New in situ generated ruthenium catalysts bearing N-heterocyclic carbene ligands for the ring-opening metathesis polymerization of cyclooctene

Article abstract of DOI:10.1002/1615-4169(200208)344:6/7<749::AID-ADSC749>3.0.CO;2-T

New 1,3-diarylimidazol(in)ium chlorides bearing phenyl, 1-naphthyl, 4-biphenyl, 2-tolyl, 2,6-dimethylphenyl, and 3,5-dimethylphenyl substituents were synthesized. They were combined with [RuCl₂ (p-cymene)] ₂ and potassium tert-butoxi

Full text of DOI:10.1002/1615-4169(200208)344:6/7<749::AID-ADSC749>3.0.CO;2-T

FULL PAPERS
New In situ Generated Ruthenium Catalysts Bearing
N-Heterocyclic Carbene Ligands for the Ring-Opening
Metathesis Polymerization of Cyclooctene
Lionel Delaude, Magdalena Szypa, Albert Demonceau, Alfred F. Noels*
¡
Center for Education and Research on Macromolecules (CERM), Institut de Chimie (B6a), University of Liege, Sart-
¡
Tilman par 4000 Liege, Belgium
Fax: ( ‡ 32)-4-366-3497, e-mail: af.noels@ulg.ac.be
Received: February 4, 2002; Accepted: May 13, 2002
Abstract: New 1,3-diarylimidazol(in)ium chlorides was not crucial to achieve high catalytic efficiencies.
bearing phenyl, 1-naphthyl, 4-biphenyl, 2-tolyl, 2,6- The presence or the absence of alkyl groups on the
dimethylphenyl, and 3,5-dimethylphenyl substituents ortho positions of the phenyl rings had a more
were synthesized. They were combined with [RuCl2 pronounced influence. Blocking all the ortho posi-
(p-cymene)]2 and potassium tert-butoxide or sodium tions was a requisite for obtaining efficient catalysts.
hydride to generate the corresponding ruthenium-N- Failure to do so probably results in the ortho-metal-
heterocyclic carbene complexes in situ. Catalyst lation of the carbene ligand, thereby altering the
precursors derived from imidazol(in)ium salts bearing coordination sphere of the ruthenium active centers.
the 2,4,6-trimethylphenyl (mesityl) and the 2,6-diiso-
propylphenyl groups were also prepared. The cata-
lytic activity of all these species in the photoinduced Keywords: N-heterocyclic carbene (NHC) ligands;
ring-opening metathesis polymerization of cyclooc- homogeneous catalysis; metathesis; polymerization;
tene was investigated. The C4-C5 double bond in the ring-opening metathesis polymerization (ROMP);
imidazole ring of the N-heterocyclic carbene ligands ruthenium
Introduction
Many groups were swift to realize that NHC×s could
further expand the potentials of the original Grubbs×
Significant advances in olefin metathesis have been catalyst. Replacing the two phosphine ligands in 1 with
achieved during the past decade using new ruthenium- NHC×s did not increase significantly itscatalytic activity[6]
based metal-carbene complexes.[1] A major break- but mixed-ligand complexes of type 2 containing one
through was achieved in the mid-1990×s by Grubbs and phosphine and one NHC proved to be highly efficient
coworkers with the discovery of the well-defined and initiators for the ROMP of low-strain cycloolefins.
fully characterized ruthenium-alkylidene complex Aromatic imidazol-2-ylidene ligands (with a formal
ˆ
RuCl2( CHPh)(PCy3)2 (1) that promotes the ring- double bond between C4 and C5 in structure 2) were
opening metathesis polymerization (ROMP) of strained first investigated,[7] but they were quickly supplanted in
and low-strain cycloolefins.[2] The 1990×s also witnessed most applications by NHC×s possessing a saturated
the experimental reality of stable nucleophilic N- backbone (i.e., a single bond between C4 and C5 in
heterocyclic carbenes (NHC×s) isolated and character- structure 2), leading to the so-called ™second generation∫
ized by Arduengo and coworkers.[3] These divalent of ruthenium-benzylidene olefin metathesis catalysts.[7c]
carbon species are neutral, two-electron ligands with a
Our own line of research focuses on a different family
negligible p-back-bonding tendency.[4] They behave as of ruthenium-arene complexes. Back in 1995, we found
phosphine mimics, yet they are better s-donors and they
form stronger bonds to metal centers than most
4
5
phosphines. Their electronic and steric properties are
liable to ample modification simply by varying the
substituents on the nitrogen atoms. Therefore, NHC×s
constitute a promising new class of ligands available for
catalyst engineering and fine tuning, and a great deal of
attention has been paid to them recently.[5]
N
N
PCy3
Cl
Cl
Cl
Cl
Ru
Ru
Ph
Ph
PCy3
PCy3
2
1
Adv. Synth. Catal. 2002, 344, No. 6+7
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1615-4150/02/3446+7-749 756 $ 17.50+.50/0
749

Lionel Delaude et al.
FULL PAPERS
that the 18-electron complex RuCl2(p-cymene)(PCy3) cyclooctene. For this purpose, new imidazolium and
(3) was a versatile and efficient promoter for the ROMP imidazolinium salts bearing aromatic substituents were
of both strained and low-strain cyclic olefins when synthesized. They were combined with [RuCl2(p-cym-
activated by a suitable carbene precursor such as ene)]2 (5) and potassium tert-butoxide or sodium
trimethylsilyldiazomethane (TMSD).[8] Recently, we hydride to generate the corresponding ruthenium-
disclosed preliminary results obtained for the ROMP NHC complexes in situ. Although less rigorous than
of cyclooctene (a typical low-strain cycloolefin) when the isolation of the free carbene followed by its reaction
the tricyclohexylphosphine in 3 was substituted by with dimer 5, we felt that this procedure was adequate
an NHC ligand (see structure 4).[9] Two major findings for the rapid screening of a large number of catalyst
emerged from this study. First, we underscored the precursors, due to its ease of set-up.
intervention of a photochemical activation step due to
visible light illumination. To the best of our knowledge,
the only precedents of photoinduced ROMP described
in the literature so far involved UV light. Furthermore,
Results and Discussion
these earlier photo-ROMP processes were applied to
highly strained cyclic olefins only (e.g., dicyclopenta- Preparation of Imidazolium and Imidazolinium Salts
diene) and afforded high molecular weight polymers
with broad polydispersities.[10] In contrast, our proce- Since our preliminary study had evidenced the inter-
dure could be carried out already at room temperature vention of a photochemical activation of catalyst
and gave quantitative yields of mainly trans-polyocte- precursor 4 in the ROMP of cyclooctene,[9] we reasoned
namer with a polydispersity index inferior to 2.[9] that the presence of highly conjugated substituents on
Second, we found that a 1:2:4 molar association of the nitrogen atoms of the imidazol-2-ylidene ligand
[RuCl2(p-cymene)]2 (5), 1,3-dimesitylimidazolium should favor the visible light absorption believed to
chloride, and potassium tert-butoxide was almost as trigger the catalytic process. This assumption was
effective as the preformed catalyst precursor 4 in the supported by UV-visible spectroscopic measurements
polymerization of cyclooctene at 60 8C. Such a combi- and by the observation that ruthenium-arene complexes
nation requires only stable and readily available com- bearing alkyl-substituted NHC ligands were devoid of
mercial reagents to generate the active catalytic species any significant catalytic activity in our test experiments.
in situ, thus leading to very simple and straightforward Thus, we decided to direct our synthetic efforts toward
polymerization procedures.[9] The recourse to a similar the preparation of imidazolium salts with aromatic
three-component catalytic system to activate the ruthe- substituents.
nium dimer 5 has also been reported by Dixneuf and
Compared to the wide range of 1,3-dialkylimidazo-
coworkers for the enyne metathesis reaction,[11] while lium or mixed alkylaryl compounds prepared lately,[13]
Nolan et al. have used NHC×s produced in situ by the 1,3-diaryl derivatives are, unfortunately, far less
deprotonation of imidazolium salts to act as ancillary accessible. Quaternization of an N-substituted imida-
ligands in various palladium-catalyzed coupling reac- zole with an aryl halide usually fails,[14] and the one-pot
tions.[12]
synthesis of an N,N'-diarylimidazole ring from acyclic
To further improve the catalyst efficiency in the precursors is often tedious. Indeed, the preparation of
photoinduced ROMP process, we have launched a 1,3-diarylimidazolium ions from glyoxal, an aromatic
detailed investigation on the role of the NHC ligand in amine, and paraformaldehyde under aqueous acidic
catalyst precursor 4. In this contribution, we describe an conditions usually leads to the concomitant formation of
extensive and systematic evaluation of the effects of highly colored ionomer by-products that are very
ligand variation on the visible light-induced ROMP of difficult to separate from the desired salt, particularly
when working on a small laboratory scale.[15] To circum-
vent these problems, a two-step procedure involving the
condensation of glyoxal and an aromatic amine
(2 equiv.) to generate the corresponding Schiff base
(6), followed by cyclization with chloromethyl ethyl
3
Ru
Cl
ether has been proposed by Arduengo and coworkers.[15]
We have applied this method to the synthesis of
compounds 7g and 7h (see Scheme 1).
PCy3
Cl
Ru
Cl
Cl
N
Attempts to prepare other N,N'-diarylethylenedii-
mines (6) to serve as intermediates in the synthesis of
new imidazolium salts gave mixed results. Thus, the
reaction of 4-aminobiphenyl (caution: see experimental
section), 2-toluidine, and 2,6-dimethylaniline afforded
the corresponding glyoxal-diimines 6d, 6e, and 6f,
Cl
N
Cl
Cl
Ru
Ru
Cl
4
5
750
Adv. Synth. Catal. 2002, 344, 749 756

In Situ Generated Ru Catalysts Bearing NHC Ligands for ROMP of Cyclooctene
FULL PAPERS
Ar
Ar
H
H
O
O
N
(CH2O)n, HCl
PrOH, H2O
N
+ 2 ArNH2
Cl-
+
8a, 9a: Ar = phenyl
H
N
N
8b, 9b: Ar = 1-naphthyl
Ar
Ar
6d - h
7d - h
8c, 9c: Ar = 3,5-dimethylphenyl
6d, 7d, 8d, 9d: Ar = 4-biphenyl
i) NaBH4
ii) HCl
6e, 7e, 8e, 9e: Ar = 2-methylphenyl (2-tolyl)
6f, 7f, 8f, 9f: Ar = 2,6-dimethylphenyl (xylyl)
6g, 7g, 8g, 9g: Ar = 2,4,6-trimethylphenyl (mesityl)
6h, 7h, 8h, 9h: Ar= 2,6-diisopropylphenyl (diip)
Ar
Ar
N
i) ∆
ii) HCl
NH·HCl
Br
HC(OEt)3, ∆
Cl-
+ 2 ArNH2
+
H
NH·HCl
Ar
N
Br
Ar
8a - h
9a - h
H3C
CH3
CH3
CH(CH3)2
Ar =
H3C
H3C
8c, 9c
CH3
CH3
CH3
6e, 7e, 8e, 9e
CH(CH3)2
6h, 7h, 8h, 9h
8a, 9a
8b, 9b
6d, 7d, 8d, 9d
6f, 7f, 8f, 9f
6g, 7g, 8g, 9g
Scheme 1.
respectively, in high yields. On the other hand, reactions dimethylphenyl derivatives 8a c. All the intermediate
with aniline, 1-naphthylamine, 3,5-dimethylaniline and diamine dihydrochlorides 8a h were then suspended in
other ortho-functionalized anilines like 2-anisidine or 2- an excess of triethyl orthoformate and the mixtures were
chloroaniline afforded only tarry materials or complex refluxed for two days. After cooling to room temper-
addition products.[16] In our hands, the cyclization step ature, a simple filtration was sufficient to isolate
using chloromethyl ethyl ether also lacked generality compounds 9a h in high yields. Attempts to use pure
and afforded deeply colored solids or molasses. Re- formic acid instead of triethyl orthoformate as the C1
course to an anhydrous acidic solution of paraformal- building block failed to afford cyclized products, maybe
dehyde as C1 building block, as recommended by because the acid and the orthoester have very different
Nolan,[17] gave cleaner products and allowed us to solvating properties.
isolate the imidazolium chlorides 7d f in satisfactory
purity.
To further enlarge our set of NHC ligands and to Polymerization of Cyclooctene
better apprehend the role of the ortho-methyl substitu-
ents in catalyst precursor 4, we have also prepared new Compounds 7d h and 9a h served as catalyst modi-
saturated imidazolinium salts 9 in parallel with the fiers in the ruthenium-catalyzed polymerization of
unsaturated imidazolium chlorides 7. Applying again a cyclooctene. In our laboratory, the ROMPof this typical
reaction scheme previously outlined by Arduengo and low-strain cycloolefin constitutes a standard test reac-
coworkers, the diimines 6d h were first reduced into tion to assess the metathetical activity of new catalyst
the corresponding diamines dihydrochlorides 8d h precursors.[8,9,18] In the present study, NHC ligands were
upon treatment with sodium borohydride followed by generated in situ by deprotonation of the corresponding
acidification with aqueous HCl.[15] Alternatively, it was imidazol(in)ium salts and combined with the ruthenium
also possible to prepare various N,N'-diarylethylenedi- dimer 5 before cyclooctene was added. Two strong
amines by nucleophilic substitution of 1,2-dibromo- bases, viz. potassium tert-butoxide and sodium hydride,
ethane with aromatic amines. This reaction is applicable were employed, and the reactions were carried out in
to a broader range of substrates than the imine route and chlorobenzene at 60 8C for 2 h. The transition metal and
allowed us to prepare the phenyl, 1-naphthyl, and 3,5- the carbene were in 1:1 stoichiometric proportions
Adv. Synth. Catal. 2002, 344, 749 756
751

Lionel Delaude et al.
FULL PAPERS
[RuCl2(p-cymene)]2 (1 equiv.)
imidazoli(ni)um salt (2 equiv.)
+
KOtBu or NaH (4 equiv.)
Cl
Ru
n
Cl
H
Cl
n
Cl–
Ru
Cl
Ru Cl Ru
PhCl, 2 h, 60
C
H
visible light illumination
Scheme 2.
10
11
(albeit 2 equivalents of base were used), while the
monomer-to-ruthenium ratio was 250. An ordinary noted, the mixture of 1,3-dimesitylimidazolium chloride
40 W ™cold white∫ fluorescent tube placed 10 cm away 7g, [RuCl2(p-cymene)]2 (5), and KO-t-Bu was almost as
from the Pyrex reaction flasks complemented the efficient as the preformed complex 4 under the exper-
experimental set-up and provided a constant and imental conditions adopted.[9] Within the time imparted
reproducible visible light source (see Scheme 2).
to the reaction, monomer conversion was essentially
Under these conditions, the formation of bimetallic quantitative and a high molecular weight polymer
hydride species instead of the expected ruthenium- containing mostly trans double bonds was formed in
carbene catalyst precursors of type 4 cannot be ruled out almost quantitative yield (see Table 1).
immediately. Indeed, Bennett et al. have prepared m-
The absence of the para-methyl group in compound 7f
hydrido complexes such as 10 and 11 by treating the did not alter its catalytic activity compared to 7g. The
ruthenium p-cymene dimer 5 with hydrogen in the polymer yield and its microstructure remained un-
presence of triethylamine. Aqueous sodium carbonate changed. Fluctuations in the molecular weight should
in isopropanol gave similar reactions, although this not be overconsidered. Unlike the Grubbs ruthenium-
procedure turned out to be non-reproducible.[19] Yet, the benzylidene complexes 1 and 2, ruthenium-arene spe-
recourse to a strong base and the release of tert-butyl cies 3 and 4 lack the alkylidene moiety required to
alcohol from potassium tert-butoxide in our system initiate a metathesis process. These catalyst precursors
prompted us to check for the possible intermediacy of afford, nevertheless, highly active species when reacted
ruthenium hydride species. This was accomplished by with cycloctene in the presence of TMSD or visible light.
monitoring the reaction of dimer 5 with the imidazolium Ill-defined mechanisms involving arene disengagement
salt 7g (2 equiv.) and KO-t-Bu or NaH (4 equiv.) in C6D5 and monomer coordination are held responsible for
Cl at 60 8C. 1H NMRspectra showed no sign of hydride
formation and supported the in situ formation of system is a poor control of the initiation step that results
complex 4. in high molecular weights and usually rather broad
these transformations. The main drawback of this
To begin our catalytic screening, we first investigated polydispersities. Because only a small number of
the activities of imidazolium chlorides 7d h in the propagating species are present in solution, slight
polymerization of cyclooctene. In the presence of the changes in the initiation efficiency from one experiment
ortho-tolyl substituted salt 7e, dimer 5, and a base, the to another lead to significant variations in the molecular
monomer did not react and was left unchanged after two weights attained. Thus, the differences in average
hours at 60 8C. The 4-biphenyl derivative 7d led to a polyoctenamer chain length obtained with imidazolium
moderate conversion of the cycloolefin (24% with KO-t- salts 7f and 7g are of minor importance, and the two
Bu and 17% with NaH, respectively) but the ring- catalyst modifiers display similar behaviors. This result
opening probably stopped at the oligomerization stage did not come as a surprise, as the structural change
and no polymer was isolated. Compounds 7f h were between the two ligands affects only a position remote
much more efficient catalyst precursors. As previously from the metal center.
Table 1. ROMP of cyclooctene in PhCl at 60 8C catalyzed by various ruthenium-NHC complexes generated in situ from
[RuCl2(p-cymene)]2 (5), an imidazolium salt (7), and a base.
10À Mn
Mw/Mn
3
[b]
[b]
Imidazolium Salt
Base
Monomer Conversion [%]
Isolated Polymer Yield [%]
sc[a]
7f
7f
7g
7g
7h
7h
KO-t-Bu
NaH
KO-t-Bu
NaH
KO-t-Bu
NaH
> 99
> 99
99
> 99
99
89
84
92
76
60
87
0.19
0.19
0.20
0.22
0.20
0.31
742
559
659
311
398
912
2.08
2.25
2.02
2.03
3.09
2.47
99
[a]
Fraction of cis double bonds within the polyoctenamer, determined by 13C NMR.
Determined by GPC in THF vs. monodisperse polystyrene standards.
[b]
752
Adv. Synth. Catal. 2002, 344, 749 756

In Situ Generated Ru Catalysts Bearing NHC Ligands for ROMP of Cyclooctene
FULL PAPERS
Switching from methyl to isopropyl groups on both butoxideanions, thus giving the advantage tothe smaller
ortho positions of the aryl substituents by replacing 7f or hydride species.
7g with 7h caused a much more significant change in
Further comparison of the data gathered in Tables 1
terms of steric occupancy around the ruthenium atom. A and 2 revealed close similarities between the results
high catalytic activity was, however, maintained, espe- obtained with catalyst precursors 7f and 9f, and 7g and
cially when the deprotonation was carried out with 9g, respectively. High yields were achieved with both
sodium hydride (vide infra). No clear-cut effect on the systems and the removal of the double bond between C4
polymer microstructure was detected, except from a and C5 in the imidazole ring only led to small increases
broadening of the molecular weight distributions.
in the values of the sc and Mw/Mn parameters. Larger
More insight into the influence of the NHC ligand discrepancies were noticed when the cyclooctene poly-
structure on the polymerization of cyclooctene came merization was accomplished in the presence of com-
from the study of imidazolinium salts 9a h in another pounds 7h and 9h, respectively. The latter displayed an
series of ROMP experiments. In addition to the five aryl inferior catalytic activity and led to polymers with a
groups already introduced on the imidazolium ring, the higher proportion of cis double bonds and a larger
phenyl, 1-naphthyl, and 3,5-dimethylphenyl units made polydispersity index. Yet, the presence of a C4-C5
their debut on the list of the nitrogen substituents. None double bond in the NHC ligand was not critical to reach
of these new catalyst modifiers afforded any polyocte- high catalytic efficiencies. The presence of alkyl groups
namer. In all three cases, monomer consumption on both ortho positions of the phenyl ring, on the
stagnated below 10%, whether potassium tert-butoxide contrary, had a crucial importance. Compounds 9a, 9b,
or sodium hydride served as the base. The N,N'-di(4- 7d, and 9d bearing phenyl, 1-naphthyl, and 4-biphenyl
biphenyl) derivative 9d and the N,N'-di(2-tolyl) salt 9e groups did not meet this criterion and were devoid of
were slightly more active, with conversions now in the any significant catalytic activity. The absence of poly-
20 30% range, but oligomers and no polymers still merization with compounds 7e and 9e, which possess
accounted for these figures. In sharp contrast with these only one methyl group blocking an ortho position was
results, satisfactory to quantitative conversions were even more revealing, as was the test reaction carried out
achieved with imidazolinium chlorides 9f h (see Ta- with the meta-disubstituted imidazolinium salt 9c.
ble 2). As far as the nature of the aryl substituent is
Although bulky ortho-substituents may influence the
concerned, the same trends were therefore observed for stability of NHC×s bearing aromatic groups and steri-
both the unsaturated imidazolium salts and their cally hinder their dimerization into the corresponding
saturated imidazolinium analogues.
tetraminoethylene derivatives, various imidazol-2-yli-
Recourse to potassium tert-butoxide to deprotonate denes bearing unsubstituted phenyl[3d] or para-substi-
the hydrochloride ligand precursors usually led to tuted aryl groups[3b] were synthesized by deprotonation
slightly better yields and conversions than the use of of imidazolium salts with potassium tert-butoxide and
sodium hydride. Molecular weight distributions were isolated. Thus, the presence of ortho-substituents is not
also narrower when KO-t-Bu was employed to generate required to generate stable carbene ligands. We believe
the catalyst precursors in situ. The lower solubility of the that the ortho-effect observed in our catalytic systems
hydride in organic media compared to the alkoxide may results from the metallation of a phenyl C-H bond
be invoked to rationalize these observations. Only with adjacent to the imidazole ring upon exposure of the
the 2,6-diisopropylphenyl-substituted imidazol(in)ium NHC ligand to the ruthenium dimer 5. Back in the late
salts 7h and 9h did NaH afford superior results 1970×s, Lappert and coworkers already reported the
compared to KO-t-Bu. In these cases, the steric hin- spontaneous ortho-metallation of 1,3-diaryl-4,5-dihy-
drance around the nitrogen atoms may restrain the droimidazol-2-ylidene ligands complexed with RuCl2
access to the acidic C-H center by the bulky tert- (PPh3)3 (see structure 12).[20] Very recently, we described
Table 2. ROMP of cyclooctene in PhCl at 60 8C catalyzed by various ruthenium-NHC complexes generated in situ from
[RuCl2(p-cymene)]2 (5), an imidazolinium salt (9), and a base.
10À Mn
Mw/Mn
3
[b]
[b]
Imidazolinium Salt Base
Monomer Conversion (%) Isolated Polymer Yield (%) sc[a]
9f
9f
9g
9g
9h
9h
KO-t-Bu 99
NaH 98
KO-t-Bu 99
NaH 91
KO-t-Bu 57
NaH 67
93
84
93
79
44
51
0.19 641
0.29 927
0.23 512
0.36 1172
0.51 838
0.42 755
2.40
2.58
2.19
2.20
2.13
3.05
[a]
[b]
Fraction of cis double bonds within the polyoctenamer, determined by 13C NMR.
Determined by GPC in THF vs. monodisperse polystyrene standards.
Adv. Synth. Catal. 2002, 344, 749 756
753

Lionel Delaude et al.
FULL PAPERS
with aHP 1037A refractive index detector and a battery of 4 PL
gel columns fitted in series (particle size: 5 mm; pore sizes:
100,000, 10,000, 1000, and 100 ä). The molecular weights (not
corrected) are reported versus monodisperse polystyrene
PPh3
Cl
R
Ru
Ru
Ph3P
Cl
N
N
N
N
1
standards used to calibrate the instrument. H and 13C NMR
N
Ph
spectra (400 and 100 MHz, respectively) were recorded at
298 K on a Bruker DPX 400 spectrometer in CDCl3 with TMS
as internal standard or in DMSO-d6 using the solvent peaks as
reference. Melting points are uncorrected. [RuCl2(p-cym-
R
(R = H, CH3, OCH3)
Ph
12
13
[22]
ene)]2 was purchased from Strem. Mesityl (6g, 7g, 8g, 9g)
a similar reaction between a stable triazolinylidene
carbene and [RuCl2(p-cymene)]2 (5). Product 13 was
isolated and characterized by X-ray diffraction.[21]
Investigations are currently under way to synthesize
and to isolate selected ruthenium-arene complexes
bearing NHC ligands on a preparative scale instead of
generating them in situ. Detailed structural analysis of
these new species should help clarify the possible
intervention of ortho-metallation in our systems and
will be reported in due course.
and 2,6-diisopropylphenyl (6h, 7h, 8h, 9h) derivatives were
prepared as described by Arduengo and coworkers.[15] Dia-
mines dihydrochlorides 8a,[23] 8b,[24] and 8c,[25] were synthesized
according to literature procedures.
Caution
4-Aminobiphenyl is listed as a known human carcinogen.[26] It
should be handled only in an efficient fume hood using
appropriate protective equipment. Spills and residues must be
disposed off according to local regulations. Other aromatic
amines are suspected carcinogen agents and should be handled
with great care.
Conclusion
By testing a wide range of imidazolium and imidazoli-
nium salts bearing aromatic substituents as catalyst
modifiers in the ROMP of cyclooctene, we were able to
establish that the presence of a C4-C5 double bond in
the NHC ligand was not critical to reach high catalytic
efficiencies. We have also underscored the existence of
Preparation of N,N'-Diarylethylenediimines (6)
A mixture of glyoxal (36.3 g of a 40% aqueous solution,
0.25 mol), 2-propanol (100 mL), and water (50 mL) was slowly
added to an arylamine (0.5 mol) in 300 mL of 2-propanol. The
reaction mixture was stirred 2.5 days at room temperature. The
resulting suspension was filtered with suction and the precip-
an ortho-effect. Blocking both ortho positions of the 1,3- itate was rinsed with water (2 Â 100 mL). It was dried under an
IRlamp and used without any further purification.
diphenyl rings with alkyl groups is a requisite for
obtaining efficient catalysts. Failure to do so probably
results in the ortho-metallation of the carbene ligand,
thereby altering the coordination sphere of the ruthe-
nium active centers. These two main structure-activity
relationships will guide us in our future endeavors to
improve the catalytic properties of ruthenium-arene
complexes for the ring-opening metathesis polymer-
ization of cyclooctene. Other fields of olefin metathesis
N,N'-Di(p-biphenyl)ethylenediimine (6d): Yellow powder;
yield: 98%; mp > 190 8C (dec.); insoluble in most organic
solvents; 1H NMR(DMSO- d6, 333 K): d ˆ7 8 (m, 18H, CHar),
8.53 (s, 2H, CH N); 13C NMR(DMSO- d6, 333 K):[27] d ˆ 114.1
ˆ
(Car), 121.7 (CHar), 126.3 (CHar), 127.3 (CHar), 128.6 (CHar),
ˆ
139.1 (Car), 139.4 (Car), 148.7 (Car), 159.5 (CH N).
N,N'-Di-(2-tolyl)ethylenediimine (6e): Yellow powder;
yield: 87%; mp 127 128 8C (lit.[16] 122 124 8C); 1H NMR
(CDCl3): d ˆ 2.39 (s, 6H, ortho-CH3), 6.98 (d, J ˆ 7.6 Hz, 2H,
could also benefitfrom these findings, although we share CHar), 7.15 7.24 (m, 6H, CHar), 8.29 (s, 2H, CH N); 13C NMR
ˆ
(CDCl3):[27] dˆ17.9(ortho-CH3),117.3(CHortho),126.9(CHmeta),
F¸rster×s view[13a] that extrapolating the results obtained
for a specific type of metathesis reaction (ring-closing
metathesis, cross-metathesis, ROMP, etc.) or even for a
specific substrate or monomer are often hazardous.
127.4 (CHpara), 130.7 (CH-C-CH3), 132.9 (C-CH3), 149.6 (Cipso),
ˆ
159.7 (CH N).
N,N'-Bis-(2,6-dimethylphenyl)ethylenediimine (6f): Yellow
1
powder; yield: 82%; mp 153 154 8C; H NMR(CDCl ): d ˆ
3
2.18 (s, 12H, ortho-CH3), 7.00 (d, 2H, CHpara), 7.09 (d, 4H,
CHmeta), 8.12 (s, 2H, CH N); 13C NMR(CDCl ):[27] d ˆ 18.4
ˆ
3
(ortho-CH3), 124.9 (CHpara), 126.6 (Cortho), 128.4 (CHmeta), 150.0
Experimental Section
ˆ
(Cipso), 163.6 (CH N).
General Remarks
Preparation of Imidazolium Chlorides (7)
Chlorobenzene was distilled over CaH2 and deoxygenated
prior to use. Polymerizations were carried out under inert
atmosphere using standard Schlenk techniques. GLC analyses
were performed on a Perkin-Elmer 8500 gas chromatograph
equipped with a RSL-150 capillary column and an FID. Gel
permeation chromatographic (GPC) analyses were performed
in THF on a Hewlett-Packard HP 1090 instrument equipped
Paraformaldehyde (0.75 g, 25 mmol) was added to 7.5 mL of a
4 N solution of HCl in dioxane (30 mmol). The mixture was
stirred and gently warmed until complete dissolution of the
solid. In a separate flask, an N,N'-diarylethylenediimine (6)
(20 mmol) was gently warmed in 40 mL of THF until a clear
solution was obtained. The mixture was cooled in an ice-water
754
Adv. Synth. Catal. 2002, 344, 749 756

In Situ Generated Ru Catalysts Bearing NHC Ligands for ROMP of Cyclooctene
FULL PAPERS
bath at 0 8C and the acidic paraformaldehyde solution was
added dropwise in 10 min. A precipitate appeared within a few
min and the resulting suspension was stirred 4 h at room
temperature. It was filtered with suction and the precipitate
was rinsed with AcOEt (3 Â 20 mL) and dried under vacuum.
1,3-Di-(p-biphenyl)imidazolium Chloride (7d): Ochre pow-
der; yield: 72%; 1H NMR(DMSO- d6): d ˆ 7.45 (t, 2H, CHar),
7.53 (t, 4H, CHar), 7.80 (d, 4H, CHar), 8.01 (d, 4H, CHar), 8.11 (d,
4H, CHar), 8.72 (m, 2H, im-H4,5), 10.67 (s, 1H, im-H2); 13C NMR
(DMSO-d6):[27] dˆ121.8 (im-CH4,5), 122.4 (CHar), 126.9 (CHar),
128.1 (CHar), 128.3 (Car), 129.1 (CHar), 133.8 (Car), 134.4 (Car),
138.4 (im-CH2), 141.5 (Car).
Preparation of Imidazolinium Chlorides (9)
An N,N'-diarylethylenediamine dihydrochloride (8) (5 mmol)
was suspended into 25 mL of triethyl orthoformate containing
2 drops of formic acid. The mixture was refluxed for 2 days in
an oil bath at 130 8C. It was then cooled to 6 8C and the
resulting suspension was filtered with suction. The precipitate
was rinsed with a small portion of Et2O and dried under
vacuum.
1,3-Diphenylimidazolinium Chloride (9a): White powder;
yield: 84%; 1H NMR(DMSO- d6): d ˆ 4.60 (s, 4H, CH2), 7.39 (t,
J ˆ 7.2 Hz, 2H, CHpara), 7.56 (t, J ˆ 8.0 Hz, 4H, CHmeta), 7.70 (d,
J ˆ 8.0 Hz, 4H, CHortho), 10.12 (s, 1H, im-H2); 13C NMR
(DMSO-d6):[27] d ˆ 48.3 (CH2), 118.5 (Cortho), 127.0 (CHpara),
129.7 (Cmeta), 136.1 (Cipso), 151.8 (im-CH2).
1,3-Di-(2-tolyl)imidazolium Chloride (7e): The crude dark
brown product (yield: 86%) was dissolved in hot acetonitrile
(50 mL) and treated with activated charcoal. The hot suspen-
sion was filtered and the solvent was evaporated to afford a
1,3-Di-(1-naphtyl)imidazolinium Chloride (9b): White
1
powder; yield: 81%; H NMR(DMSO- d6): d ˆ 4.79 (s, 4H,
1
beige powder; yield: 32%; H NMR(DMSO- d6): d ˆ 2.37 (s,
CH2), 7.32 (t, 4H, CHar) , 7.80 (t, 2H, CHar), 8.01 (d, 2H, CHar),
6H, ortho-CH3), 7.50 (d, 2H, CHar), 7.56 (s, 4H, CHar), 7.67 (s,
2H, CHar), 8.35 (s, 2H, im-H4,5), 9.94 (s, 1H, im-H2); 13C NMR
(DMSO-d6):[27] d ˆ 17.1 (ortho-CH3), 124.1 (im-CH4,5), 126.7
(CHar), 127.3 (CHar), 130.7 (CHar), 131.6 (CHar), 133.4 (C-CH3),
134.1 (Cipso), 137.8 (im-CH2).
8.16 (t, 4H, CHar), 8.43 (d, 2H, CHar), 9.55 (s, 1H, im-H2);
13
C NMR(DMSO- d6):[27] d ˆ 53.7 (CH2), 122.4 (CHar), 124.6
(CHar), 125.7 (CHar), 127.2 (CHar), 127.8 (CHar), 128.1 (Car),
128.6 (CHar), 130.1 (CHar), 132.4 (Car), 133.9 (Cipso), 159.7 (im-
CH2).
1,3-Bis-(2,6-dimethylphenyl)imidazolium Chloride (7f):
1,3-Bis-(3,5-dimethylphenyl)imidazolinium Chloride (9c):
1
Pinkish powder; yield: 55%; H NMR(DMSO- d6): d ˆ 2.18
1
White powder; yield: 80%; H NMR(DMSO- d6): d ˆ 2.35 (s,
(s, 12H, ortho-CH3), 7.39 (d, 4H, CHmeta), 7.49 (t, 2H, CHpara),
8.38 (s, 2H, im-H4,5), 9.88 (s, 1H, im-H2); 13C NMR(DMSO- d6):[27]
d ˆ 17.0 (ortho-CH3), 124.7 (im-CH4,5), 128.9 (CHmeta), 130.8
(CHpara), 133.4 (Cipso), 134.7 (Cortho), 138.5 (im-CH2).
12H, meta-CH3), 4.54 (s, 4H, CH2), 7.03 (s, 2H, CHpara), 7.29 (s,
4H, CHortho), 9.88 (s, 1H, im-H2); 13C NMR(DMSO- d6):[27] d ˆ
20.9 (ortho-CH3), 48.2 (CH2), 115.9 (CHortho), 128.3 (CHpara),
139.1 (Cmeta), 151.3 (Cipso), 162.9 (im-CH2).
1,3-Di-(p-biphenyl)imidazolinium Chloride (9d): Yellow
1
powder; yield: 80%; H NMR(DMSO- d6): d ˆ 4.67 (s, 4H,
CH2), 7.40 (t, 2H, CHar) , 7.50 (t, 4H, CHar), 7.75 (d, 4H, CHar),
7.82 7.90 (m, 8H, CHar), 10.33 (s, 1H, im-H2); 13C NMR
(DMSO-d6):[27] d ˆ 48.3 (CH2), 119.0 (CHar), 126.6 (CHar),
127.2 (CHar), 127.8 (Car), 129.0 (CHar), 135.4 (Car), 138.5 (Car),
138.7 (Car), 151.5 (im-CH2).
Preparation of N,N'-Diarylethylenediamines
Dihydrochlorides (8)
A solution of an N,N×-diarylethylenediimine (6) (50 mmol) in
200 mL of THF was cooled to 0 8C before 8.0 g of sodium
borohydride (211 mmol) were added in small portions over a
1 h period. The reaction mixture was slowly brought back to
room temperature and stirred overnight (16 21 h). It was then
refluxed for 2 h. After cooling to room temperature, 200 g of
ice were added in small portions over a 1 h period followed by
150 mL of cold 3 M aqueous HCl. The resulting suspension was
filtered with suction and the precipitate was rinsed with 50 mL
of cold water. It was dried under vacuum and used without any
further purification.
1,3-Di-(2-tolyl)imidazolinium Chloride (9e): Beige powder;
yield: 55%; 1H NMR(DMSO- d6): d ˆ 2.46 (s, 6H, ortho-CH3),
4.54 (s, 4H, CH2), 7.42 (m, 6H, CHar), 7.60 (m, 2H, CHar), 9.30 (s,
1H, im-H2); 13C NMR(DMSO- d6):[27] d ˆ 17.4 (ortho-CH3),
52.1 (CH2), 125.8 (CHar), 127.2 (CHar), 129.4 (CHar), 131.6
(CHar), 133.6 (C-CH3), 135.0 (Cipso), 158.2 (im-CH2).
1,3-Bis-(2,6-dimethylphenyl)imidazolinium Chloride (9f):
1
White powder; yield: 79%; H NMR(DMSO- d6): d ˆ 2.42 (s,
12H, ortho-CH3), 4.53 (s, 4H, CH2), 7.28-7.36 (m, 6H, CHar),
9.19 (s, 1H, im-H2); 13C NMR(DMSO- d6):[27] d ˆ 17.2 (ortho-
CH3), 50.8 (CH2), 128.9 (CHmeta), 129.9 (CHpara), 133.3 (Cipso),
135.7 (Cortho), 160.1 (im-CH2).
N,N'-Di-(p-biphenyl)ethylenediamine
Dihydrochloride
1
(8d): Off-white powder; yield: 66%; H NMR(DMSO- d6):
d ˆ 3.50 (s, 4H, CH2), 7.17 (d, 4H, CHar), 7.30 (t, 2H, CHar), 7.43
(t, 4H, CHar) , 7.62 (d, 8H, CHar); 13C NMR(DMSO- d6):[27] d ˆ
43.9 (CH2), 125.6 (CHar), 126.0 (CHar), 127.6 (CHar), 128.8
(CHar), 139.8 (Cipso).
Typical Polymerization Procedure
N,N'-Di-(2-toly)ethylenediamine Dihydrochloride (8e):
1
Light gray powder; yield: 58%; H NMR(DMSO- d6): d ˆ
A mixture of [RuCl2(p-cymene)]2 (5) (0.0092 g, 1.5 Â 10À5 mol),
an imidazoli(ni)um salt (3 Â 10À5 mol), and 95 99% potassium
tert-butoxide or 95% sodium hydride (6 Â 10À5 mol) was placed
in a 25 mL round-bottomed flask containing a magnetic
stirring bar and capped with a three-way stopcock. The reactor
was purged of air (three vacuum/argon cycles) before dry
chlorobenzene (5 mL) was added. The solution was warmed to
60 8C in a thermostatted oil bath and irradiated by a 40 W ∫cold
white∫ fluorescent tube placed 10 cm away from the Pyrex
reaction flask. Cyclooctene (1 mL, 7.5 Â 10À3 mol) was added
via a syringe and the reaction mixture was stirred 2 h at 60 8C.
2.32 (s, 6H, ortho-CH3), 3.55 (s, 4H, CH2), 7.02 (m, 2H, CHar),
7.19 (m, 6H, CHar); 13C NMR(DMSO- d6):[27] d ˆ 17.4 (ortho-
CH3), 44.1 (CH2), 117.1 (CHortho), 123.6 (CHpara), 127.1 (CHmeta),
127.8 (C-CH3), 131.3 (CH-C-CH3), 139.4 (Cipso).
N,N'-Bis-(2,6-dimethylphenyl)ethylenediamine Dihydro-
chloride (8f): White pinkish powder; yield: 59%; H NMR
(DMSO-d6): d ˆ 2.34 (s, 12H, ortho-CH3), 3.53 (s, 2H, CH2),
7.02 (s, 6H, CHar); 13C NMR(DMSO- d6):[27] d ˆ 18.2 (ortho-
CH3), 46.3 (CH2), 127.1 (CHpara), 129.7 (Cortho), 131.5 (CHmeta),
136.5 (Cipso).
1
Adv. Synth. Catal. 2002, 344, 749 756
755

Lionel Delaude et al.
FULL PAPERS
The conversion was monitored by gas chromatography using
the cyclooctane impurity of cyclooctene as an internal stand-
ard. The resulting gel was diluted with CHCl3 (2 Â 10 mL) and
slowly poured into 500 mL of MeOH under vigorous stirring.
The precipitated polyoctenamer was dried under high vacuum
and characterized by NMRand GPC.
[9] L. Delaude, A. Demonceau, A. F. Noels, Chem. Com-
mun. 2001, 986 987.
[10] a) P. A. van der Schaaf, A. Hafner, A. M¸hlebach,
Angew. Chem. Int. Ed. 1996, 35, 1845 1847; b) A.
Hafner, A. M¸hlebach, P. A. van der Schaaf, Angew.
Chem. Int. Ed. 1997, 36, 2121 2124.
[11] a) L. Ackermann, C. Bruneau, P. H. Dixneuf, Synlett
¬
¬
2001, 397 399; b) D. Semeril, M. Cleran, C. Bruneau,
P. H. Dixneuf, Adv. Synth. Catal. 2001, 343, 184 187.
[12] a) C. Zhang, J. Huang, M. L. Trudell, S. P. Nolan, J. Org.
Chem. 1999, 64, 3804 3805; b) J. Huang, S. P. Nolan, J.
Am. Chem. Soc. 1999, 121, 9889 9890; c) G. A. Grasa,
M. S. Viciu, J. Huang, S. P. Nolan, J. Org. Chem. 2001, 66,
7729 7737; d) G. A. Grasa, S. P. Nolan, Org. Lett. 2001,
3, 119 122.
[13] See, for example: a) A. F¸rstner, L. Ackermann, B.
Gabor, R. Goddard, C. W. Lehmann, R. Mynott, F.
Stelzer, O. R. Thiel, Chem. Eur. J. 2001, 7, 3236 3253;
b) A. A. D. Tulloch, A. A. Danopoulos, S. Winston, S.
Kleinhenz, G. Eastham, J. Chem. Soc. Dalton Trans.
2000, 4499 4506; c) M. T. Powell, D. -R. Hou, M. C.
Perry, X. Cui, K. Burgess, J. Am. Chem. Soc. 2001, 123,
8878 8879.
[14] K. J. Harlow, A. F. Hill, T. Welton, Synthesis 1996, 697
698.
[15] A. J. Arduengo III, R. Krafczyk, R. Schmutzler, H. A.
Craig, J. N. Goerlich, W. J. Marshall, M. Unverzagt,
Tetrahedron 1999, 55, 14523 14534.
[16] J. M. Kliegman, R. K. Barnes, J. Org. Chem. 1970, 35,
3140 3143.
[17] S. P. Nolan, personal communication, Fourteenth Inter-
national Symposium on Olefin Metathesis, M.I.T. (Cam-
bridge, Massachusetts), 2001.
Acknowledgements
The financial support of the EU, through grant TMR-HPRN CT
2000-10 ™Polycat∫, is gratefully acknowledged.
References and Notes
[1] a) K. J. Ivin, J. C. Mol, Olefin Metathesis and Metathesis
Polymerization, Academic Press, San Diego, 1997;
b) T. M. Trnak, R. H. Grubbs, Acc. Chem. Res. 2001,
34, 18 29.
[2] a) P. Schwab, M. B. France, J. W. Ziller, R. H. Grubbs,
Angew. Chem. Int. Ed. 1995, 34, 2039 2041; b) P.
Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc.
1996, 118, 100 110; c) D. M. Lynn, S. Kanaoka, R. H.
Grubbs, J. Am. Chem. Soc. 1996, 118, 784 790.
[3] a) A. J. Arduengo III, R. L. Harlow, M. Kline, J. Am.
Chem. Soc. 1991, 113, 361 363; b) A. J. Arduengo III,
H. V. Rasika Dias, R. L. Harlow, M. Kline, J. Am. Chem.
Soc. 1992, 114, 5530 5534; c) A. J. Arduengo III, F.
Davidson, H. V. R. Dias, J. R. Goerlich, D. Khasnis, W. J.
Marshall, T. K. Prakasha, J. Am. Chem. Soc. 1997, 119,
12742 12749; d) A. J. Arduengo III, J. R. Goerlich, R.
Krafczyk, W. J. Marshall, Angew. Chem. Int. Ed. 1998,
37, 1963 1965; e) A. J. Arduengo III, Acc. Chem. Res.
1999, 32, 913 921.
[18] D. Jan, L. Delaude, F. Simal, A. Demonceau, A. F. Noels,
J. Organomet. Chem. 2000, 606, 55 64.
[4] D. Bourissou, O. Guerret, F. P. GabbaÔ, G. Bertrand,
Chem. Rev. 2000, 100, 39 91.
[19] a) M. A. Bennett, T.-N. Huang, T. W. Turney, Chem.
Commun. 1979, 312 314; b) M. A. Bennett, J. P. Ennett,
K. I. Gell, J. Organomet. Chem. 1982, 233, C17 C20;
c) M. A. Bennett, J. P. Ennett, Organometallics 1984, 3,
1365 1374.
[20] P. B. Hitchcock, M. F. Lappert, P. L. Pye, S. Thomas, J.
Chem. Soc, Dalton Trans. 1979, 1929 1942.
[21] F. Simal, D. Jan, L. Delaude, A. Demonceau, M.-R.
Spirlet, A. F. Noels, Can. J. Chem. 2001, 79, 529 535.
[22] M. A. Bennett, A. K. Smith, J. Chem. Soc. Dalton Trans.
1974, 233 241.
[23] A. E. Schouten, Recl. Trav. Chim. Pays-Bas 1937, 56,
541 561.
[24] G. B. Russell, G. J. Sutherland, R. D. Topsom, J. Vaugh-
an, J. Org. Chem. 1962, 27, 4375 4377.
[25] R. Knorr, R. Zˆlch, K. Poldorn, Heterocycles 1995, 40,
559 576.
[26] IARC Monographs on the Evaluation of the Carcino-
genic Risk of Chemicals to Humans, Vol. 1, IARC Press,
Lyon, 1972, pp. 74 79.
[5] a) W. A. Herrmann, C. Kˆcher, Angew. Chem. Int. Ed.
1997, 36, 2162 2187; b) T. Weskamp, V. P. W. Bˆhm,
W. A. Herrmann, J. Organomet. Chem. 2000, 600, 12 22;
c) L. Jafarpour, S. P. Nolan, Adv. Organomet. Chem.
2001, 46, 181 222; d) W. A. Herrmann, Angew. Chem.
Int. Ed. 2002, 41, 1290 1309.
[6] a) T. Weskamp, W. C. Schattenmann, M. Spiegler, W. A.
Herrmann, Angew. Chem. Int. Ed. 1998, 37, 2490 2493;
b) T. Weskamp, W. C. Schattenmann, M. Spiegler, W. A.
Herrmann, Angew. Chem. Int. Ed. 1999, 38, 262.
[7] a) U. Frenzel, T. Weskamp, F. J. Kohl, W. C. Schatten-
mann, O. Nuyken, W. A. Herrmann, J. Organomet.
Chem. 1999, 586, 263 265; b) T. Weskamp, F. J. Kohl,
W. Hieringer, D. Gleich, W. A. Herrmann, Angew.
Chem. Int. Ed. 1999, 38, 2416 2419; c) C. W. Bielawski,
R. H. Grubbs, Angew. Chem. Int. Ed. 2000, 39, 2903
2906.
[8] a) A. W. Stumpf, E. Saive, A. Demonceau, A. F. Noels,
Chem. Commun. 1995, 1127 1128; b) A. Demonceau,
A. W. Stumpf, E. Saive, A. F. Noels, Macromolecules
1997, 30, 3127 3136.
[27] Tentative assignments. Relative positions on the aro-
matic cycles (ortho, meta, para) refer to the carbon
attached to the imidazole ring (labelled as ipso).
756
Adv. Synth. Catal. 2002, 344, 749 756