5) proton resonances in the case of N-alkyl substitution.
This indicates a different magnetic anisotropy of the ring system, which is consistent with a decrease in 6x-electron delocalizati~n.[~'] Isoelectronic Congeners
x Backdonation also contributes to the stability of isoelectronic stable nitrenium and phosphenium ions LN+ and LP+
(L = N,N-di-tert-butyl-I ,4-diaza-2-butene or -butane) .I4',431
Isoelectronic Group 13 congeners, for example N-heterocyclic
gallium and aluminum complexes containing the fragments
Angew. Chem. Int. Ed. Eizgl. 1991,36,2162-2187
N-Heterocyclic Carbenes
NR,Ga(R)NR, and NR,Al(R)NR,, were experimentally verified with 1,4-diazabutadiene (DAB) l i g a n d ~ . [ ~ ~They
. ~ ' ] are either stabilized by intramolecular coordination of a hemilabile
dimethylamino side group or by dimerization. Alkyl- and arylsubstituted diyls :E[CH(SiMe,),], and :E(2,4,6-tBu3C,H,),
(E = Ge and Sn), which are monomeric in solution, were studied earlier ;[46.471 in the meantime numerous compounds are
known.['411Stability and resistance to dimerization or insertion
reactions once again reach a maximum for N-heterocyclic ring
systems containing the DAB fragment and a Group 13-15 element,"O. 1'1 Solid-state Structures
1.2. Monomer or Dimer? The Singlet-Triplet Gap
1.2.1. C - C-Unsaturated Carbenes
After carbon monoxide and isocyanides, unsaturated Nheterocyclic carbenes have the largest singlet- triplet gap (about
85 kcalmol-' according to ab initio calculations) of any divalent carbon compounds." 'I They show no tendency to dimerize,
which is in contrast to their C-C-saturated analogues. Despite
the predicted (thermodynamic) lability for tetraazafulvalene,
the formal dimer 14 of 2,3-dihydro-lH-imidazol-2-ylidenes
a genuine double bond was isolated by Chen et al. by deprotonation of a doubly trimethylene-bridged diimidazolium salt
[Eq. (S)] and structurally
The bond energy was
The solid-state structures of 2,3-dihydro-1H-imidazol-2-yli- .
denes and representatives of C-C-saturated and acyclic car@CH
@-c@ N
benes were extensively studied by X-ray and neutron diffraction
techniques. In the case of 2,3-dihydro-lH-imidazol-2-ylidenes,
the N-C-N angles at the divalent carbon atom are all within a
narrow range (k1"). For ylidenes with methyl, p-tolyl, or
adamantyl substituents R'(R3) on nitrogen, the carbenic angles
are 101.5(1), 101 2(1), and 102.2(2)", r e ~ p e c t i v e l y . [ ~ ~ * ~ ~ ' ~ ~ ]
These data are in excellent agreement with predictions for prototypical singlet ('A') carbenes bearing Ir-donor substituents
(CF, : F-C-F 102").[481These angles are significantly smaller
than the typical range known for imidazolium salts (108.5109.7").[491It was established for 1,3-diadamantyl-2,3-dihydro14
1H-imidazol-2-ylidene that the C2-N'(N3) bonds (1.37 A) are
estimated to be 4 k 3 kcal mol - '. The tetramethylene-bridged
longer than in the imidazolium cation (1.33 A). Considering this
analogue, however, readily converts into the corresponding ditogether with a slight elongation of the N'(N')-C4(C5) bonds,
carbene. Monobridged and untethered imidazolium salts exist
it was concluded that 7~ delocalization is less pronounced
in the monomeric ylidene form after deprotonation and show no
than in imidazolium
in 2,3-dihydro-lH-imidazol-2-ylidenes
tendency to dimerize. Thummel et al. independently reported
that 14 is accessible by bulk electrolysis (- 1.6 V) of the corresponding doubly tethered diimidazolium salt in acet~nitrile.['~~
Doubly tethered benzo- and naphtho-2,3-dihydro-lH-imida1.1.5. Acyclic carbenes
zol-2-ylidene dimers which are reluctant to undergo C-C bond
The success stor!! of stable carbenes culminated in 1996 when
cleavage were also obtained through bulk electrolyis. It has
Alder et al. synthesized bis(diisopropy1amino)carbene(13), the
long been recognized that enetetramines are strong reducing
first stable acylic carbene, by C-deprotonation of N,N,N',N'-teH-imidazolyliagents. For example, 2,2'-dibenzo-2,3-dihydro-l
traisopropylformamidinium chloride with lithium diisopropyldene shows the first ionization potential at about 6 eV and is
amide (LDA) in THF [Eq. (7)].[s01It is more sensitive to oxygen
easily oxidized to the corresponding dicati~n.['~]
1.2.2. C-C-Saturated Carbenes
The singlet- triplet gap becomes smaller upon saturation
of the C-C bond, since five-center six-electron (5c-6e)
a delocalization as a stabilizing factor is no longer possible. As
of saturated carbenes, electron-rich olefins 8 a were used
by Lappert et al. to prepare imidazolidine-carbene metal complexes. In 1995 Arduengo et al. succeeded in synthesizing imidaand moisture than the cyclic congeners. The N-C-N (121.0')
zolidine-2-ylidene 15, a stable and crystalline monomer of the
angle is significantly larger than in 1,3-dimesityl-1H-imidazolelectron-rich olefins [Eq. (9)].I5']
idin-2-ylidene (104.7") and ylidene 11 (102.2'). The barrier to
N-C rotation gives proof of substantial double-bond characI
ter.["' The " C NMR chemical shift of the divalent carbon
atom in 13 is shifted downfield by about 40ppm relative
-KCI. -Hz
to that for the divalant carbon atom in 2,3-dihydro-1HI
k S
imidazol-2-ylidenesand downfield by 10 ppm relative to that for
15 (see below).
Angeu. Chem. Inr. Ed. Er!gl. 1997, 36. 2162-2187
W. A. Herrmann and C. Kocher
ylidene 15, and the
acyclic diaminocarbene 13 were synthesized by deprotonation
of amidinium salts. This showed that the different degree of
aromaticity in the precursor has no first-order effect on deprotonation in this special case. Semiempirical calculations (AM 1
method) suggest that the proton affinities of similarly substituted representatives of all three systems vary by less than 1 YOof
the absolute values.[501
isomerized, probably by an intermolecular pathway, to thiazole
under further irradiation or upon heating to 60 K. Dissociative
electron ionization (EI) of acetyl-2-thiazole yielded the 2,3-dihydrothiazol-2-ylidene radical cation, from which neutral ylidene
17 was generated by neutralization-reionization mass spectroscopy [Eq. (12); Schwarz and Terlouw et al.].[591
1.2.3. Triazole-Derived Carbenes
The first crystalline triazole-derived carbene, 1,3,4-triphenyl4,Sdihydro-I H-l,2,4-triazol-5-yIidene,
was not prepared by deprotonation of the highly protic triazolium salt, because the
latter forms-in
contrast to imidazolium salts-(isolable)
5-methoxytriazoles upon treatment with NaOMe.r5614,5-Dihydro-1 ff-l,2,4-triazol-S-ylidene16 was obtained at 80 "C
and 0.1 mbar by endothermic elimination of methanol from the
corresponding 5-methoxytriazole in the solid state [Eq. (1 0)] .
The bond lengths N'-C and N3-C (1.351(3) and 1.373(4) A)
are much shorter than expected for a single bond. This result
and ab initio calculations indicate strong interaction of the filled
nitrogen 2p orbitals with the unoccupied carbon 2p orbital. As
in the case of imidazole-derived carbenes, no dimers are formed
in solution. Ylidene 16 decomposes unspecifically above 150 "C
(differential scanning calorimetry/thermogravimetry (DSC/
H-l,2,4-triazol-5TG) studies). 1,3,4-Triphenyl-4,5-dihydro-l
ylidene is the first carbene to be commercially available
(ACROS). Cyclic voltammetry and ESR measurements demonstrated that 16 can be reversibly reduced to the corresponding
radical anion.r571Density functional theory (DFT) calculations on the spin density in the most stable structure of the
radical anion suggested that the single electron is delocalized
between the N2-C3 double bond and the adjacent phenyl substituent.
1.2.4. Thiazole-Derived Carbenes
Substituted 2,3-dihydrothiazol-2-ylidenesas well as the unsubstituted parent compound as a thiazole tautomer were investigated experimentally and by quantum chemical calculations only very recently. These compounds are close analogues
of the intermediate involved in enzymatic C-C bond forming
and breaking transformations catalyzed by thiamin (vitamin B,) . The parent species were characterized by matrix-isolation techniques as well as in the gas phase: Prolonged irradiation (254 nm) of thiazol-2-caboxylic acid in an argon matrix
(12 K) yielded unsubstituted 2,3-dihydrothiazol-2-ylidene(17)
with loss of CO, [Maier et al.; Eq. (ll)].[581The analysis of
calculated and measured IR spectra together with I3C labeling
experiments confirmed the formation of 17.C0,. Ylidene 17
The geometric parameters and energies of neutral thiazole,
2,3-dihydrothiazol-2-ylidene,and the corresponding radical
cations were calculated using the hybrid DFT option. Ylidene
17 is 31.5 k c a l m o l ~' less stable than thiazole; the energy barrier
is 72.4 kcalmol-'. The calculated geometry of planar 17
(lA')[59] is in perfect agreement with the X-ray diffraction data
of 3-(2,6-diisopropylphenyl)-2,3-dihydrothiazol-2-y~idene
the first stable crystalline 2,3-dihydrothiazol-2-ylidenesynthesized by Arduengo et al. through deprotonation of the corresponding thiazolium chloride with potassium hydride in THF
[Eq. (13)].[601With 104.2' the angle at the carbene carbon
atom (S-C,,,,,,,-N)
is again smaller than in the azolium ion
(112.0"). As for other singlet carbenes, the bond lengths between
C2 and the heteroatoms are greater than those in azolium salts.
Monomeric 2,3-dihydrothiazol-2-ylideneswith sterically less
demanding groups on nitrogen can be observed in solution at
low temperature but are prone to fast dimerization. With 18
an equilibrium between a stable carbene and its dimer was
observed for the first time in the presence of a protic acid
1.2.5. Functionalized Chiral and Multidentate Carbenes
N-Heterocyclic carbenes form a plethora of metal complexes,
either over carbenes, which are formed by deprotonation of
CH-acidic imidazolium salts, or by in situ deprotonation by
metal-bound basic ligands (e. g. amides, alkoxides, or acetate;
see Section 5). Free carbenes with oxygen, nitrogen, and diarylalkylphosphino donors in the side chain(s) or with chiral
residues were synthesized in our group using the particularly
beneficial liquid-ammonia route (Figure 2) .r34,613621 When the
Angew. Chem. Inr. Ed. Engl. 1997,36, 2162-2187
N-Heterocyclic Carbenes
19: R = Me
Figure 2. Side chain functionallzed, chiral, and multidentate 2,3-dihydro-lHimidazol-2-ylidenes.
imidazolium salts were deprotonated in a mixture of liquid
ammonia and aprotic polar solvents such as THF, the desired
carbene was obtained in exellent yield and purity; in most
cases the reaction went to completion within about 30 min at
-40 "C. Many otherwise inaccessible carbenes have thus been
As two-electron ligands, carbenes are related to ethers,
amines, isonitriles, and phosphanes with regard to their coordination chemistry. Chelating derivatives are of special interest.
Asymmetric precursors are accessible from commercially available starting materials in four-step syntheses with yields of up to
90%.[631Because of the rigid geometry of the planar five-membered rings, the optimum number of atoms forming an appropriate chelate ligartd does not match that found with diphosphanes or diamines. Methylene-bridged 2,3-dihydro-1H-imidazol-2-ylidenes and 1,2,4-1H-triazol-Sylidenes form six-membered rings upon metal complexation. However, the natural bite
angle of 78-79" found in transition metal complexes (W, Pd) of
these ligands is exactly that of 1,l'-bis(dipheny1phosphino)methane (four-membered ring) . [ 6 3 9641 At rhodium centers
this dicarbene exhibits a bridging coordination mode.[651In
ethylene-bridged 2,3-dihydro-IH-imidazol-2-ylidenes an unfavorable eclipsed conformation of the hydrogen atoms in the
ethylene bridge also favors dinuclear structures.[371With an
o-xylylene backbone (C,), however, bidentate ligands well suited for chelating metal coordination on Rh were obtained.[661In
addition to phenyl- and naphthylethyl-substituted carbenes,
other chiral 1,3-di-R-2,3-dihydroimidazol-2-ylidenes
and complexes thereof were prepared in our group. Enantiomerically
pure acetylated I ,l'-bis( 1-hydroxyethyI)ferrocene reacted with
imidazole under retention of configuration at the pseudobenzylic positions. Chiral diimidazolium salts were obtained after
alkylation. Differrocenyl-substituted imidazolium salts were
synthesized from enantiomerically pure I-ferrocenylamine by a
ring-closure synthesis.[671Complexes of chelating oxazoline/
carbene donors are available by a simple general route from
2-amino alcohols.[' 19]
A tripodal carbene ligand of the Trofimenko type, tris(2,3-dihydro-I H-imidazol-2-ylidene)borate,resulted from deprotonation of the corresponding precursor salt with nBuLi/n-hexane in
T H E The homoleptic hexacarbene -iron complex 19 resulted
from subsequent treatment with FeCl, (Figure 3).[681Another
tridentate carbene ligand, [I ,3,5-{tris(3-tert-butyl-2,3-dihydro1H-imidazol-2-ylidene)methy}-2,4,6-trimethylbenzene] (20),
was isolated as a solid. Inter- or intramolecular interactions of
the carbene groups with each other were not
Angeu. Chem Int. Ed. Eiigl. 1997, 36, 2162-2187
Figure 3. Tridentate carbenes as ligands in the hexacarbene complex 19 and as free
carbene 20.
1.3. Nomenclature of Ylidenes
The suffixes of the systematical names of heterocyclic compounds are determined by the kind of heteroatoms in the ring.
The first differentiation is made between a ring system containing at least one nitrogen atom and all other heterocyclic rings.
Five-membered rings containing at least one nitrogen atom
have the suffixes -olidine (saturated ring systems), -oline (ring
containing one double bond), and -ole (ring containing the maximum number of double bonds).
The N-heterocyclic ylidenes first isolated by Arduengo are
formal tautomers of 1H-imidazole a. Protonation of IH-imidazole at the basic N3 position and subsequent deprotonation of
the resulting 1,3-dihydroimidazolium ion b at the C2 center
yields the 1,3-dihydrosubstituted ylidene c. The number of
n electrons (six) remains unchanged. The addendum -ylidene
refers to compounds in which two hydrogen substituents are
replaced by two electrons or one pair of electrons. The parent
compound of c (d) is unequivocally described as 2,3-dihydro1H-imidazole. Thus, c is called 2,3-dihydro-l H-imidazol-2-ylidene. As d represents a ring system with only one double bond,
it may also by called an imidazoline (see above). Although the
W. A. Herrmann and C. Kocher
name imidazoline by itself comprises other tautomers, 1,3-di-Rimidazoline-2-ylidene is unequivocal for compounds of type c.
The names 1,3-di-R-2,3-dihydro-l H-imidazol-2-ylidene and
are synonymous here because of
the substitution pattern of imidazole-derived carbenes with (organic) substituents in positions 1 and 3, and the carbenic center
at the C2 atom. Definitely not correct, however, is the name
N-heterocyclic ylidenes of type e
with a saturated C-C bond are correctly called 1,3-di-R2,3,4,5-tetrahydro-I H-imidazol-2-ylidenes or 1,3-di-R-imidazolidine-2-ylidenes. Triazole-derived carbenes of type f are
1,3,4-tri-(R)-4,5-dihydro-lH-triazol-5-ylidenes,and thiazolederived g 2,3-dihydro-thiazol-2-ylidenes.
2. Electronic Structure of N-Heterocyclic Carbenes,
Silylenes, and Germylenes
2.1. Singlet-Triplet Gap in Methylene
Due to a small HOMO-LUMO gap, the energy difference
between the singlet and triplet states can be very small or even
negative because of the energy gain from increased exchange
interaction in the triplet state. It was experimentally verified that
the parent carbene :CH, has a triplet ground state. This fact has
served as a reference for various ab initio methods, since the
contribution of electron correlation in singlet and triplet states
is significantly different, and an accurate value for the singlettriplet gap can be calculated only by using large basis sets and
extensive configuration interaction (CI) .c7O1
Delocalization in N-Heterocyclic Carbenes
In their first theoretical work Dixon and Arduengo calculated
singlet- triplet splitting and proton affinity of unsubstituted 2,3dihydro-1H-imidazol-2-ylidene by ab initio methods (singlet:
TCSCF, triplet: ROHF).[711With 79.4 kcalmol-I the energy
gap is nearly twice as large as that of :CF, (calcd: 45, found:
z 57 kcal mol- I ) . On the other hand, the proton affinity is estimated to be about 250 kcalmol- making 2,3-dihydro-IH-imidazol-2-ylidenes one of the
strongest organic Brmsted bases
known. Interestingly, this value is
larger then that for :CH,
(205 kcalmol-I), whereas :CF,
has a lower proton affinity
(172 kcalmol-I). It was concluded that a) A is similar to :CF2
and b) ylidic resonance structures do not dominate. An analysis
of the MP2 wavefunction using the localized orbital approach
provided a similar c o n c l ~ s i o n The
: ~ ~ stability
of A is due to
odonation of charge to the nitrogen atoms, whereas
71 backdonation plays only a minor role.
This point of view was reconfirmed.[731The electron distribution of perdeuterated tetramethyl-2,3-dihydro-lH-imidazol-2ylidene was determined experimentally by neutron and X-ray
diffraction. In addition, the electron density was calculated by
DFT methods using a gradient-corrected function. Experimental and theoretical electron distributions matched each other
very well. Based on contour line diagrams of the electron density, the authors reestablished their picture of similarity between
:CF2 and A. The “egg-shaped’’ density at the carbenic carbon
resembles that of :CF, and was interpreted as significant electron deficiency above and below the molecular plane. Accordingly, contour plots 70 pm above the molecular plane showed
separate maxima for the two nitrogen atoms and one for the
C-C double bond, which was interpreted as a lack of
71 delocalization in the N-C-N fragment and a separated double
bond. The authors concluded that the stability of 2,3-dihydro1H-imidazol-2-ylidenes must be due to kinetic hindrance because of electrostatic repulsion between the localized 71 electrons
and an attacking nucleophile.
The above-mentioned theoretical studies judged N-C-N
n delocalization to be of minor importance. However, it is a
necessary prerequisite for an aromatic type of delocalization in
the 671-electron system in A. The appartent difference in the
stability of isolated A and B provoked skepticism of this point
of view.
Avoiding the question of aromaticity, Thiel et al. attributed
the difference between A and B to the larger singlet-triplet
separation in A entailing a higher barrier to d i m e r i ~ a t i o n . ~ ~ ~ ]
However, the reason for the higher singlet-triplet gap in the
unsaturated system A remains unexplained.
2.3. Aromaticity
On the basis of (hypothetical) isostructural and isodesmic
reactions, the relative thermodynamic stability of a given set of
species can be estimated. Sauers therefore verified the influence
of heteroatoms in heterocyclic 6n-electron c a r b e n e ~ . [ The
aromatic stabilization energy (ASE) is the negative of the
isodesmic reaction energy in Equation (14) (including the zero
point energy correction).
Significant stabilization for the 671-electron system can be
achieved with two nitrogen atoms. In contrast, two oxygen
atoms, with higher electronegativity than N and S, afford the
lowest stabilization. This rules out stabilization merely by
o effects, as proposed by C i o s l o w ~ k i .Consequently,
as long
as 6n-electron delocalization plays a role, N-C-N
71 delocalization must also be present.
Two profound theoretical studies by Heinemann et al.[76]as
well as Frenking and Bohme[771came to essentially the same
conclusion. It was demonstrated that, even in systems such as
pyridine or pyrrole, the analysis of electron-density contour
plots 70 pm above the molecular plane exhibit the same features of “separated” 71 electron density above the individual
Therefore, the method of density mapping employed
by Arduengo et al. is not very reliable for analyzing the presence
of electron delocalization.
The isodesmic reaction energies in Equation (1 5) are
very interesting.[761For R = NH, (the plane of the NH, frag-
Angew. Chem. In[.Ed. Engi. 1997, 36, 2162-2181
N-Heterocyclic Carbenes
ments is perpendicular to the carbene plane) the reaction energy
of about 20 kcalmol- demonstrates the stabilization with respect to singlet methylene due to the higher electronegativity of
the nitrogen atoms. However, in a completely planar conformation, where x delocalization from the nitrogen lone pairs into
the “empty” carbene p-type orbital is possible, this stabilization
energy is increased by 70 kcal mol- Thus, x backdonation
dominates carbene stabilization. Addition of a CH,-CH,
bridge increases the stabilization by only 6 kcal mol- whereas
the unsaturated C‘H=CH linkage effects an additional stabilization of 26 kcalmo-’ with respect to the planar aminocarbene.
This demonstrates a significant stabilization effect of the double
bond and indicates the presence of a certain degree of cyclic
electron delocalization. On the other hand, this stabilization is
small compared to that originating from N-C-N x delocalization.
Thermodynamic calculations led to two main conclusions :
a) The major contribution to carbene stabilization is due to
px-p7c electron donation from the two nitrogen atoms to the
carbene carbon atom, and b) the unsaturated cyclic carbene A
is about 20 kcal mol- more stable than the saturated analogue B.
2.4. Heterocyclic Silylenes and Germylenes
Because of their similarity, silylenes and germylenes were
studied in parallel to the carbene by a number of methods mentioned above.[ll,il.73,76-791Arduengo et al. concluded that
the electron-density distribution of silylenes and germylenes differs significantly from that of carbenes, and resembles chelation
of a Group-I4 element by a DAB ligand.[731However, numerous thermodynamic calculations performed in various studies
led directly to the conclusion that also silylenes and germylenes
are stabilized by pn-px d e l o c a l i ~ a t i o n .771
~ ~ On
~ . the other
hand, 6x-electron delocalization was judged to be present but
less pronounced than in carbene A. As an example, an isodesmic
reaction energy affords an increased thermodynamic stability of
about 9 kcal mol- for the unsaturated silylene with respect to
the cyclic saturated analogue (about 20 kcalmol- for the carbene) .r761
2.5. Magnetic Properties
The thermodynamic calculations clearly demonstrated the increased stability of the 6x-electron system. However, this was
only an indication for aromaticity but no proof. A large anisotropy between in-plane and out-of-plane magnetic susceptibility is known to be a good indicator for ring currents, and
therefore a good measure for what is meant by “aromaticity”.
It should be noted here that there are at least two noncompatible
scales of aromaticity.[801
Heinemann et al. as well as Frenking and Bohme calculated
magnetic susceptibilities. They pointed out that the absolute
values of susceptibility are not very accurate, but the relative
parallel and perpendicular components should be reliable.[77]A
reference value of Ax = 50 was given for pyridine. As expected
the 6melectron carhene A is “less aromatic” than pyridine with
Angeu. Chem. hi.Ed. Ecgl. 1997, 36. 2162-2187
an anisotropy Ax of only 28, which is nevertheless much larger
than the value of Ax = 9 for B. The protonated imidazolium
cation has a slightly higher anisotropy of Ax = 31. It is reasonable to conclude that cyclic electron delocalization in A is less
developed as in, for example, benzene or pyridine, as expected
for five-membered heteroaromatic rings.[”] However, there is a
distinct difference between A and B, which suggests a relation
between cyclic electron delocalization and increased thermodynamic stability.
2.6. Chemical Shielding Tensor
In addition to these studies that focus on the electronic structure of 2,3-dihydro-1H-imidazol-2-ylidene, a number of theoretical investigations concerning experimental properties were
performed. The chemical shielding tensor of carbenes was calculated based on the local density functional and Hartree-Fock
and compared with the experimental CPMAS solid-state NMR spectra of 1,3,4,5-tetramethyI-2,3-dihydro-1 H-imidazol-2-ylidene. The large chemical shielding anisotropy at the carbene carbon atom could be reproduced by all
theoretical methods. Because of a large deshielding component
of the tensor in the molecular plane orthogonal to the lone pair,
the resonance for the carbene carbon atom appears at very low
field. This tensor element is dominated by a paramagnetic
(deshielding) contribution arising from a low-lying n + n* transition. This large deshielding paramagnetic contribution is missing in the shielding tensors of the doubly bonded carbon atoms
C4 and C5 of A and also for the protonated carbene carbon Cz
in the imidazolium cation AH’, which indicates the absence of
x delocalization in the N-C-N fragment (carbenic structure).
However, this reasoning is somewhat misleading, since the presence of such a large paramagnetic contribution is not necessarily
due to different x* energy levels. The most significant difference
between the carbenic carbon atom Cz in A and the carbon atoms
C4 and Cs as well as the C 2 center in the imidazolium cation is
the fact that the lone pair is replaced by a significantly stabilized
electron pair forming the C-H bond. The extent of paramagnetic contribution is proportional to the inverse of the energy
difference of the related transition. Therefore, this contribution
will be much larger for carbene carbon atoms, for which the
n + x* transition has a low energy gap, than for other cases for
which the equivalent o + x* transition out of the C-H bond will
have a much larger energy gap and therefore a much smaller
paramagnetic contribution. Exactly this kind of simplified picture was established to explain the difference between the I3C
NMR chemical shielding tensors of benzene and the ips0 carbon
atom of phenyllithium.[s’]
2.7. Photoelectron Spectroscopy
Since photoelectron spectra can shed light on the electronic
structure of molecules, the whole series of isolated, cyclic 68electron carbenes, silylenes, and germylenes was investigated.rzllDensity functional calculations were employed to explain
the experimental data. Based on Koopman’s theorem on the
Kohn- Sham orbitals and using first-order time-dependent per2171
W. A. Herrmann and C. Kocher
turbation theory, the spectra could be simulated very well. The
most important feature is the fact that, in contrast to the Si and
Ge homologues, the lowest-energy ionization process is the removal of an electron from the in-plane lone-pair orbital which
appears as the HOMO in the DFT calculation. The next two
bands correspond to %-typeorbitals of the five-membered ring.
The H F orbital energies for all carbenes, silylenes, and
gennylenes show a n-type HOMO, and that the second highest
moIecular orbital (HOMO - 1) is that for the in-plane lone
pair.1761The question of why the classical Koopman’s theorem
is violated only in the case of the 6%-electroncarbene then appears. In fact, it implies that the errors in neglecting electron
correlation in the HF wavefunction and electronic relaxation
upon ionization cancel out. This apparently does not apply to
carbenes. These deficiencies can be overcome by using correlated ab initio methods to calculate also the energy of the radical
cation: Ionization energies of A were reproduced very well with
respect to both order and absolute value at the PUMP4/
6-31G(d) or CCSD(T)/6-31G(d) level of theory. The analysis
revealed that significant relaxation of the wavefunction occurs
upon removing an electron from the in-plane lone pair, which
results in increased electron delocalization. This is in line with
the increased aromaticity of the cationic imidazolium, for which
charge density of the former lone pair is diminished by “adduct
formation” with a proton. The authors concluded that the inplane electron density acts as a barrier that reduces the extent of
cyclic electron delocalization. Because of the energy gained
upon removing an electron from the lone pair, this ionization
appears to be the lowest-energy ionization process. This fact was
confirmed by the Kohn-Sham orbital energies.
3. Reactivity
confirmed by 2,3-dihydro-IH-imidazol-2-ylidenes,
which do
not undergo electrophilic reactions like insertions or cycloadditions; furthermore, they can be handled in typical donor solvents including THF, liquid ammonia, or acetonitrile. 2,3-Dihydro-I H-imidazol-2-ylidene adducts are also reluctant to add to
nucleophiles. This shows that the PIXLUMO lies rather high in
energy as a result of nitrogen lone pair conjugation, thus decreasing the electrophilic character of the carbene carbon atom.
The stabiiity of 2,3-dihydro-I H-imidazoi-2-yiidenes with respect to dimerization (nucleophilic perpendicular attack of one
singlet carbene at the vacant out-of-plane orbital of another
singlet carbene) suggests that the term “carbene” is misleading.c3’1
3.1. Bertrand’s Carbene
Ab initio calculations indicate that Bertrand’s phosphanylsilylcarbenes [(R,N),P](SiMe,)C: (R = iPr (21), c-C,H,,) and
the more temperature-sensitive phosphanylphosphoniumcarbenes [{(iPr,N),P}{P(iPr,N),H}C:]X are best regarded as
1’-phosphaacetylenes with polar P +-C- multiple bonds
[Eq. (16 ) ] ; reactivity typicai for ,13-phosphanylcarbenes reiPr2N,
Whereas triplet carbenes exhibit radical-like reactivity, singlet
carbenes are expected to show nuleophilic and electrophilic reactivity due to the a-type lone pair and the vacant p orbital. As
a result of small HOMO-LUMO gaps, carbenes are very reactive. One of the most intriguing and controversial questions
regarding the reactivity and stability of 2,3-dihydro-I H-imidazol-2-ylidenes, imidazolidine-2-ylidene, and bis(diisopropy1amino)ylidene focuses on their electronic structure. The theoretical investigations discussed above, however, concluded that
unsaturated as well as saturated N-heterocyclic carbenes of the
imidazole type benefit very well from n donation from the adjacent amino groups; this is equivalent to a contribution of the
two ylidic mesomeric structures to the electronic structure of
It is doubtful whether much can be learned about the behavior of transient, highly reactive singlet carbenes by studying the
stable ones. Transient carbenes generally react with unsaturated
nucleophiles such as allenes, acetylenes, arenes, or olefins. Following Skell’s theory the spin multiplicity of the carbene correlates with the stereochemistry of cyclopropanes resulting from
[1+ 21 cycloadditions to olefins.[82*
831 However, it has long been
known that nucleophilic carbenes in which the singlet state is
stabilized by interaction of the vacant TC orbital with a delocalized 7c system do not react with nucleophilic olefins (e. g. hexene)
but with electrophilic olefins such as fumaric esters. This rule is
s ~ l t s . ~84,
~ ’ * The P-C-P angle in [ (iPr,N),PCP((iPr,N),H)]+
(165”) is significantly larger than that expected for singlet carbenes. A singlet-triplet energy seperation of only 3 kcalmol-’
was computed. However, reactivity typical for carbene is
known, for example, for intramolecular insertion into C-H
bonds, stereospecific cyclopropanation with electron-poor
olefins (dimethyl fumarate), addition to aldehydes to yield oxiranes, [I + I] addition to isocyanides to yield ketene imines, and
concerted or stepwise (cyc1o)additions to nitriles to give 2Hazirins and 1,2-LS-azaphosphetes, respectively.[861
3.2. Basicity and Nucleophilicity
The basicity following the concept of Br~rnsted-Lowry and
the nucleophilicity of 2,3-dihydro-I H-imidazol-2-ylidenes were
examined by Alder et al. With pK, = 24 in [DJDMSO (fluorene: pK, = 22.9; 2,3-benzofluorene: pK, = 23.5) 1,3-diisopropyl-4,5-dimethyl-2,3-dihydro-lH-imidazol-2-ylidene
is a
stronger base than 1,5-diazabicyclo[4.3.O]non-5-ene(DBN),
1,8-diazabicyclo[5.4.O]undec-7-ene (DBU) , and proton sponges
but weaker than phosphazene bases.[871The rate of proton
transfer between a free carbene and its conjugate acid is close to
the exchange limit at room temperature, but the carbene is too
Angen. Chem. Int. Ed. Engl. 1997, 36, 2162-2187
N-Heterocyclic Carbenes
nucleophilic for synthetic applications as a strong base due to a
sterically unhindered conformer which is only slightly higher in
energy than the preferred congested conformer. The ratio of
elimination to substitution in the reaction with 2-bromopropane
(20:80) is similar to that for DBN and DBU. Proton exchange
between the acyclic bis(diisopropy1amino)carbene and its conjugate acid is much slower due to steric hindrance; this may also
explain the stability of the acyclic carbene, for example, towards
oxidative benzoin
forrnoin reaction
Scheme 1. Organic transformations of aldehydes catalyzed by N-heterocyclic
Michael- Stetter reactions were achieved using chiral triazolium
salts as catalyst p r e s ~ r s o r s . [ ~ ~ ]
3.3. N-Heterocyclic Carbenes as Nudeophilic Catalysts
N-Heterocyclic carbenes acting as nucleophiles have manifold catalytic applications in synthetic chemistry as well as in
nature: Breslow showed in the context of ylidene-catalyzed benzoin condensations[881that the vitamin B, enzyme cofactor thiamin (22), a naturally occurring thiazolium salt, plays a key role
in biochemistry (Figure 4) .[891 As thiamin diphosphate (the
3.4. N-Heterocyclic Carbenes as Lewis Bases
Figure 4. Structure of thiamin (22) as well as the framework of the corresponding
catalytically active 1,3-dihydrothiazo1-2-ylidene
coenzyme of pyruvate decarboxylase), it catalyzes the decarboxylation of pyruvic acid to active acetaldehyde as well as the
benzoin condensation of benzaldehyde. In basic aqueous
buffers the active catalyst of this reaction is a 2,3-dihydrothiazol-2-ylidene of the form 23. Stetter recognized that aliphatic
aldehydes that undergo aldol condensation in the presence of
cyanides are activated for a benzoin-type condensation upon
addition of ylidene catalyst^.[^^-^^^
2,3-Dihydro-1 H-imidazol-2-ylidenes and 4,s-diyhdro-I HI ,2,4-triazol-S-ylidenes also found catalytic applications in organic synthesis, for example in the benzoin condensation of
higher aldehydes to cc-hydroxyketones, the oxidative benzoin
condensation of aldehydes, alkohols, and aromatic nitro compounds to yield esters, and selective formoin reactions of
The catalytic
formaldehyde to give C, to C, building
activity and selectivity of a given system composed of a 1,3-heteroazolium salt and a base correlated nicely with the kinetic
acidity of the 1,3-heteroazolium salt. The transient formation of
dimers was ruled out. Nucleophilic ylidenes also served as catalysts in the Michael-Stetter reaction (Scheme 1). In this
“umpolung” reaction yielding 1,Cdicarbonyl compounds, a nucleophilic heterocyclic ylidene attacks an aldehyde; the generated acyl carbanion equivalent then adds to a,B-unsaturated ketones in a 1,4 manner. Intramolecular variants of this reaction
provided furanones, pyranones, and chromanones. Asymmetric
variants of ylidene-catalyzed benzoin condensations and
Angew. Chem. Inr. Ed Engl. 1997,36, 2162-2187
The transition metal coordination chemistry of 2,3-dihydro1H-imidazol-2-ylidenes is discussed in detail in Section 5. 2,3Dihydro-1 H-imidazol-2-ylidenes form stable adducts with soft
and weak Lewis acids, such as iodine, in which the carbenes,[’. 9s1just like tertiary phosphanes, act as pronouncedly
basic 0 donors. The nearly linear arrangement together with
other structural and spectroscopic data of the imidazole moiety
suggest a hypervalent (10e-I-2c) central iodine atom (I-) with an
expanded 1 . . . I bond and a positively charged imidazole moiety
of slightly augmented 7~ delocalization. 2-Iodo-l,3-dimesitylimidazolium iodide further reacted with 1,3-dimesityl-2,3-dihydro1H-imidazol-2-ylidene to give almost linear cationic bis(carbene) adducts of hypervalent iodine with only slightly different
C-I bond distances [Eq. (17)] .[961
[)>;-I+<(] N
X:(]I-’. - N
MI e s l
Adducts (1 :1) of 2,3-dihydro-lH-imidazol-2-ylidenes
also display a characteristic linear
arrangement [Eq. (18); R = adamantyl]. Increased 7~ delocaliza-
W. A. Herrmann and C. Kocher
tion in the imidazole ring indicates a formal charge assignment
opposite to that in halonium methylides. These halonium
methylides (RX+ t -CR,) with a small C-X-C angle are typically formed from reactions of electrophilic carbenes with halogen
Upon complexation to a Lewis acid, electron density at C2
decreases, and ylidic mesomeric structures become more important. In this sense a (hetero)aromatic imidazolium salt is an
adduct of free carbene with the strongest Lewis acid, a proton.
By allowing 1,3-dimesityl-2,3-dihydro-l
to react with the corresponding imidazolium salt, a linear 3c-4e
C-H-C hydrogen bridge is formed [172.5”; Eq. (19)].[971This
“proton complex” has two significantly different C-H bond
lengths (202.6(45) versus 115.9(45) pm, X-ray diffraction). It
was suggested that mesityl substituents favor the formation of
homoleptic bis(carbene) adducts because of the “cavity” seen in
the structure of the 2,3-dihydro-I H-imidazol-2-ylidene.
There is only one example of backbone functionalization of
N-heterocyclic carbenes. A new type of bimetallic complex was
obtained by cis-osmylation of the isolated double bond of 1,3dimethyl-2,3-dihydro-lH-imidazol-2-ylidene in [M(CO),(carbene)] (M = Cr, Mo, W) with osmium tetraoxide [Eq. (20)J.
Upon oxidation the ylidene ligand becomes slightly more nucleophilic. There is no interaction between 0s’” and Cr’, as seen
from cyclic v~ltarnmetry.[~*J
(alkylidene) ligand can be transferred to other substrates;
cyclopropanation and metathesis of olefins are probably the
most prominent examples.[31A general reaction scheme is not
easy to produce since particularly the nature of the metal influences the reactivity of the carbene. However, metal-carbene
complexes often react with cleavage of the metal-carbon bond.
For this reason the common carbenes :CR1R2 (R’, R2 = H,
alkyl, aryl, alkoxy, amino) are not considered as iigands that
would survive the standard reaction conditions of organometallic catalysis.
The quite distinctive role of diamino-substituted carbenes
(singlet ground state) was anticipated in coordination chemistry, and the classification into “Fischer-” and “Schrock-type”
complexes ([Cr(CO),{C(OMe)Ph)] and [(q’-C,H,),Ta(CH,)(CH,)], respectively) does not hold for N-heterocylic carbenes
of the imidazole and imidazolidine series. At first glance these
may be classified as Fischer-type compounds which are formally
derived from rather “soft” nucleophilic carbenes and are susceptible to nuleophilic attack at the carbene carbon.
Conventional carbenes are rather weak cs donors. Their
bonding to metals depends on pronounced x backbonding;
metals in low formal oxidation states therefore prevail in this
domain of coordination chemistry. The CO stretching frequencies in metal-carbonyl complexes are a sensitive probe for electronic ligand properties. The differences between single-faced
two-electron ligands in a series of isoelectronic compounds
[M(CO),L] can be derived from the IR spectra. The ratio of
cs donors to 71 acceptors (that is, the electron density induced at
the metal center) of Fe(CO),-bonded heteroatom-substituted
carbenes increases in the order :C(OR)R <:C(NR)R < :C(NR),
e imidazolidine-2-ylideneG 2,3-dihydro-I H-imidazol-2-ylidene
(see Table 2).
The stretching frequencies of the trans-CO ligands in carbene
complexes [M(CO),L] are identical within the limit of error with
those of imidazolidine-2-ylidene and 2,3-dihydro-I H-imidazol2-ylidene complexes (Table 1 and 2). The latter induces a
remarkably higher electron density at the metal center than the
basic trimethylphosphane.r12ar
Table 1. The v(C0) absorptions [cm-‘1 in carbene complexes [Cr(CO),L] [139].
M = Cr. Mo, W
3.5. Electronic Properties of N-Heterocyclic
Carbene Ligands
It has been standard textbook knowledge for several decades
that divalent carbon species :CR‘R2 exhibit o-donor and z-acceptor properties upon binding to transition metals. In principle, a “double bond” situation represented in formula 1 results, depending somewhat on the nature of R’ and R2. In all
cases, however, metal-carbon bonds shorter than single bonds
in metal -alkyl complexes are present. Unfortunately, there is
only little thermodynamic data on metal-carbene bonds. With
regard to the reactivity of the metal-bonded carbon fragment,
the literature discriminates between electrophilic (Fischer) and
nucleophilic (Schrock) carbenesL2]In many cases the carbene
2.3-dihydro-I H-imidazol-2-ylidene
covered by E
Table 2. The v(C0) absorptions [cm-‘1 in carbene complexes [Fe(CO),L] [139].
2,3-dihydro-l H-imidazol-2-ylidene
1931, 1925
The x-acceptor ability of N-heterocyclic carbenes is negligible, and lies between that of nitriles and pyridine, as shown for
the examples in Table 3. The ri-acceptor character decreases
Angen Chem. I n t . Ed. Engl. 1997.36, 2162-2187
N-Heterocyclic Carbenes
in the order NO > CO > RNC > PF, > P(OR), > P(aryl), >
P(alkyl), > RCN 2-N-heterocyclic carbenes > py. Tables 4 and 5
compare further 113 characteristics of the complexes discussed
4.1. Alkaline Earth Metals and Zinc
The chemistry of beryllium shows unique aspects among the
alkaline earth metals, since Be2+ is one of the hardest Lewis
acids known. Tetrahedral adducts of the formula BeCl;L,
formed upon treatment of BeCl,, a polymer in the solid state,
with donors. Use of the sterically unencumbered 1,3-dimethyl2,3-dihydro-1H-imidazol-2-ylidene (L’‘) resulted in the formation of complexes BeCI,(LMe), and [BeCl(LMe),]Cl (Figure 5).’”] An X-ray diffraction analysis of the latter showed
[a] L and L-L are heterocyclic mono- and dicarbene ligands, respectively
Table 4. The v(C0) absorptions in trans-[RhL’L2(CO)X] complexes [65]
L’, L2[a]
t [cm-’1 (medium)
1924 (KBr)
1929 (KBr)
1939 (benzene)
1957, 1958, 1960 (benzene)
1968 (KBr)
1983 (benzene)
1994 (KBr)
2003 (benzene)
2018, 2020 (benzene)
LM‘, LM‘
LC’, LC’
PPh,, PPh,
R W J , ,P(C,Fs),
P(OPh),. P(OPh),
CI, Br, I
C1, Br
Lcy = 1,3-bis(cyclopa] LM‘= 1,3-Dirnethyl-2,3-dihydro-lH-imidazol-2-ylidene,
= 1,3-bis(drphenylmethy1)-2,3dihydro-l H-imidazol-2-ylidene.
Table 5. Average CO force constants.f[Nm-’] in [M(CO).L,] complexes [12].
L: co
[M0fCO)s LI
[a] 1,3-Dimethyl-2,3-dihydro-lH-imrdazol-2-yIidene
“Conventional” carbene ligands such as alkoxy(alky1)carbenes are similar in their bonding properties (with respect to
low-valent metals) to carbon monoxide, and N,O-heterocyclic
carbenes may also act as n:
N-heterocyclic carbenes, however, very much resemble electron-rich alkylphosb1
phanes and N donors such as nitriles and pyridine.112a3
4. Main Group Element and Rare Earth
Metal Complexes
With this background it was not surprising that metals incapable of n: backbonding also coordinate these N-heterocyclic
carbenes as nucleophilic two-electron ligands, which are in this
sense similar to the distant relatives amines and ethers. Both the
genesis of these complexes and their structural data suggest that
they could be viewed as donor adducts, just like ammonia and
ether complexes.
Angew. Chem Int Ed Engl 1997, 36.2162-2187
Figure 5. Structure of the cation [BeCI(LMe),]’ in the chloride salt
distorted tetrahedral coordination at the beryllium atom. With
1.822(3) (1.807(3) A) the Be-Cearbenedistances are in the range of
Be-alkyl or Be-aryl single bonds (1.708(6)-1.85(3) A), and
under consideration of the different ionic radii they can be compared to Be-0 and Be-N bonds. The Be-Cl bond (2.083 A) is
significantly longer than in other tetravalent complexes and in
BeCl,. An excess of free carbene did not displace the second
chloride ion. Complete solvation to yield dicationic complexes
was only observed for [Be(H,O),]Z+ and [Be(NH,),]’+.
Monomeric and dimeric 1 : 1 adducts of Mg2+and Zn2+were
obtained from the reaction of diethyl zinc and diethyl magnesium with 1,3-diadarnantyl- and 1,3-dimesity1-2,3-dihydro-l
Himidazol-2-ylidene (LMes,LAd) [Eq. (21)].llool The solid-state
structure of (C,H,),Zn .LAdshows trigonal-planar coordination at zinc. The angle between the coordination sphere of zinc
and the imidazole plane is 81.6”. Mesityl groups do not provide
as much steric protection as adamantyl groups, as seen in the
dimeric magnesium adduct (C,H,),Mg. LMeswith bridging ethyl
groups. The 13C NMR resonances of the imidazole C2 nuclei
are shifted upfield by 25- 30 ppm. There is evidence that in THF
the trivalent adducts are coordinatively saturated by THE
The metal-donor bonds are predominately ionic and more
labile in complexes of the heavier analogues of beryllium and
W. A. Herrmann and C. Kocher
magnesium. 2,3-Dihydro-lH-imidazol-2-ylidene
complexes of
calcium, strontium, and barium were nevertheless obtained for
bis(trimethylsily1)amides [M{N(SiMe,),},(thf),] (M = Ca, Sr,
Ba) upon displacement of the two THF molecules by two
equivalents of 1,3-dimethyl-2,3-dihydro-l
[Eq. (22)J. The solubility and stability of the adducts decrease
adducts imply alternating bond lengths and only weak
interactions between the N-C-N moiety and the C-C
71 systems within the five-membered rings.["']
When 1,3-dialkyl-2,3-dihydro-l H-imidazol-2-ylidenes were allowed to react
with 2-bromo-2,3-dihydro-I H-l,3,2-diazaboroles, borolylimidazolium salts formed under displacement of the halide.['43]
Borane adducts are also known for N,O-heterocyclic carb e n e ~ . ~A' ~silylene~]
borane adduct was reported r e ~ e n t l y . [ ' ~ ~ l
When 1,3-bis(2,4,6-trimethylphenyl)-2,3-dihydro-lH-imidaR-N,
201-2-ylidene was allowed to react with H,AI.NMe,, a 1 : 1 adIM
duct with AlH, formed.['031The upfield shift of the I3C NMR
<"N(SiMe3)z (22) signal for the carbenoid carbon center is, in agreement with
I! I
71 delocalization, even more pronounced (6 = 175.3; parent carRI
L N ,
bene: 6 = 219.7; imidazolium salt: 6 % 136). 1,3-DiisopropylR
M = Ca. Sr, Ba; R = Me, tBu
adducts with
Me,A1 or Me,Ga are also stable. The metal-C,,,,,,,
bonds in
signififrom calcium to barium. Whereas the calcium and strontium
adducts can be isolated in the solid state at - 36 "C, the barium
2.124(6) A; AI-C,,,hyl 1.940(5) and 2.062(7)
In the
adduct is stable only in solution. For the barium adduct a
dissociation equilibrium is observed. Upon complexation to
strontium and barium the resonances of the carbene centers are
only slightly shifted to higher field (6 = 198 and 203, respectively) .[1011
4.3. Carbon and Higher Homologues
4.2. Boron, Aluminum, and Gallium
Borane adducts with (formally) neutral carbon donors such
as carbon monoxide, isonitriles, or phosphane ylides are well
established. Since conventional carbene complexes show pronounced metal-to-ligand 71 backdonation, electron-poor (coordination) fragments are not well suited for stabilizing carbene
ligands. However, N-heterocyclic carbenes do not depend on
backdonation due to delocalization of the nitrogen lone pair;
I : 1 adducts with BH, or BF, (Scheme 2) are thermally stable
The pronounced ability of the N-heterocyclic fragment to
take on a formal positive charge is demonstrated by ylidic
olefins (exocyclic CYl,,-C bond), which form upon deprotonation of pentamethylimidazolium salts [Eq. (23)]. The chemical
M = Al, Ga
, iPr;
Scheme 2. Adducts of 2,3-dihydro-lH-imidazol-2-ylidenes
and Lewis acids MR,
(M = AI, Ga; R = H, Me), BH,, and BF,.
and can even be sublimed without decomposition. The "B
NMR chemical shift of 1,2,3,4-tetramethyl-2,3-dihydro-l
H-imidazol-2-ylidene + BH, at 6 = - 3S is typical for ylide-borane
adducts, and between that of amine- and phosphane- borane
adducts. The 'J('H,"B) coupling constant, however, is very
similar to that of the anionic analogue [(C,H,)BH,]-. Both
X-ray analysis and ab initio calculations on these borane
reactivity of the methylene carbon atom in C8H14N, and the
strong shielding in the I3C NMR signal (6 = 40.2) reveal high
electron density on the methylene carbon atom and extended
charge separation.['05a1The exocyclic double bond is extremely
polarized due to a combination of the enamine structure and
x delocaiization; it acts as a strong ylidic donor towards boranes
and metal-carbonyl c o m p l e ~ e s . [1' ' ~In~ ~
~ complexes the
exocyclic double bond is clearly bound to the metal in the
otherwise unusual end-on mode. This ylidic olefin also allowed
the preparation of mono adducts with 4f elements such as La"',
Nd"', and Y'I' from Ln[N(Me,Si),], (Ln = La, Nd) and
[(C,H,)Y(C,Me,)], respectively. Whereas coordination to
tow-valent metal-carbonyl complexes ([M(CO),(C,H 14NZ)],
M = Mo, W) entails a significant high-field shift of the
Angew. Chem. Int. Ed. Engl. 1997, 36, 2162-2187
N-Heterocyclic Carbenes
methylene carbon resonance (Ad ~ 4 5 ) ,there is almost no
change in the exocyclic ylidic bond upon coordination to
lanthanide ions."05d' Like these ylidic olefins, the 2-iminoimida~olines['~~"'
with exocyclic C,,,,-NH
bonds display interesting chemistry.['47b3
Similar to organophosphanes, N-heterocyclic carbenes form
stable adducts with carbon disulfide. In the crystal structure the
plane of the CS, fragment is perpendicular to the five-membered
ring (decoupled R systems).rlOse*fJ
Strongly polarized structures are most clzarly observed in the mono adducts Ge1;L
(24, L = 1,3-dimesityl-2,3-dihydro-l
H-imidazol-2-ylidene) and
SnR, L (25, L = 1,3-diisopropyl-4,5-dimethyl-2,3-dihydro-l
Mes ti= 95.4O
(R= Me, Et, iPr)
imidazol-2-ylidene; R = C1, 2,4,6-iPr,C6H,), which were used
as models for the dimerization of singlet carbenes along a nonleast-motion pathway.['06*
The heavier main group elements show pronounced pyramidalization of the coordination
sphere (sum of angles at Ge 290°, at Sn 282", C-Ge-I 95.4 and
100.1') and the two halves of the adducts are twisted with respect to one another. The I3C NMR resonance of the carbene
carbon atom in the GeI, adduct is shifted upfield by 61 ppm,
and the N-C-N angle opened from 101.4 to 107.0". Long C-Ge
and C-Sn distances (2.102 and 2.290(5) A, respectively) suggest
that the adducts be described as Lewis acid- base adducts. Tetravalent dicarbene adducts of the less Lewis acidic SnCl, with
sterically unencumbered carbenes are very insoluble.
The chemistry of' hypervalent silicon compounds is dominated by complex anions with sixfold coordination. Substituents
are in most cases negatively charged organic alkyl or aryl
residues or fluoro, alkoxy, or hydrido ligands. A new way of
forming element -carbon bonds was opened with the isolation
of nucleophilic carbenes. The reaction of carbenes with main
group element halides may yield several possible products
(Scheme 3): a) neutral or ionic compounds resulting from
simple adduct formation or displacement of a halide ion, or
b) reduction of the central element with formation of 2-haloimidazolium ions. Lewis acidic silanes such as tetrachlorosilane
as well as diaryl- or dialkyldichlorosilanes form monomeric
and neutral pentavalent 1 : 1 adducts with 2.3-dihydro-1Himidazol-2-ylidenes. The SiCl, adducts are rather labile. The
less Lewis acidic trimethylsilanes Me,SiCX are inert or, such as
Me,SiI, form ionic 2-(trimethylsilyl)imidazolium salts.[107]
The tendency to form complexes is more pronounced for
tetravalent tin than for silicon. Pentavalent tin adducts are prepared from diphenyldichlorostannane. The solid-state structure
of Sn(C,H,),Cl, .L(L = 1,3-diisopropy1-4,5-dimethyl-2,3-dihydro-1 H-imidazol-2-ylidene) shows that the three organic substituents adopt the equatorial positions of a distorted trigonal
bipyramid. With 2.179(3) A the Sn-Ccarbenebond is longer than
the Sn-Ph bonds (2.122(5) and 2.139(3) A). The 'I9Sn NMR
resonance is shifted to higher field by about 30 ppm relative to
that of the phosphane a d d ~ c t s . [ ' ~Allowing
1,3-dimesityl-2,3dihydro-I H-imidazol-2-ylidene to react with suitable phosphorous precursors yielded adducts with two- and six-coordinate
phoshorus centers (LMes~PF,Ph/LMes~PPh)
which were also
structurally ~haracterized."~']
4.4. Sulfur, Selenium, and Tellurium
Ylidic structures with a positively charged imidazolium group
have spotlighted a new perspective in organotellurium chemistry. In general the existence of tellurocarbonyl compounds
such as telluroaldehydes or telluroketones can only be proven
by stabilization with metals. In contrast, 2,3-dihydro-IH-imidazol-2-ylidenes readily react with elemental sulfur, selenium, and
tellurium to form stable adducts [Eq. (24)] .['OS1 According to
Scheme 3. Possible reactions of 4,S-dimethyl-2,3-d1hydro-l
H-im1dazol-2-ylidene with main
group elements. R = Me, Et, iPr.
Chrm Inr Ed. Eng! 1997, 36, 2162-2187
X = S, Se, Te
the crystal structure analysis the formal C,,,,-Te
double bond is almost as long as a single bond
(2.087(4) A). The stability of this compound, in contrast to the analogous selenium compound, results
from the decrease in double-bond character. The signals in the lZ5Teand 77SeNMR spectra of the telluroand selenoimidazolines, respectively, resonate upfield
with respect to that of other chalcocarbonyl compounds. This shielding is explained by 7~ delocalization. Oxidation of the sulfur imidazoline with iodine
yields a charge-transfer complex; in the cases of the
seleno and telluroimidazolines, however, structures
with covalent bonds are formed; the latter exists as a
dimer. Chalcogenoimidazolines also serve as ligands
in transition metal c o r n p ~ e x e ~ . [ ' ~ ~ ]
21 77
W. A. Herrmann and C. Kocher
H-l,2,4-triazol-5-ylidenesimilarly reacts with chalcogenes such as oxygen, sulfur and selenium to give the respective triazolinone, triazothion, and triazoselenone. The free triazolylidene also inserts into the 0-H, N-H,
and S-H bonds of alcohols, amines, and thiols, but no case of
insertion into C-H bonds has been reported yet.[109]Reaction
with protonation or alkylation occurs with acids and trialkyoxonium salts, respectively. The products formed from reaction
with activated double bonds such as those in fumaric and maleic
esters, amides, and nitriles suggest [2+ I] cycloaddition with
subsequent 1,2-hydrogen shift. Nonactivated double bonds are
not attacked. Heterocumulenes (e. g. CS,, C,H,N=C=O, and
C,H,N=C=S) are also susceptible to nucleophilic attack by the
triazol-derived carbene.
M = Sm"; R' = Me, R2 = Me
M = Sm"; R' = P r , R2 = Et
M = Yb"; R'
= Me, P r ; R2 = Me, Et
4.5 Rare Earth Metals
Until a few years ago isonitriles were the only neutral ligands
in organolanthanide chemistry that coordinate through just one
carbon atom. In view of the isolation of stable alkaline earth
metal complexes of nucleophilic carbenes, it was not surprising
that rare earth metals, which are nearly incapable of n backbonding, also coordinate N-heterocyclic carbenes (Scheme 4).
2,3-Dihydro-I H-imdidazol-2-ylidenes are not reduced by cyclopentadienyl precursors of Yb" or Sm": 1,3,4,5-Tetraalkyl2,3-dihydro-1H-imidazol-2-ylidenes
readily replace the THF
molecule in [($-C,R,),M"(thf)]
(M = Yb, Sm). The monoand bis(carbene) adducts obtained [($-C,R,),Yb"(L)]
[(q5-C5R5),Srn1'(L),,](n =1 or 2) are more stable to air and
moisture than the THF adducts.["OIn the solid-state I3C NMR spectra of the diamagnetic and
thermally very stable ytterbium complexes, the resonances for
the ligating carbon atom are shifted to higher field by only
10ppm with respect to that of the free carbenes. Solid-state
structures were determined in two cases. With 2.552(4) and
2.598(3) A the Yb-C,,,,,,,
bond lengths in the pseudotrigonal
complexes are in the range of "elongated single bonds"."
The analogous monocarbene samarium(I1) complexes are also
high-melting stable solids. Addition of a second equivalent of
1,3,4,5-tetraalkyl-2,3-dihydro-l~-imidazol-2-ylidene to [($C,Me,),Sm"(L)]
yielded the bis(carbene) adduct [($C,Me,),Sm"(L),] . The X-ray structure analysis of this complex
shows pseudotetrahedral coordination and rather long metalCcarbene
bonds (2.837(7) and 2.845(7) A). With 101.1(6) and
101.9(6)" the internal valence angles N-C,,,,-N are very close to
that for free carbenes.["']
Seven-coordinate monocarbene adducts of Eu"' and Yrrr
prepared from the dionato precursors [Eu(thd),] and [Y(thd),]
(thd = 2,2,6,6-tetramethylheptane-3,5-dionate)
.[l lo] In the 13C
NMR spectrum the former Coarbene
in [Y(thd),L] resonates at
6 = 199.4 with a 'J(Y,C) coupling constant of 33 Hz. Although
the metal-carbene bond lengths are longer for lanthanides than
for s-, p-, and d-block metals, the carbene does not dissociate on
the NMR time-scale. With 2.663(4) A the Eu-Ccarbenebond
length in [Eu(thd),L] is similar to Pr"'-C and Sm"'-C distances
in isonitrile complexes.
Lanthanide amides (with or without additional donors such
as THF) are also versatile starting compounds for carbene comb
M = Y , R = SiHMe2
M = La, R = SiMea
Scheme 4. Complexes of 2,3-dihydro-l H-imidazol-2-ylidenes and rare earth
plexes. Addition of one or two equivalents of 1,3-dimethyl-2,3dihydro-1 H-imidazol-2-ylidene to [Y{N(SiHMe,),} ,(thf),]
yielded [Y{N(SiHMe,),},(LMe)] and [Y{N(SiHMe2),),(LM'),1,
which were both structurally characterized. The Y -Ccarbene
distances are 2.55(1) and 2.648(8)/2.671(9) A, and in the range of
elongated Y -C single bonds. The bond lengths compare well to
other lanthanide-carbene distances. Monocarbene adducts
of lanthanum [La(N(SiMe,),),(LMe)] are available from
[La{N(SiMe,),),] .[I 14*
5. Transition Metal Complexes
Transition metal complexes of heterocyclic carbenes are
accessible by three different routes (Scheme 5 ) . Most common is the treatment of simple metal salts or coordination compounds (neutral, cationic, or anionic) with azolium precursors
(route A). The coordination properties of the anion X - determine whether it enters the inner coordination sphere of the
resulting carbene complex ;the most commonly employed azolium iodides normally yield iodo complexes in which the iodide
Angew. Chem. Inr. Ed. Engl. 1997,36, 2162-2187
N-Heterocyclic Carbenes
homoleptic mercury(I1) carbene complex and Ofele's chromium(o) complex were already mentioned [see Eqs. (2) and (I)].
Palladium(n) complexes were prepared similarly from the metal
acetate [Eqs. (25) and (26)].1'3*1161
I- + Pd(0AC)z
Scheme 5 . Synthesis of transition metal carbene complexes by A) in situ deprotonation of azolium salts, B) reaction with free N-heterocyclic carbenes, or
C) reaction with anionic metal-cyano complexes
has replaced other ligands (e. g. chloride, bromide, acetate, carbon monoxide).
This is certainly not the case when non- or weakly coordinating anions are employed, for example tetraphenylborate or
hexafluorophosphate. In some cases the metal complexes act as
a reducing agent for imidazolium salts (evolution of hydrogen).
The scope of heterocyclic carbenes which can be directly generated at metal centers is more varied following route A because
of the broad variety of CH-acidic, cationic heteroaromatic compounds (Scheme 6).
Ofele pointed out that this procedure is not limited to imidazolium salts. Also triazolium, tetrazolium, benzimidazolium,
and even the less acidic pyrazolium salts yield complexes of the
respective N-heterocyclic carbene ligands. Other metal bases
from elements of Groups 6- 8 were identified as suitable precursors, for example metalates of chromium, molybdenum, and
tungsten, some of which contain p-hydroxo or p-alkoxo bridges
(Table 6 ) . In the case of the di- and tetranuclear complexes, the
Table 6. Multinuclear Group-6 metalates capable of reducing azolium salts with
formation of carbene ligands.
Metal M
Product [a]
[a] L = carbene ligands derived from imidazole, triazole, tetrazole, benzimidazole,
or pyrazole.
acidic azolium protons are not transferred to the hydroxy or
methoxy ligands but reduced to hydrogen [Eq. (27)]
X = S. thiazolium
X = 0. oxazolium
Scheme 6. Heteroaromatic azolium salts which can be deprotonated in situ to give
coordinated N-heterocyclic carbenes.
4 [Cr&CO),( p-OH)3l3-+ 12[LHl*
-3 H2
6 cis-[(L)zCr0(CO)4]c2 *'Cr3+*'
It is useful to have a basic ligand in the precursor that effects
deprotonation of the azolium salt. C-Deprotonation of azolium
salts is also feasible with (anionic) carbonylmetalates, which are
normally generated through typical Hieber base reactions. Both
variants were reported in 1968 for the first time: Wanzlick's
K[Re2(CO),(p2-OR),] was successfully applied to synthesize
dicarbene complexes fac-[Re(CO),(L),I], and Na,[Fe,(CO),]
was a suitable precursor for carbene complexes of the type trans[Fe(CO),L,] and cis-[Fe(CO),(L-L)]. Again the metal precursor was partially oxidized during the synthesis.[63]
Chem. Inr Ed. Engl. 1997, 36, 2162-2187
W. A. Herrmann and C. Kocher
A variety of rhodium and iridium complexes were synthesized
from easily available p-alkoxy complexes and azolium salts.
Carbene-complex formation is exclusively effected by proton
transfer to the metal-bound ethoxy group (Scheme 7). The reac-
[M(CN)(CO)s]-+ HBF., + CNR' + RZR3C=0
3 cc
- BF4
[Cp(CN)2(CO)Fe]-+ CNBU + PhCHO + PhNH3CI
tions seem to be thermodynamically controlled, as further acidic
protons in the azolium unit did not inhibit carbene-complex
formation. Even when an excess of base and azolium salt was
employed, neutral monocarbene complexes of the type [M(v41,S-cod)LX] (M = Rh, Ir; cod = cyclooctadiene) are formed selectively in almost quantitative yield.[651On the other hand, free
carbenes displaced the halide ligands to yield cationic complexes
of the type [Rh(q4-1,5-cod)L,]+X- (Scheme 7). The acidic protons are therefore transferred to metal-bound ethoxide groups
exclusively. However, cationic dicarbene complexes are also
accessible when the imidazolium halides are first transferred to
imidazolium ethoxides salts with potassium ethoxide in ethanol
(2-ethoxyimidazoles are not observed) .I6']
According to route B in Scheme 5, nucleophilic N-heterocyclic carbenes may replace other ligands, for example a twoelectron ligand (such as halide, carbon monoxide, or acetonitrile) in the complex. From carbonyl complexes such as
[M(CO),] (M = Cr, Mo, W), [Fe(CO),], or [Ni(CO),], one or
two molecules of carbon monoxide are displaced. Further substitution requires photolytic conditions.["". b1
Two-electron donors are displaced by 2,3-dihydro-I Himidazol-2-ylidenes in smooth reactions from the adducts
[(Me,S)MCl] (M = Cu, Ag, Au), [MCl,(thf),] (M = Ti, Zr,
Hf), [MCl,(py),] (M = Nb, Ta), and [VCl,(tmeda),].[1501Free
N-heterocyclic carbenes thus act as conventional two-electron
donors, and react with a broad variety of precursors. Bridging
moieties of dinuclear precursor complexes are split by the carbene. Dinuclear halo- or acetato-bridged complexes of Ru, Os,
Rh, Ir, and Pd yield mononuclear carbene complexes in excellent yie.ds.113.34, 37.611
Acyclic carbenes with heteroatoms can be generated at the
metal centers by transforming other C-bound ligands, for example isonitriles [Eqs. (28), (29)]. Nucleophilic attack of an alcohol or an amine at the electrophilic carbon atom starts the
reaction sequence." "1 The four-component condensation
following route C, established by Fehhammer et
yielded the ligand framework of 2,3-dihydro-l H-imidazolz-ylidene, Metal complexes of hydrogen cyanide
Scheme 7. Mono- and diylidene complexes of Rh and Ir are accesslble a) directly
from metal alkoxides and azolium salts or b) from free N-heterocyclic ylidenes and
dinuclear p-chloro complexes.
t ? / "
M = Rh. Ir
mono- and
dicarbene complexes
generated from anionic metal-cyano precursors) were treated
successively with an aldehyde, an isocyanide and an amine to
form (ring-substituted) N-heterocyclic carbene complexes.
Because of the isolobal analogy between an oxygen atom and
a ML, d6 transition metal complex fragment, this route
represents the organometallic analogue of Ugi's 4CC reaction
to hydantoins.1"81 The same principle can be employed to
generate oxazolidine-derived carbenes bound to transition
metals by a 3CC [2+ 1 21 cycloaddition of hydrogen isocyanide metal complexes, isocyanides, and carbonyl compounds [see Eq. (28)].['192 1201
Less general routes were mentioned in the recent literature.
For example, Matsumara reported that palladium und rhodium
complexes of six-membered 2,6-diazocarbenes result in a most
peculiar manner from 10-S-3-tetraazapentalene derivatives
[Eq. (30)].[1211Sellmann et al. obtained nickel and platinum
complexes with tridentate dibenzenethiolate-imidazolidine-2ylidene ligands according to Equation (31) .["']
Furthermore, Ghosh discussed the analogy between the "N-disordered
porphyrines" [Eq. (32), Ar = p-Tolyl] and Arduengo carbenes.[1231
6. Structural Chemistry
The structure of bis(l,3-dimethyl-2,3-dihydro-l
H-imidazol2-yIidene)mercury(11) perchlorate was determined" 511 three
years after its synthesis by Wanzlick in 1968.[81In the twenty
years since 1971 only a few solid-state structures of this
type were reported in the literature. On the other hand, there
are many examples of structurally characterized complexes
with C-C-saturated carbene ligands not considered in this
The data rather suggest a bonding order near, or even
below, one. In comparison to Schrock- and Fischer-type
carbene complexes, imidazole-derived ylidenes show long
metal-carbon bonds. This is expected from the contribution
of the ylidic resonance structures to the free 2,3-dihydro-I Himidazol-2-ylidenes. Significant metal-to-carbon backdonation
can only be observed in linear homoleptic bis( 1,3-dimesityl2,3-difiydro-lH-imidazol-2-ylidene)
complexes [M(LMeSh1of
of zero-valent nickel and platinum. The Ni-C,,,,,,,
are 0.15 8, shorter than in [Ni(CO),(LMe')21,which cannot be
N-Heterocyclic Carbenes
Figure 6. Averaged values of the C
bond lengths d (circles) and the NC,.,,,,,-N angles a (squares) as a function of the metal from the transition metal
series x . The first column (I= 0) lists the values for free carbenes [I 151.
Whereas isolated data are not meaningful, the trend in Figure 6
is more elucidating: Angle N-C,,,,,,,-N increases for the heavier
metals, while the Ccarbene-N
bond length decreases. This supports the general rule that metal-carbon bonds become
stronger when going from 3d to 4d and 5d transition metals.
7. Catalysis
explained with the decrease in coordination number from four
to two. A barrier to rotation has not been observed for the
metal-carbene bond, which is in agreement with single-bond
character. The position of the carbene ligands relative to the
remaining complex fragment is strongly influenced by packing
effects: For [W(CO),L] three different rotamers with an
eclipsed, a staggered, and an intermediate arrangement are
structurally characterized for nearly identical carbene ligands.l6'I
The structures of free 2,3-dihydro-l H-imidazol-2-ylidenes, as
determined by X-ray and neutron diffraction, all show nearly
identical geometries (Table 7). Considering a 3 0 limit of error
for X-ray and neutron diffraction data, the only difference
worth discussing is an increase of about 2" in the ring angle
N-C,,,,,,,-N upon metal coordination. Whereas the geometric
details of free and metal-bound carbenes hardly differ from each
other, the structures of azolium cations are very different. In
accordance with an increase in K delocalization in the five-membered ring upon protonation, the N-C bond lengths decrease
and the C=C borid lengths increase. The angles in the ring
become more similar. Apart from the C-C double bond all
structural parameters of the carbene ligands are between those
of the two extreme structures.
A dependence of the ring parameters on the metal is seen
when the average bond lengths are considered (Figure 6 ) .
Applications of C-C-unsaturated N-heterocyclic carbenes as
ligands in metal-containing catalysts were disclosed in a series of
patents (1994). Previous investigations on catalytic properties
were limited to C-C-saturated derivatives, but no advantage over the standard metal-phosphane catalysts was
achieved.['24, 1251 A recent publication on the Heck olefination
sparked renewed interest in the catalysis-related aspect of carbenes.cL3]Reactions such as hydroformylation, hydrogenation,
isomerization, furan synthesis,"261 and olefin metathesis were
investigated. The synthesis of chiral 2,3-dihydro-I H-imidazol2-ylidene complexes and their application in asymmetric catalysis was another important development.
7.1. Hydrosilylation of Alkenes and Alkynes
The hydrosilylation of terminal alkenes catalyzed by rhodium-carbene complexes such as [RhC1(q4-1,5-cod)L],[RhCl(PPh,),L], and [RhCI(CO)(PPh,)L] led to selective antiMarkovnikov addition of the silane [Eq. (33)].[124,1251 The
hydrosilylation products were obtained in yields of up to 98%,
depending on the silane and the rhodium-carbene catalyst.
Table 7. Average hondin: parameters of the 2,3-dihydro-lH-imidazol-2-ylidene
system (see text; distances in
free carbene
carbene ligand[d]
imidazolium ion
Angen. Chem. In!. Ed. Ergl 1997,36,2162-2187
A, angles in ") [115].
108 2
W. A. Herrmann and C. Kocher
Other rhodium(1) catalysts, for example [RhCl(PPh,),], gave
lower yields. Conjugated dienes with rhodium-carbene catalysts yielded mixtures of 1,4- and 1,2-addition products, while
[RhCl(PPh,),] led to pure 1,4-isomers.
In the hydrosilylation of alkyne~['~']
[Eq. (34)] the product
distribution depended on the reaction conditions and the cata-
SiEt3H, cat
lyst. With the rhodium(1) carbene complexes [RhCl(q4-I,5cod)L] and [RhCl(PPh,),L], silane addition gave mixtures of cis
and trans adducts. Upon further heating in the presence of both
the catalyst and the silane, complete cis + trans isomerization
The rates of the hydrosilylation of alkynes increased in the
presence of air and/or ultraviolet light. For example, irradiation
of a mixture of phenylacetylene, triethylsilane, and catalyst
[RhCl(q4-1,5-cod)L] effected quantitative conversion within
one hour, whereas in the absence of ultraviolet irradiation or the
catalyst no reaction was observed. No such effects were seen for
[RhCl(PPh,),] .
7.2. Hydrosilylation of Ketones
The hydrosilylation of ketones according to Equation (35) is
of synthetic interest, since this reaction followed by hydrolysis
PhCOMe +SiEt3H
of the resulting silylether is a mild route for reducing ketones to
secondary alcohols. The use of triethylsilane requires relatively
high temperatures. [RhCl(q4-I ,5-cod)L], [RhCl(PPh,),L],
[RhCl(CO)L,], and several ruthenium-carbene complexes were
tested as
The rhodium complexes proved to be
efficient catalysts. Only the dicarbene complex [RhCl(CO)L,]
was rather unreactive; the mixed phosphane-carbene derivative [RhCI(PPh,),L] showed the best performance (98 % conversion, 4 h, 100 "C). The nature of the N substituent on the
carbene ligand also effected the catalytic activity, indicating that
the carbene ligand is indeed involved in the catalysis. Ruthenium(@ complexes are less active.
The reaction temperatures could be significantly lowered if
diphenylsilane was used as the silylating agent [Eq. (36)]. How-
ever, with some catalysts considerable amounts of the silyl enol
ether side product formed.['251The best results were obtained
with the monocarbene complex [RhCl(q4-1,5-cod)L] (N-phenyl
derivative, 98 % conversion, 24 h, 25 "C). Increasing number of
carbene ligands lowered the yields.
Treating silanes with a rhodium(1) carbene complex in the
absence of another substrate resulted in coupling to disilanes.
This reaction was exploited for the synthesis of di~ilanes.['~~'
The asymmetric version of the catalytic hydrosilylation is particularly interesting. For example, prochiral ketones are converted into optically active secondary alcohols. Although chiral
carbene complexes were synthesized by Lappert et al. in a multistep approach, they failed to generate significant enantiomeric
excesses. Therefore, our recently published asymmetric hydrosilylation of acetophenone [Eq. (37)] represents the first successful asymmetric catalysis using chiral metal-carbene complexes,1128.
- 3OoC, THF
=. 90% chemical yield
> 30% optical yield
cat. =
The synthesis of the chiral carbene complex is straightforward: the C,-symmetrical imidazolium salt is synthesized in one
step from enantiomerically pure a-naphthylethylamine and directly converted into the catalytically active rhodium-carbene
complex (alkoxo route). With this chiral rhodium complex as
catalyst, the hydrosilylation of ketones proceeds smoothly, even
at low temperatures (- 30 "C); asymmetric inductions over
30% (not optimized) are obtained.['281
Although the enantiomeric excesses are subject to improvement, they are not influenced by the amount of catalyst or the
degree of conversion. It should be mentioned that a free carbene
ligand reacts with a hydrosilane with insertion into the Si-H
bond. If there were free ligands present in the reaction mixture,
they would be irreversibly destroyed and the enantiomeric excess would then decrease in the course of the reaction. It can
therefore be taken for granted that the metal-carbene bond is
kinetically inert under the reaction conditions.
7.3. Hydrogenation of Olefins
Mixed carbene-phosphane complexes, for example [RhCI(PPh,),L] and [Ru(CI)L(PPh,),], are so far the most active of
the carbene-containing catalysts for the hydrogenation of
olefins. Hydrogenation of C-C double bonds proceeds
smoothly under atmospheric pressure [Eq. (38)] .[1251
Angew. Chem. Int. Ed. Engl. 1991, 36, 2162-2187
N-Heterocyclic Carbenes
The hydrogenation of olefins catalyzed by rhodium($) carbene complexes is being investigated further in our
Preliminary results suggest that carbenes are best combined with
other ligands that lower the electron density at the metal center
and dissociate in solution to generate coordination sites for the
respective substrate (“hemilabile coordination”). This is plausible, since the activation of H, proceeds more readily when
rather strong p-accepting phosphanes, for example pyrollyl
phosphane, are employed.[’31]Combined phosphane-carbene
complexes, however, exhibit pronounced stability under the
conditions of catalysis.
7.4. Heck Olefination
The application of palladium complexes [PdL,I,j as catalysts
for the Heck olefination of aryl halides according to Equation (39) has been subject to intense research.[’31The new cata-
branched products, a large excess of phosphanes (up to 1000fold) is required. Stoichiometric rhodium triphenylphosphane
complexes slowly decompose under the conditions of a catalytic
In contrast, since the long-term stability of rhodium(r) carbene complexes-for example [RhC1(q4-1,5-cod)L], [RhCl(PPh,),Lj, or [RhCl(CO)L,]-is much higher , they can be used
without ligand excess; this is quite an advantage over conventional catalysts. However, the higher electron density at the
metal entails reduced activities in comparison with triphenylphosphane complexes (e. g. [RhH(CO)(PPh,),]). Higher activities in combination with long-term stability were observed in the
case of mixed rhodium carbene phosphane complexes.[’321All
experimental observations lead to the conclusion that the carbene ligands remain at the metal center during the catalytic
7.6. Furan Synthesis
In 1996 Dixneuf et al. reported an interesting application of
ruthenium(I1) carbene complexes of the general type [Ru(pcymene)Cl,L] .[1261 In a catalytic reaction (Z)-3-methylpent-2en-4-yn-1-01 was converted into 2,3-dimethylfuran [Eq. (40).
lysts are extraordinarily stable with respect to heat, moisture,
and oxygen. Kinetic studies show that they become very active
after a short induction period : Treatment with reducing agents
yielded Pd(o) derivatives which are the true active spiecies (Figure 7) and have remarkable long-term stability in the Heck
The reactivity of the catalyst is strongly dependent on the nature
of the carbene ligand. Best results were obtained with a benz2,3-dihydro-1H-imidazol-2-ylidene derivative.
7.7. Olefin Metathesis
Figure 7. Palladium(0) carbene complexes as active species in the Heck olefination
with the example of the reaction of 4-bromoacetophenone (squares) with n-butyl
acrylate to form n-butyl-(E)-4-acetylcinnamate (circles). Hydrazine hydrate was
added after 67 min; s = concentration.
olefination. No colloidal palladium was observed. These properties make carbene complexes of this type suitable for the activation of chloroarenes.
Enders et al. reported the synthesis of chiral palladium-carbene complexes and their application in asymmetric Heck reactions. However, they have not yet obtained significant optical
inductions.“ 1 6 ]
Ruthenium@) carbene complexes were successfully employed
as catalysts in olefin metathesis. The bond between ruthenium
and the N-heterocyclic carbene carbon atom does not react with
strained cyclic olefins; norbornene is not polymerized with
[Ru(p-cymene)Cl,L]. This suggests that the carbene acts as a
stabilizing “spectator ligand”. The ruthenium catalyst has a
high activity in the metathesis of norbornene (turnover number
(TON) at room temperature > 8000) only when an alkylidene
moiety is generated with trimethylsilyldiazomethane. Functional groups as in exo/endo-bicyclo-[2.2.l]hept-5-en-2-acetate
are tolerated (TON > 100).[621 Ruthenium(I1) complexes with
N-heterocyclic carbenes are similar to analogous complexes
with basic phosphanes, for example [Ru(p-cymene)Cl,(PCy,),],
in their catalytical performance. Further work in our laboratories is in progress.[’33’
7.5. Hydroformylation
7.8. Conclusions
Rhodium(1) phosphane complexes are commonly used as catalysts in the hydroformylation of olefins. To avoid decomposition of the complex and to obtain a good ratio of linear to
Angen. Cliem. In!. Ed. Engl. 1997,36, 2162-2187
N-Heterocyclic carbene complexes exhibit promising properties for a number of catalytic reactions in organic chemistry. The
W. A. Herrmann and C. Kocher
cheap synthesis of imidazolium salts makes them attractive for
industrial applications. Since the strong o-donating abilities of
the carbene ligands lead to increased electron density at the
metal centers, heterocyclic carbenes can generally be seen as
alternatives and extensions to more basic phosphanes.
Since this new class of catalysts feature strong metal-carbene
bonds, immobilization techniques have a realistic chance for
catalytic applications; the first results are now available. A number of methods to immobilize either the imidazolium salt or a
preformed catalyst have been tested. For example, the coupling
of a functionalized rhodium(1) carbene complex with a Merrifield resin works well [Eq. (4111. Many different functionalized
amorphous a-Ge
2 [I;Ge:
470 K
+ cN>Ge:
Scheme 8. Germylenes as a source for depositing a-Ge films by CVD.
metal-carbene complexes with amino, ester, ether, hydroxy,
and keto groups were synthesized for the purpose of subsequent
attachment to organic and inorganic
8. Chemical Vapor Deposition (CVD)
of Germanium
Photoelectron spectroscopy (PES) and thermolytic studies
opened an interesting application of the C-C-saturated cyclogermylene (B, Scheme 8) in materials science!". 1341 The
compound yielded [Ge/GeH] layers, isobutylene, HCN, and
(possibly) isobutane but no hydrogen upon heating to 900 K,
and [Ge/GeH] layers, A, and N,N'-bis-tert-butylethylenediamine upon heating to 470 K (Scheme 8).
A comprehensive CVD study by Veprek et al. established that
these cyclogermylenes are suitable precursors for generating
thin films of amorphic a-germanium by MOCVD (sublimation
at about 40 OCjO.25 mbar).[135-1371Depending somewhat on
the nature of the support meterial, germanium coatings began
growing around 245 "C. At lower temperatures selective deposition of germanium on silicon patterned with SiO, was possible.
To demonstrate the selectivity, silicon wafers covered with 1.8 p
thick patterned SiO, were used. The contact holes were filled at
a substrate temperature of 230 "C with germanium without any
deposition onto the SiO, surface. X-ray diffraction analysis
showed that all films were amorphous without detectable crystallite (sensitivity about 1 ~ 0 1 % )The
dark conductivity of the
germanium film on SiO, ("ball-like film") reached values of
6.6 x
Scm-'. This relatively high value and low activation
energy of 0.20 eV are due to the dendritic-like nature of these
films. The compact films deposited on clean SiO, surfaces had
a relatively high dark conductivity (7.2 x
Scm-', 25°C)
and activation energy (0.25 eV). Hydrogen contents were
around 0.2 atom %.
The decomposition of cyclogermylene B follows a bimolecular mechanism: One germanium atom is deposited, while an
equivalent amount of the C-C-unsaturated derivative A is
formed together with the diamine. The precision of this decomposition seems to arise from a predissociative bond lengthening
along with the occupation of antibonding orbitals residing at
germanium approaching the surface. This corresponds to a
weakening of the Ge-N bonds and consequently promotes decomposition. The same feature was recognized in the case of
the arsenic-cobalt heterobimetallic compound C (Scheme 8),
which formed thin films of cobalt arsenide under CVD conditions at temperatures as low as 210°C.[1381Electron micrographs showed uniform films, energy dispersive X-ray spectroscopy and electron spectroscopy confirmed their bulk
composition, and powder diffraction patterns and conductivity
measurements proved the crystallinity and electrical properties.
As in the case of germanium above, impurities due to carbon,
oxygen, and nitrogen were below 1 %.
9. N-Heterocyclic carbenes-Quo
Heterocyclic carbenes of the imidazole, triazole, and thiazole
families are no longer laboratory curiosities, since general
avenues for synthesis have been opened following Arduengo's
discovery of stable, isolable carbenes. A specific strength of
these compounds is their universal capability to coordinate to
metal centers, ranging from electron-rich transition metals (e. g.
PdO, Rh') to electron-poor main group metal cations (e. g. BeZ+)
and high oxidation state metals such as Ti", NbV, and Re"".
Angew. Chem. Int. Ed. Engl.
N-Heterocyclic Ca-benes
Considering this versatility, N-heterocyclic carbenes surpass
even phosphanes in that pure G donation suffices to form stable
“adducts” with certain metals and nonmetals-just as for
amines and ethers. In addition, unexpectedly stable bonds are
formed with catalytically relevant metals, as exemplified by the
palladium(0) carbene complexes that easily survive temperatures as high as 140 “C under the harsh conditions of catalytic
Heck coupling reactions.
These specific ligand properties make N-heterocyclic carbenes ideally suited for redox-dependent catalytic cycles because
they tolerate rathe: different electronic situations. The future is
thus seen in homogeneous catalysis. Effects to be expected from
ligand functionalization are far from being explored. Side chains
containing hemilabile phosphorus, nitrogen, and oxygen functionalities should trigger the chemistry at the metal center and,
at the same time, enhance the catalyst’s stability.
Chelating dicarbenes have already been verified in the free
state and attached to metal centers. Water-soluble and immobilized derivatives may bridge the gap between homogeneous and
heterogeneous catalysis. Finally, chiral derivatives will exploit
the advantageous flexibility of composition, not least because
simple ring syntheses are available. Chiral carbenes synthesized
from a-phenylethylamine, glyoxal, and formaldehyde work reasonably well as part of a rhodium(1) complex in asymmetric
hydrosilylation, but can certainly be improved by generating a
fixed geometry (no rotation around the N-C bonds). We do not
think that the planar core geometry of this class of ligands, as
compared to the conical geometry of phosphanes, will reduce
the efficiency of chirality transfer, not least because of the
chance to synthesize bulky, chelating derivatives. It is reasonable to assume that N-heterocyclic carbenes will extend the wellestablished phosphanes as catalyst ligands, both alone and in
combination with the latter.
What have we learned from the foregoing discussion? There
are three major reasons why N-heterocyclic carbenes enter a
new era in chemi,stry thirty years after the landmark publications of Wanzlick[81and Ofele:‘71a) They are available as free
compounds from an efficient general route (deprotonation of
azolium salts in liquid ammonia), thus making a plethora of
substituted, functionalized, chiral, or immobilized derivatives
important in catalytic applications accessible. b) They are compatible with IiteralIy any type of main group and transition
metal element of the periodic table, both in low and high formal
oxidation states. However, the syntheses of catalytically relevant metal complexes does not depend on the availability of the
free carbenes, but rather rests on several indirect ways to convert
an azolium precursor into the metal-bound carbene. c) They
form highly efficient catalysts with certain metals; palladium(I1)
complexes of the type [PdL,X,J were first used successfully in
the Heck-type C -C coupling,[’31 and another example is the
enantioselective hydrosilylation.[’281
We conclude that N-heterocyclic carbenes have long been an
overlooked class of catalytically relevant ligands in coordination chemistry. Although Wanzlick pointed out in his very first
publication (1968) that nucleophilic carbenes are “close relatives of isonitriles”(!), they have always been viewed as curiosities. Their renaissance in the past several years, however, justifies the assessment that they will play a major role in tomorrow’s
organometallic chemistry.
AnRew. Chem. Inr. Ed. Engl. 1997, 36, 2162-2187
The authors are grateful to Dr. K . Ofele (Technische Universitat Miinchen) for his continuous participation in our research
related to N-heterocyclic carbenes. We acknodedge generous
support by the Deutsche Forschungsgemeinschaft and the Fonds
der Chemischen Industrie. The (former) students Dr. M . Elison,
Dr. J. Fischer, Dr. G. R . J. Artus, Dr. D. Miiialios, C.-P.
Reisinger, G. M . Lobmaier, Dr. 0. Runte, M . Steinbeck, 7:
Weskamp, R. Eckl, M . Prinz, and G. Gerstberger deserve our
gratitude for their experimentally skillful contributions to our
work on the coordination chemistry of N-heterocyclic carbenes
and their applications in catalysis. We thank Dr. R. Schmidt and
Dr. L. J. Goosen for contributions to the manuscript.
Received: May 12, 1997 [A230IE]
German version: Angew. Chem. 1997, 109,2256-2282
111 P. S. Skell, S. R. Sandler, J Am. Chem. Soc. 1958, 80, 2024.
[2] a) E. 0. Fischer, A. Maasbol, Angew. Chem. 1964,76,645: Angew. Chem. Int.
Ed. Engl. 1964, 580; b) E. 0. Fischer, ibid. 1974,86. 651 (Nobel lecture).
[3] Applied Homogeneous Catalysis with Organometallic Cumplrxes (Eds. :
B. Cornils, W. A. Herrmann), VCH, Weinheim. 1996.
[4] R. H Grubbs in Comprehensive Organometallic Chemistry, Yol. 8 (Eds.: G.
Wilkinson, F. G. A. Stone, E. W. Abel), Pergamon, Oxford, 1982, p. 499.
[S] M. Brookhart, W. B. Studabaker, Chem. Rev. 1987,87, 411.
[6] K. H. Dotz in OrganometaNics in Organic Synthesis (Eds.: A. de Meijere,
H. tom Dieck), Springer, Berlin, 1988.
[7] K Ofele, J. Organomet. Chem. 1968, 12, P42
[8] H:W. Wanzlick, H.-J. Schonherr, Angew. Chem. 1968,80.154;Angew. Chem.
Inr. Ed. Engl. 1968, 7, 141.
191 A. J. Arduengo 111, M. Kline, J. C. Calabrese, F. Davidson, J Am. Chem. SOC.
1991, 113,9704.
1101 W A. Herrmann, M. Denk, J. Behm, W. Scherer, F. R. Klingan, H. Bock, B.
Solouki, M. Wagner, Angew. Chem. 1992, 104, 1489; Angew. Chem. Inr. Ed
Eng/ 1992, 31,485
[ l l ] a) M. Denk, R. Lennon, R. Hayashi, R. West, A. V. Belyakov, H. P. Verne,
A Haaland, M. Wagner, N. Metzler, J Am. Chem. Soc. 1994,116,2691;b) B.
Gehrhus, M. F. Lappert, J. Heinicke, R. Boese, D. Blaser. J. Chem. SOC.Chem.
Commun. 1995, 1931; c) B. Gehrhus, P. B. Hitchcock, M. F. Lappert,
J. Heinicke, R. Boese, D. Blaser, J Organomer. Chem. 1996, 521, 21 I .
[12] a) K. Ofele, C. G. Kreiter, Chem. Ber. 1972, 105, 529; b) K Ofele, M. Herberhold, 2. Narurforsch. 1973, 28b, 306; c) C G. Kreiter, K. Ofele, G . W.
Wieser, Chem. Ber. 1976, 109, 1749; d) K. Ofele, M. Herberhold. Angew.
Chem. 1970,82, 775; Angew. Chem. h t . Ed. Engl. 1970, Y, 739; e) K. Ofele,
E. Roos, M. Herberhold, Z. Naturforsch. 1976,31b, 1070; f) K. Ofele, W. A.
Herrmann, D. Mihalios, M. Elison, E. Herdtweck, W. Scherer, J. Mink, J
Organomet. Chem. 1993,459,177;g) N. Kuhn, T. Kratz. R Boese, D. Blaser,
ibrd. 1994, 470, C8; h) ibid. 1994, 479, C32; i) F. E. Hahn, L. Imhof,
Organometallics1997,16,763;j) F. E. Hahn, M. Tamm, J Organomet. Chem.
1993, 4S6, C l 1 ; k) J Chem. Soc. Chem. Commun. 1993.42.
[13] W A. Herrmann, M. Elison, J. Fischer, C. Kocher, G. R. J. Artus, Angew.
Chem. 1995, 107, 2602; Angew. Chem. Int. Ed. Engl. 1995, 34, 2371.
[14] a) M. Gomberg, Chem. Ber. 1900, 33, 3150; b) A. R. Forrester, J. M. Hay,
R. H. Thomso, Organic Chemistry of Stable Radicals, Academic Press, London, 1968.
[lS] D. Griller, K. U. Ingold, Acc. Chem. Res. 1976, 9, 13.
1161 D. Seyferth, Acc. Chem. Res. 1971, 5, 65.
[17] W. E. Parham, E. E. Schweizer, Organic Reactions, Vol. 13, Wiley, New York,
[18] W. Kirmse, Carbene Chemislry, Academic Press, New York, 1971.
[I91 K. Hirai, K. Komatsu, H. Tomioka, Chem. Lett. 1994. 503.
[20] H. Tomioka, T.Watanabe, K. Hirai, K Furukawa, T. Takuj. K . Itoh, J Am.
Chem. Sor. 1995, 117,6376.
1211 A. J. Arduengo 111, H. Bock, H. Chen, M. Denk, D. A. Dixon, J. C. Green,
W. A. Herrmann, N. L. Jones, M. Wagner, R. West, J Am. Chem. SOC.1994,
(221 E. A. Carter, W A. Goddard, J. Phys. Chem. 1986, 90. 998.
[23] S. S. Krishnamurthy, Curr. Sci. 1991, 60, 619.
1241 R. A. MOSS,T. Zdrojewski, G. J. Ho, J Chem. So<.Chem. Commun. 1991.946.
[25] M. Elsaidi, K. Kassam, D. L. Pole, T. Tadey, J. Warkentin, J Am. Chem. Soc.
1992, 114,8751.
[26] Y Cheng, S. Goon, 0. Methcohn, Chem. Commun. 1996, 1395.
[27] a) H. W. Wanzlick, E. Schikora, Angew. Chem. 1960, 72,494; b) M. F. Lappert, R. W. McCabe, J. J. MacQuitty, P. L. Pye, P. I. Riley, J Chem. So<.
Dalron Trans. 1980, 90; c) M. F. Lappert, P. L. Pye, J Chem. SOC.Dalton
Trans. 1978, 837, and references therein.
[28] H. Quast, S . Hiinig, Angew. Chem. 1964,76,989;Angew. Chem. I n / . Ed. Engl.
1964,3, 800.
[29] H.-W. Wanzlick, H.-J. Schonherr, Justus Liebigs Ann. Chem. 1970, 731, 176.
[30] M. Regitz, Angew. Chem. 1991, 103, 691; Angew. Chem. Int. Ed. Engl. 1991,
30, 674.
[31] R. Gleiter, R. Hoffmann, J Am. Chem. Soc. 1968, 90, 5457.
[32] J. V. Nef, Justus Lzebigs Ann. Chem. 1895, 287, 359.
[33] A. J. Arduengo 111, R. L. Harlow, M. Kline,J. Am. Chem. Soc. 1991,113,361.
[34] W. A. Herrmann, C. Kocher, L. Goossen, G. R. J. Artus, Chem. Eur. J 1996,
2, 1627.
[35] N. Kuhn, T. Kratz, Synthesis 1993, 561.
[36] A. J. Arduengo III, H. V. R. Dias, R L. Harlow, M. Kline, J. Am. Chem. Soc.
1992, 114, 5530.
[37] W A. Herrmann, M. Elison, J. Fischer, C. Kocher, Chem. Eur. 1 1996,2,772.
[38] J. Fuller, R. T. Carlin, H. C. DeLong, D. Haworth, J. Chem. Soc. Chem.
Commun. 1994,299.
[39] Y Chauvin, L. Mussmann, H. Olivier, Angew. Chem. 1995,107,2941; Angew,.
Chem. In[. Ed. Engl. 1995, 34, 2698.
[40] Y Chauvin, H. Olivier-Bourbigou, Chemtech 1995, 26.
A. J. Arduengo 111, D. A. Dixon, K. K. Kumashiro, C. Lee, W P. Power,
K. W. ZiIm, J. Am. Chem. Soc. 1994, 116, 6361.
N. Burford, P. Losier, C. Macdonald, V. Kyrimis, P. K. Bakshi, T. S.
Cameron, Inorg. Chem. 1994,33, 1434.
M. K. Denk, S . Gupta, R. Ramachandran, TetrahedronLett. 1996,37, 9025.
D. S. Brown, A. Decken, A. H. Cowley, J. Am. Chem. Soc. 1995, 117, 5421.
D. S. Brown, A. Decken, C. A. Schnee, A. H. Cowley, Inorg. Chem. 1995,34,
D. E. Goldberg, P. B. Hitchcock, M. L. Lappert, K. M. Thomas, A. J.
Thorne, T. Fjeldberg, A. Haaland, B. E. R. Schilling. J. Chem. Soc. Dalton
Trans. 1986, 2387.
P. J. Davidson, D. H. Harris, M. F. Lappert, .
Chem. Soc. Dalton Trans. 1976,
C. W Bauschlicher, Jr., H. F. Schaefer 111, P. S. Bagus, J Am. Chem. Soc.
1977,99, 7106.
V. Langer. K. Huml, G. Reck, Acta Crystallogr. Sect. B 1982,38, 298.
R. W. Alder, P. R. Allen, M. Murray, A. G. Orpen, Angew. Chem. 1996,108,
1211; Angew. Chem. Int. Ed. Engl. 1996, 35, 1121.
D. A. Dixon, K. D. Dobbs, A. J. Arduengo III, G. Bertrand, J. Am. Chem.
Soc. 1991,113, 8782.
T. A. Taton, P. Chen, Angew. Chem. 1996,108,1098; Angew. Chem. In[. Ed.
Engl. 1996,35, 1011.
2. Q. Shi, V. Goulle, R. P. Thummel, Tetrahedron Lett. 1996, 37, 2357.
E. Cetinkaya, P. B. Hitchcock, H. Kiicukbay, M. F. Lappert, S . AI-Juaid, J.
Organornet. Chem. 1994,481, 89.
A. J. Arduengo 111, J. R. Goerlich, W. J. Marshall, J Am. Chem. Soc. 1995,
117, 11027.
D. Enders, K. Breuer, G. Raabe, J. Runsink, J. H. Teles, J. P. Melder, K. Ebel,
S. Brode, Angew. Chem. 1995,107, 1119; Angew. Chem. Int. Ed. Engl. 1995,
34, 1021.
D. Enders, K. Breuer, G. Raabe, J. Sirnonet, A. Ghanirni, H. B. Stegmann,
J. H Teles, personal communication, 1996.
G. Maier, J. Endres, H. P. Reisenauer. Angew. Chem. 1997,109,3788; Angew
Chem. Int. Ed. Engl. 1997. 36, 1709.
G. A. McGibbon, J. Hrusak, D. J. Lavorato, H. Schwarz, J. K. Terlouw,
Chem. Eur. J 1997, 3, 232.
A. J. Arduengo 111, J. R. Goerlich, W. J. Marshall, Liebigs Ann. 1997, 365.
W. A. Herrmann, L. J. Goossen, G. R. J. Artus, C. Kocher, Organometallics
1997, 16, 2472.
C. Kocher, Dissertation, Technische Universitat Munchen, 1997.
D. Michalios, Dissertation, Technische Universitat Miinchen, 1992.
W. A. Herrmann, C. P. Reisinger, unpublished results. 1997.
C. Kocher, W. A. Herrmann, J. Organomet. Chem. 1997,532, 261.
W. A. Herrmann, T. Weskamp, unpublished results, 1997.
W. A. Herrmann, T. Weskamp, R. Eckl, unpublished results, 1996.
U. Kernbach, M. Ramm, P. Luger, W. P. Fehlhammer, Angew. Chem. 1996,
108, 333; Angew. Chem. Int. Ed. Engl. 1996,35, 310.
H. V. R. Dias, W. Jin, Tetrahedron Lett. 1994, 35, 1365.
a) K. K. Irikura, W. A. Goddard 111, J. L. Beauchamp, J. Am. Chem. Soc.
1992, 114, 48; b) C. W. Bauschlicher, Jr., S. R. Langhoff, P. R. Taylor,
J. Chem. Phys. 1987, 87, 387; c) S . K. Shin, W. A. Goddard III, J. L.
Beauchamp, ibid. 1990, 93, 4986; d) [bid. 1990, 94, 6963.
D. A. Dixon, A. J. Arduengo 111, J Phys. Chem. 1991,95,4180.
J. Cioslowski, Int. J Quantum Chem. 1993, 27, 309.
A. J. Arduengo 111, H. V. R. Dias, D. A. Dixon, R. L. Harlow, W. T. Klooster,
T. F. Koetzle, J. Am. Chem. Soc. 1994, 116, 6812.
C. Heinemann, W. Thiel, Chem. Phys. Lett. 1994, 217, 11.
R. R. Sauers, Tetrahedron Lett. 1996, 37, 149.
C. Heinemann, T. Muller, Y Apeloig, H. Schwarz, J. Am. Chem. Soc. 1996,
C. Boehme, G. Frenking, J Am. Chem. Soc. 1996, 118, 2039.
C. Heinemann, W. A. Herrmann, W. Thiel, J. Organornet. Chem. 1994,475,73.
W. A. Herrmann and C. Kocher
1791 M. Denk, J. C. Green, N. Metzler, M. Wagner, J Chem. Soc. Dalton Trans.
1994, 2405.
[SO] A. R. Katritzky, M. Karelson, N. Malhotra, Hererocycles 1991, 32, 127.
[811 S . Berger, U. Fleischer, C. Geletneky, J. C. W. Lohrenz, Chem Ber. 1995,128,
[82] R. Hoffmann, J. Am. Chem Soc. 1968, 90, 1475.
I831 R. Hoffmann, R. B. Woodward, Ace. Chem. Res. 1968, 1, 17.
[84] G. Bertrand, R. Reed, Coord. Chem. Rev. 1994, 137, 323
[85] G. Alcaraz, U. Wecker, A. Baceiredo, F. Dahan, G. Bertrand, Angew. Chem.
1995, 107, 1358; Angew. Chem. Int. Ed. Engl. 1995, 34, 1246.
I861 P. Dyer, A. Baceiredo, G. Bertrand, Inorg. Chem. 1996, 35, 46.
[87] R. W. Alder, P. R. Allen, S. J. Williams, 1 Chem. Soc. Chem. Commun. 1995,
[SS] R. Breslow, J Am. Chem. Soc. 1958, 80, 3719.
[89] R. Kluger, Chem. Rev. 1987,87, 863.
[90] W. Kreiser, Nachr. Chem. Tech. Lab. 1981, 29, 172.
[91] H. Stetter, Angew. Chem. 1976,84695; Angew. Chem. Int. Ed. Engl. 1976.15,
[92] H. Stetter, H. Kuhlmann, Org. React. ( N Y ) 1991, 40, 407.
[93] J. H. Teles, J.-P. Melder, K. Ebel, R. Schneider, E. Gehrer, W. Harder,
S. Brode, D. Enders, K. Breuer, G. Raabe, Helv. Chim. Acta 1996,
79, 61.
[94] a) D. Enders, K. Breuer, J. H. Teles, Helv. Chim. Acta 1996, 79, 1217; b) D.
Enders, K. Breuer, J. Runsink, J. H. Teles, ibid. 1996, 79, 1899.
[95] N. Kuhn, T. Kratz, G. Henkel, J Chem. Soc. Chem. Commun. 1993, 1778.
[96] A. J. Arduengo 111, M. Tamm, J. C. Calabrese, J Am. Chem. Soc. 1994, 116,
[97] A. J. Arduengo 111, S . F. Gamper, M. Tamm, J. C. Calabrese, F. Davidson,
H. A. Craig, J Am. Chem. SOC.1995, 117,572.
[98] W A. Herrmann, P. W. Roesky, M. Elison, G. R. J. Artus, K. Ofele,
Organometallics 1995, 14, 1085.
W. A. Herrmann, 0. Runte, G. R. J. Artus, J Organornet. Chem. 1995, 501,
A. J. Arduengo 111, H. V. R. Dias, F. Davidson, R. L. Harlow, J. Organomet.
Chem. 1993,462,13.
0. Runte, Dissertation, Technische Universitat Miinchen, 1997.
N. Kuhn, G. Henkel, T. Kratz, J. Kreutzberg, R. Boese, A. H. Maulitz, Chem.
Ber. 1993, 126, 2041.
A. J. Arduengo 111, H. V. R. Dias, J. C. Calabrese, F. Davidson, J. Am. Chem.
Soc. 1992, 114, 9724.
X.-W. Li, J. Su, G. H. Robinson, Chem. Commun. 1996, 2683.
a) N. Kuhn, H. Bohnen, J. Kreutzberg, D. Blaser, R. Boese, J. Chem. Soc.
Chem. Commun. 1993, 1136; b) N. Kuhn, H. Bohnen, D. Blaser, R. Boese,
Chem. Ber. 1994, 127, 1405; c) N. Kuhn, H. Bohnen, G. Henkel, I.
Kreutzberg, 2. Naturforsch. B 1996,51, 1267; d) H. Schumann, M. Glanz, J.
Winterfeld, H. Hemling, N. Kuhn, H. Bohnen, D. Blaser, R. Boese, J
Organomet. Chem. 1995,493, C14; e) N. Kuhn, H. Bohnen, G . Henkel, 2.
Naturforsch. B 1994, 49, 1473; f) N. Kuhn, G. Weyers, G. Henkel, Chem.
Commun. 1997,627.
A. J. Arduengo III, H. V. R. Dias, J. C. Calabrese, F. Davidson, Inorg. Chem.
1993, 32, 1541.
a) N. Kuhn, T. Kratz, D. Blaser, R. Boese, Chem. Ber. 1995, 128, 245; b) A.
Schafer, M. Weidenbruch, W. Saak, S . Pohl, J. Chem. Soc. Chem. Commun.
1995, 1157.
Se: a) N. Kuhn, G. Henkel, T. Kratz, 2. Naturforsch. B 1993, 48, 973;
b) D. J. Williams, M. R. Fawcett-Brown, R. R. Raye, D. VanDerveer, Y T.
Pang, R. L. Jones, K. L. Bergbauer, Heteroat. Chem. 1993,4,409; Te: c) N.
Kuhn, G. Henkel, T. Kratz, Chem. Ber. 1993, 126, 2047.
D. Enders, K. Breuer, J. H. Teles, K. Ebel, Org. Synth. 1997, submitted.
A. J. Arduengo III, M. Tamm, S. J. McLain, C. J. Calabrese, F. Davidson,
W. J. Marshall, J. Am. Chem. Soc. 1994, 116, 7927.
H. Schumann, M. Glanz. J. Winterfeld, H. Hemling, N Kuhn, T. Kratz,
Angem. Chem. 1994, 106, 1829; Angew. Chem. Int. Ed. Engl. 1994,33, 1733.
H. Schumann, M. Glanz, J. Winterfeld, H. Hemling, N. Kuhn, T. Kratz,
Chem. Ber. 1994, 127, 2369.
R. D. Fischer, Angew. Chem. 1994, 106, 2253; Angew. Chem. Int. Ed. Engl.
1994, 33, 2165.
F. Munck, Dissertation. Technische UniversitLt Munchen, 1996.
G. R. J. Artus, Dissertation, Technische Universitat Munchen, 1996.
D. Enders, H. Gielen, G. Raabe, J. Runsink, J. H. Teles, Chem. Ber. 1996, 129,
W. P. Fehlhammer, K. Bartel, A. Volkl, D. Achatz, 2. Naturforsch. B 1982,
37, 1044.
I. Ugi. K. Offermann, Isonitrile Chemistry, Vol. 20, Academic Press, London,
D. Rieger, S . D. Lotz, U. Kernbach, C. Andre, J. Bertran-Nadal, W. P.
Fehlhammer, J Organornet. Chem 1995, 491, 135.
W. P. Fehlhammer, M. Fritz, Chem. Rev. 1993, 93, 1243.
N. Matsumura, J.-I. Kawano, N. Fukunishi, H Inoue, J Am Chem. Soc.
1995, 117, 3623.
[122] D. Sellmann, W Prechtel, F. Knoch, M. Moll, Inorg. Chem. 1993, 32, 538.
Angew. Chem. I n t . Ed. Engl. 1997, 36,2162-2187
N-Heterocyclic Carbenes
11231 A. Ghosh, Angcw. Chem. 1995, 107,1117; Angen. Chem. Int. Ed. Engl. 1995,
11241 J. E. Hill. T. A. Nile, J. Organomer. Chem. 1977, 137, 293.
11251 M. F. Lappert in Transirion Meral Chemistry (Eds.: A. Miiller, E. Diemann),
Verlag Chemie. Weinheim, 1981.
11261 B. Cetinkaya, 1. Ozdemir, C. Bruneau, P. H. Dixneuf, J. Mol. Catal. A 1997,
118, L1.
11271 M. F. Lappert, 3.. K. Maskell, J Organomer Chem. 1984,264,217.
11281 W. A. Herrmdnn, L. J. Goossen, C. Kocher, G. R. I. Artus, Angew. Chem.
1996, 108. 2980, Angew. Chem. Inr. Ed. Engl. 1996,35, 2805.
11291 L. Goossen, Dissertation, Technische Universitat Munchen, 1997.
[130] W A. Herrmann, M. Steinbeck, unpublished results, 1997.
[131] A. M. Trzeciak, J. J. Ziolkowski, Abstr. Pup. XX. Colloquy on Organometalhc
ChemDfri..Germany-Poland (Halle-Wittenberg), 1996, p. 11
11321 C. Kocher, Dissertation, Technische Universitat Miinchen, 1997.
1133) W. A. Herrmann, unpublished results.
11341 F.-R Klingdn, I)issertation, Technische Universitat Munchen, 1995.
[I351 J. Procop. R Merica, F. Glatz, S . Veprek, F.-R. Klingan, W. A. Herrmann, J
Non-Cryst Solids 1996, 2638.
11361 S . Veprek, J. Procop, R. Menca, F. Glatz, F-R. Klingan, W. A. Herrmann,
Mac. RES.SOC.Symp. Proc. 1994, 336, 541
11371 S. Veprek, J. Procop, R. Merica, F. Glatz, F:R. Klingan, W. A. Herrmann,
A b s w Pap. CVD-I3 Amerzcan Chemical Society Sprlng Meefing,(Los Angeles, CA) 1996
11381 F.-R. Klingdn, P,. Miehr, R. A. Fischer, W. A. Herrmann, Appl. Phys. L e r r .
1995.67. 822
[139] M. Elison, Dissertation, Technische Universitat Munchen, 1995.
I1401 C. Elschenbroich, A. Salzer, Orgunomerallchemie, 3. Aufl., Teubner, Stuttgart, 1990, p. 279.
[I411 M. Driess, H. Griitzmacher, Angew. Chem. 1996,108,900; Angew. Chem. Inr.
Ed. Engl. 1996,35, 828.
[I421 R. Boese, unpublished results.
11431 L. Weber, E. Dobbert, H.-G. Stammler, B. Neumann, R. Boese, D. Blaser,
Chem. Ber. 1997, 130, 705.
11441 M. Tamm, T. Lugger, F. E. Hahn, Organometallics 1996, 15, 1251.
[I451 N. Metzler, M. Denk, Chem. Commun. 1996, 2657.
11461 a) N. Kuhn, R. Fawzi, M. Steimann, J. Wiethoff, D. Blaser, R. Boese, Z.
Narurforsch. B 1995,50, 1779; b) N. Kuhn, R. Fawn, M. Steimann, J. Wiethoff, 2. Anorg. Allg. Chem. 1997, 623, 769; c) N. Kuhn, R. Fawzi, M.
Steimann, J. Wiethoff, Chem. Ber. 1996, 129, 479.
[147] A. J. Arduengo 111, R Krafczyk, W. J. Marshall, R. Schmutzler, J. Am. Chem.
Soc. 1997,119, 3381.
[148] a) N. Kuhn, T. Kratz, G. Henkel, Chem. Ber. 1994, 127,849; b) N. Kuhn, T.
Kratz, G. Henkel, Z . Naturforsch. B 1996, 51, 295.
[149] N. Kuhn, R. Fawzi, T. Kratz, M. Stelmann, Phosphorus Sulfur Silicon Relal.
Elem. 1996, 108, 107.
[150] a) W. A. Herrmann, K. Ofele, M. Elison, F E. Kiihn, P. W. Roesky, J.
Organomer. Chem. 1994,480, C7; b) N. Kuhn, T. Kratz. D. Blaser, R. Boese,
Inorg. Chim. A m 1995, 238, 179.
[151] P. Luger, G. Ruban, Acra Crystallogr. Serr. B 1971, 27, 2276
Deposition of Data from X-Ray Structure Analyses
In order to make life easier for authors and referees the Cambridge Crystallographic Data
Centre (CCDC) and the Fachinformationszentrum Karlsruhe (FIZ) have unified their procedures for the deposition of data from single-crystal X-ray structure analyses.
Prior to submitting your manuscripts please deposit the data for your compound(s) electronically
at the appropriate data base, that is, at the CCDC for organic and organometallic compounds
and at the FIZ for inorganic compounds. Both data bases will be pleased to provide help (see
our Notice to Authors in the first issue of this year). In general, you will receive a depository
number from the data base within two working days after electronic deposition; please include
this number with the appropriate standard text (see our Notice to Authors) in your manuscript.
This will enable the referees to retrieve the structure data quickly and efficiently if they need this
information to reach their decision.
This is now the uniform procedure for manuscripts submitted to the journals Advanced
Materials, .Angewandte Chemie, Chemische BerichtelRecueil, Chemistry-A European Journal,
and Liebigs AnnalenlRecueil.
Angen,. Cllem. Int. Ed Engl 1997, 36, 2162-2187
21 87
Без категории
Размер файла
2 989 Кб
carbene, heterocyclic
Пожаловаться на содержимое документа