N,N′-dialkyl- and N-aIkyl-N-mesityl-substituted N-heterocyclic carbenes as ligands in Grubbs catalysts

Article abstract of DOI:10.1002/chem.200600064

Various symmetrically and asymmetrically substituted N-heterocyclic carbene (NHC) ligands bearing aliphatic nitrogen-containing side groups have been synthesised. In our attempts to isolate the corresponding second-generation Grubbs catalysts, we were unsuccessful when using the symmetrical aliphatic NHC ligands. For the asymmetrical ligands bearing an aliphatic moiety on one side and an aromatic mesityl group on the other side, substitution of a phosphine ligand was achieved. The performance of a soformed series of Ru-based metathesis initiators has been evaluated for the ring-opening metathesis polymerisation (ROMP) of cycloocta-1,5-diene and the ring-dosing metathesis (RCM) of diethyl diallylmalonate.

Full text of DOI:10.1002/chem.200600064

DOI: 10.1002/chem.200600064
N,N’-Dialkyl- and N-Alkyl-N-mesityl-Substituted N-Heterocyclic Carbenes
as Ligands in Grubbs Catalysts
Nele Ledoux,* Bart Allaert, Siegfried Pattyn, Hans Vander Mierde, Carl Vercaemst, and
Francis Verpoort[a]
Abstract: Various symmetrically and
asymmetrically substituted N-heterocy-
clic carbene (NHC) ligands bearing ali-
phatic nitrogen-containing side groups
have been synthesised. In our attempts
to isolate the corresponding second-
generation Grubbs catalysts, we were
unsuccessful when using the symmetri-
cal aliphatic NHC ligands. For the
asymmetrical ligands bearing an ali-
phatic moiety on one side and an aro-
matic mesityl group on the other side,
substitution of a phosphine ligand was
achieved. The performance of a so-
formed series of Ru-based metathesis
initiators has been evaluated for the
ring-opening metathesis polymerisation
(ROMP) of cycloocta-1,5-diene and
the ring-closing metathesis (RCM) of
diethyl diallylmalonate.
Keywords:
carbene ligands
·
metathesis · ruthenium
Introduction
lyst design was the substitution of one phosphine ligand by a
bulky N-heterocyclic carbene (NHC) ligand, leading to in-
creased activity and stability. The H2IMes-substituted
(H2IMes=1,3-bis-(mesityl)-4,5-dihydroimidazol-2-ylidene)
complex 2 is nowadays widely known as the Grubbs second-
generation catalyst.[4]
In spite of their booming success, variation of NHC li-
gands has remained rather unexplored for Grubbs second-
generation type catalysts, while it is evident that this varia-
tion could have a significant influence on the catalytic activi-
ty and selectivity.[5] Symmetrical dihydro NHC ligands seem
to be limited to aromatic N-substituents (e.g., mesityl,[4] 2,6-
diisopropylphenyl[6]). Aiming at further improvement of the
application profile of Grubbs second-generation catalysts,
we decided to pursue the coordination of dihydro NHC li-
gands bearing aliphatic groups with different steric bulk.
However, a previous attempt towards coordination of an ali-
phatic NHC ligand by Mol et al. was not promising. They
described the synthesis of 1,3-di(1-adamantyl)-4,5-dihydro-
imidazolinium chloride [H2IAd(H)][Cl] and the unsymmet-
Olefin metathesis has evolved towards a powerful tool for
the formation of carbon–carbon bonds, resulting in the
recent Nobel Prize awarding of Y. Chauvin, R. R. Schrock
and R. H. Grubbs, who made groundbreaking contributions
to the development of olefin metathesis reactions and intro-
duced a large number of catalytic initiators.[1,2] The intensive
research effort has culminated in the discovery of well-de-
fined ruthenium–carbene catalysts, such as the Grubbs first-
generation catalyst 1.[3] An important breakthrough in cata-
rical
1-(1-adamantyl)-3-mesityl-4,5-dihydroimidazolinium
chloride [H2IAdMes(H)][Cl].[7] It should be noted that only
H2IAdMes reacted in a favourable manner to give its
second-generation analogue 3. Failure of the reaction with
H2IAd was assigned to the bulkiness of the adamantyl
moiety which could not take place directly over the benzyl-
idene unit. The X-ray structure of 3 showed that the only
isomer formed had the mesityl group above the benzylidene
moiety. Surprisingly, complex 3 displayed only negligible
metathesis activity, which was ascribed to steric blocking.
[a] N. Ledoux, B. Allaert, S. Pattyn, H. V. Mierde, C. Vercaemst,
Prof. Dr. F. Verpoort
Department of Inorganic and Physical Chemistry
Laboratory of Organometallic Chemistry and Catalysis
Ghent University, Krijgslaan 281 (S-3), 9000 Ghent (Belgium)
Fax : (+32)9-264-4983
E-mail: nele.ledoux@UGent.be
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. The Supporting
Information contains further synthetic details for all NHC precursors.
4654
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4654 – 4661

FULL PAPER
We were hoping that a reduced bulkiness of the N-substitu-
ents would allow easier NHC coordination and have a posi-
tive effect on the metathesis activity of the resulting cata-
lysts. Higher electron density at the carbenic centre of satu-
rated NHCs compared to their unsaturated analogues[8] in
combination with even further enhancement of this electron
donation caused by alkyl N-substituents[9] would allow an in-
crease in catalyst activity. However this constitutes a subject
of discussion, since Nolan et al. reported unexpected weaker
À
Pd C(NHC) bonds for electron-donating alkyl-substituted
G
NHCs.[10] Even more noteworthy was their study of the CO
stretching frequencies in [Ni(CO)3
(NHC)] complexes.[11]
U
Alkyl-substituted NHCs were found to be only marginally
more electron donating than aryl-substituted ones. Further-
more saturated NHCs turned out slightly less electron do-
nating than their unsaturated counterparts; this result is not
in line with the common assumption that metal complexes
bearing a saturated NHC perform better in catalytic reac-
tions because of a higher electron donation. This demon-
strates that considering only ligand basicity would be an
oversimplification of the metal–NHC bonding properties.[12]
In this respect, it is worth mentioning that Plenio et al. re-
cently concluded from cyclic voltammetry experiments that
the donor properties of aryl-substituted NHC ligands are
characterised by s-donation from the carbene carbon atom
and by transfer of electron density between the aromatic
NHC side groups and the Ru=CHPh unit.[13] It should also
be noted that there is so far no straightforward explanation
for the enhanced reactivity of Grubbs second-generation
catalysts, although a lot of research is ongoing in this partic-
ular area.[14]
Figure 1. 31P spectrum showing partial reaction of 1 with ligand 4a.
Results and Discussion
The synthesised NHC precursors include 1,3-diisopinocam-
pheyl-4,5-dihydroimidazolinium chloride (4a), 1,3-di-tert-
butyl-4,5-dihydroimidazolinium chloride (4b), 1,3-dicyclo-
hexyl-4,5-dihydroimidazolinium chloride (4c), 1,3-di-n-octyl-
4,5-dihydroimidazolinium chloride (4d) and the pinane-
based imidazolinium chloride 4e. Ligands 4a and 4e could
induce some enantioselectivity,[15] which might be driven to
a higher extent by modification of the pinane-derived
moiety. As a base to deprotonate the imidazolium chloride,
we used potassium bis(trimethylsilyl)amide (KHMDS). This
base liberates the free carbene at room temperature and its
steric bulkis high enough to prevent fast reaction with
Grubbs catalyst 1.[16]
Reaction of one equivalent of 4a with one equivalent of
base and 1 in dry toluene did not allow substitution of the
phosphine, even when the reaction mixture was heated or
stirred for several hours. An excess of NHC ligand
(1.5 equiv) led to the observation of a new benzylidene a-
proton at d=20.61 ppm with a conversion of approximately
25%. The 31P NMR spectrum showed a new signal at d=
20.14 ppm as well as a signal corresponding to free PCy3
(Figure 1). This was assigned to NHC coordination. Howev-
er, the downfield shift of the benzylidene signal surprised us,
since generally for Grubbs second-generation complexes
this peakis situated more upfield. The addition of more
than a twofold excess of ligand allowed full conversion, but
at this point difficulties to remove the ligand excess and free
phosphine arose. Effort to isolate pure compound by precip-
itation remained unsuccessful due to high solubility in all
common organic solvents. The complex decomposed upon
chromatography on silica gel.
Several attempts to exchange one phosphine with H2ItBu
4b, applying an excess of ligand as well as prolonged reac-
tion times, were unsuccessful. Hahn et al. recently reported
the synthesis of a H2ItBu substituted rhodium(i) complex,
which showed low stability in solution. This was rationalised
by the steric demand of the N-tert-butyl substituents, result-
ing in a weakbond between the metal and the carbene
carbon atom.[17] Likewise we assume to have encountered
steric obstruction, and we decided to synthesise 4c–4e,
hoping that their different geometry would allow coordina-
tion. Our endeavour involving ligand precursor 4c led to the
observation of a new 1H benzylidene resonance at d=
Chem. Eur. J. 2006, 12, 4654 – 4661
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4655

N. Ledoux et al.
20.28 ppm, and a new 31P signal at d=27.49 ppm. The
13C NMR spectrum of the crude mixture showed a small
doublet at d=210.2 ppm, which corresponds to the NHC
carbene carbon atom coordinated to ruthenium. Various at-
tempts were undertaken to achieve isolation in order to get
full characterisation, but the new complex was found to be
too unstable. Reaction involving 4e gave rise to a new small
1H peakat d=20.08 ppm; this is a little downfield to the
original benzylidene signal. Efforts to obtain full conversion
of 1 by applying longer reaction times led to decomposition
of most of the catalytic species, which made it impossible to
obtain pure product. Also ligand 4d derived from a primary
amine, did not allow the isolation of a NHC-substituted
complex and induced decomposition of the catalytic system.
Given these observations, we decided to synthesise the
unsymmetrical analogues of our NHC precursors, that is, 1-
mesityl-3-isopinocampheyl-4,5-dihydroimidazolium chloride
(5a), 1-mesityl-3-tert-butyl-4,5-dihydroimidazolium chloride
(5b), 1-mesityl-3-cyclohexyl-4,5-dihydroimidazolium chlo-
Compared to the unsymmetrical ligands bearing one mesi-
tyl moiety, the symmetrical ligands 4a–4e were much more
difficult to coordinate to the Grubbs parent complex 1. We
observed coordination of ligands 4a, 4c and 4e, but ligand
excesses were required and the so-formed complexes
showed limited stability, which prevented their isolation.
Therefore, we assume that p interactions between the ben-
zylidene and the amino side group might be of significant
importance. This is confirmed by the observation that the
carbene derived from 5c coordinates in such a way that the
mesityl group is oriented towards the benzylidene moiety
(single-crystal X-ray analysis of complex 6c: Figure 2,
ACHTREUNG
ACHTREUNG
ACHTREUNG
NHCs was straightforward following a synthetic pathway de-
scribed by Grubbs et al.[18] Condensation of ethyl chloroox-
oacetate and 2,4,6-trimethylaniline affords the desired oxa-
nilic ethylester, which is then treated with the aliphatic
amine to provide the corresponding oxalamide. Reduction
and subsequent addition of HCl results in the dihydrochlo-
A
hydroimidazolium chloride in reaction with triethyl ortho-
formate. Exposure of the Grubbs catalyst 1 to these asym-
metrically substituted NHC precursors and KHMDS as a
base afforded the complexes 6a–6e under mild reaction
conditions (Scheme 1)
Figure 2. ORTEP diagram of 6c. For clarity hydrogen atoms have been
omitted.
Table 1) Both aromatic rings are nearly coplanar, allowing
p–p interactions. The observation that intramolecular p–p
stacking between the benzylidene carbene unit and the N-
aryl substituents on the NHC residue might constitute a
strong structural element in second-generation metathesis
Table 1. Selected bond lengths [Š] and angles [8] and sructural compari-
son of 6c, 2 and 3.
6c[a]
2[b]
3{c}
Ru=C
1.830(6)
2.060(5)
2.419(3)
2.376(3)
2.481(3)
1.835(2)
2.085(2)
2.3988(5)
2.3912(5)
2.4245(5)
1.851(5)
2.083(5)
2.427(1)
2.398(1)
2.521(1)
À
Ru CNN
À
Ru Cl(1)
À
Ru Cl(2)
À
Ru P
À
À
Cl Ru Cl
167.73(6)
162.42(14)
98.9(2)
98.63(18)
85.58(15)
89.18(16)
137.2(4)
167.71(2)
163.73(6)
100.24(8)
95.98(6)
83.26(5)
94.55(5)
167.98(5)
171.1(1)
96.9(2)
91.5(2)
84.7(1)
88.1(1)
137.0(4)
À
À
N2C Ru P
À
N2C Ru=C
À
P Ru=C
À
À
N2C Ru Cl(1)
À
À
N2C Ru Cl(2)
Ru=C Ph
À
136.98(16)
Scheme 1. Synthetic pathway towards complexes 6a–6e.
[a] This work. [b] Reference [20]. [c] Reference [7].
4656
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4654 – 4661

Grubbs Catalysts
FULL PAPER
catalysts was previously reported by Fürstner et al. for unsa-
turated NHC entities.[5,19]
In contrast to the Molꢁs complex 3, most of our complexes
showed nice olefin metathesis activities. Their catalytic per-
formance in the ring-opening metathesis polymerisation
(ROMP) of cycloocta-1,5-diene (COD) was compared with
the reactivity of Grubbs catalysts 1 and 2, using different
solvents and monomer/catalyst ratios (Figures 3 and 4). Co-
ordination of the NHC ligand led to a positive effect on the
ROMP actvity for complexes 2, 6a, 6c, 6d and 6e. In con-
trast, for complexes 3 and 6b, NHC coordination induced
activity loss. These last two complexes required heating and
low COD/catalyst ratio in order to obtain high conversion.
(Table 2, entries 14 and 15)
Figure 4. Monitoring ROMP of COD by 1H NMR spectroscopy (208C).
Conditions: monomer/catalyst=3000, catalyst concentration=0.452 mm.
~
^
*
;
Top: solvent=CDCl3. Bottom: solvent=C6D6. 2: &; 6a: ; 6c: ; 6d:
*
.
6e:
It is worth noting that the catalytic activity of complex 2
strongly depends on the solvent used, while a solvent effect
is less significant for complexes 6a–6e. The catalysts 6a, 6c,
6d and 6e show a slightly higher ROMP-activity in CDCl3
than in C6D6 when a monomer/catalyst ratio of 300 is used.
Their activity is somewhat higher in C6D6 when the mono-
mer/catalyst ratio equals 3000, which is due to an induction
period and loss of activity in the CDCl3 polymerisation.
Complex 2 is unambiguously more active in C6D6. Such an
increased reactivity in aromatic solvents was also observed
Figure 3. Monitoring ROMP of COD by 1H NMR spectroscopy (208C).
Conditions: monomer/catalyst=300, catalyst concentration=4.52 mm.
by Fürstner et al. for [RuCl
2(=CHPh)
T
E
^
Top: solvent=CDCl3. Bottom: solvent=C6D6. 1: ”; 2: &; 6a: ; 6b: *;
~
*
*
.
6c: ; 6d: ; 6e:
was assigned to competing interactions of the N-mesityl
Chem. Eur. J. 2006, 12, 4654 – 4661
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4657

N. Ledoux et al.
Table 2. ROMP of COD.
6a–6e, no rotation of the NHC ligand is possible due to the
high steric demand of the N-alkyl substituents.
Entry Catalyst T [8C] COD/cat. t [min] Conver
cis [%][a]
sion [%]
Furthermore it is evident that when high substrate/catalyst
ratios are used, decomposition of the catalyst system is not
negligible, leading to incomplete conversion as observed for
catalyst 6a.
1
1
RT
3000
30
60
30
30
30
30
30
30
30
23
55
100
0
75
71
9
2
3
4
5
6
7
8
9
2
3
RT
RT
RT
RT
RT
RT
RT
RT
3000
3000
3000
3000
3000
3000
3000
30000
–
13C NMR spectroscopy allowed the determination of the
cis fraction of the newly formed double bonds in the poly-
mer chains (Table 2).[21] The cis/trans ratio can be seen as
the primary microstructural characteristic, having a well-es-
tablished relationship with the solid state and solution prop-
eries of the ROMP polymer.[1a] Grubbs second-generation
catalyst 2 gave rise to a ROMP polymer with a predomi-
nately trans-olefin content, while other catalyst systems gen-
erally led to a higher cis value. A high trans content for 2
was also observed by Grubbs et al., as it could be expected
for an equilibrium-controlled polymerisation in which secon-
dary chain transfer occurs.[22] Entries 1–8 in Table 2 demon-
strate that all NHC-bearing complexes (except 6b) show a
significantly higher trans-content than the first-generation
Grubbs complex 1. For a higher COD/catalyst ratio (en-
tries 9–13) all catalysts but 2 show a predominately cis-
olefin content; these results indicate that for 6a, 6c, 6d and
6e considerably less chain transfer occurred, probably
caused by a quicker decomposition of the catalyst systems.
Itꢁs worth pointing out that complexes 6b and 3 are both
bright green complexes, while the more active complexes 2,
6a, 6c, 6d and 6e are all pinkish. We assume this is the
result of higher steric requirements of the tert-butyl and ada-
mantyl groups. These NHCs are the only ones in the series
in which the first carbon atom adjacent to the amino group
is bonded to three other carbon atoms. Whereas the other
NHC entities can orient their side groups perpendicularly to
the imidazoline plane in order to minimise steric interac-
tions, such an orientation cannot be obtained for the ada-
mantyl- and tert-butyl-bearing ligands.[11] Mol et al. reported
the possibility of an interaction of a b-carbon atom of the
adamantyl group with the metal centre.[7] Analogously 6b
might show a similar interaction resulting in the green
colour and reduced activity of the catalyst.
6a
6b
6c
6d
6e
2
100
0
17
–
100
100
100
86
100
7
30
48
71
74
94
10
26
34
69
72
37
51
46
12
13
–
80
78
76
82
83
–
30
[b]
10
11
12
13
6a
6c
6d
6e
RT
RT
RT
RT
30000
30000
30000
30000
30
[b]
30
[b]
30
[b]
60
120
80
80
75
77
[b]
14
15
3
6b
70
70
300
300
120
120
[a] Percent olefin with cis configuration in the polymer backbone; ratio
based on data from 1H and 13C NMR spectra (13C NMR spectroscopy:
d=32.9 ppm allylic carbon trans; d=27.6 ppm allylic carbon cis).
[b] Overnight.
group with the solvent molecules, which reduce the intramo-
lecular p–p interactions with the benzylidene moiety.[5] As
mentioned earlier, we expect 6a, 6c, 6d and 6e to have sim-
ilar p–p interactions, but when lower COD/catalyst ratios
were applied, we did not observe such a competition effect.
The increased activity of 2 in aromatic solvents might be as-
signed to a p–p interaction between the mesityl group that
is not coplanar with the benzylidene moiety and the aromat-
ic solvent. (Scheme 2) An alternative explanation for the
The RCM activity of the new complexes was tested on
the standard RCM substrate diethyl diallylmalonate.
(Figure 5) A significant dependence of the reactivity on the
bulkiness of the NHC entities was observed. The most
crowded NHCs correspond to the lowest RCM activity,
while activity increases considerably for complexes bearing
less bulky NHCs. The most active catalyst system was found
to be complex 6e, bearing an NHC ligand with a small
methyl amino moiety. This complex was substantially more
active than the Grubbs complex 2. It is therefore undeniable
that the steric bulkof the amino side group is of great im-
portance. (Scheme 3) During the course of our investigation,
Blechert et al. preceded us with a report on the synthesis of
complex 6e. Complex 6e was found to give a better diaste-
Scheme 2. Increased activity of 2 in aromatic solvents.
different response of complex 2 on the solvent change might
be that a rotation of the NHC ligand allows two populations
to be stabilised by p–p interactions, compared to only one
for catalysts 6a–6e. This would also explain its longer induc-
tion period and high reactivity after the retarding p–p stack-
ing effect is lost. It is, however, taken for granted that for
reoselective RCM and significantly different E/Z ratios in
cross metathesis.[23] This study constitues more evidence of
the interesting characteristics of these novel metathesis ini-
4658
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4654 – 4661

Grubbs Catalysts
FULL PAPER
mary or secondary groups on the nitrogen atom. Further-
more the observation that the least steric complex 6e is the
most active for RCM, clearly demonstrates that modifica-
tion of the NHC ligand can induce substantial changes in
the reactivity pattern of the corresponding catalysts and that
systematic variation of the N-substituents may eventually
allow fine tuning.
Experimental Section
All reactions and manipulations involving organometallic compounds
were conducted in oven-dried glassware under argon atmosphere using
standard Schlenktechniques and dried, distilled and degassed solvents.
Starting chemicals were purchased from commercial sources and used as
received. 1H, 13C and 31P NMR measurements were performed with a
Varian Unity-300 spectrometer.
One-pot procedure for synthesis of complexes 3, 6a and 6c: [RuCl2
A
CHPh)(PCy3)2] (1 equiv), NHC chloride salt (1.5 equiv) and KHMDS
AHCTREUNG
Figure 5. Monitoring RCM of diethyl diallylmalonate by 1H NMR spec-
troscopy (208C). Conditions: diethyl diallylmalonate/catalyst=200, cata-
(1.5 equiv, 0.5m in toluene) were dissolved in dry toluene and stirred at
room temperature for 1 h. Toluene was evaporated under vacuum and a
small amount of MeOH was added under vigorous stirring. The precipi-
tate was filtered off, washed with MeOH, and dried. The spectroscopic
and analytical data of the complexes prepared by this method are com-
piled below.
~
;
^
lyst concentration=4.52 mm, solvent=CD2Cl2. 2: &; 6a: ; 6b: *; 6c:
*
*
.
6d: ; 6e:
A
G
(IAdMesH2)
ACHTREUNG
A
N
ACHTREUNG
1H NMR (CDCl3): d=19.20 (s, 1H; Ru=CHPh), 8.91 (brs, 1H; o-C5H6),
7.42 (t, 1H; p-C5H6), 7.15 (m, 2H; m-H), 6.64 and 5.81 (2”brs, 2H;
o-C5H6 and C6H2Me3), 5.19 (s, 1H; C6H2Me3), 4.09 (m, 1H; N-CaH),
3.76
(m,
2H;
iPCampNCH2CH2NMes),
3.63
(m,
2H;
iPCampNCH2CH2NMes), 2.89 (brs, 1H), 2.52–1.05 ppm (several peaks);
Scheme 3. Importance of the steric bulkof the amino side groups.
31P NMR (CDCl3): d=22.41 ppm; 13C NMR (CDCl3): d=297.2 (Ru=
CHPh), 217.1 (d,
J(P,C)=77.0 Hz, iPCampNCNMes), 151.1 (iC6H5),
A
137.8, 137.3, 131.8–126.8 (several peaks), 58.6 (Ca iPCamp), 50.9 (iP-
CampNCH2CH2NMes), 48.0 (iPCampNCH2CH2NMes), 44.3, 41.9, 40.7,
38.7, 35.9, 35.1, 33.8, 32.8, 28.3–26.5 (several peaks), 24.2, 21.2 (p-Me),
18.6 ppm (o-Me); elemental analysis calcd (%) for C47H71N2Cl2PRu
(867.04): C 65.11, H 8.25, N 3.23; found: C 64.73 H 8.25 N 3.19.
tiators and we agree with Blechert et al. that further investi-
gation to gain a better understanding of the selectivity ef-
fects is welcome.
A
G
N
found to be a sticky compound, which dissolved only slowly in toluene, it
was treated with KHMDS before addition of 1. A solution of KHMDS in
toluene (0.5m, 1.7 mL, 0.850 mmol) was added to [ItBuMesH2]Cl
(0.231 g, 0.823 mmol) in dry toluene (5 mL). The resulting suspension
was stirred for 30 min. Complex 1 (0.34 g, 0.414 mmol) was added and
the reaction mixture was stirred for another 1.5 h to reach full conversion
of the Ru precursor. The solution was filtered and evaporated. Since the
desired complex dissolved in MeOH and hexane, acetone was used to
precipitate the catalyst. The desired complex was filtered off as a bright
green powder in 56% yield. 1H NMR (CDCl3): d=19.09 (s, 1H; Ru=
CHPh), 9.16 (brs, 1H; o-C5H6), 7.38 (t, 1H; p-C5H6), 7.24 (m, 2H; m-
C6H5), 6.95 (brs, 1H; o-C6H5), 6.71 (s, 1H; C6H2Me3), 5.80 (s, 1H;
C6H2Me3), 3.91–3.64 (m, 4H; tBuNCH2CH2NMes), 2.42 (s, 6H; o-CH3),
2.37 (s, 3H; p-CH3), 2.06 (s, 9H; (CH3)3C), 2.09, 1.94, 1.90, 1.70, 1.56,
1.27, 1.10 ppm (all PCy3 protons); 31P NMR (CDCl3): d=22.92 ppm;
Conclusion
A series of NHC ligands bearing aliphatic amino side
groups were synthesised and reacted with the Grubbs first-
generation catalyst. Reactions involving symmetrical NHCs
did not allow us to isolate any pure NHC-substituted com-
plexes due to their instability. Unsymmetrical NHCs with a
planar mesityl group on one amino side chain reacted in a
favourable manner, and the resulting complexes were stable
enough to be isolated. X-ray crystallographic analysis dem-
onstrated that the mesityl group is coplanar with the phenyl
ring of the benzylidene; this result indicates that p–p inter-
action between the mesityl arm and the benzylidene moiety
might constitute an important structural element and that
this needs to be considered in future catalyst design of
Grubbs second-generation analogues. Catalysts 6a, 6c, 6d
and 6e were found to surpass the parent complex 1 for the
ROMP of cycloocta-1,5-diene. Catalyst 6b, substituted with
an NHC derived from tBu-NH2 was considerably less meta-
thesis active than the catalysts derived from amines with pri-
13C NMR (CDCl3): d=300.3 (Ru=CHPh), 217.5 (d,
tBuNCNMes), 152.4 (iC6H5), 138.2, 137.9, 137.8, 137.2, 132.7, 131.3,
129.5, 129.3, 128.8, 128.6, 127.2, 57.3 ((CH3)3C), 51.2
J(P,C)=77.3 Hz,
G
(tBuNCH2CH2NMes), 46.1 (tBuNCH2CH2NMes), 35.5, 33.2, 30.3, 28.9,
27.9, 27.2, 27.1, 26.7, 26.5, 21.1 (p-CH3), 19.0, 18.6 ppm (o-CH3); elemen-
tal analysis calcd (%) for C41H63N2Cl2PRu (786.91): C 62.58, H 8.07, N
3.56; found C 62.09, H 7.77, N 3.40.
A
2(=CHPh)
G
N
1H NMR (CDCl3): d=19.10 (s, 1H; Ru=CHPh), 8.73 (brs, 1H; o-C5H6),
7.39 (t, 1H; p-C5H6), 7.12 (m, 2H; m-C6H5), 6.62 (brs, 1H; o-C6H5), 6.00
Chem. Eur. J. 2006, 12, 4654 – 4661
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4659

N. Ledoux et al.
(brs, 2H; C6H2Me3), 4.53 (m, 1H; N-CH), 3.89 (m, 2H; CyN-
CH2CH2NMes), 3.72 (m, 2H; CyNCH2CH2NMes), 3.47 (s, 1H), 2.46
(brs, 1H), 2.22, 1.89, 1.60–1.11 ppm (several peaks); 31P NMR (CDCl3):
of allylic protons of the ring-closed product and of the disappearing sub-
strate.
Representative procedure for ROMP tests (Table 2): Small oven-dried
glass vials with septum were charged with a stirring bar and the appropri-
ate amount of catalyst taken from a CH2Cl2 stocksolution. The dichloro-
methane was subsequently evaporated, and the glass vials with solid cata-
lyst were kept under argon atmosphere. To start the ROMP test, toluene
(200 mL) was added in order to dissolve the catalyst. The appropriate
amount of COD monomer was then transferred to the vial by syringe
under vigorous stirring at room temperature for 2, 6a, 6c, 6d and 6e and
at 708C for 3 and 6b. After a certain time span, a small quantity of the
reaction mixture, which had become viscous, was taken out of the vial
and dissolved in CDCl3. Conversion was then determined by 1H NMR
spectroscopy.
d=28.13; 13C NMR (CDCl3): d=295.7 (Ru=CHPh), 215.5 (d, J
(P,C)=
A
76.9 Hz, CyNCNMes), 151.2 (iC6H5), 137.7, 137.6, 136.7, 130.7–128.1 (sev-
eral peaks), 58.2 (C1 Cy), 50.7 (CyNCH2CH2NMes), 43.8 (CyN-
CH2CH2NMes), 32.3, 31.1, 30.5, 29.4–25.0 (several peaks), 21.1 (p-Me),
18.7 ppm (o-Me); elemental analysis calcd (%) for C43H65N2Cl2PRu
(812.95): C 63.53, H 8.06, N 3.45; found C 63.20, H 7.99, N 3.40; single
crystals suitable for X-ray analysis were grown by slow evaporation of a
solution of 6c in CH2Cl2 (Figure 2).
A
G
N
(0.29 g, 0.861 mmol) was stirred with an equimolar quantity of KHMDS
in toluene (0.5m, 1.722 mL) for 15 min. Complex 1 (0.4 g, 0.487 mmol)
was added and the resulting solution was allowed to stir at room temper-
ature for 1 h, during which the mixture changed colour from purple to
darkred. The solution was filtered to remove salts and the solvent was
evaporated. The crude mixture was loaded onto a column of silica gel
and the product was eluted by flash chromatography (9:1 hexane/Et2O).
Complex 6d was obtained as a pure pinkish compound with 49% yield.
1H NMR (CDCl3): d=18.99 (s, 1H; Ru=CHPh), 7.96 (brs, 1H; o-C5H6),
7.37 (m, 1H; p-C6H5), 7.09 (m, 2H; m-C6H5), 6.88 (brs, 1H; o-C6H5),
6.86 (brs, 1H; C6H2Me3), 6.22 (brs, 1H; C6H2Me3), 4.17 (t, 2H;
NCH2CH2N-CH2), 3.89 (t, 2H; n-octNCH2CH2NMes), 3.74 (t, 2H; n-
octNCH2CH2NMes), 2.27, 2.16, 1.87, 1.61, 1.30, 1.11, 0.90 ppm (57H);
31P NMR (CDCl3): d=32.61 ppm; 13C NMR (CDCl3): d=294.5–293.3 (br,
CCDC-295190 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgements
The authors gratefully acknowledge the FWO-Flanders and the research
Fund of Ghent University for generous financial support during the prep-
aration of this manuscript. We wish to thankJacques PØcaut of DRFMC
(DØpartement de Recherche Fondamentale sur la Mati›re CondensØe—
Laboratoire du CEA, Grenoble) for collecting the crystallographic data
set and Olivier F. Grenelle and Marc G. Proot of Chevron Technology,
Ghent for elemental analysis. Also special thanks to Marek Figlus and
Prof. S. W. Markowicz (Institute of Organic Chemistry, Technical Univer-
sity of Lódz, Poland) for generous donation of the pinane-derived amine
used to synthesise imidazolinium chloride 4e.
Ru=CHPh), 217.1 (d, J(P,C)=75.2 Hz, n-octNCNMes), 149.9 (iC6H5),
H
136.5, 136.2, 135.4, 129.4 (br), 128.4, 127.8, 127.4, 127.0, 126.8, 125.5, 49.9,
49.7 (n-octNCH2CH2NMes and C1-n-octyl), 47.3 (n-octNCH2CH2NMes),
34.8, 34.0, 30.9, 30.8, 30.5, 28.8, 28.3, 27.4, 26.8, 26.7, 26.1, 25.9, 25.5, 25.4,
25.2, 24.4, 21.7 (p-CH3), 19.8, 17.4 (o-CH3), 13.1 ppm (C8-n-octyl); ele-
mental analysis calcd (%) for C45H71N2Cl2PRu (843.02): C 64.11, H 8.49,
N 3.32; found: C 63.27 H 8.38 N 3.28.
A
(=CHPh)
G
N
(0.091 g, 0.381 mmol) was stirred with an equimolar quantity of KHMDS
in toluene (0.5m, 0.762 mL) for 30 min. Complex 1 (0.2 g, 0.24 mmol) was
added and the resulting solution was allowed to stir at room temperature
for 1 h. The solution was filtered to remove salts and evaporated in
vacuo. Precipitation of pure pinkproduct was achieved by addition of
hexane to a concentrated solution of the complex in CH2Cl2. Yield: 77%;
1H NMR (CDCl3): d=18.89 (s, 1H; Ru=CHPh), 7.81 (brs, 1H; o-C5H6),
7.37 (t, 1H; p-C6H5), 7.10 (m, 2H; m-C6H5), 6.90 (s, 1H; o-C6H5), 6.82
(brs, 1H; C6H2Me3), 6.28 (brs, 1H; C6H2Me3), 3.95 (m, 2H;
MeNCH2CH2NMes), 3.82 (s, 3H; NCH3), 3.49 (m, 2H;
MeNCH2CH2NMes), 2.32, 2.17, 1.89, 1.61, 1.27, 1.12, 0.88 ppm; 31P NMR
(CDCl3): d=34.92 ppm; 13C NMR (CDCl3): d=294.3–239.4 (br, Ru=
[1] The 2005 Nobel Prize in Chemistry is shared by Y. Chauvin (Institut
FranÅais du PØtrole, Rueil-Malmaison, France), R. R. Schrock(Mas-
sachusetts Institute of Technology, Cambridge, Massachusetts, USA)
and R. H. Grubbs (California Institute of Technology, Pasadena,
California, USA).
[2] a) M. Schuster, S. Blechert, Angew. Chem. 1997, 109, 2124–2144;
Angew. Chem. Int. Ed. 1997, 36, 2036–2056; b) A. Fürstner, Angew.
Chem. 2000, 112, 3140–3172; Angew. Chem. Int. Ed. 2000, 39, 3012–
3043; c) R. R. Schrock, J. Chem. Soc. Dalton Trans. 2001, 2541–
2550; d) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18–
29; e) R. H. Grubbs, Handbook of Metathesis, Wiley-VCH, Wein-
heim, 2003; f) R. R. Schrock, A. H. Hoveyda, Angew. Chem. 2003,
115, 4740–4782; Angew. Chem. Int. Ed. 2003, 42, 4592–4633; g) J. C.
Mol, J. Mol. Catal. A 2004, 213, 39–45; h) R. H. Grubbs, Tetrahe-
dron 2004, 60, 7117–7140.
[3] a) P. Schwab, M. B. France, J. W. Ziller, R. H. Grubbs, Angew.
Chem. 1995, 107, 2179–2181; Angew. Chem. Int. Ed. Engl. 1995, 34,
2039–2041; b) P. Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem.
Soc. 1996, 118, 100–110.
[4] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953–
956.
[5] A. Fürstner, L. Ackermann, B. Gabor, R. Goddard, C. W. Lehmann,
R. Mynott, F. Stelzer, O. R. Thiel, Chem. Eur. J. 2001, 7, 3236–3253.
[6] M. B. Dinger, J. C. Mol, Adv. Synth. Catal. 2002, 344, 671–677.
[7] M. B. Dinger, P. Nieczypor, J. C. Mol, Organometallics 2003, 22,
5291–5296.
CHPh), 219.4 (d, J(P,C)=74.4 Hz, MeNCNMes), 151.0 (iC6H5), 137.8,
T
137.2, 136.5, 130.5, 130.0 (br), 129.0, 128.3, 127.9, 52.4
(MeNCH2CH2NMes), 51.4 (MeNCH2CH2NMes), 37.6, 35.9, 35.1, 31.7,
31.5, 31.4, 30.6, 30.1, 29.6, 28.0, 27.1, 26.8, 22.8, 21.1 (p-CH3), 18.4 ppm
(o-CH3); elemental analysis calcd (%) for C38H57N2Cl2PRu (744.83):
C 61.28, H 7.71, N 3.76; found C 60.98, H 7.55, N 3.60.
1
Note: For all of these complexes only one 31P signal and one single H a-
benzylidene signal were found, suggesting that only one single isomer
had been formed.
Monitoring ROMP of COD (Figures 3 and 4): After charging an NMR
tube with the appropriate amount of catalyst dissolved in dry, deuterated
solvent (CDCl3 or C6D6), COD was injected into the tube. The polymeri-
sation reaction was monitored as a function of time at 208C by integrat-
1
ing olefinic H signals of the formed polymer and the disappearing mono-
mer.
Monitoring RCM of diethyl diallylmalonate (Figure 5): An NMR tube
was charged with of a solution of the catalyst in CD2Cl2 (0.6 mL; 4.52 mm
or 0.002712 mmol catalyst). Diethyl diallylmalonate (200 equiv or
0.13 mL) was added and the NMR tube was closed. The ethylene gener-
ated during the reaction process was not removed so that the RCM reac-
tions were carried out under equilibrium conditions. The progress of the
ring-closing reaction was monitored at 208C by integration of 1H signals
[8] a) A. C. Hillier, W. J. Sommer, B. S. Yong, J. L. Petersen, L. Cavallo,
S. P. Nolan, Organometallics 2003, 22, 4322–4326; b) K. Denk, P.
Sirsch, W. A. Herrmann, J. Organomet. Chem. 2002, 649, 219–224;
c) A. M. Magill, K. J. Cavell, B. F. Yates, J. Am. Chem. Soc. 2004,
126, 8717–8724.
[9] J. Huang, L. Jafarpour, A. C. Hillier, E. D. Stevens, S. P. Nolan, Or-
ganometallics 2001, 20, 2878–2882.
4660
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 4654 – 4661

Grubbs Catalysts
FULL PAPER
[10] M. S. Viciu, O. Navarro, R. F. Germananeau, R. A. Kelly III, W.
Sommer, N. Marion, E. D. Stevens, L. Cavallo, S. P. Nolan, Organo-
metallics 2004, 23, 1629–1635.
[11] R. Dorta, E. D. Stevens, N. M. Scott, C. Costabile, L. Cavallo, C. D.
Hoff, S. P. Nolan, J. Am. Chem. Soc. 2005, 127, 2485–2495.
[12] L. Cavallo, A. Correa, C. Costabile, H. Jacobsen, J. Organomet.
Chem. 2005, 690, 5407–5413.
with MeOH (2 mL) and dried under vacuum. Yield: 93% (higher
than when KOtBu was used)
[17] M. Paas, B. Wibbeling, R. Frçhlich, F. E. Hahn, Eur. J. Inorg. Chem.
2006, 158–162.
[18] a) A. W. Waltman, R. H. Grubbs, Organometallics 2004, 23, 3105–
3107; b) H. Clavier, L. Coutable, J.-C. Guillemin, M. Mauduit, M.
Tetrahedron Asymm. 2005, 16, 921–924.
[13] M. Sübner, H. Plenio, Chem. Commun. 2005, 5417–5419.
[14] a) M. S. Sanford, M. Ulman, R. H. Grubbs, J. Am. Chem. Soc. 2001,
123, 749–750; b) M. S. Sanford, J. A. Love, R. H. Grubbs, J. Am.
Chem. Soc. 2001, 123, 6543–6554; c) C. Adlhart, P. Chen, J. Am.
Chem. Soc. 2004, 126, 3496–3510; d) B. F. Straub, Angew. Chem.
2005, 117, 6129–6132; Angew. Chem. Int. Ed. 2005, 44, 5974–5978.
[15] S. Lee, J. F. Hartwig, J. Org. Chem. 2001, 66, 3402–3415.
[16] Deprotonation with KHMDS allowed us to synthesise Grubbs
second-generation catalyst 2 in less than one hour with very high
yields. A one-pot procedure could be followed and heating of the re-
action mixture was not required: dry toluene (20 mL) was added to
[19] S. Prühs, C. W. Lehmann, A. Fürstner, Organometallics 2004, 23,
280–287.
[20] a) J. A. Love, M. S. Sanford, M. W. Day, R. H. Grubbs, J. Am.
Chem. Soc. 2003, 125, 10103–10109; b) T. M. Trnka, J. P. Morgan,
M. S. Sanford, T. E. Wilhelm, M. Scholl, T.-L. Choi, S. Ding, M. W.
Day, R. H. Grubbs, J. Am. Chem. Soc. 2003, 125, 2546–2558.
[21] Differences in the NoE effect are minimal for similar carbon atoms.
1H NMR showed different chemical shift values for the cis and trans
olefinic protons, but the distinct signals were not resolved enough to
allow quantitative determination of the cis/trans ratio.
[22] C. W. Bielawski, R. H. Grubbs, Angew. Chem. 2000, 112, 3025–
3028; Angew. Chem. Int. Ed. 2000, 39, 2903–2905.
an oven-dried flaskcharged with
1
(1.32 g, 1.6 mmol),
[H2IMes(H)]Cl (0.658 g, 19.2 mmol, 1.2 equiv) and a solution of
KHMDS in toluene (0.5m, 3.84 mL). The resulting reaction mixture
was stirred at room temperature during 1 h. Solvent was removed
under vacuum and of MeOH (10 mL) was added under vigorous
stirring. The resulting pinkish complex 2 was filtered off, washed
[23] K. Vehlow, S. Maechling, S. Blechert, Organometallics 2006, 25, 25–
28.
Received: January 16, 2006
Published online: March 15, 2006
Chem. Eur. J. 2006, 12, 4654 – 4661
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4661